From 2a0168fea37b68ad14e2cb60bf215111e49d4870 Mon Sep 17 00:00:00 2001 From: xleroy Date: Wed, 23 Jul 2014 08:54:56 +0000 Subject: Merge of "newspilling" branch: - Support single-precision floats as first-class values - Introduce chunks Many32, Many64 and types Tany32, Tany64 to support saving and restoring registers without knowing the exact types (int/single/float) of their contents, just their sizes. - Memory model: generalize the opaque encoding of pointers to apply to any value, not just pointers, if chunks Many32/Many64 are selected. - More properties of FP arithmetic proved. git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2537 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e --- .depend | 23 +- Makefile | 3 +- arm/Archi.v | 18 +- arm/Asm.v | 79 +- arm/Asmgen.v | 95 +- arm/Asmgenproof.v | 9 +- arm/Asmgenproof1.v | 255 +++-- arm/ConstpropOp.vp | 30 +- arm/ConstpropOpproof.v | 68 +- arm/Machregs.v | 28 +- arm/NeedOp.v | 9 +- arm/Op.v | 112 +- arm/PrintAsm.ml | 121 +- arm/SelectOp.vp | 34 +- arm/SelectOpproof.v | 76 ++ arm/ValueAOp.v | 17 + arm/eabi/Conventions1.v | 52 +- arm/hardfloat/Conventions1.v | 97 +- backend/Allocation.v | 116 +- backend/Allocproof.v | 227 ++-- backend/CMtypecheck.ml | 27 +- backend/CSE.v | 3 +- backend/CSEproof.v | 1 - backend/Cminor.v | 62 +- backend/Constprop.v | 1 + backend/Constpropproof.v | 2 + backend/IRC.ml | 53 +- backend/Inliningproof.v | 1 + backend/Lineartyping.v | 786 +++---------- backend/Locations.v | 61 +- backend/NeedDomain.v | 101 +- backend/PrintCminor.ml | 24 +- backend/PrintXTL.ml | 2 + backend/RTLtyping.v | 88 +- backend/Regalloc.ml | 60 +- backend/SelectDiv.vp | 12 + backend/SelectDivproof.v | 16 + backend/SelectLong.vp | 4 + backend/SelectLongproof.v | 28 + backend/Selection.v | 15 + backend/Selectionproof.v | 15 + backend/Stacking.v | 8 +- backend/Stackingproof.v | 111 +- backend/ValueAnalysis.v | 4 +- backend/ValueDomain.v | 317 +++-- cfrontend/C2C.ml | 18 +- cfrontend/Cexec.v | 19 +- cfrontend/Clight.v | 6 +- cfrontend/Cminorgen.v | 2 + cfrontend/Cminorgenproof.v | 24 +- cfrontend/Cop.v | 274 +++-- cfrontend/Csharpminor.v | 4 +- cfrontend/Cshmgen.v | 110 +- cfrontend/Cshmgenproof.v | 131 ++- cfrontend/Ctypes.v | 21 +- cfrontend/Initializers.v | 4 +- cfrontend/Initializersproof.v | 21 +- cfrontend/PrintClight.ml | 4 + cfrontend/PrintCsyntax.ml | 2 + cfrontend/SimplExpr.v | 3 + cfrontend/SimplExprspec.v | 4 + cfrontend/SimplLocals.v | 5 +- cfrontend/SimplLocalsproof.v | 57 +- common/AST.v | 68 +- common/Events.v | 15 +- common/Globalenvs.v | 6 +- common/Memdata.v | 529 +++++---- common/Memory.v | 325 +++--- common/Memtype.v | 20 +- common/PrintAST.ml | 4 + common/Values.v | 139 ++- driver/Interp.ml | 2 +- exportclight/ExportClight.ml | 6 +- flocq/Appli/Fappli_IEEE_bits.v | 51 + ia32/Archi.v | 15 +- ia32/Asm.v | 99 +- ia32/Asmgen.v | 101 +- ia32/Asmgenproof.v | 15 +- ia32/Asmgenproof1.v | 213 +++- ia32/ConstpropOp.vp | 14 +- ia32/ConstpropOpproof.v | 56 +- ia32/Machregs.v | 30 +- ia32/NeedOp.v | 10 +- ia32/Op.v | 134 ++- ia32/PrintAsm.ml | 103 +- ia32/PrintOp.ml | 2 + ia32/SelectOp.vp | 35 +- ia32/SelectOpproof.v | 102 +- ia32/Unusedglob1.ml | 6 +- ia32/ValueAOp.v | 13 + ia32/standard/Conventions1.v | 74 +- lib/Camlcoq.ml | 9 +- lib/Fappli_IEEE_extra.v | 1506 ++++++++++++++++++++++++ lib/Floats.v | 2482 ++++++++++++++++------------------------ powerpc/Archi.v | 14 +- powerpc/Asm.v | 60 +- powerpc/Asmgen.v | 79 +- powerpc/Asmgenproof.v | 14 +- powerpc/Asmgenproof1.v | 135 +-- powerpc/ConstpropOp.vp | 14 +- powerpc/ConstpropOpproof.v | 56 +- powerpc/Machregs.v | 37 +- powerpc/NeedOp.v | 8 +- powerpc/Op.v | 137 +-- powerpc/PrintAsm.ml | 77 +- powerpc/SelectOp.vp | 29 +- powerpc/SelectOpproof.v | 124 +- powerpc/Unusedglob1.ml | 4 + powerpc/ValueAOp.v | 9 + powerpc/eabi/Conventions1.v | 64 +- test/c/fftsp.c | 12 +- test/regression/NaNs.c | 48 +- test/spass/Makefile.bak | 13 - 113 files changed, 6831 insertions(+), 4267 deletions(-) create mode 100644 lib/Fappli_IEEE_extra.v delete mode 100644 test/spass/Makefile.bak diff --git a/.depend b/.depend index 0529e871..5c27a2cb 100644 --- a/.depend +++ b/.depend @@ -8,7 +8,8 @@ lib/Ordered.vo lib/Ordered.glob lib/Ordered.v.beautified: lib/Ordered.v lib/Coql lib/Iteration.vo lib/Iteration.glob lib/Iteration.v.beautified: lib/Iteration.v lib/Axioms.vo lib/Coqlib.vo lib/Wfsimpl.vo lib/Integers.vo lib/Integers.glob lib/Integers.v.beautified: lib/Integers.v lib/Coqlib.vo $(ARCH)/Archi.vo $(ARCH)/Archi.glob $(ARCH)/Archi.v.beautified: $(ARCH)/Archi.v flocq/Appli/Fappli_IEEE.vo flocq/Appli/Fappli_IEEE_bits.vo -lib/Floats.vo lib/Floats.glob lib/Floats.v.beautified: lib/Floats.v lib/Axioms.vo lib/Coqlib.vo lib/Integers.vo flocq/Appli/Fappli_IEEE.vo flocq/Appli/Fappli_IEEE_bits.vo flocq/Core/Fcore.vo flocq/Calc/Fcalc_round.vo flocq/Calc/Fcalc_bracket.vo flocq/Prop/Fprop_Sterbenz.vo $(ARCH)/Archi.vo +lib/Fappli_IEEE_extra.vo lib/Fappli_IEEE_extra.glob lib/Fappli_IEEE_extra.v.beautified: lib/Fappli_IEEE_extra.v flocq/Core/Fcore.vo flocq/Core/Fcore_digits.vo flocq/Calc/Fcalc_digits.vo flocq/Calc/Fcalc_ops.vo flocq/Calc/Fcalc_round.vo flocq/Calc/Fcalc_bracket.vo flocq/Prop/Fprop_Sterbenz.vo flocq/Appli/Fappli_IEEE.vo flocq/Appli/Fappli_rnd_odd.vo flocq/Appli/Fappli_IEEE_bits.vo +lib/Floats.vo lib/Floats.glob lib/Floats.v.beautified: lib/Floats.v lib/Coqlib.vo lib/Integers.vo flocq/Appli/Fappli_IEEE.vo flocq/Appli/Fappli_IEEE_bits.vo lib/Fappli_IEEE_extra.vo flocq/Core/Fcore.vo $(ARCH)/Archi.vo lib/Parmov.vo lib/Parmov.glob lib/Parmov.v.beautified: lib/Parmov.v lib/Axioms.vo lib/Coqlib.vo lib/UnionFind.vo lib/UnionFind.glob lib/UnionFind.v.beautified: lib/UnionFind.v lib/Coqlib.vo lib/Wfsimpl.vo lib/Wfsimpl.glob lib/Wfsimpl.v.beautified: lib/Wfsimpl.v lib/Axioms.vo @@ -29,13 +30,13 @@ common/Switch.vo common/Switch.glob common/Switch.v.beautified: common/Switch.v common/Determinism.vo common/Determinism.glob common/Determinism.v.beautified: common/Determinism.v lib/Coqlib.vo common/AST.vo lib/Integers.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo common/Behaviors.vo common/Unityping.vo common/Unityping.glob common/Unityping.v.beautified: common/Unityping.v lib/Coqlib.vo lib/Maps.vo common/Errors.vo backend/Cminor.vo backend/Cminor.glob backend/Cminor.v.beautified: backend/Cminor.v lib/Coqlib.vo lib/Maps.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Events.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Smallstep.vo common/Switch.vo -$(ARCH)/Op.vo $(ARCH)/Op.glob $(ARCH)/Op.v.beautified: $(ARCH)/Op.v lib/Coqlib.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo +$(ARCH)/Op.vo $(ARCH)/Op.glob $(ARCH)/Op.v.beautified: $(ARCH)/Op.v lib/Axioms.vo lib/Coqlib.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo backend/CminorSel.vo backend/CminorSel.glob backend/CminorSel.v.beautified: backend/CminorSel.v lib/Coqlib.vo lib/Maps.vo common/AST.vo lib/Integers.vo common/Events.vo common/Values.vo common/Memory.vo backend/Cminor.vo $(ARCH)/Op.vo common/Globalenvs.vo common/Switch.vo common/Smallstep.vo $(ARCH)/SelectOp.vo $(ARCH)/SelectOp.glob $(ARCH)/SelectOp.v.beautified: $(ARCH)/SelectOp.v lib/Coqlib.vo driver/Compopts.vo common/AST.vo lib/Integers.vo lib/Floats.vo $(ARCH)/Op.vo backend/CminorSel.vo backend/SelectDiv.vo backend/SelectDiv.glob backend/SelectDiv.v.beautified: backend/SelectDiv.v lib/Coqlib.vo driver/Compopts.vo common/AST.vo lib/Integers.vo lib/Floats.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo backend/SelectLong.vo backend/SelectLong.glob backend/SelectLong.v.beautified: backend/SelectLong.v lib/Coqlib.vo common/AST.vo lib/Integers.vo lib/Floats.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo common/Errors.vo backend/Selection.vo backend/Selection.glob backend/Selection.v.beautified: backend/Selection.v lib/Coqlib.vo common/AST.vo common/Errors.vo lib/Integers.vo common/Globalenvs.vo backend/Cminor.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo backend/SelectDiv.vo backend/SelectLong.vo -$(ARCH)/SelectOpproof.vo $(ARCH)/SelectOpproof.glob $(ARCH)/SelectOpproof.v.beautified: $(ARCH)/SelectOpproof.v lib/Coqlib.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo backend/Cminor.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo +$(ARCH)/SelectOpproof.vo $(ARCH)/SelectOpproof.glob $(ARCH)/SelectOpproof.v.beautified: $(ARCH)/SelectOpproof.v lib/Coqlib.vo lib/Maps.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo backend/Cminor.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo backend/SelectDivproof.vo backend/SelectDivproof.glob backend/SelectDivproof.v.beautified: backend/SelectDivproof.v lib/Coqlib.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo backend/Cminor.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo $(ARCH)/SelectOpproof.vo backend/SelectDiv.vo backend/SelectLongproof.vo backend/SelectLongproof.glob backend/SelectLongproof.v.beautified: backend/SelectLongproof.v lib/Coqlib.vo common/AST.vo common/Errors.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo backend/Cminor.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo $(ARCH)/SelectOpproof.vo backend/SelectLong.vo backend/Selectionproof.vo backend/Selectionproof.glob backend/Selectionproof.v.beautified: backend/Selectionproof.v lib/Coqlib.vo lib/Maps.vo common/AST.vo common/Errors.vo lib/Integers.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo backend/Cminor.vo $(ARCH)/Op.vo backend/CminorSel.vo $(ARCH)/SelectOp.vo backend/SelectDiv.vo backend/SelectLong.vo backend/Selection.vo $(ARCH)/SelectOpproof.vo backend/SelectDivproof.vo backend/SelectLongproof.vo @@ -59,20 +60,20 @@ $(ARCH)/ValueAOp.vo $(ARCH)/ValueAOp.glob $(ARCH)/ValueAOp.v.beautified: $(ARCH) backend/ValueAnalysis.vo backend/ValueAnalysis.glob backend/ValueAnalysis.v.beautified: backend/ValueAnalysis.v lib/Coqlib.vo lib/Maps.vo driver/Compopts.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo lib/Lattice.vo backend/Kildall.vo backend/Registers.vo $(ARCH)/Op.vo backend/RTL.vo backend/ValueDomain.vo $(ARCH)/ValueAOp.vo backend/Liveness.vo lib/Axioms.vo $(ARCH)/ConstpropOp.vo $(ARCH)/ConstpropOp.glob $(ARCH)/ConstpropOp.v.beautified: $(ARCH)/ConstpropOp.v lib/Coqlib.vo driver/Compopts.vo common/AST.vo lib/Integers.vo lib/Floats.vo $(ARCH)/Op.vo backend/Registers.vo backend/ValueDomain.vo backend/Constprop.vo backend/Constprop.glob backend/Constprop.v.beautified: backend/Constprop.v lib/Coqlib.vo lib/Maps.vo common/AST.vo lib/Integers.vo lib/Floats.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo lib/Lattice.vo backend/Kildall.vo backend/Liveness.vo backend/ValueDomain.vo $(ARCH)/ValueAOp.vo backend/ValueAnalysis.vo $(ARCH)/ConstpropOp.vo -$(ARCH)/ConstpropOpproof.vo $(ARCH)/ConstpropOpproof.glob $(ARCH)/ConstpropOpproof.v.beautified: $(ARCH)/ConstpropOpproof.v lib/Coqlib.vo driver/Compopts.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo backend/ValueDomain.vo $(ARCH)/ConstpropOp.vo +$(ARCH)/ConstpropOpproof.vo $(ARCH)/ConstpropOpproof.glob $(ARCH)/ConstpropOpproof.v.beautified: $(ARCH)/ConstpropOpproof.v lib/Coqlib.vo driver/Compopts.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo backend/ValueDomain.vo $(ARCH)/ConstpropOp.vo backend/Constpropproof.vo backend/Constpropproof.glob backend/Constpropproof.v.beautified: backend/Constpropproof.v lib/Coqlib.vo lib/Maps.vo driver/Compopts.vo common/AST.vo lib/Integers.vo common/Values.vo common/Events.vo common/Memory.vo common/Globalenvs.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo lib/Lattice.vo backend/Kildall.vo backend/ValueDomain.vo $(ARCH)/ValueAOp.vo backend/ValueAnalysis.vo $(ARCH)/ConstpropOp.vo backend/Constprop.vo $(ARCH)/ConstpropOpproof.vo backend/CSEdomain.vo backend/CSEdomain.glob backend/CSEdomain.v.beautified: backend/CSEdomain.v lib/Coqlib.vo lib/Maps.vo common/AST.vo common/Values.vo common/Memory.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo $(ARCH)/CombineOp.vo $(ARCH)/CombineOp.glob $(ARCH)/CombineOp.v.beautified: $(ARCH)/CombineOp.v lib/Coqlib.vo common/AST.vo lib/Integers.vo $(ARCH)/Op.vo backend/CSEdomain.vo backend/CSE.vo backend/CSE.glob backend/CSE.v.beautified: backend/CSE.v lib/Coqlib.vo lib/Maps.vo common/Errors.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo backend/ValueDomain.vo backend/ValueAnalysis.vo backend/CSEdomain.vo backend/Kildall.vo $(ARCH)/CombineOp.vo -$(ARCH)/CombineOpproof.vo $(ARCH)/CombineOpproof.glob $(ARCH)/CombineOpproof.v.beautified: $(ARCH)/CombineOpproof.v lib/Coqlib.vo lib/Integers.vo common/Values.vo common/Memory.vo $(ARCH)/Op.vo backend/RTL.vo backend/CSEdomain.vo $(ARCH)/CombineOp.vo +$(ARCH)/CombineOpproof.vo $(ARCH)/CombineOpproof.glob $(ARCH)/CombineOpproof.v.beautified: $(ARCH)/CombineOpproof.v lib/Coqlib.vo common/AST.vo lib/Integers.vo common/Values.vo common/Memory.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo backend/CSEdomain.vo $(ARCH)/CombineOp.vo backend/CSEproof.vo backend/CSEproof.glob backend/CSEproof.v.beautified: backend/CSEproof.v lib/Coqlib.vo lib/Maps.vo common/AST.vo common/Errors.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo backend/Kildall.vo backend/ValueDomain.vo $(ARCH)/ValueAOp.vo backend/ValueAnalysis.vo backend/CSEdomain.vo $(ARCH)/CombineOp.vo $(ARCH)/CombineOpproof.vo backend/CSE.vo backend/NeedDomain.vo backend/NeedDomain.glob backend/NeedDomain.v.beautified: backend/NeedDomain.v lib/Coqlib.vo lib/Maps.vo lib/IntvSets.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo lib/Lattice.vo backend/Registers.vo backend/ValueDomain.vo $(ARCH)/Op.vo backend/RTL.vo $(ARCH)/NeedOp.vo $(ARCH)/NeedOp.glob $(ARCH)/NeedOp.v.beautified: $(ARCH)/NeedOp.v lib/Coqlib.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo $(ARCH)/Op.vo backend/NeedDomain.vo backend/RTL.vo backend/Deadcode.vo backend/Deadcode.glob backend/Deadcode.v.beautified: backend/Deadcode.v lib/Coqlib.vo common/Errors.vo lib/Maps.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Memory.vo backend/Registers.vo $(ARCH)/Op.vo backend/RTL.vo lib/Lattice.vo backend/Kildall.vo backend/ValueDomain.vo backend/ValueAnalysis.vo backend/NeedDomain.vo $(ARCH)/NeedOp.vo backend/Deadcodeproof.vo backend/Deadcodeproof.glob backend/Deadcodeproof.v.beautified: backend/Deadcodeproof.v lib/Coqlib.vo common/Errors.vo lib/Maps.vo lib/IntvSets.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo lib/Lattice.vo backend/Kildall.vo backend/ValueDomain.vo backend/ValueAnalysis.vo backend/NeedDomain.vo $(ARCH)/NeedOp.vo backend/Deadcode.vo -$(ARCH)/Machregs.vo $(ARCH)/Machregs.glob $(ARCH)/Machregs.v.beautified: $(ARCH)/Machregs.v lib/Coqlib.vo lib/Maps.vo common/AST.vo lib/Integers.vo $(ARCH)/Op.vo +$(ARCH)/Machregs.vo $(ARCH)/Machregs.glob $(ARCH)/Machregs.v.beautified: $(ARCH)/Machregs.v lib/Coqlib.vo lib/Maps.vo common/AST.vo $(ARCH)/Op.vo backend/Locations.vo backend/Locations.glob backend/Locations.v.beautified: backend/Locations.v lib/Coqlib.vo lib/Maps.vo lib/Ordered.vo common/AST.vo common/Values.vo $(ARCH)/Machregs.vo -$(ARCH)/$(VARIANT)/Conventions1.vo $(ARCH)/$(VARIANT)/Conventions1.glob $(ARCH)/$(VARIANT)/Conventions1.v.beautified: $(ARCH)/$(VARIANT)/Conventions1.v lib/Coqlib.vo common/AST.vo backend/Locations.vo +$(ARCH)/$(VARIANT)/Conventions1.vo $(ARCH)/$(VARIANT)/Conventions1.glob $(ARCH)/$(VARIANT)/Conventions1.v.beautified: $(ARCH)/$(VARIANT)/Conventions1.v lib/Coqlib.vo common/AST.vo common/Events.vo backend/Locations.vo backend/Conventions.vo backend/Conventions.glob backend/Conventions.v.beautified: backend/Conventions.v lib/Coqlib.vo common/AST.vo backend/Locations.vo $(ARCH)/$(VARIANT)/Conventions1.vo backend/LTL.vo backend/LTL.glob backend/LTL.v.beautified: backend/LTL.v lib/Coqlib.vo lib/Maps.vo common/AST.vo lib/Integers.vo common/Values.vo common/Events.vo common/Memory.vo common/Globalenvs.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Locations.vo backend/Conventions.vo backend/Allocation.vo backend/Allocation.glob backend/Allocation.v.beautified: backend/Allocation.v lib/FSetAVLplus.vo $(ARCH)/Archi.vo lib/Coqlib.vo lib/Ordered.vo common/Errors.vo lib/Maps.vo lib/Lattice.vo common/AST.vo lib/Integers.vo common/Memdata.vo $(ARCH)/Op.vo backend/Registers.vo backend/RTL.vo backend/Kildall.vo backend/Locations.vo backend/Conventions.vo backend/RTLtyping.vo backend/LTL.vo @@ -80,7 +81,7 @@ backend/Allocproof.vo backend/Allocproof.glob backend/Allocproof.v.beautified: b backend/Tunneling.vo backend/Tunneling.glob backend/Tunneling.v.beautified: backend/Tunneling.v lib/Coqlib.vo lib/Maps.vo lib/UnionFind.vo common/AST.vo backend/LTL.vo backend/Tunnelingproof.vo backend/Tunnelingproof.glob backend/Tunnelingproof.v.beautified: backend/Tunnelingproof.v lib/Coqlib.vo lib/Maps.vo lib/UnionFind.vo common/AST.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Locations.vo backend/LTL.vo backend/Tunneling.vo backend/Linear.vo backend/Linear.glob backend/Linear.v.beautified: backend/Linear.v lib/Coqlib.vo common/AST.vo lib/Integers.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Locations.vo backend/LTL.vo backend/Conventions.vo -backend/Lineartyping.vo backend/Lineartyping.glob backend/Lineartyping.v.beautified: backend/Lineartyping.v lib/Coqlib.vo lib/Ordered.vo lib/Maps.vo lib/Iteration.vo common/AST.vo lib/Integers.vo common/Values.vo common/Events.vo $(ARCH)/Op.vo $(ARCH)/Machregs.vo backend/Locations.vo backend/Conventions.vo backend/LTL.vo backend/Linear.vo common/Globalenvs.vo common/Memory.vo +backend/Lineartyping.vo backend/Lineartyping.glob backend/Lineartyping.v.beautified: backend/Lineartyping.v lib/Coqlib.vo common/AST.vo lib/Integers.vo common/Values.vo common/Globalenvs.vo common/Memory.vo common/Events.vo $(ARCH)/Op.vo $(ARCH)/Machregs.vo backend/Locations.vo backend/Conventions.vo backend/LTL.vo backend/Linear.vo backend/Linearize.vo backend/Linearize.glob backend/Linearize.v.beautified: backend/Linearize.v lib/Coqlib.vo lib/Maps.vo lib/Ordered.vo common/AST.vo common/Errors.vo $(ARCH)/Op.vo backend/Locations.vo backend/LTL.vo backend/Linear.vo backend/Kildall.vo lib/Lattice.vo backend/Linearizeproof.vo backend/Linearizeproof.glob backend/Linearizeproof.v.beautified: backend/Linearizeproof.v lib/Coqlib.vo lib/Maps.vo lib/Ordered.vo lib/Lattice.vo common/AST.vo lib/Integers.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Errors.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Locations.vo backend/LTL.vo backend/Linear.vo backend/Linearize.vo backend/CleanupLabels.vo backend/CleanupLabels.glob backend/CleanupLabels.v.beautified: backend/CleanupLabels.v lib/Coqlib.vo lib/Ordered.vo backend/Linear.vo @@ -91,10 +92,10 @@ $(ARCH)/$(VARIANT)/Stacklayout.vo $(ARCH)/$(VARIANT)/Stacklayout.glob $(ARCH)/$( backend/Stacking.vo backend/Stacking.glob backend/Stacking.v.beautified: backend/Stacking.v lib/Coqlib.vo common/Errors.vo common/AST.vo lib/Integers.vo $(ARCH)/Op.vo backend/Locations.vo backend/Linear.vo backend/Bounds.vo backend/Mach.vo backend/Conventions.vo $(ARCH)/$(VARIANT)/Stacklayout.vo backend/Lineartyping.vo backend/Stackingproof.vo backend/Stackingproof.glob backend/Stackingproof.v.beautified: backend/Stackingproof.v lib/Coqlib.vo common/Errors.vo common/AST.vo lib/Integers.vo common/Values.vo $(ARCH)/Op.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo backend/Locations.vo backend/LTL.vo backend/Linear.vo backend/Lineartyping.vo backend/Mach.vo backend/Bounds.vo backend/Conventions.vo $(ARCH)/$(VARIANT)/Stacklayout.vo backend/Stacking.vo $(ARCH)/Asm.vo $(ARCH)/Asm.glob $(ARCH)/Asm.v.beautified: $(ARCH)/Asm.v lib/Coqlib.vo lib/Maps.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo backend/Locations.vo $(ARCH)/$(VARIANT)/Stacklayout.vo backend/Conventions.vo -$(ARCH)/Asmgen.vo $(ARCH)/Asmgen.glob $(ARCH)/Asmgen.v.beautified: $(ARCH)/Asmgen.v lib/Coqlib.vo common/Errors.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Memdata.vo $(ARCH)/Op.vo backend/Locations.vo backend/Mach.vo $(ARCH)/Asm.vo +$(ARCH)/Asmgen.vo $(ARCH)/Asmgen.glob $(ARCH)/Asmgen.v.beautified: $(ARCH)/Asmgen.v lib/Coqlib.vo common/Errors.vo common/AST.vo lib/Integers.vo lib/Floats.vo $(ARCH)/Op.vo backend/Locations.vo backend/Mach.vo $(ARCH)/Asm.vo backend/Asmgenproof0.vo backend/Asmgenproof0.glob backend/Asmgenproof0.v.beautified: backend/Asmgenproof0.v lib/Coqlib.vo lib/Intv.vo common/AST.vo common/Errors.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo common/Events.vo common/Smallstep.vo backend/Locations.vo backend/Mach.vo $(ARCH)/Asm.vo $(ARCH)/Asmgen.vo backend/Conventions.vo -$(ARCH)/Asmgenproof1.vo $(ARCH)/Asmgenproof1.glob $(ARCH)/Asmgenproof1.v.beautified: $(ARCH)/Asmgenproof1.v lib/Coqlib.vo common/AST.vo common/Errors.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo $(ARCH)/Op.vo backend/Locations.vo backend/Mach.vo $(ARCH)/Asm.vo $(ARCH)/Asmgen.vo backend/Asmgenproof0.vo backend/Conventions.vo -$(ARCH)/Asmgenproof.vo $(ARCH)/Asmgenproof.glob $(ARCH)/Asmgenproof.v.beautified: $(ARCH)/Asmgenproof.v lib/Coqlib.vo common/Errors.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Locations.vo backend/Mach.vo backend/Conventions.vo $(ARCH)/Asm.vo $(ARCH)/Asmgen.vo backend/Asmgenproof0.vo $(ARCH)/Asmgenproof1.vo +$(ARCH)/Asmgenproof1.vo $(ARCH)/Asmgenproof1.glob $(ARCH)/Asmgenproof1.v.beautified: $(ARCH)/Asmgenproof1.v lib/Coqlib.vo common/Errors.vo lib/Maps.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Globalenvs.vo $(ARCH)/Op.vo backend/Locations.vo backend/Mach.vo $(ARCH)/Asm.vo $(ARCH)/Asmgen.vo backend/Conventions.vo backend/Asmgenproof0.vo +$(ARCH)/Asmgenproof.vo $(ARCH)/Asmgenproof.glob $(ARCH)/Asmgenproof.v.beautified: $(ARCH)/Asmgenproof.v lib/Coqlib.vo lib/Maps.vo common/Errors.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo common/Events.vo common/Globalenvs.vo common/Smallstep.vo $(ARCH)/Op.vo backend/Locations.vo backend/Conventions.vo backend/Mach.vo $(ARCH)/Asm.vo $(ARCH)/Asmgen.vo backend/Asmgenproof0.vo $(ARCH)/Asmgenproof1.vo cfrontend/Ctypes.vo cfrontend/Ctypes.glob cfrontend/Ctypes.v.beautified: cfrontend/Ctypes.v lib/Coqlib.vo common/AST.vo common/Errors.vo $(ARCH)/Archi.vo cfrontend/Cop.vo cfrontend/Cop.glob cfrontend/Cop.v.beautified: cfrontend/Cop.v lib/Coqlib.vo common/AST.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/Memory.vo cfrontend/Ctypes.vo cfrontend/Csyntax.vo cfrontend/Csyntax.glob cfrontend/Csyntax.v.beautified: cfrontend/Csyntax.v lib/Coqlib.vo lib/Integers.vo lib/Floats.vo common/Values.vo common/AST.vo cfrontend/Ctypes.vo cfrontend/Cop.vo diff --git a/Makefile b/Makefile index 9cad0e83..6bbf114b 100644 --- a/Makefile +++ b/Makefile @@ -66,7 +66,8 @@ FLOCQ=\ # General-purpose libraries (in lib/) LIB=Axioms.v Coqlib.v Intv.v Maps.v Heaps.v Lattice.v Ordered.v \ - Iteration.v Integers.v Archi.v Floats.v Parmov.v UnionFind.v Wfsimpl.v \ + Iteration.v Integers.v Archi.v Fappli_IEEE_extra.v Floats.v \ + Parmov.v UnionFind.v Wfsimpl.v \ Postorder.v FSetAVLplus.v IntvSets.v # Parts common to the front-ends and the back-end (in common/) diff --git a/arm/Archi.v b/arm/Archi.v index e6935413..5657f31d 100644 --- a/arm/Archi.v +++ b/arm/Archi.v @@ -25,12 +25,24 @@ Definition big_endian := false. Notation align_int64 := 8%Z (only parsing). Notation align_float64 := 8%Z (only parsing). -Program Definition default_pl : bool * nan_pl 53 := (false, nat_iter 51 xO xH). +Program Definition default_pl_64 : bool * nan_pl 53 := + (false, nat_iter 51 xO xH). -Definition choose_binop_pl (s1: bool) (pl1: nan_pl 53) (s2: bool) (pl2: nan_pl 53) := +Definition choose_binop_pl_64 (s1: bool) (pl1: nan_pl 53) (s2: bool) (pl2: nan_pl 53) := (** Choose second NaN if pl2 is sNaN but pl1 is qNan. In all other cases, choose first NaN *) (Pos.testbit (proj1_sig pl1) 51 && negb (Pos.testbit (proj1_sig pl2) 51))%bool. -Global Opaque big_endian default_pl choose_binop_pl. +Program Definition default_pl_32 : bool * nan_pl 24 := + (false, nat_iter 22 xO xH). + +Definition choose_binop_pl_32 (s1: bool) (pl1: nan_pl 24) (s2: bool) (pl2: nan_pl 24) := + (** Choose second NaN if pl2 is sNaN but pl1 is qNan. + In all other cases, choose first NaN *) + (Pos.testbit (proj1_sig pl1) 22 && + negb (Pos.testbit (proj1_sig pl2) 22))%bool. + +Global Opaque big_endian + default_pl_64 choose_binop_pl_64 + default_pl_32 choose_binop_pl_32. diff --git a/arm/Asm.v b/arm/Asm.v index 9d3ba5b9..f054db04 100644 --- a/arm/Asm.v +++ b/arm/Asm.v @@ -143,6 +143,7 @@ Inductive instruction : Type := | Pcmp: ireg -> shift_op -> instruction (**r integer comparison *) | Peor: ireg -> ireg -> shift_op -> instruction (**r bitwise exclusive or *) | Pldr: ireg -> ireg -> shift_addr -> instruction (**r int32 load *) + | Pldr_a: ireg -> ireg -> shift_addr -> instruction (**r any32 load to int register *) | Pldrb: ireg -> ireg -> shift_addr -> instruction (**r unsigned int8 load *) | Pldrh: ireg -> ireg -> shift_addr -> instruction (**r unsigned int16 load *) | Pldrsb: ireg -> ireg -> shift_addr -> instruction (**r signed int8 load *) @@ -155,6 +156,7 @@ Inductive instruction : Type := | Porr: ireg -> ireg -> shift_op -> instruction (**r bitwise or *) | Prsb: ireg -> ireg -> shift_op -> instruction (**r integer reverse subtraction *) | Pstr: ireg -> ireg -> shift_addr -> instruction (**r int32 store *) + | Pstr_a: ireg -> ireg -> shift_addr -> instruction (**r any32 store from int register *) | Pstrb: ireg -> ireg -> shift_addr -> instruction (**r int8 store *) | Pstrh: ireg -> ireg -> shift_addr -> instruction (**r int16 store *) | Psdiv: instruction (**r signed division *) @@ -172,15 +174,31 @@ Inductive instruction : Type := | Pfsubd: freg -> freg -> freg -> instruction (**r float subtraction *) | Pflid: freg -> float -> instruction (**r load float constant *) | Pfcmpd: freg -> freg -> instruction (**r float comparison *) - | Pfcmpzd: freg -> instruction (**r float comparison with 0.0 *) + | Pfcmpzd: freg -> instruction (**r float comparison with 0.0 *) | Pfsitod: freg -> ireg -> instruction (**r signed int to float *) | Pfuitod: freg -> ireg -> instruction (**r unsigned int to float *) | Pftosizd: ireg -> freg -> instruction (**r float to signed int *) | Pftouizd: ireg -> freg -> instruction (**r float to unsigned int *) - | Pfcvtsd: freg -> freg -> instruction (**r round to singled precision *) + | Pfabss: freg -> freg -> instruction (**r float absolute value *) + | Pfnegs: freg -> freg -> instruction (**r float opposite *) + | Pfadds: freg -> freg -> freg -> instruction (**r float addition *) + | Pfdivs: freg -> freg -> freg -> instruction (**r float division *) + | Pfmuls: freg -> freg -> freg -> instruction (**r float multiplication *) + | Pfsubs: freg -> freg -> freg -> instruction (**r float subtraction *) + | Pflis: freg -> float32 -> instruction (**r load float constant *) + | Pfcmps: freg -> freg -> instruction (**r float comparison *) + | Pfcmpzs: freg -> instruction (**r float comparison with 0.0 *) + | Pfsitos: freg -> ireg -> instruction (**r signed int to float *) + | Pfuitos: freg -> ireg -> instruction (**r unsigned int to float *) + | Pftosizs: ireg -> freg -> instruction (**r float to signed int *) + | Pftouizs: ireg -> freg -> instruction (**r float to unsigned int *) + | Pfcvtsd: freg -> freg -> instruction (**r round to single precision *) + | Pfcvtds: freg -> freg -> instruction (**r expand to double precision *) | Pfldd: freg -> ireg -> int -> instruction (**r float64 load *) + | Pfldd_a: freg -> ireg -> int -> instruction (**r any64 load to FP reg *) | Pflds: freg -> ireg -> int -> instruction (**r float32 load *) | Pfstd: freg -> ireg -> int -> instruction (**r float64 store *) + | Pfstd_a: freg -> ireg -> int -> instruction (**r any64 store from FP reg *) | Pfsts: freg -> ireg -> int -> instruction (**r float32 store *) (* Pseudo-instructions *) @@ -392,7 +410,7 @@ Definition compare_int (rs: regset) (v1 v2: val) (m: mem) := #CC <- (Val.cmpu (Mem.valid_pointer m) Cge v1 v2) #CV <- (Val.sub_overflow v1 v2). -(** Semantics of [fcmpd] instruction: +(** Semantics of [fcmp] instructions: << == N=0 Z=1 C=1 V=0 < N=1 Z=0 C=0 V=0 @@ -415,6 +433,20 @@ Definition compare_float (rs: regset) (v1 v2: val) := #CV <- Vundef end. +Definition compare_float32 (rs: regset) (v1 v2: val) := + match v1, v2 with + | Vsingle f1, Vsingle f2 => + rs#CN <- (Val.of_bool (Float32.cmp Clt f1 f2)) + #CZ <- (Val.of_bool (Float32.cmp Ceq f1 f2)) + #CC <- (Val.of_bool (negb (Float32.cmp Clt f1 f2))) + #CV <- (Val.of_bool (negb (Float32.cmp Ceq f1 f2 || Float32.cmp Clt f1 f2 || Float32.cmp Cgt f1 f2))) + | _, _ => + rs#CN <- Vundef + #CZ <- Vundef + #CC <- Vundef + #CV <- Vundef + end. + (** Testing a condition *) Definition eval_testcond (c: testcond) (rs: regset) : option bool := @@ -523,6 +555,8 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out Next (nextinstr (rs#r1 <- (Val.xor rs#r2 (eval_shift_op so rs)))) m | Pldr r1 r2 sa => exec_load Mint32 (Val.add rs#r2 (eval_shift_addr sa rs)) r1 rs m + | Pldr_a r1 r2 sa => + exec_load Many32 (Val.add rs#r2 (eval_shift_addr sa rs)) r1 rs m | Pldrb r1 r2 sa => exec_load Mint8unsigned (Val.add rs#r2 (eval_shift_addr sa rs)) r1 rs m | Pldrh r1 r2 sa => @@ -551,6 +585,8 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out Next (nextinstr (rs#r1 <- (Val.sub (eval_shift_op so rs) rs#r2))) m | Pstr r1 r2 sa => exec_store Mint32 (Val.add rs#r2 (eval_shift_addr sa rs)) r1 rs m + | Pstr_a r1 r2 sa => + exec_store Many32 (Val.add rs#r2 (eval_shift_addr sa rs)) r1 rs m | Pstrb r1 r2 sa => exec_store Mint8unsigned (Val.add rs#r2 (eval_shift_addr sa rs)) r1 rs m | Pstrh r1 r2 sa => @@ -606,19 +642,48 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out Next (nextinstr (rs #FR6 <- Vundef #r1 <- (Val.maketotal (Val.intoffloat rs#r2)))) m | Pftouizd r1 r2 => Next (nextinstr (rs #FR6 <- Vundef #r1 <- (Val.maketotal (Val.intuoffloat rs#r2)))) m + | Pfabss r1 r2 => + Next (nextinstr (rs#r1 <- (Val.absfs rs#r2))) m + | Pfnegs r1 r2 => + Next (nextinstr (rs#r1 <- (Val.negfs rs#r2))) m + | Pfadds r1 r2 r3 => + Next (nextinstr (rs#r1 <- (Val.addfs rs#r2 rs#r3))) m + | Pfdivs r1 r2 r3 => + Next (nextinstr (rs#r1 <- (Val.divfs rs#r2 rs#r3))) m + | Pfmuls r1 r2 r3 => + Next (nextinstr (rs#r1 <- (Val.mulfs rs#r2 rs#r3))) m + | Pfsubs r1 r2 r3 => + Next (nextinstr (rs#r1 <- (Val.subfs rs#r2 rs#r3))) m + | Pflis r1 f => + Next (nextinstr (rs#r1 <- (Vsingle f))) m + | Pfcmps r1 r2 => + Next (nextinstr (compare_float32 rs rs#r1 rs#r2)) m + | Pfcmpzs r1 => + Next (nextinstr (compare_float32 rs rs#r1 (Vsingle Float32.zero))) m + | Pfsitos r1 r2 => + Next (nextinstr (rs#r1 <- (Val.maketotal (Val.singleofint rs#r2)))) m + | Pfuitos r1 r2 => + Next (nextinstr (rs#r1 <- (Val.maketotal (Val.singleofintu rs#r2)))) m + | Pftosizs r1 r2 => + Next (nextinstr (rs #FR6 <- Vundef #r1 <- (Val.maketotal (Val.intofsingle rs#r2)))) m + | Pftouizs r1 r2 => + Next (nextinstr (rs #FR6 <- Vundef #r1 <- (Val.maketotal (Val.intuofsingle rs#r2)))) m | Pfcvtsd r1 r2 => Next (nextinstr (rs#r1 <- (Val.singleoffloat rs#r2))) m + | Pfcvtds r1 r2 => + Next (nextinstr (rs#r1 <- (Val.floatofsingle rs#r2))) m | Pfldd r1 r2 n => exec_load Mfloat64 (Val.add rs#r2 (Vint n)) r1 rs m + | Pfldd_a r1 r2 n => + exec_load Many64 (Val.add rs#r2 (Vint n)) r1 rs m | Pflds r1 r2 n => exec_load Mfloat32 (Val.add rs#r2 (Vint n)) r1 rs m | Pfstd r1 r2 n => exec_store Mfloat64 (Val.add rs#r2 (Vint n)) r1 rs m + | Pfstd_a r1 r2 n => + exec_store Many64 (Val.add rs#r2 (Vint n)) r1 rs m | Pfsts r1 r2 n => - match exec_store Mfloat32 (Val.add rs#r2 (Vint n)) r1 rs m with - | Next rs' m' => Next (rs'#FR6 <- Vundef) m' - | Stuck => Stuck - end + exec_store Mfloat32 (Val.add rs#r2 (Vint n)) r1 rs m (* Pseudo-instructions *) | Pallocframe sz pos => let (m1, stk) := Mem.alloc m 0 sz in diff --git a/arm/Asmgen.v b/arm/Asmgen.v index 2513a5e8..fa4faa60 100644 --- a/arm/Asmgen.v +++ b/arm/Asmgen.v @@ -185,6 +185,18 @@ Definition transl_cond | Cnotcompfzero cmp, a1 :: nil => do r1 <- freg_of a1; OK (Pfcmpzd r1 :: k) + | Ccompfs cmp, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; + OK (Pfcmps r1 r2 :: k) + | Cnotcompfs cmp, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; + OK (Pfcmps r1 r2 :: k) + | Ccompfszero cmp, a1 :: nil => + do r1 <- freg_of a1; + OK (Pfcmpzs r1 :: k) + | Cnotcompfszero cmp, a1 :: nil => + do r1 <- freg_of a1; + OK (Pfcmpzs r1 :: k) | _, _ => Error(msg "Asmgen.transl_cond") end. @@ -241,6 +253,10 @@ Definition cond_for_cond (cond: condition) := | Cnotcompf cmp => cond_for_float_not_cmp cmp | Ccompfzero cmp => cond_for_float_cmp cmp | Cnotcompfzero cmp => cond_for_float_not_cmp cmp + | Ccompfs cmp => cond_for_float_cmp cmp + | Cnotcompfs cmp => cond_for_float_not_cmp cmp + | Ccompfszero cmp => cond_for_float_cmp cmp + | Cnotcompfszero cmp => cond_for_float_not_cmp cmp end. (** Translation of the arithmetic operation [r <- op(args)]. @@ -261,6 +277,9 @@ Definition transl_op | Ofloatconst f, nil => do r <- freg_of res; OK (Pflid r f :: k) + | Osingleconst f, nil => + do r <- freg_of res; + OK (Pflis r f :: k) | Oaddrsymbol s ofs, nil => do r <- ireg_of res; OK (Ploadsymbol r s ofs :: k) @@ -400,9 +419,30 @@ Definition transl_op | Odivf, a1 :: a2 :: nil => do r <- freg_of res; do r1 <- freg_of a1; do r2 <- freg_of a2; OK (Pfdivd r r1 r2 :: k) + | Onegfs, a1 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; + OK (Pfnegs r r1 :: k) + | Oabsfs, a1 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; + OK (Pfabss r r1 :: k) + | Oaddfs, a1 :: a2 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; do r2 <- freg_of a2; + OK (Pfadds r r1 r2 :: k) + | Osubfs, a1 :: a2 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; do r2 <- freg_of a2; + OK (Pfsubs r r1 r2 :: k) + | Omulfs, a1 :: a2 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; do r2 <- freg_of a2; + OK (Pfmuls r r1 r2 :: k) + | Odivfs, a1 :: a2 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; do r2 <- freg_of a2; + OK (Pfdivs r r1 r2 :: k) | Osingleoffloat, a1 :: nil => do r <- freg_of res; do r1 <- freg_of a1; OK (Pfcvtsd r r1 :: k) + | Ofloatofsingle, a1 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; + OK (Pfcvtds r r1 :: k) | Ointoffloat, a1 :: nil => do r <- ireg_of res; do r1 <- freg_of a1; OK (Pftosizd r r1 :: k) @@ -415,6 +455,18 @@ Definition transl_op | Ofloatofintu, a1 :: nil => do r <- freg_of res; do r1 <- ireg_of a1; OK (Pfuitod r r1 :: k) + | Ointofsingle, a1 :: nil => + do r <- ireg_of res; do r1 <- freg_of a1; + OK (Pftosizs r r1 :: k) + | Ointuofsingle, a1 :: nil => + do r <- ireg_of res; do r1 <- freg_of a1; + OK (Pftouizs r r1 :: k) + | Osingleofint, a1 :: nil => + do r <- freg_of res; do r1 <- ireg_of a1; + OK (Pfsitos r r1 :: k) + | Osingleofintu, a1 :: nil => + do r <- freg_of res; do r1 <- ireg_of a1; + OK (Pfuitos r r1 :: k) | Ocmp cmp, _ => do r <- ireg_of res; transl_cond cmp args @@ -440,31 +492,34 @@ Definition loadind_int (base: ireg) (ofs: int) (dst: ireg) (k: code) := indexed_memory_access (fun base n => Pldr dst base (SAimm n)) mk_immed_mem_word base ofs k. Definition loadind (base: ireg) (ofs: int) (ty: typ) (dst: mreg) (k: code) := - match ty with - | Tint => - do r <- ireg_of dst; OK (loadind_int base ofs r k) - | Tfloat => - do r <- freg_of dst; - OK (indexed_memory_access (Pfldd r) mk_immed_mem_float base ofs k) - | Tsingle => - do r <- freg_of dst; + match ty, preg_of dst with + | Tint, IR r => + OK (indexed_memory_access (fun base n => Pldr r base (SAimm n)) mk_immed_mem_word base ofs k) + | Tany32, IR r => + OK (indexed_memory_access (fun base n => Pldr_a r base (SAimm n)) mk_immed_mem_word base ofs k) + | Tsingle, FR r => OK (indexed_memory_access (Pflds r) mk_immed_mem_float base ofs k) - | Tlong => + | Tfloat, FR r => + OK (indexed_memory_access (Pfldd r) mk_immed_mem_float base ofs k) + | Tany64, FR r => + OK (indexed_memory_access (Pfldd_a r) mk_immed_mem_float base ofs k) + | _, _ => Error (msg "Asmgen.loadind") end. Definition storeind (src: mreg) (base: ireg) (ofs: int) (ty: typ) (k: code) := - match ty with - | Tint => - do r <- ireg_of src; + match ty, preg_of src with + | Tint, IR r => OK (indexed_memory_access (fun base n => Pstr r base (SAimm n)) mk_immed_mem_word base ofs k) - | Tfloat => - do r <- freg_of src; - OK (indexed_memory_access (Pfstd r) mk_immed_mem_float base ofs k) - | Tsingle => - do r <- freg_of src; + | Tany32, IR r => + OK (indexed_memory_access (fun base n => Pstr_a r base (SAimm n)) mk_immed_mem_word base ofs k) + | Tsingle, FR r => OK (indexed_memory_access (Pfsts r) mk_immed_mem_float base ofs k) - | Tlong => + | Tfloat, FR r => + OK (indexed_memory_access (Pfstd r) mk_immed_mem_float base ofs k) + | Tany64, FR r => + OK (indexed_memory_access (Pfstd_a r) mk_immed_mem_float base ofs k) + | _, _ => Error (msg "Asmgen.storeind") end. @@ -546,7 +601,7 @@ Definition transl_load (chunk: memory_chunk) (addr: addressing) transl_memory_access_float Pflds mk_immed_mem_float dst addr args k | Mfloat64 => transl_memory_access_float Pfldd mk_immed_mem_float dst addr args k - | Mint64 => + | _ => Error (msg "Asmgen.transl_load") end. @@ -567,7 +622,7 @@ Definition transl_store (chunk: memory_chunk) (addr: addressing) transl_memory_access_float Pfsts mk_immed_mem_float src addr args k | Mfloat64 => transl_memory_access_float Pfstd mk_immed_mem_float src addr args k - | Mint64 => + | _ => Error (msg "Asmgen.transl_store") end. diff --git a/arm/Asmgenproof.v b/arm/Asmgenproof.v index cfe4f542..341f6a01 100644 --- a/arm/Asmgenproof.v +++ b/arm/Asmgenproof.v @@ -237,16 +237,15 @@ Hint Resolve indexed_memory_access_label. Remark loadind_label: forall base ofs ty dst k c, loadind base ofs ty dst k = OK c -> tail_nolabel k c. Proof. - intros. destruct ty; monadInv H. - unfold loadind_int; TailNoLabel. - TailNoLabel. - TailNoLabel. + unfold loadind, loadind_int; intros; + destruct ty, (preg_of dst); inv H; TailNoLabel. Qed. Remark storeind_label: forall base ofs ty src k c, storeind src base ofs ty k = OK c -> tail_nolabel k c. Proof. - intros. destruct ty; monadInv H; TailNoLabel. + unfold storeind; intros; + destruct ty, (preg_of src); inv H; TailNoLabel. Qed. Remark transl_cond_label: diff --git a/arm/Asmgenproof1.v b/arm/Asmgenproof1.v index 3e002175..a0d67526 100644 --- a/arm/Asmgenproof1.v +++ b/arm/Asmgenproof1.v @@ -405,22 +405,6 @@ Proof. split. Simpl. intros; Simpl. Qed. -(* -Lemma loadind_float_correct: - forall (base: ireg) ofs dst (rs: regset) m v k, - Mem.loadv Mfloat64al32 m (Val.add rs#base (Vint ofs)) = Some v -> - exists rs', - exec_straight ge fn (loadind_float base ofs dst k) rs m k rs' m - /\ rs'#dst = v - /\ forall r, r <> PC -> r <> IR14 -> r <> dst -> rs'#r = rs#r. -Proof. - intros; unfold loadind_float. apply indexed_memory_access_correct; intros. - econstructor; split. - apply exec_straight_one. simpl. unfold exec_load. rewrite H0; rewrite H; eauto. auto. - split. Simpl. intros; Simpl. -Qed. -*) - Lemma loadind_correct: forall (base: ireg) ofs ty dst k c (rs: regset) m v, loadind base ofs ty dst k = OK c -> @@ -430,56 +414,32 @@ Lemma loadind_correct: /\ rs'#(preg_of dst) = v /\ forall r, r <> PC -> r <> IR14 -> r <> preg_of dst -> rs'#r = rs#r. Proof. - unfold loadind; intros. - destruct ty; monadInv H. + unfold loadind; intros. destruct ty; destruct (preg_of dst); inv H; simpl in H0. - (* int *) - erewrite ireg_of_eq by eauto. apply loadind_int_correct; auto. + apply loadind_int_correct; auto. - (* float *) - erewrite freg_of_eq by eauto. simpl in H0. apply indexed_memory_access_correct; intros. econstructor; split. apply exec_straight_one. simpl. unfold exec_load. rewrite H. rewrite H0. eauto. auto. split. Simpl. intros; Simpl. - (* single *) - erewrite freg_of_eq by eauto. simpl in H0. apply indexed_memory_access_correct; intros. econstructor; split. apply exec_straight_one. simpl. unfold exec_load. rewrite H. rewrite H0. eauto. auto. split. Simpl. intros; Simpl. -Qed. - -(** Indexed memory stores. *) - -(* -Lemma storeind_int_correct: - forall (base: ireg) ofs (src: ireg) (rs: regset) m m' k, - Mem.storev Mint32 m (Val.add rs#base (Vint ofs)) (rs#src) = Some m' -> - src <> IR14 -> - exists rs', - exec_straight ge fn (storeind_int src base ofs k) rs m k rs' m' - /\ forall r, r <> PC -> r <> IR14 -> rs'#r = rs#r. -Proof. - intros; unfold storeind_int. apply indexed_memory_access_correct; intros. +- (* any32 *) + apply indexed_memory_access_correct; intros. econstructor; split. - apply exec_straight_one. simpl. unfold exec_store. - rewrite H1. rewrite H2; auto with asmgen. rewrite H; eauto. auto. - intros; Simpl. -Qed. - -Lemma storeind_float_correct: - forall (base: ireg) ofs (src: freg) (rs: regset) m m' k, - Mem.storev Mfloat64al32 m (Val.add rs#base (Vint ofs)) (rs#src) = Some m' -> - exists rs', - exec_straight ge fn (storeind_float src base ofs k) rs m k rs' m' - /\ forall r, r <> PC -> r <> IR14 -> rs'#r = rs#r. -Proof. - intros; unfold storeind_float. apply indexed_memory_access_correct; intros. + apply exec_straight_one. simpl. unfold exec_load. rewrite H. rewrite H0. eauto. auto. + split. Simpl. intros; Simpl. +- (* any64 *) + apply indexed_memory_access_correct; intros. econstructor; split. - apply exec_straight_one. simpl. unfold exec_store. - rewrite H0. rewrite H1; auto with asmgen. rewrite H; eauto. auto. - intros; Simpl. + apply exec_straight_one. simpl. unfold exec_load. rewrite H. rewrite H0. eauto. auto. + split. Simpl. intros; Simpl. Qed. -*) + +(** Indexed memory stores. *) Lemma storeind_correct: forall (base: ireg) ofs ty src k c (rs: regset) m m', @@ -490,29 +450,38 @@ Lemma storeind_correct: /\ forall r, r <> PC -> r <> IR14 -> preg_notin r (destroyed_by_setstack ty) -> rs'#r = rs#r. Proof. unfold storeind; intros. - destruct ty; monadInv H; simpl in H0. + assert (NOT14: preg_of src <> IR IR14) by eauto with asmgen. + destruct ty; destruct (preg_of src); inv H; simpl in H0. - (* int *) - erewrite ireg_of_eq in H0 by eauto. apply indexed_memory_access_correct; intros. econstructor; split. apply exec_straight_one. simpl. unfold exec_store. - rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. - assert (IR x <> IR IR14) by eauto with asmgen. congruence. - auto. intros; Simpl. + rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. auto. + intros; Simpl. - (* float *) - erewrite freg_of_eq in H0 by eauto. apply indexed_memory_access_correct; intros. econstructor; split. apply exec_straight_one. simpl. unfold exec_store. - rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. - auto. intros; Simpl. + rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. auto. + intros; Simpl. - (* single *) - erewrite freg_of_eq in H0 by eauto. apply indexed_memory_access_correct; intros. econstructor; split. apply exec_straight_one. simpl. unfold exec_store. - rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. - auto. intros; Simpl. + rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. auto. + intros; Simpl. +- (* any32 *) + apply indexed_memory_access_correct; intros. + econstructor; split. + apply exec_straight_one. simpl. unfold exec_store. + rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. auto. + intros; Simpl. +- (* any64 *) + apply indexed_memory_access_correct; intros. + econstructor; split. + apply exec_straight_one. simpl. unfold exec_store. + rewrite H. rewrite H1; auto with asmgen. rewrite H0; eauto. auto. + intros; Simpl. Qed. (** Translation of shift immediates *) @@ -775,6 +744,112 @@ Proof. exfalso; eapply Float.cmp_lt_gt_false; eauto. Qed. +Lemma compare_float32_spec: + forall rs f1 f2, + let rs1 := nextinstr (compare_float32 rs (Vsingle f1) (Vsingle f2)) in + rs1#CN = Val.of_bool (Float32.cmp Clt f1 f2) + /\ rs1#CZ = Val.of_bool (Float32.cmp Ceq f1 f2) + /\ rs1#CC = Val.of_bool (negb (Float32.cmp Clt f1 f2)) + /\ rs1#CV = Val.of_bool (negb (Float32.cmp Ceq f1 f2 || Float32.cmp Clt f1 f2 || Float32.cmp Cgt f1 f2)). +Proof. + intros. intuition. +Qed. + +Lemma compare_float32_inv: + forall rs v1 v2, + let rs1 := nextinstr (compare_float32 rs v1 v2) in + forall r', data_preg r' = true -> rs1#r' = rs#r'. +Proof. + intros. unfold rs1, compare_float32. + assert (nextinstr (rs#CN <- Vundef #CZ <- Vundef #CC <- Vundef #CV <- Vundef) r' = rs r'). + { repeat Simplif. } + destruct v1; destruct v2; auto. + repeat Simplif. +Qed. + +Lemma compare_float32_nextpc: + forall rs v1 v2, + nextinstr (compare_float32 rs v1 v2) PC = Val.add (rs PC) Vone. +Proof. + intros. unfold compare_float32. destruct v1; destruct v2; reflexivity. +Qed. + +Lemma cond_for_float32_cmp_correct: + forall c n1 n2 rs, + eval_testcond (cond_for_float_cmp c) + (nextinstr (compare_float32 rs (Vsingle n1) (Vsingle n2))) = + Some(Float32.cmp c n1 n2). +Proof. + intros. + generalize (compare_float32_spec rs n1 n2). + set (rs' := nextinstr (compare_float32 rs (Vsingle n1) (Vsingle n2))). + intros [A [B [C D]]]. + unfold eval_testcond. rewrite A; rewrite B; rewrite C; rewrite D. + destruct c; simpl. +(* eq *) + destruct (Float32.cmp Ceq n1 n2); auto. +(* ne *) + rewrite Float32.cmp_ne_eq. destruct (Float32.cmp Ceq n1 n2); auto. +(* lt *) + destruct (Float32.cmp Clt n1 n2); auto. +(* le *) + rewrite Float32.cmp_le_lt_eq. + destruct (Float32.cmp Clt n1 n2); destruct (Float32.cmp Ceq n1 n2); auto. +(* gt *) + destruct (Float32.cmp Ceq n1 n2) eqn:EQ; + destruct (Float32.cmp Clt n1 n2) eqn:LT; + destruct (Float32.cmp Cgt n1 n2) eqn:GT; auto. + exfalso; eapply Float32.cmp_lt_gt_false; eauto. + exfalso; eapply Float32.cmp_gt_eq_false; eauto. + exfalso; eapply Float32.cmp_lt_gt_false; eauto. +(* ge *) + rewrite Float32.cmp_ge_gt_eq. + destruct (Float32.cmp Ceq n1 n2) eqn:EQ; + destruct (Float32.cmp Clt n1 n2) eqn:LT; + destruct (Float32.cmp Cgt n1 n2) eqn:GT; auto. + exfalso; eapply Float32.cmp_lt_eq_false; eauto. + exfalso; eapply Float32.cmp_lt_eq_false; eauto. + exfalso; eapply Float32.cmp_lt_gt_false; eauto. +Qed. + +Lemma cond_for_float32_not_cmp_correct: + forall c n1 n2 rs, + eval_testcond (cond_for_float_not_cmp c) + (nextinstr (compare_float32 rs (Vsingle n1) (Vsingle n2)))= + Some(negb(Float32.cmp c n1 n2)). +Proof. + intros. + generalize (compare_float32_spec rs n1 n2). + set (rs' := nextinstr (compare_float32 rs (Vsingle n1) (Vsingle n2))). + intros [A [B [C D]]]. + unfold eval_testcond. rewrite A; rewrite B; rewrite C; rewrite D. + destruct c; simpl. +(* eq *) + destruct (Float32.cmp Ceq n1 n2); auto. +(* ne *) + rewrite Float32.cmp_ne_eq. destruct (Float32.cmp Ceq n1 n2); auto. +(* lt *) + destruct (Float32.cmp Clt n1 n2); auto. +(* le *) + rewrite Float32.cmp_le_lt_eq. + destruct (Float32.cmp Clt n1 n2) eqn:LT; destruct (Float32.cmp Ceq n1 n2) eqn:EQ; auto. +(* gt *) + destruct (Float32.cmp Ceq n1 n2) eqn:EQ; + destruct (Float32.cmp Clt n1 n2) eqn:LT; + destruct (Float32.cmp Cgt n1 n2) eqn:GT; auto. + exfalso; eapply Float32.cmp_lt_gt_false; eauto. + exfalso; eapply Float32.cmp_gt_eq_false; eauto. + exfalso; eapply Float32.cmp_lt_gt_false; eauto. +(* ge *) + rewrite Float32.cmp_ge_gt_eq. + destruct (Float32.cmp Ceq n1 n2) eqn:EQ; + destruct (Float32.cmp Clt n1 n2) eqn:LT; + destruct (Float32.cmp Cgt n1 n2) eqn:GT; auto. + exfalso; eapply Float32.cmp_lt_eq_false; eauto. + exfalso; eapply Float32.cmp_lt_eq_false; eauto. + exfalso; eapply Float32.cmp_lt_gt_false; eauto. +Qed. + Ltac ArgsInv := repeat (match goal with | [ H: Error _ = OK _ |- _ ] => discriminate @@ -881,6 +956,37 @@ Local Opaque compare_float. simpl. apply cond_for_float_not_cmp_correct. Local Opaque compare_float. simpl. apply cond_for_float_not_cmp_correct. exact I. apply compare_float_inv. +- (* Ccompfs *) + econstructor. + split. apply exec_straight_one. simpl. eauto. apply compare_float32_nextpc. + split. destruct (Val.cmpfs_bool c0 (rs x) (rs x0)) eqn:CMP; auto. + destruct (rs x); try discriminate. destruct (rs x0); try discriminate. + simpl in CMP. inv CMP. apply cond_for_float32_cmp_correct. + apply compare_float32_inv. +- (* Cnotcompfs *) + econstructor. + split. apply exec_straight_one. simpl. eauto. apply compare_float32_nextpc. + split. destruct (Val.cmpfs_bool c0 (rs x) (rs x0)) eqn:CMP; auto. + destruct (rs x); try discriminate. destruct (rs x0); try discriminate. + simpl in CMP. inv CMP. +Local Opaque compare_float32. simpl. apply cond_for_float32_not_cmp_correct. + exact I. + apply compare_float32_inv. +- (* Ccompfszero *) + econstructor. + split. apply exec_straight_one. simpl. eauto. apply compare_float32_nextpc. + split. destruct (Val.cmpfs_bool c0 (rs x) (Vsingle Float32.zero)) eqn:CMP; auto. + destruct (rs x); try discriminate. + simpl in CMP. inv CMP. apply cond_for_float32_cmp_correct. + apply compare_float32_inv. +- (* Cnotcompfzero *) + econstructor. + split. apply exec_straight_one. simpl. eauto. apply compare_float32_nextpc. + split. destruct (Val.cmpfs_bool c0 (rs x) (Vsingle Float32.zero)) eqn:CMP; auto. + destruct (rs x); try discriminate. simpl in CMP. inv CMP. + simpl. apply cond_for_float32_not_cmp_correct. + exact I. + apply compare_float32_inv. Qed. (** Translation of arithmetic operations. *) @@ -1039,6 +1145,18 @@ Transparent destroyed_by_op. (* floatofintu *) econstructor; split. apply exec_straight_one; simpl. rewrite H0; simpl. eauto. auto. intuition Simpl. + (* intofsingle *) + econstructor; split. apply exec_straight_one; simpl. rewrite H0; simpl. eauto. auto. + simpl. intuition Simpl. + (* intuofsingle *) + econstructor; split. apply exec_straight_one; simpl. rewrite H0; simpl. eauto. auto. + simpl. intuition Simpl. + (* singleofint *) + econstructor; split. apply exec_straight_one; simpl. rewrite H0; simpl. eauto. auto. + intuition Simpl. + (* singleofintu *) + econstructor; split. apply exec_straight_one; simpl. rewrite H0; simpl. eauto. auto. + intuition Simpl. (* Ocmp *) contradiction. Qed. @@ -1234,6 +1352,8 @@ Proof. discriminate. eapply transl_load_float_correct; eauto. eapply transl_load_float_correct; eauto. + discriminate. + discriminate. Qed. Lemma transl_store_correct: @@ -1256,13 +1376,10 @@ Proof. - eapply transl_store_int_correct; eauto. - eapply transl_store_int_correct; eauto. - discriminate. -- unfold transl_memory_access_float in H. monadInv H. rewrite (freg_of_eq _ _ EQ) in *. - eapply transl_memory_access_correct; eauto. - intros. econstructor; split. apply exec_straight_one. - simpl. unfold exec_store. rewrite H. rewrite H2; eauto with asmgen. - rewrite H1. eauto. auto. intros. Simpl. - simpl; auto. - eapply transl_store_float_correct; eauto. +- eapply transl_store_float_correct; eauto. +- discriminate. +- discriminate. Qed. End CONSTRUCTORS. diff --git a/arm/ConstpropOp.vp b/arm/ConstpropOp.vp index 2b658a4e..4f4bf5aa 100644 --- a/arm/ConstpropOp.vp +++ b/arm/ConstpropOp.vp @@ -70,6 +70,22 @@ Nondetfunction cond_strength_reduction if Float.eq_dec n2 Float.zero then (Cnotcompfzero c, r1 :: nil) else (cond, args) + | Ccompfs c, r1 :: r2 :: nil, FS n1 :: v2 :: nil => + if Float32.eq_dec n1 Float32.zero + then (Ccompfszero (swap_comparison c), r2 :: nil) + else (cond, args) + | Ccompfs c, r1 :: r2 :: nil, v1 :: FS n2 :: nil => + if Float32.eq_dec n2 Float32.zero + then (Ccompfszero c, r1 :: nil) + else (cond, args) + | Cnotcompfs c, r1 :: r2 :: nil, FS n1 :: v2 :: nil => + if Float32.eq_dec n1 Float32.zero + then (Cnotcompfszero (swap_comparison c), r2 :: nil) + else (cond, args) + | Cnotcompfs c, r1 :: r2 :: nil, v1 :: FS n2 :: nil => + if Float32.eq_dec n2 Float32.zero + then (Cnotcompfszero c, r1 :: nil) + else (cond, args) | _, _, _ => (cond, args) end. @@ -164,18 +180,19 @@ Definition make_xorimm (n: int) (r: reg) := else (Oxorimm n, r :: nil). Definition make_mulfimm (n: float) (r r1 r2: reg) := - if Float.eq_dec n (Float.floatofint (Int.repr 2)) + if Float.eq_dec n (Float.of_int (Int.repr 2)) then (Oaddf, r :: r :: nil) else (Omulf, r1 :: r2 :: nil). +Definition make_mulfsimm (n: float32) (r r1 r2: reg) := + if Float32.eq_dec n (Float32.of_int (Int.repr 2)) + then (Oaddfs, r :: r :: nil) + else (Omulfs, r1 :: r2 :: nil). + Definition make_cast8signed (r: reg) (a: aval) := if vincl a (Sgn 8) then (Omove, r :: nil) else (Ocast8signed, r :: nil). Definition make_cast16signed (r: reg) (a: aval) := if vincl a (Sgn 16) then (Omove, r :: nil) else (Ocast16signed, r :: nil). -Definition make_singleoffloat (r: reg) (a: aval) := - if vincl a Fsingle && generate_float_constants tt - then (Omove, r :: nil) - else (Osingleoffloat, r :: nil). Nondetfunction op_strength_reduction (op: operation) (args: list reg) (vl: list aval) := @@ -207,10 +224,11 @@ Nondetfunction op_strength_reduction | Oshl, r1 :: r2 :: nil, v1 :: I n2 :: nil => make_shlimm n2 r1 r2 | Oshr, r1 :: r2 :: nil, v1 :: I n2 :: nil => make_shrimm n2 r1 r2 | Oshru, r1 :: r2 :: nil, v1 :: I n2 :: nil => make_shruimm n2 r1 r2 - | Osingleoffloat, r1 :: nil, v1 :: nil => make_singleoffloat r1 v1 | Ocmp c, args, vl => make_cmp c args vl | Omulf, r1 :: r2 :: nil, v1 :: F n2 :: nil => make_mulfimm n2 r1 r1 r2 | Omulf, r1 :: r2 :: nil, F n1 :: v2 :: nil => make_mulfimm n1 r2 r1 r2 + | Omulfs, r1 :: r2 :: nil, v1 :: FS n2 :: nil => make_mulfsimm n2 r1 r1 r2 + | Omulfs, r1 :: r2 :: nil, FS n1 :: v2 :: nil => make_mulfsimm n1 r2 r1 r2 | _, _, _ => (op, args) end. diff --git a/arm/ConstpropOpproof.v b/arm/ConstpropOpproof.v index 00ea8bc0..597c9602 100644 --- a/arm/ConstpropOpproof.v +++ b/arm/ConstpropOpproof.v @@ -80,6 +80,10 @@ Ltac SimplVM := let E := fresh in assert (E: v = Vfloat n) by (inversion H; auto); rewrite E in *; clear H; SimplVM + | [ H: vmatch _ ?v (FS ?n) |- _ ] => + let E := fresh in + assert (E: v = Vsingle n) by (inversion H; auto); + rewrite E in *; clear H; SimplVM | [ H: vmatch _ ?v (Ptr(Gl ?id ?ofs)) |- _ ] => let E := fresh in assert (E: Val.lessdef v (Genv.symbol_address ge id ofs)) by (eapply vmatch_ptr_gl; eauto); @@ -123,6 +127,18 @@ Proof. - destruct (Float.eq_dec n2 Float.zero); simpl; auto. subst n2; auto. rewrite H1; auto. +- destruct (Float32.eq_dec n1 Float32.zero). + subst n1. simpl. destruct (rs#r2); simpl; auto. rewrite Float32.cmp_swap. auto. + simpl. rewrite H1; auto. +- destruct (Float32.eq_dec n2 Float32.zero). + subst n2. simpl. auto. + simpl. rewrite H1; auto. +- destruct (Float32.eq_dec n1 Float32.zero). + subst n1. simpl. destruct (rs#r2); simpl; auto. rewrite Float32.cmp_swap. auto. + simpl. rewrite H1; auto. +- destruct (Float32.eq_dec n2 Float32.zero); simpl; auto. + subst n2; auto. + rewrite H1; auto. - auto. Qed. @@ -335,7 +351,7 @@ Lemma make_mulfimm_correct: exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulf rs#r1 rs#r2) v. Proof. intros; unfold make_mulfimm. - destruct (Float.eq_dec n (Float.floatofint (Int.repr 2))); intros. + destruct (Float.eq_dec n (Float.of_int (Int.repr 2))); intros. simpl. econstructor; split. eauto. rewrite H; subst n. destruct (rs#r1); simpl; auto. rewrite Float.mul2_add; auto. simpl. econstructor; split; eauto. @@ -348,13 +364,40 @@ Lemma make_mulfimm_correct_2: exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulf rs#r1 rs#r2) v. Proof. intros; unfold make_mulfimm. - destruct (Float.eq_dec n (Float.floatofint (Int.repr 2))); intros. + destruct (Float.eq_dec n (Float.of_int (Int.repr 2))); intros. simpl. econstructor; split. eauto. rewrite H; subst n. destruct (rs#r2); simpl; auto. rewrite Float.mul2_add; auto. rewrite Float.mul_commut; auto. simpl. econstructor; split; eauto. Qed. +Lemma make_mulfsimm_correct: + forall n r1 r2, + rs#r2 = Vsingle n -> + let (op, args) := make_mulfsimm n r1 r1 r2 in + exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulfs rs#r1 rs#r2) v. +Proof. + intros; unfold make_mulfsimm. + destruct (Float32.eq_dec n (Float32.of_int (Int.repr 2))); intros. + simpl. econstructor; split. eauto. rewrite H; subst n. + destruct (rs#r1); simpl; auto. rewrite Float32.mul2_add; auto. + simpl. econstructor; split; eauto. +Qed. + +Lemma make_mulfsimm_correct_2: + forall n r1 r2, + rs#r1 = Vsingle n -> + let (op, args) := make_mulfsimm n r2 r1 r2 in + exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulfs rs#r1 rs#r2) v. +Proof. + intros; unfold make_mulfsimm. + destruct (Float32.eq_dec n (Float32.of_int (Int.repr 2))); intros. + simpl. econstructor; split. eauto. rewrite H; subst n. + destruct (rs#r2); simpl; auto. rewrite Float32.mul2_add; auto. + rewrite Float32.mul_commut; auto. + simpl. econstructor; split; eauto. +Qed. + Lemma make_cast8signed_correct: forall r x, vmatch bc rs#r x -> @@ -383,21 +426,6 @@ Proof. econstructor; split; simpl; eauto. Qed. -Lemma make_singleoffloat_correct: - forall r x, - vmatch bc rs#r x -> - let (op, args) := make_singleoffloat r x in - exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.singleoffloat rs#r) v. -Proof. - intros; unfold make_singleoffloat. - destruct (vincl x Fsingle && generate_float_constants tt) eqn:INCL. - InvBooleans. exists rs#r; split; auto. - assert (V: vmatch bc rs#r Fsingle). - { eapply vmatch_ge; eauto. apply vincl_ge; auto. } - inv V; simpl; auto. rewrite Float.singleoffloat_of_single by auto. auto. - econstructor; split; simpl; eauto. -Qed. - Lemma op_strength_reduction_correct: forall op args vl v, vl = map (fun r => AE.get r ae) args -> @@ -459,14 +487,16 @@ Proof. InvApproxRegs; SimplVM. inv H0. apply make_shrimm_correct; auto. (* shru *) InvApproxRegs; SimplVM. inv H0. apply make_shruimm_correct; auto. -(* singleoffloat *) - InvApproxRegs; SimplVM; inv H0. apply make_singleoffloat_correct; auto. (* cmp *) inv H0. apply make_cmp_correct; auto. (* mulf *) InvApproxRegs; SimplVM; inv H0. rewrite <- H2. apply make_mulfimm_correct; auto. InvApproxRegs; SimplVM; inv H0. fold (Val.mulf (Vfloat n1) rs#r2). rewrite <- H2. apply make_mulfimm_correct_2; auto. +(* mulfs *) + InvApproxRegs; SimplVM; inv H0. rewrite <- H2. apply make_mulfsimm_correct; auto. + InvApproxRegs; SimplVM; inv H0. fold (Val.mulfs (Vsingle n1) rs#r2). + rewrite <- H2. apply make_mulfsimm_correct_2; auto. (* default *) exists v; auto. Qed. diff --git a/arm/Machregs.v b/arm/Machregs.v index b55259be..791ccbb6 100644 --- a/arm/Machregs.v +++ b/arm/Machregs.v @@ -44,14 +44,10 @@ Global Opaque mreg_eq. Definition mreg_type (r: mreg): typ := match r with - | R0 => Tint | R1 => Tint | R2 => Tint | R3 => Tint - | R4 => Tint | R5 => Tint | R6 => Tint | R7 => Tint - | R8 => Tint | R9 => Tint | R10 => Tint | R11 => Tint - | R12 => Tint - | F0 => Tfloat | F1 => Tfloat | F2 => Tfloat | F3 => Tfloat - | F4 => Tfloat| F5 => Tfloat | F6 => Tfloat | F7 => Tfloat - | F8 => Tfloat | F9 => Tfloat | F10 => Tfloat | F11 => Tfloat - | F12 => Tfloat | F13 => Tfloat | F14 => Tfloat | F15 => Tfloat + | R0 | R1 | R2 | R3 | R4 | R5 | R6 | R7 + | R8 | R9 | R10 | R11 | R12 => Tany32 + | F0 | F1 | F2 | F3 | F4 | F5 | F6 | F7 + | F8 | F9 | F10 | F11 | F12 | F13 | F14 | F15 => Tany64 end. Open Scope positive_scope. @@ -84,18 +80,14 @@ Definition is_stack_reg (r: mreg) : bool := false. Definition destroyed_by_op (op: operation): list mreg := match op with | Odiv | Odivu => R0 :: R1 :: R2 :: R3 :: R12 :: nil - | Ointoffloat | Ointuoffloat => F6 :: nil + | Ointoffloat | Ointuoffloat | Ointofsingle | Ointuofsingle => F6 :: nil | _ => nil end. Definition destroyed_by_load (chunk: memory_chunk) (addr: addressing): list mreg := nil. -Definition destroyed_by_store (chunk: memory_chunk) (addr: addressing): list mreg := - match chunk with - | Mfloat32 => F6 :: nil - | _ => nil - end. +Definition destroyed_by_store (chunk: memory_chunk) (addr: addressing): list mreg := nil. Definition destroyed_by_cond (cond: condition): list mreg := nil. @@ -106,16 +98,10 @@ Definition destroyed_by_jumptable: list mreg := Definition destroyed_by_builtin (ef: external_function): list mreg := match ef with | EF_memcpy sz al => if zle sz 32 then nil else R2 :: R3 :: R12 :: nil - | EF_vstore Mfloat32 => F6 :: nil - | EF_vstore_global Mfloat32 _ _ => F6 :: nil | _ => nil end. -Definition destroyed_by_setstack (ty: typ): list mreg := - match ty with - | Tsingle => F6 :: nil - | _ => nil - end. +Definition destroyed_by_setstack (ty: typ): list mreg := nil. Definition destroyed_at_function_entry: list mreg := R12 :: nil. diff --git a/arm/NeedOp.v b/arm/NeedOp.v index 3fb0d721..e91ea64d 100644 --- a/arm/NeedOp.v +++ b/arm/NeedOp.v @@ -42,6 +42,7 @@ Definition needs_of_operation (op: operation) (nv: nval): list nval := | Omove => nv::nil | Ointconst n => nil | Ofloatconst n => nil + | Osingleconst n => nil | Oaddrsymbol id ofs => nil | Oaddrstack ofs => nil | Ocast8signed => op1 (sign_ext 8 nv) @@ -74,8 +75,11 @@ Definition needs_of_operation (op: operation) (nv: nval): list nval := | Oshrximm _ => op1 (default nv) | Onegf | Oabsf => op1 (default nv) | Oaddf | Osubf | Omulf | Odivf => op2 (default nv) - | Osingleoffloat => op1 (singleoffloat nv) + | Onegfs | Oabsfs => op1 (default nv) + | Oaddfs | Osubfs | Omulfs | Odivfs => op2 (default nv) + | Ofloatofsingle | Osingleoffloat => op1 (default nv) | Ointoffloat | Ointuoffloat | Ofloatofint | Ofloatofintu => op1 (default nv) + | Ointofsingle | Ointuofsingle | Osingleofint | Osingleofintu => op1 (default nv) | Omakelong => op2 (default nv) | Olowlong | Ohighlong => op1 (default nv) | Ocmp c => needs_of_condition c @@ -87,7 +91,6 @@ Definition operation_is_redundant (op: operation) (nv: nval): bool := | Ocast16signed => sign_ext_redundant 16 nv | Oandimm n => andimm_redundant nv n | Oorimm n => orimm_redundant nv n - | Osingleoffloat => singleoffloat_redundant nv | _ => false end. @@ -180,7 +183,6 @@ Proof. - apply notint_sound; auto. - apply notint_sound. apply needs_of_shift_sound; auto. - apply needs_of_shift_sound; auto. -- apply singleoffloat_sound; auto. Qed. Lemma operation_is_redundant_sound: @@ -195,7 +197,6 @@ Proof. - apply sign_ext_redundant_sound; auto. omega. - apply andimm_redundant_sound; auto. - apply orimm_redundant_sound; auto. -- apply singleoffloat_redundant_sound; auto. Qed. End SOUNDNESS. diff --git a/arm/Op.v b/arm/Op.v index b50a7b00..e7971f0b 100644 --- a/arm/Op.v +++ b/arm/Op.v @@ -57,11 +57,14 @@ Inductive condition : Type := | Ccompushift: comparison -> shift -> condition (**r unsigned integer comparison *) | Ccompimm: comparison -> int -> condition (**r signed integer comparison with a constant *) | Ccompuimm: comparison -> int -> condition (**r unsigned integer comparison with a constant *) - | Ccompf: comparison -> condition (**r floating-point comparison *) + | Ccompf: comparison -> condition (**r 64-bit floating-point comparison *) | Cnotcompf: comparison -> condition (**r negation of a floating-point comparison *) | Ccompfzero: comparison -> condition (**r floating-point comparison with 0.0 *) - | Cnotcompfzero: comparison -> condition. (**r negation of a floating-point comparison with 0.0 *) - + | Cnotcompfzero: comparison -> condition (**r negation of a floating-point comparison with 0.0 *) + | Ccompfs: comparison -> condition (**r 32-bit floating-point comparison *) + | Cnotcompfs: comparison -> condition (**r negation of a floating-point comparison *) + | Ccompfszero: comparison -> condition (**r floating-point comparison with 0.0 *) + | Cnotcompfszero: comparison -> condition. (**r negation of a floating-point comparison with 0.0 *) (** Arithmetic and logical operations. In the descriptions, [rd] is the result of the operation and [r1], [r2], etc, are the arguments. *) @@ -69,7 +72,8 @@ Inductive condition : Type := Inductive operation : Type := | Omove: operation (**r [rd = r1] *) | Ointconst: int -> operation (**r [rd] is set to the given integer constant *) - | Ofloatconst: float -> operation (**r [rd] is set to the given float constant *) + | Ofloatconst: float -> operation (**r [rd] is set to the given 64-bit float constant *) + | Osingleconst: float32 -> operation (**r [rd] is set to the given 32-bit float constant *) | Oaddrsymbol: ident -> int -> operation (**r [rd] is set to the the address of the symbol plus the offset *) | Oaddrstack: int -> operation (**r [rd] is set to the stack pointer plus the given offset *) (*c Integer arithmetic: *) @@ -113,12 +117,23 @@ Inductive operation : Type := | Osubf: operation (**r [rd = r1 - r2] *) | Omulf: operation (**r [rd = r1 * r2] *) | Odivf: operation (**r [rd = r1 / r2] *) + | Onegfs: operation (**r [rd = - r1] *) + | Oabsfs: operation (**r [rd = abs(r1)] *) + | Oaddfs: operation (**r [rd = r1 + r2] *) + | Osubfs: operation (**r [rd = r1 - r2] *) + | Omulfs: operation (**r [rd = r1 * r2] *) + | Odivfs: operation (**r [rd = r1 / r2] *) | Osingleoffloat: operation (**r [rd] is [r1] truncated to single-precision float *) + | Ofloatofsingle: operation (**r [rd] is [r1] expanded to double-precision float *) (*c Conversions between int and float: *) | Ointoffloat: operation (**r [rd = signed_int_of_float(r1)] *) | Ointuoffloat: operation (**r [rd = unsigned_int_of_float(r1)] *) | Ofloatofint: operation (**r [rd = float_of_signed_int(r1)] *) | Ofloatofintu: operation (**r [rd = float_of_unsigned_int(r1)] *) + | Ointofsingle: operation (**r [rd = signed_int_of_single(r1)] *) + | Ointuofsingle: operation (**r [rd = unsigned_int_of_single(r1)] *) + | Osingleofint: operation (**r [rd = single_of_signed_int(r1)] *) + | Osingleofintu: operation (**r [rd = single_of_unsigned_int(r1)] *) (*c Manipulating 64-bit integers: *) | Omakelong: operation (**r [rd = r1 << 32 | r2] *) | Olowlong: operation (**r [rd = low-word(r1)] *) @@ -160,6 +175,7 @@ Definition eq_operation (x y: operation): {x=y} + {x<>y}. Proof. generalize Int.eq_dec; intro. generalize Float.eq_dec; intro. + generalize Float32.eq_dec; intro. assert (forall (x y: ident), {x=y}+{x<>y}). exact peq. generalize eq_shift; intro. generalize eq_condition; intro. @@ -203,6 +219,10 @@ Definition eval_condition (cond: condition) (vl: list val) (m: mem): | Cnotcompf c, v1 :: v2 :: nil => option_map negb (Val.cmpf_bool c v1 v2) | Ccompfzero c, v1 :: nil => Val.cmpf_bool c v1 (Vfloat Float.zero) | Cnotcompfzero c, v1 :: nil => option_map negb (Val.cmpf_bool c v1 (Vfloat Float.zero)) + | Ccompfs c, v1 :: v2 :: nil => Val.cmpfs_bool c v1 v2 + | Cnotcompfs c, v1 :: v2 :: nil => option_map negb (Val.cmpfs_bool c v1 v2) + | Ccompfszero c, v1 :: nil => Val.cmpfs_bool c v1 (Vsingle Float32.zero) + | Cnotcompfszero c, v1 :: nil => option_map negb (Val.cmpfs_bool c v1 (Vsingle Float32.zero)) | _, _ => None end. @@ -213,6 +233,7 @@ Definition eval_operation | Omove, v1::nil => Some v1 | Ointconst n, nil => Some (Vint n) | Ofloatconst n, nil => Some (Vfloat n) + | Osingleconst n, nil => Some (Vsingle n) | Oaddrsymbol s ofs, nil => Some (Genv.symbol_address genv s ofs) | Oaddrstack ofs, nil => Some (Val.add sp (Vint ofs)) | Ocast8signed, v1::nil => Some (Val.sign_ext 8 v1) @@ -254,11 +275,22 @@ Definition eval_operation | Osubf, v1::v2::nil => Some(Val.subf v1 v2) | Omulf, v1::v2::nil => Some(Val.mulf v1 v2) | Odivf, v1::v2::nil => Some(Val.divf v1 v2) + | Onegfs, v1::nil => Some(Val.negfs v1) + | Oabsfs, v1::nil => Some(Val.absfs v1) + | Oaddfs, v1::v2::nil => Some(Val.addfs v1 v2) + | Osubfs, v1::v2::nil => Some(Val.subfs v1 v2) + | Omulfs, v1::v2::nil => Some(Val.mulfs v1 v2) + | Odivfs, v1::v2::nil => Some(Val.divfs v1 v2) | Osingleoffloat, v1::nil => Some(Val.singleoffloat v1) + | Ofloatofsingle, v1::nil => Some(Val.floatofsingle v1) | Ointoffloat, v1::nil => Val.intoffloat v1 | Ointuoffloat, v1::nil => Val.intuoffloat v1 | Ofloatofint, v1::nil => Val.floatofint v1 | Ofloatofintu, v1::nil => Val.floatofintu v1 + | Ointofsingle, v1::nil => Val.intofsingle v1 + | Ointuofsingle, v1::nil => Val.intuofsingle v1 + | Osingleofint, v1::nil => Val.singleofint v1 + | Osingleofintu, v1::nil => Val.singleofintu v1 | Omakelong, v1::v2::nil => Some(Val.longofwords v1 v2) | Olowlong, v1::nil => Some(Val.loword v1) | Ohighlong, v1::nil => Some(Val.hiword v1) @@ -281,8 +313,6 @@ Ltac FuncInv := match goal with | H: (match ?x with nil => _ | _ :: _ => _ end = Some _) |- _ => destruct x; simpl in H; try discriminate; FuncInv - | H: (match ?v with Vundef => _ | Vint _ => _ | Vfloat _ => _ | Vptr _ _ => _ end = Some _) |- _ => - destruct v; simpl in H; try discriminate; FuncInv | H: (Some _ = Some _) |- _ => injection H; intros; clear H; FuncInv | _ => @@ -303,13 +333,18 @@ Definition type_of_condition (c: condition) : list typ := | Cnotcompf _ => Tfloat :: Tfloat :: nil | Ccompfzero _ => Tfloat :: nil | Cnotcompfzero _ => Tfloat :: nil + | Ccompfs _ => Tsingle :: Tsingle :: nil + | Cnotcompfs _ => Tsingle :: Tsingle :: nil + | Ccompfszero _ => Tsingle :: nil + | Cnotcompfszero _ => Tsingle :: nil end. Definition type_of_operation (op: operation) : list typ * typ := match op with | Omove => (nil, Tint) (* treated specially *) | Ointconst _ => (nil, Tint) - | Ofloatconst f => (nil, if Float.is_single_dec f then Tsingle else Tfloat) + | Ofloatconst f => (nil, Tfloat) + | Osingleconst f => (nil, Tsingle) | Oaddrsymbol _ _ => (nil, Tint) | Oaddrstack _ => (nil, Tint) | Ocast8signed => (Tint :: nil, Tint) @@ -351,11 +386,22 @@ Definition type_of_operation (op: operation) : list typ * typ := | Osubf => (Tfloat :: Tfloat :: nil, Tfloat) | Omulf => (Tfloat :: Tfloat :: nil, Tfloat) | Odivf => (Tfloat :: Tfloat :: nil, Tfloat) + | Onegfs => (Tsingle :: nil, Tsingle) + | Oabsfs => (Tsingle :: nil, Tsingle) + | Oaddfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Osubfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Omulfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Odivfs => (Tsingle :: Tsingle :: nil, Tsingle) | Osingleoffloat => (Tfloat :: nil, Tsingle) + | Ofloatofsingle => (Tsingle :: nil, Tfloat) | Ointoffloat => (Tfloat :: nil, Tint) | Ointuoffloat => (Tfloat :: nil, Tint) | Ofloatofint => (Tint :: nil, Tfloat) | Ofloatofintu => (Tint :: nil, Tfloat) + | Ointofsingle => (Tsingle :: nil, Tint) + | Ointuofsingle => (Tsingle :: nil, Tint) + | Osingleofint => (Tint :: nil, Tsingle) + | Osingleofintu => (Tint :: nil, Tsingle) | Omakelong => (Tint :: Tint :: nil, Tlong) | Olowlong => (Tlong :: nil, Tint) | Ohighlong => (Tlong :: nil, Tint) @@ -390,7 +436,6 @@ Proof with (try exact I). intros. destruct op; simpl; simpl in H0; FuncInv; try subst v... congruence. - destruct (Float.is_single_dec f); red; auto. unfold Genv.symbol_address. destruct (Genv.find_symbol genv i)... destruct sp... destruct v0... @@ -433,9 +478,20 @@ Proof with (try exact I). destruct v0; destruct v1... destruct v0; destruct v1... destruct v0; destruct v1... - destruct v0... simpl. apply Float.singleoffloat_is_single. - destruct v0; simpl in H0; inv H0. destruct (Float.intoffloat f); simpl in H2; inv H2... - destruct v0; simpl in H0; inv H0. destruct (Float.intuoffloat f); simpl in H2; inv H2... + destruct v0... + destruct v0... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0... + destruct v0... + destruct v0; simpl in H0; inv H0. destruct (Float.to_int f); simpl in H2; inv H2... + destruct v0; simpl in H0; inv H0. destruct (Float.to_intu f); simpl in H2; inv H2... + destruct v0; simpl in H0; inv H0... + destruct v0; simpl in H0; inv H0... + destruct v0; simpl in H0; inv H0. destruct (Float32.to_int f); simpl in H2; inv H2... + destruct v0; simpl in H0; inv H0. destruct (Float32.to_intu f); simpl in H2; inv H2... destruct v0; simpl in H0; inv H0... destruct v0; simpl in H0; inv H0... destruct v0; destruct v1... @@ -503,6 +559,10 @@ Definition negate_condition (cond: condition): condition := | Cnotcompf c => Ccompf c | Ccompfzero c => Cnotcompfzero c | Cnotcompfzero c => Ccompfzero c + | Ccompfs c => Cnotcompfs c + | Cnotcompfs c => Ccompfs c + | Ccompfszero c => Cnotcompfszero c + | Cnotcompfszero c => Ccompfszero c end. Lemma eval_negate_condition: @@ -520,6 +580,10 @@ Proof. repeat (destruct vl; auto). destruct (Val.cmpf_bool c v v0); auto. destruct b; auto. repeat (destruct vl; auto). repeat (destruct vl; auto). destruct (Val.cmpf_bool c v (Vfloat Float.zero)); auto. destruct b; auto. + repeat (destruct vl; auto). + repeat (destruct vl; auto). destruct (Val.cmpfs_bool c v v0); auto. destruct b; auto. + repeat (destruct vl; auto). + repeat (destruct vl; auto). destruct (Val.cmpfs_bool c v (Vsingle Float32.zero)); auto. destruct b; auto. Qed. (** Shifting stack-relative references. This is used in [Stacking]. *) @@ -772,6 +836,8 @@ Ltac InvInject := inv H; InvInject | [ H: val_inject _ (Vfloat _) _ |- _ ] => inv H; InvInject + | [ H: val_inject _ (Vsingle _) _ |- _ ] => + inv H; InvInject | [ H: val_inject _ (Vptr _ _) _ |- _ ] => inv H; InvInject | [ H: val_list_inject _ nil _ |- _ ] => @@ -804,6 +870,10 @@ Proof. inv H3; inv H2; simpl in H0; inv H0; auto. inv H3; simpl in H0; inv H0; auto. inv H3; simpl in H0; inv H0; auto. + inv H3; inv H2; simpl in H0; inv H0; auto. + inv H3; inv H2; simpl in H0; inv H0; auto. + inv H3; simpl in H0; inv H0; auto. + inv H3; simpl in H0; inv H0; auto. Qed. Ltac TrivialExists := @@ -869,11 +939,27 @@ Proof. inv H4; inv H2; simpl; auto. inv H4; inv H2; simpl; auto. inv H4; inv H2; simpl; auto. + + inv H4; simpl; auto. + inv H4; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; simpl; auto. + inv H4; simpl; auto. + + inv H4; simpl in H1; inv H1. simpl. destruct (Float.to_int f0); simpl in H2; inv H2. + exists (Vint i); auto. + inv H4; simpl in H1; inv H1. simpl. destruct (Float.to_intu f0); simpl in H2; inv H2. + exists (Vint i); auto. + inv H4; simpl in *; inv H1. TrivialExists. + inv H4; simpl in *; inv H1. TrivialExists. - inv H4; simpl in H1; inv H1. simpl. destruct (Float.intoffloat f0); simpl in H2; inv H2. + inv H4; simpl in H1; inv H1. simpl. destruct (Float32.to_int f0); simpl in H2; inv H2. exists (Vint i); auto. - inv H4; simpl in H1; inv H1. simpl. destruct (Float.intuoffloat f0); simpl in H2; inv H2. + inv H4; simpl in H1; inv H1. simpl. destruct (Float32.to_intu f0); simpl in H2; inv H2. exists (Vint i); auto. inv H4; simpl in *; inv H1. TrivialExists. inv H4; simpl in *; inv H1. TrivialExists. diff --git a/arm/PrintAsm.ml b/arm/PrintAsm.ml index 7e6827ef..2ad91142 100644 --- a/arm/PrintAsm.ml +++ b/arm/PrintAsm.ml @@ -161,9 +161,10 @@ let distance_to_emit_constants () = (* Associate labels to floating-point constants and to symbols *) let float_labels = (Hashtbl.create 39 : (int64, int) Hashtbl.t) +let float32_labels = (Hashtbl.create 39 : (int32, int) Hashtbl.t) let label_float f = - let bf = camlint64_of_coqint(Floats.Float.bits_of_double f) in + let bf = camlint64_of_coqint(Floats.Float.to_bits f) in try Hashtbl.find float_labels bf with Not_found -> @@ -173,6 +174,17 @@ let label_float f = max_pos_constants := min !max_pos_constants (!currpos + 1024); lbl' +let label_float32 f = + let bf = camlint_of_coqint(Floats.Float32.to_bits f) in + try + Hashtbl.find float32_labels bf + with Not_found -> + let lbl' = new_label() in + Hashtbl.add float32_labels bf lbl'; + size_constants := !size_constants + 4; + max_pos_constants := min !max_pos_constants (!currpos + 1024); + lbl' + let symbol_labels = (Hashtbl.create 39 : (ident * Integers.Int.int, int) Hashtbl.t) @@ -188,6 +200,7 @@ let label_symbol id ofs = let reset_constants () = Hashtbl.clear float_labels; + Hashtbl.clear float32_labels; Hashtbl.clear symbol_labels; size_constants := 0; max_pos_constants := max_int @@ -200,6 +213,10 @@ let emit_constants oc = and bflo = Int64.logand bf 0xFFFF_FFFFL in fprintf oc ".L%d: .word 0x%Lx, 0x%Lx\n" lbl bflo bfhi) float_labels; + Hashtbl.iter + (fun bf lbl -> + fprintf oc ".L%d: .word 0x%lx\n" lbl bf) + float32_labels; Hashtbl.iter (fun (id, ofs) lbl -> fprintf oc ".L%d: .word %a\n" @@ -348,8 +365,7 @@ let print_builtin_vload_common oc chunk args res = fprintf oc " ldr %a, [%a, #0]\n" ireg res2 ireg addr end; 2 | Mfloat32, [IR addr], [FR res] -> - fprintf oc " flds %a, [%a, #0]\n" freg_single res ireg addr; - fprintf oc " fcvtds %a, %a\n" freg res freg_single res; 2 + fprintf oc " flds %a, [%a, #0]\n" freg_single res ireg addr; 1 | Mfloat64, [IR addr], [FR res] -> fprintf oc " fldd %a, [%a, #0]\n" freg res ireg addr; 1 | _ -> @@ -379,8 +395,7 @@ let print_builtin_vstore_common oc chunk args = fprintf oc " str %a, [%a, #0]\n" ireg src2 ireg addr; fprintf oc " str %a, [%a, #4]\n" ireg src1 ireg addr; 2 | Mfloat32, [IR addr; FR src] -> - fprintf oc " fcvtsd %a, %a\n" freg_single FR6 freg src; - fprintf oc " fsts %a, [%a, #0]\n" freg_single FR6 ireg addr; 2 + fprintf oc " fsts %a, [%a, #0]\n" freg_single src ireg addr; 1 | Mfloat64, [IR addr; FR src] -> fprintf oc " fstd %a, [%a, #0]\n" freg src ireg addr; 1 | _ -> @@ -409,11 +424,11 @@ let align n a = (n + a - 1) land (-a) let rec next_arg_location ir ofs = function | [] -> Int32.of_int (ir * 4 + ofs) - | (Tint | Tsingle) :: l -> + | (Tint | Tsingle | Tany32) :: l -> if ir < 4 then next_arg_location (ir + 1) ofs l else next_arg_location ir (ofs + 4) l - | (Tfloat | Tlong) :: l -> + | (Tfloat | Tlong | Tany64) :: l -> if ir < 3 then next_arg_location (align ir 2 + 2) ofs l else next_arg_location ir (align ofs 8 + 8) l @@ -525,11 +540,9 @@ module FixupEABI = struct let fixup_single oc dir f i = match dir with - | Incoming -> (* f <- i; f <- double_of_single f *) - fprintf oc " fmsr %a, %a\n" freg_single f ireg i; - fprintf oc " fcvtds %a, %a\n" freg f freg_single f - | Outgoing -> (* f <- single_of_double f; i <- f *) - fprintf oc " fcvtsd %a, %a\n" freg_single f freg f; + | Incoming -> (* f <- i *) + fprintf oc " fmsr %a, %a\n" freg_single f ireg i + | Outgoing -> (* i <- f *) fprintf oc " fmrs %a, %a\n" ireg i freg_single f let fixup_conventions oc dir tyl = @@ -537,11 +550,11 @@ module FixupEABI = struct if i >= 4 then 0 else match tyl with | [] -> 0 - | Tint :: tyl' -> + | (Tint | Tany32) :: tyl' -> fixup (i+1) tyl' | Tlong :: tyl' -> fixup (((i + 1) land (-2)) + 2) tyl' - | Tfloat :: tyl' -> + | (Tfloat | Tany64) :: tyl' -> let i = (i + 1) land (-2) in if i >= 4 then 0 else begin fixup_double oc dir (freg_param i) (ireg_param i) (ireg_param (i+1)); @@ -549,7 +562,7 @@ module FixupEABI = struct end | Tsingle :: tyl' -> fixup_single oc dir (freg_param i) (ireg_param i); - 2 + fixup (i+1) tyl' + 1 + fixup (i+1) tyl' in fixup 0 tyl let fixup_arguments oc dir sg = @@ -577,8 +590,8 @@ module FixupHF = struct let rec fixup_actions used fr tyl = match tyl with | [] -> [] - | (Tint | Tlong) :: tyl' -> fixup_actions used fr tyl' - | Tfloat :: tyl' -> + | (Tint | Tlong | Tany32) :: tyl' -> fixup_actions used fr tyl' + | (Tfloat | Tany64) :: tyl' -> if fr >= 8 then [] else begin let dr = find_double used 0 in assert (dr < 8); @@ -599,7 +612,7 @@ module FixupHF = struct 1 + fixup_outgoing oc act end | (fr, Single, sr) :: act -> - fprintf oc " fcvtsd s%d, d%d\n" sr fr; + fprintf oc " fcpys s%d, s%d\n" sr (2*fr); 1 + fixup_outgoing oc act let rec fixup_incoming oc = function @@ -612,8 +625,10 @@ module FixupHF = struct end | (fr, Single, sr) :: act -> let n = fixup_incoming oc act in - fprintf oc " fcvtds d%d, s%d\n" fr sr; + if fr = sr then n else begin + fprintf oc " fcpys s%d, s%d\n" (2*fr) sr; 1 + n + end let fixup_arguments oc dir sg = if sg.sig_cc.cc_vararg then @@ -628,14 +643,8 @@ module FixupHF = struct let fixup_result oc dir sg = if sg.sig_cc.cc_vararg then FixupEABI.fixup_result oc dir sg - else begin - match proj_sig_res sg, dir with - | Tsingle, Outgoing -> - fprintf oc " fcvtsd s0, d0\n"; 1 - | Tsingle, Incoming -> - fprintf oc " fcvtds d0, s0\n"; 1 - | _ -> 0 - end + else + 0 end let (fixup_arguments, fixup_result) = @@ -703,7 +712,7 @@ let print_instruction oc = function fprintf oc " cmp %a, %a\n" ireg r1 shift_op so; 1 | Peor(r1, r2, so) -> fprintf oc " eor %a, %a, %a\n" ireg r1 ireg r2 shift_op so; 1 - | Pldr(r1, r2, sa) -> + | Pldr(r1, r2, sa) | Pldr_a(r1, r2, sa) -> fprintf oc " ldr %a, [%a, %a]\n" ireg r1 ireg r2 shift_addr sa; 1 | Pldrb(r1, r2, sa) -> fprintf oc " ldrb %a, [%a, %a]\n" ireg r1 ireg r2 shift_addr sa; 1 @@ -727,7 +736,7 @@ let print_instruction oc = function fprintf oc " orr %a, %a, %a\n" ireg r1 ireg r2 shift_op so; 1 | Prsb(r1, r2, so) -> fprintf oc " rsb %a, %a, %a\n" ireg r1 ireg r2 shift_op so; 1 - | Pstr(r1, r2, sa) -> + | Pstr(r1, r2, sa) | Pstr_a(r1, r2, sa) -> fprintf oc " str %a, [%a, %a]\n" ireg r1 ireg r2 shift_addr sa; begin match r1, r2, sa with | IR14, IR13, SAimm n -> cfi_rel_offset oc "lr" (camlint_of_coqint n) @@ -786,19 +795,53 @@ let print_instruction oc = function | Pftouizd(r1, r2) -> fprintf oc " ftouizd %a, %a\n" freg_single FR6 freg r2; fprintf oc " fmrs %a, %a\n" ireg r1 freg_single FR6; 2 + | Pfabss(r1, r2) -> + fprintf oc " fabss %a, %a\n" freg_single r1 freg_single r2; 1 + | Pfnegs(r1, r2) -> + fprintf oc " fnegs %a, %a\n" freg_single r1 freg_single r2; 1 + | Pfadds(r1, r2, r3) -> + fprintf oc " fadds %a, %a, %a\n" freg_single r1 freg_single r2 freg_single r3; 1 + | Pfdivs(r1, r2, r3) -> + fprintf oc " fdivs %a, %a, %a\n" freg_single r1 freg_single r2 freg_single r3; 1 + | Pfmuls(r1, r2, r3) -> + fprintf oc " fmuls %a, %a, %a\n" freg_single r1 freg_single r2 freg_single r3; 1 + | Pfsubs(r1, r2, r3) -> + fprintf oc " fsubs %a, %a, %a\n" freg_single r1 freg_single r2 freg_single r3; 1 + | Pflis(r1, f) -> + (* We could make good use of the fconsts instruction, but it's available + in VFD v3 and up, not in v1 nor v2 *) + let lbl = label_float32 f in + fprintf oc " flds %a, .L%d @ %.12g\n" freg_single r1 lbl (camlfloat_of_coqfloat32 f); 1 + | Pfcmps(r1, r2) -> + fprintf oc " fcmps %a, %a\n" freg_single r1 freg_single r2; + fprintf oc " fmstat\n"; 2 + | Pfcmpzs(r1) -> + fprintf oc " fcmpzs %a\n" freg_single r1; + fprintf oc " fmstat\n"; 2 + | Pfsitos(r1, r2) -> + fprintf oc " fmsr %a, %a\n" freg_single r1 ireg r2; + fprintf oc " fsitos %a, %a\n" freg_single r1 freg_single r1; 2 + | Pfuitos(r1, r2) -> + fprintf oc " fmsr %a, %a\n" freg_single r1 ireg r2; + fprintf oc " fuitos %a, %a\n" freg_single r1 freg_single r1; 2 + | Pftosizs(r1, r2) -> + fprintf oc " ftosizs %a, %a\n" freg_single FR6 freg_single r2; + fprintf oc " fmrs %a, %a\n" ireg r1 freg_single FR6; 2 + | Pftouizs(r1, r2) -> + fprintf oc " ftouizs %a, %a\n" freg_single FR6 freg_single r2; + fprintf oc " fmrs %a, %a\n" ireg r1 freg_single FR6; 2 | Pfcvtsd(r1, r2) -> - fprintf oc " fcvtsd %a, %a\n" freg_single r1 freg r2; - fprintf oc " fcvtds %a, %a\n" freg r1 freg_single r1; 2 - | Pfldd(r1, r2, n) -> + fprintf oc " fcvtsd %a, %a\n" freg_single r1 freg r2; 1 + | Pfcvtds(r1, r2) -> + fprintf oc " fcvtds %a, %a\n" freg r1 freg_single r2; 1 + | Pfldd(r1, r2, n) | Pfldd_a(r1, r2, n) -> fprintf oc " fldd %a, [%a, #%a]\n" freg r1 ireg r2 coqint n; 1 | Pflds(r1, r2, n) -> - fprintf oc " flds %a, [%a, #%a]\n" freg_single r1 ireg r2 coqint n; - fprintf oc " fcvtds %a, %a\n" freg r1 freg_single r1; 2 - | Pfstd(r1, r2, n) -> + fprintf oc " flds %a, [%a, #%a]\n" freg_single r1 ireg r2 coqint n; 1 + | Pfstd(r1, r2, n) | Pfstd_a(r1, r2, n) -> fprintf oc " fstd %a, [%a, #%a]\n" freg r1 ireg r2 coqint n; 1 | Pfsts(r1, r2, n) -> - fprintf oc " fcvtsd %a, %a\n" freg_single FR6 freg r1; - fprintf oc " fsts %a, [%a, #%a]\n" freg_single FR6 ireg r2 coqint n; 2 + fprintf oc " fsts %a, [%a, #%a]\n" freg_single r1 ireg r2 coqint n; 1 (* Pseudo-instructions *) | Pallocframe(sz, ofs) -> fprintf oc " mov r12, sp\n"; @@ -927,10 +970,10 @@ let print_init oc = function | Init_int64 n -> fprintf oc " .quad %Ld\n" (camlint64_of_coqint n) | Init_float32 n -> - fprintf oc " .word 0x%lx %s %.15g \n" (camlint_of_coqint (Floats.Float.bits_of_single n)) + fprintf oc " .word 0x%lx %s %.15g \n" (camlint_of_coqint (Floats.Float32.to_bits n)) comment (camlfloat_of_coqfloat n) | Init_float64 n -> - fprintf oc " .quad %Ld %s %.18g\n" (camlint64_of_coqint (Floats.Float.bits_of_double n)) + fprintf oc " .quad %Ld %s %.18g\n" (camlint64_of_coqint (Floats.Float.to_bits n)) comment (camlfloat_of_coqfloat n) | Init_space n -> if Z.gt n Z.zero then diff --git a/arm/SelectOp.vp b/arm/SelectOp.vp index e4005c92..ad8a9454 100644 --- a/arm/SelectOp.vp +++ b/arm/SelectOp.vp @@ -321,6 +321,12 @@ Definition addf (e1 e2: expr) := Eop Oaddf (e1 ::: e2 ::: Enil). Definition subf (e1 e2: expr) := Eop Osubf (e1 ::: e2 ::: Enil). Definition mulf (e1 e2: expr) := Eop Omulf (e1 ::: e2 ::: Enil). +Definition negfs (e: expr) := Eop Onegfs (e ::: Enil). +Definition absfs (e: expr) := Eop Oabsfs (e ::: Enil). +Definition addfs (e1 e2: expr) := Eop Oaddfs (e1 ::: e2 ::: Enil). +Definition subfs (e1 e2: expr) := Eop Osubfs (e1 ::: e2 ::: Enil). +Definition mulfs (e1 e2: expr) := Eop Omulfs (e1 ::: e2 ::: Enil). + (** ** Comparisons *) Nondetfunction compimm (default: comparison -> int -> condition) @@ -370,6 +376,9 @@ Nondetfunction compu (c: comparison) (e1: expr) (e2: expr) := Definition compf (c: comparison) (e1: expr) (e2: expr) := Eop (Ocmp (Ccompf c)) (e1 ::: e2 ::: Enil). +Definition compfs (c: comparison) (e1: expr) (e2: expr) := + Eop (Ocmp (Ccompfs c)) (e1 ::: e2 ::: Enil). + (** ** Integer conversions *) Definition cast8unsigned (e: expr) := andimm (Int.repr 255) e. @@ -391,21 +400,38 @@ Nondetfunction cast16signed (e: expr) := (** ** Floating-point conversions *) Definition singleoffloat (e: expr) := Eop Osingleoffloat (e ::: Enil). +Definition floatofsingle (e: expr) := Eop Ofloatofsingle (e ::: Enil). + Definition intoffloat (e: expr) := Eop Ointoffloat (e ::: Enil). Definition intuoffloat (e: expr) := Eop Ointuoffloat (e ::: Enil). Nondetfunction floatofint (e: expr) := match e with - | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.floatofint n)) Enil + | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.of_int n)) Enil | _ => Eop Ofloatofint (e ::: Enil) end. Nondetfunction floatofintu (e: expr) := match e with - | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.floatofintu n)) Enil + | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.of_intu n)) Enil | _ => Eop Ofloatofintu (e ::: Enil) end. +Definition intofsingle (e: expr) := Eop Ointofsingle (e ::: Enil). +Definition intuofsingle (e: expr) := Eop Ointuofsingle (e ::: Enil). + +Nondetfunction singleofint (e: expr) := + match e with + | Eop (Ointconst n) Enil => Eop (Osingleconst (Float32.of_int n)) Enil + | _ => Eop Osingleofint (e ::: Enil) + end. + +Nondetfunction singleofintu (e: expr) := + match e with + | Eop (Ointconst n) Enil => Eop (Osingleconst (Float32.of_intu n)) Enil + | _ => Eop Osingleofintu (e ::: Enil) + end. + (** ** Recognition of addressing modes for load and store operations *) (** We do not recognize the [Aindexed2] and [Aindexed2shift] modes @@ -424,6 +450,8 @@ Definition can_use_Aindexed2 (chunk: memory_chunk): bool := | Mint64 => false | Mfloat32 => false | Mfloat64 => false + | Many32 => true + | Many64 => false end. Definition can_use_Aindexed2shift (chunk: memory_chunk): bool := @@ -436,6 +464,8 @@ Definition can_use_Aindexed2shift (chunk: memory_chunk): bool := | Mint64 => false | Mfloat32 => false | Mfloat64 => false + | Many32 => true + | Many64 => false end. Nondetfunction addressing (chunk: memory_chunk) (e: expr) := diff --git a/arm/SelectOpproof.v b/arm/SelectOpproof.v index 1dd2c200..c68d2277 100644 --- a/arm/SelectOpproof.v +++ b/arm/SelectOpproof.v @@ -598,6 +598,31 @@ Proof. red; intros; TrivialExists. Qed. +Theorem eval_negfs: unary_constructor_sound negfs Val.negfs. +Proof. + red; intros. TrivialExists. +Qed. + +Theorem eval_absfs: unary_constructor_sound absfs Val.absfs. +Proof. + red; intros. TrivialExists. +Qed. + +Theorem eval_addfs: binary_constructor_sound addfs Val.addfs. +Proof. + red; intros; TrivialExists. +Qed. + +Theorem eval_subfs: binary_constructor_sound subfs Val.subfs. +Proof. + red; intros; TrivialExists. +Qed. + +Theorem eval_mulfs: binary_constructor_sound mulfs Val.mulfs. +Proof. + red; intros; TrivialExists. +Qed. + Section COMP_IMM. Variable default: comparison -> int -> condition. @@ -693,6 +718,12 @@ Proof. intros; red; intros. unfold compf. TrivialExists. Qed. +Theorem eval_compfs: + forall c, binary_constructor_sound (compfs c) (Val.cmpfs c). +Proof. + intros; red; intros. unfold compfs. TrivialExists. +Qed. + Theorem eval_cast8signed: unary_constructor_sound cast8signed (Val.sign_ext 8). Proof. red; intros until x. unfold cast8signed; case (cast8signed_match a); intros. @@ -724,6 +755,11 @@ Proof. red; intros. unfold singleoffloat. TrivialExists. Qed. +Theorem eval_floatofsingle: unary_constructor_sound floatofsingle Val.floatofsingle. +Proof. + red; intros. unfold floatofsingle. TrivialExists. +Qed. + Theorem eval_intoffloat: forall le a x y, eval_expr ge sp e m le a x -> @@ -764,6 +800,46 @@ Proof. TrivialExists. Qed. +Theorem eval_intofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.intofsingle x = Some y -> + exists v, eval_expr ge sp e m le (intofsingle a) v /\ Val.lessdef y v. +Proof. + intros; unfold intofsingle. TrivialExists. +Qed. + +Theorem eval_singleofint: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.singleofint x = Some y -> + exists v, eval_expr ge sp e m le (singleofint a) v /\ Val.lessdef y v. +Proof. + intros until y; unfold singleofint. case (singleofint_match a); intros. + InvEval. simpl in H0. TrivialExists. + TrivialExists. +Qed. + +Theorem eval_intuofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.intuofsingle x = Some y -> + exists v, eval_expr ge sp e m le (intuofsingle a) v /\ Val.lessdef y v. +Proof. + intros; unfold intuofsingle. TrivialExists. +Qed. + +Theorem eval_singleofintu: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.singleofintu x = Some y -> + exists v, eval_expr ge sp e m le (singleofintu a) v /\ Val.lessdef y v. +Proof. + intros until y; unfold singleofintu. case (singleofintu_match a); intros. + InvEval. simpl in H0. TrivialExists. + TrivialExists. +Qed. + Theorem eval_addressing: forall le chunk a v b ofs, eval_expr ge sp e m le a v -> diff --git a/arm/ValueAOp.v b/arm/ValueAOp.v index b3123618..a14d6b98 100644 --- a/arm/ValueAOp.v +++ b/arm/ValueAOp.v @@ -44,6 +44,10 @@ Definition eval_static_condition (cond: condition) (vl: list aval): abool := | Cnotcompf c, v1 :: v2 :: nil => cnot (cmpf_bool c v1 v2) | Ccompfzero c, v1 :: nil => cmpf_bool c v1 (F Float.zero) | Cnotcompfzero c, v1 :: nil => cnot (cmpf_bool c v1 (F Float.zero)) + | Ccompfs c, v1 :: v2 :: nil => cmpfs_bool c v1 v2 + | Cnotcompfs c, v1 :: v2 :: nil => cnot (cmpfs_bool c v1 v2) + | Ccompfszero c, v1 :: nil => cmpfs_bool c v1 (FS Float32.zero) + | Cnotcompfszero c, v1 :: nil => cnot (cmpfs_bool c v1 (FS Float32.zero)) | _, _ => Bnone end. @@ -61,6 +65,7 @@ Definition eval_static_operation (op: operation) (vl: list aval): aval := | Omove, v1::nil => v1 | Ointconst n, nil => I n | Ofloatconst n, nil => if propagate_float_constants tt then F n else ftop + | Osingleconst n, nil => if propagate_float_constants tt then FS n else ftop | Oaddrsymbol id ofs, nil => Ptr (Gl id ofs) | Oaddrstack ofs, nil => Ptr (Stk ofs) | Ocast8signed, v1 :: nil => sign_ext 8 v1 @@ -102,11 +107,22 @@ Definition eval_static_operation (op: operation) (vl: list aval): aval := | Osubf, v1::v2::nil => subf v1 v2 | Omulf, v1::v2::nil => mulf v1 v2 | Odivf, v1::v2::nil => divf v1 v2 + | Onegfs, v1::nil => negfs v1 + | Oabsfs, v1::nil => absfs v1 + | Oaddfs, v1::v2::nil => addfs v1 v2 + | Osubfs, v1::v2::nil => subfs v1 v2 + | Omulfs, v1::v2::nil => mulfs v1 v2 + | Odivfs, v1::v2::nil => divfs v1 v2 | Osingleoffloat, v1::nil => singleoffloat v1 + | Ofloatofsingle, v1::nil => floatofsingle v1 | Ointoffloat, v1::nil => intoffloat v1 | Ointuoffloat, v1::nil => intuoffloat v1 | Ofloatofint, v1::nil => floatofint v1 | Ofloatofintu, v1::nil => floatofintu v1 + | Ointofsingle, v1::nil => intofsingle v1 + | Ointuofsingle, v1::nil => intuofsingle v1 + | Osingleofint, v1::nil => singleofint v1 + | Osingleofintu, v1::nil => singleofintu v1 | Omakelong, v1::v2::nil => longofwords v1 v2 | Olowlong, v1::nil => loword v1 | Ohighlong, v1::nil => hiword v1 @@ -185,6 +201,7 @@ Proof. unfold eval_operation, eval_static_operation; intros; destruct op; InvHyps; eauto with va. destruct (propagate_float_constants tt); constructor. + destruct (propagate_float_constants tt); constructor. rewrite Int.add_zero_l; eauto with va. fold (Val.sub (Vint i) a1). auto with va. apply of_optbool_sound. eapply eval_static_condition_sound; eauto. diff --git a/arm/eabi/Conventions1.v b/arm/eabi/Conventions1.v index c02af1aa..c26d29ee 100644 --- a/arm/eabi/Conventions1.v +++ b/arm/eabi/Conventions1.v @@ -15,6 +15,7 @@ Require Import Coqlib. Require Import AST. +Require Import Events. Require Import Locations. (** * Classification of machine registers *) @@ -178,13 +179,13 @@ Proof. Qed. Lemma int_callee_save_type: - forall r, In r int_callee_save_regs -> mreg_type r = Tint. + forall r, In r int_callee_save_regs -> mreg_type r = Tany32. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. Lemma float_callee_save_type: - forall r, In r float_callee_save_regs -> mreg_type r = Tfloat. + forall r, In r float_callee_save_regs -> mreg_type r = Tany64. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. @@ -235,12 +236,21 @@ Qed. Definition loc_result (s: signature) : list mreg := match s.(sig_res) with | None => R0 :: nil - | Some Tint => R0 :: nil - | Some (Tfloat | Tsingle) => F0 :: nil + | Some (Tint | Tany32) => R0 :: nil + | Some (Tfloat | Tsingle | Tany64) => F0 :: nil | Some Tlong => R1 :: R0 :: nil end. -(** The result location is a caller-save register or a temporary *) +(** The result registers have types compatible with that given in the signature. *) + +Lemma loc_result_type: + forall sig, + subtype_list (proj_sig_res' sig) (map mreg_type (loc_result sig)) = true. +Proof. + intros. unfold proj_sig_res', loc_result. destruct (sig_res sig) as [[]|]; auto. +Qed. + +(** The result locations are caller-save registers *) Lemma loc_result_caller_save: forall (s: signature) (r: mreg), @@ -292,12 +302,12 @@ Definition sreg_param (n: Z) : mreg := Fixpoint loc_arguments_rec (tyl: list typ) (ofs: Z) {struct tyl} : list loc := match tyl with | nil => nil - | Tint :: tys => - (if zle 0 ofs then S Outgoing ofs Tint else R (ireg_param ofs)) + | (Tint | Tany32) as ty :: tys => + (if zle 0 ofs then S Outgoing ofs ty else R (ireg_param ofs)) :: loc_arguments_rec tys (ofs + 1) - | Tfloat :: tys => + | (Tfloat | Tany64) as ty :: tys => let ofs := align ofs 2 in - (if zle 0 ofs then S Outgoing ofs Tfloat else R (freg_param ofs)) + (if zle 0 ofs then S Outgoing ofs ty else R (freg_param ofs)) :: loc_arguments_rec tys (ofs + 2) | Tsingle :: tys => (if zle 0 ofs then S Outgoing ofs Tsingle else R (sreg_param ofs)) @@ -321,8 +331,8 @@ Definition loc_arguments (s: signature) : list loc := Fixpoint size_arguments_rec (tyl: list typ) (ofs: Z) {struct tyl} : Z := match tyl with | nil => ofs - | (Tint | Tsingle) :: tys => size_arguments_rec tys (ofs + 1) - | (Tfloat | Tlong) :: tys => size_arguments_rec tys (align ofs 2 + 2) + | (Tint | Tsingle | Tany32) :: tys => size_arguments_rec tys (ofs + 1) + | (Tfloat | Tlong | Tany64) :: tys => size_arguments_rec tys (align ofs 2 + 2) end. Definition size_arguments (s: signature) : Z := @@ -404,6 +414,19 @@ Proof. split. omega. split. omega. congruence. apply sreg_param_caller_save. exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. intuition omega. +- (* Tany32 *) + destruct H. + subst l. destruct (zle 0 ofs). + split. omega. split. omega. congruence. + apply ireg_param_caller_save. + exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. intuition omega. +- (* Tany64 *) + assert (ofs <= align ofs 2) by (apply align_le; omega). + destruct H. + subst l. destruct (zle 0 (align ofs 2)). + split. omega. split. auto. congruence. + apply freg_param_caller_save. + exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. intuition omega. Qed. Lemma loc_arguments_acceptable: @@ -433,6 +456,9 @@ Proof. assert (ofs <= align ofs 2) by (apply align_le; omega). apply Zle_trans with (align ofs 2 + 2); auto; omega. apply Zle_trans with (ofs + 1); auto; omega. + apply Zle_trans with (ofs + 1); auto; omega. + assert (ofs <= align ofs 2) by (apply align_le; omega). + apply Zle_trans with (align ofs 2 + 2); auto; omega. Qed. Lemma size_arguments_above: @@ -474,6 +500,10 @@ Proof. eapply Zle_trans. 2: apply H0. simpl typesize; omega. omega. - (* Tsingle *) destruct H1; auto. destruct (zle 0 ofs0); inv H1. apply H0. omega. + - (* Tany32 *) + destruct H1; auto. destruct (zle 0 ofs0); inv H1. apply H0. omega. + - (* Tany64 *) + destruct H1; auto. destruct (zle 0 (align ofs0 2)); inv H1. apply H0. omega. } unfold size_arguments. apply H1. auto. Qed. diff --git a/arm/hardfloat/Conventions1.v b/arm/hardfloat/Conventions1.v index e3875e7d..40a761cc 100644 --- a/arm/hardfloat/Conventions1.v +++ b/arm/hardfloat/Conventions1.v @@ -15,6 +15,7 @@ Require Import Coqlib. Require Import AST. +Require Import Events. Require Import Locations. (** * Classification of machine registers *) @@ -178,13 +179,13 @@ Proof. Qed. Lemma int_callee_save_type: - forall r, In r int_callee_save_regs -> mreg_type r = Tint. + forall r, In r int_callee_save_regs -> mreg_type r = Tany32. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. Lemma float_callee_save_type: - forall r, In r float_callee_save_regs -> mreg_type r = Tfloat. + forall r, In r float_callee_save_regs -> mreg_type r = Tany64. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. @@ -235,12 +236,21 @@ Qed. Definition loc_result (s: signature) : list mreg := match s.(sig_res) with | None => R0 :: nil - | Some Tint => R0 :: nil - | Some (Tfloat | Tsingle) => F0 :: nil + | Some (Tint | Tany32) => R0 :: nil + | Some (Tfloat | Tsingle | Tany64) => F0 :: nil | Some Tlong => R1 :: R0 :: nil end. -(** The result location is a caller-save register or a temporary *) +(** The result registers have types compatible with that given in the signature. *) + +Lemma loc_result_type: + forall sig, + subtype_list (proj_sig_res' sig) (map mreg_type (loc_result sig)) = true. +Proof. + intros. unfold proj_sig_res', loc_result. destruct (sig_res sig) as [[]|]; auto. +Qed. + +(** The result locations are caller-save registers *) Lemma loc_result_caller_save: forall (s: signature) (r: mreg), @@ -286,15 +296,15 @@ Fixpoint loc_arguments_rec (tyl: list typ) (ir fr ofs: Z) {struct tyl} : list loc := match tyl with | nil => nil - | Tint :: tys => + | (Tint | Tany32) as ty :: tys => if zlt ir 4 then R (ireg_param ir) :: loc_arguments_rec tys (ir + 1) fr ofs - else S Outgoing ofs Tint :: loc_arguments_rec tys ir fr (ofs + 1) - | Tfloat :: tys => + else S Outgoing ofs ty :: loc_arguments_rec tys ir fr (ofs + 1) + | (Tfloat | Tany64) as ty :: tys => if zlt fr 8 then R (freg_param fr) :: loc_arguments_rec tys ir (fr + 1) ofs else let ofs := align ofs 2 in - S Outgoing ofs Tfloat :: loc_arguments_rec tys ir fr (ofs + 2) + S Outgoing ofs ty :: loc_arguments_rec tys ir fr (ofs + 2) | Tsingle :: tys => if zlt fr 8 then R (freg_param fr) :: loc_arguments_rec tys ir (fr + 1) ofs @@ -325,12 +335,12 @@ Fixpoint loc_arguments_vararg (tyl: list typ) (ofs: Z) {struct tyl} : list loc := match tyl with | nil => nil - | Tint :: tys => - (if zlt ofs 0 then R (ireg_param (ofs + 4)) else S Outgoing ofs Tint) + | (Tint|Tany32) as ty :: tys => + (if zlt ofs 0 then R (ireg_param (ofs + 4)) else S Outgoing ofs ty) :: loc_arguments_vararg tys (ofs + 1) - | Tfloat :: tys => + | (Tfloat|Tany64) as ty :: tys => let ofs := align ofs 2 in - (if zlt ofs 0 then R (freg_param (ofs + 4)) else S Outgoing ofs Tfloat) + (if zlt ofs 0 then R (freg_param (ofs + 4)) else S Outgoing ofs ty) :: loc_arguments_vararg tys (ofs + 2) | Tsingle :: tys => (if zlt ofs 0 then R (freg_param (ofs + 4)) else S Outgoing ofs Tsingle) @@ -356,11 +366,11 @@ Definition loc_arguments (s: signature) : list loc := Fixpoint size_arguments_rec (tyl: list typ) (ir fr ofs: Z) {struct tyl} : Z := match tyl with | nil => ofs - | Tint :: tys => + | (Tint|Tany32) :: tys => if zlt ir 4 then size_arguments_rec tys (ir + 1) fr ofs else size_arguments_rec tys ir fr (ofs + 1) - | Tfloat :: tys => + | (Tfloat|Tany64) :: tys => if zlt fr 8 then size_arguments_rec tys ir (fr + 1) ofs else size_arguments_rec tys ir fr (align ofs 2 + 2) @@ -378,8 +388,8 @@ Fixpoint size_arguments_rec (tyl: list typ) (ir fr ofs: Z) {struct tyl} : Z := Fixpoint size_arguments_vararg (tyl: list typ) (ofs: Z) {struct tyl} : Z := match tyl with | nil => Zmax 0 ofs - | (Tint | Tsingle) :: tys => size_arguments_vararg tys (ofs + 1) - | (Tfloat | Tlong) :: tys => size_arguments_vararg tys (align ofs 2 + 2) + | (Tint | Tsingle | Tany32) :: tys => size_arguments_vararg tys (ofs + 1) + | (Tfloat | Tlong | Tany64) :: tys => size_arguments_vararg tys (align ofs 2 + 2) end. Definition size_arguments (s: signature) : Z := @@ -469,6 +479,19 @@ Proof. eapply IHtyl; eauto. subst. split; [omega | congruence]. eapply INCR. eapply IHtyl; eauto. omega. +- (* any32 *) + destruct (zlt ir 4); destruct H. + subst. left; apply ireg_param_in_params. + eapply IHtyl; eauto. + subst. split; [omega | congruence]. + eapply INCR. eapply IHtyl; eauto. omega. +- (* any64 *) + destruct (zlt fr 8); destruct H. + subst. right; apply freg_param_in_params. + eapply IHtyl; eauto. + subst. split. apply Zle_ge. apply align_le. omega. congruence. + eapply INCR. eapply IHtyl; eauto. + apply Zle_trans with (align ofs 2). apply align_le; omega. omega. Qed. Remark loc_arguments_vararg_charact: @@ -530,6 +553,20 @@ Proof. right; apply freg_param_in_params. split. xomega. congruence. eapply INCR. eapply IHtyl; eauto. omega. +- (* any32 *) + destruct H. + destruct (zlt ofs 0); subst l. + left; apply ireg_param_in_params. + split. xomega. congruence. + eapply INCR. eapply IHtyl; eauto. omega. +- (* any64 *) + set (ofs' := align ofs 2) in *. + assert (ofs <= ofs') by (apply align_le; omega). + destruct H. + destruct (zlt ofs' 0); subst l. + right; apply freg_param_in_params. + split. xomega. congruence. + eapply INCR. eapply IHtyl; eauto. omega. Qed. Lemma loc_arguments_acceptable: @@ -569,6 +606,10 @@ Proof. apply Zle_trans with (align ofs0 2 + 2); auto; omega. destruct (zlt fr 8); eauto. apply Zle_trans with (ofs0 + 1); eauto. omega. + destruct (zlt ir 4); eauto. apply Zle_trans with (ofs0 + 1); auto; omega. + destruct (zlt fr 8); eauto. + apply Zle_trans with (align ofs0 2). apply align_le; omega. + apply Zle_trans with (align ofs0 2 + 2); auto; omega. Qed. Remark size_arguments_vararg_above: @@ -582,6 +623,8 @@ Proof. assert (ofs0 <= align ofs0 2) by (apply align_le; omega). xomega. assert (ofs0 <= align ofs0 2) by (apply align_le; omega). xomega. xomega. + xomega. + assert (ofs0 <= align ofs0 2) by (apply align_le; omega). xomega. Qed. Lemma size_arguments_above: @@ -626,6 +669,18 @@ Proof. eauto. inv H. apply size_arguments_rec_above. eauto. +- (* any32 *) + destruct (zlt ir 4); destruct H. + discriminate. + eauto. + inv H. apply size_arguments_rec_above. + eauto. +- (* any64 *) + destruct (zlt fr 8); destruct H. + discriminate. + eauto. + inv H. apply size_arguments_rec_above. + eauto. Qed. Lemma loc_arguments_vararg_bounded: @@ -656,6 +711,14 @@ Proof. destruct H. destruct (zlt ofs0 0); inv H. apply size_arguments_vararg_above. eauto. +- (* any32 *) + destruct H. + destruct (zlt ofs0 0); inv H. apply size_arguments_vararg_above. + eauto. +- (* any64 *) + destruct H. + destruct (zlt (align ofs0 2) 0); inv H. apply size_arguments_vararg_above. + eauto. Qed. Lemma loc_arguments_bounded: diff --git a/backend/Allocation.v b/backend/Allocation.v index f4fcd6e4..919843b5 100644 --- a/backend/Allocation.v +++ b/backend/Allocation.v @@ -609,6 +609,20 @@ Definition subst_loc (l1 l2: loc) (e: eqs) : option eqs := (EqSet2.elements_between (select_loc_l l1) (select_loc_h l1) (eqs2 e)) (Some e). +(** [loc_type_compat env l e] checks that for all equations [r = l] in [e], + the type [env r] of [r] is compatible with the type of [l]. *) + +Definition sel_type (k: equation_kind) (ty: typ) : typ := + match k with + | Full => ty + | Low | High => Tint + end. + +Definition loc_type_compat (env: regenv) (l: loc) (e: eqs) : bool := + EqSet2.for_all_between + (fun q => subtype (sel_type (ekind q) (env (ereg q))) (Loc.type l)) + (select_loc_l l) (select_loc_h l) (eqs2 e). + (** [add_equations [r1...rN] [m1...mN] e] adds to [e] the [N] equations [ri = R mi [Full]]. Return [None] if the two lists have different lengths. *) @@ -759,18 +773,25 @@ Definition destroyed_by_move (src dst: loc) := | _, _ => destroyed_by_op Omove end. +Definition well_typed_move (env: regenv) (dst: loc) (e: eqs) : bool := + match dst with + | R r => true + | S sl ofs ty => loc_type_compat env dst e + end. + (** Simulate the effect of a sequence of moves [mv] on a set of equations [e]. The set [e] is the equations that must hold after the sequence of moves. Return the set of equations that must hold before the sequence of moves. Return [None] if the set of equations [e] cannot hold after the sequence of moves. *) -Fixpoint track_moves (mv: moves) (e: eqs) : option eqs := +Fixpoint track_moves (env: regenv) (mv: moves) (e: eqs) : option eqs := match mv with | nil => Some e | (src, dst) :: mv => - do e1 <- track_moves mv e; + do e1 <- track_moves env mv e; assertion (can_undef_except dst (destroyed_by_move src dst)) e1; + assertion (well_typed_move env dst e1); subst_loc dst src e1 end. @@ -803,89 +824,90 @@ Definition kind_second_word := if Archi.big_endian then Low else High. equations that must hold "before" these instructions, or [None] if impossible. *) -Definition transfer_aux (f: RTL.function) (shape: block_shape) (e: eqs) : option eqs := +Definition transfer_aux (f: RTL.function) (env: regenv) + (shape: block_shape) (e: eqs) : option eqs := match shape with | BSnop mv s => - track_moves mv e + track_moves env mv e | BSmove src dst mv s => - track_moves mv (subst_reg dst src e) + track_moves env mv (subst_reg dst src e) | BSmakelong src1 src2 dst mv s => let e1 := subst_reg_kind dst High src1 Full e in let e2 := subst_reg_kind dst Low src2 Full e1 in assertion (reg_unconstrained dst e2); - track_moves mv e2 + track_moves env mv e2 | BSlowlong src dst mv s => let e1 := subst_reg_kind dst Full src Low e in assertion (reg_unconstrained dst e1); - track_moves mv e1 + track_moves env mv e1 | BShighlong src dst mv s => let e1 := subst_reg_kind dst Full src High e in assertion (reg_unconstrained dst e1); - track_moves mv e1 + track_moves env mv e1 | BSop op args res mv1 args' res' mv2 s => - do e1 <- track_moves mv2 e; + do e1 <- track_moves env mv2 e; do e2 <- transfer_use_def args res args' res' (destroyed_by_op op) e1; - track_moves mv1 e2 + track_moves env mv1 e2 | BSopdead op args res mv s => assertion (reg_unconstrained res e); - track_moves mv e + track_moves env mv e | BSload chunk addr args dst mv1 args' dst' mv2 s => - do e1 <- track_moves mv2 e; + do e1 <- track_moves env mv2 e; do e2 <- transfer_use_def args dst args' dst' (destroyed_by_load chunk addr) e1; - track_moves mv1 e2 + track_moves env mv1 e2 | BSload2 addr addr' args dst mv1 args1' dst1' mv2 args2' dst2' mv3 s => - do e1 <- track_moves mv3 e; + do e1 <- track_moves env mv3 e; let e2 := remove_equation (Eq kind_second_word dst (R dst2')) e1 in assertion (loc_unconstrained (R dst2') e2); assertion (can_undef (destroyed_by_load Mint32 addr') e2); do e3 <- add_equations args args2' e2; - do e4 <- track_moves mv2 e3; + do e4 <- track_moves env mv2 e3; let e5 := remove_equation (Eq kind_first_word dst (R dst1')) e4 in assertion (loc_unconstrained (R dst1') e5); assertion (can_undef (destroyed_by_load Mint32 addr) e5); assertion (reg_unconstrained dst e5); do e6 <- add_equations args args1' e5; - track_moves mv1 e6 + track_moves env mv1 e6 | BSload2_1 addr args dst mv1 args' dst' mv2 s => - do e1 <- track_moves mv2 e; + do e1 <- track_moves env mv2 e; let e2 := remove_equation (Eq kind_first_word dst (R dst')) e1 in assertion (reg_loc_unconstrained dst (R dst') e2); assertion (can_undef (destroyed_by_load Mint32 addr) e2); do e3 <- add_equations args args' e2; - track_moves mv1 e3 + track_moves env mv1 e3 | BSload2_2 addr addr' args dst mv1 args' dst' mv2 s => - do e1 <- track_moves mv2 e; + do e1 <- track_moves env mv2 e; let e2 := remove_equation (Eq kind_second_word dst (R dst')) e1 in assertion (reg_loc_unconstrained dst (R dst') e2); assertion (can_undef (destroyed_by_load Mint32 addr') e2); do e3 <- add_equations args args' e2; - track_moves mv1 e3 + track_moves env mv1 e3 | BSloaddead chunk addr args dst mv s => assertion (reg_unconstrained dst e); - track_moves mv e + track_moves env mv e | BSstore chunk addr args src mv args' src' s => assertion (can_undef (destroyed_by_store chunk addr) e); do e1 <- add_equations (src :: args) (src' :: args') e; - track_moves mv e1 + track_moves env mv e1 | BSstore2 addr addr' args src mv1 args1' src1' mv2 args2' src2' s => assertion (can_undef (destroyed_by_store Mint32 addr') e); do e1 <- add_equations args args2' (add_equation (Eq kind_second_word src (R src2')) e); - do e2 <- track_moves mv2 e1; + do e2 <- track_moves env mv2 e1; assertion (can_undef (destroyed_by_store Mint32 addr) e2); do e3 <- add_equations args args1' (add_equation (Eq kind_first_word src (R src1')) e2); - track_moves mv1 e3 + track_moves env mv1 e3 | BScall sg ros args res mv1 ros' mv2 s => let args' := loc_arguments sg in let res' := map R (loc_result sg) in - do e1 <- track_moves mv2 e; + do e1 <- track_moves env mv2 e; do e2 <- remove_equations_res res (sig_res sg) res' e1; assertion (forallb (fun l => reg_loc_unconstrained res l e2) res'); assertion (no_caller_saves e2); do e3 <- add_equation_ros ros ros' e2; do e4 <- add_equations_args args (sig_args sg) args' e3; - track_moves mv1 e4 + track_moves env mv1 e4 | BStailcall sg ros args mv1 ros' => let args' := loc_arguments sg in assertion (tailcall_is_possible sg); @@ -893,9 +915,9 @@ Definition transfer_aux (f: RTL.function) (shape: block_shape) (e: eqs) : option assertion (ros_compatible_tailcall ros'); do e1 <- add_equation_ros ros ros' empty_eqs; do e2 <- add_equations_args args (sig_args sg) args' e1; - track_moves mv1 e2 + track_moves env mv1 e2 | BSbuiltin ef args res mv1 args' res' mv2 s => - do e1 <- track_moves mv2 e; + do e1 <- track_moves env mv2 e; let args' := map R args' in let res' := map R res' in do e2 <- remove_equations_res res (sig_res (ef_sig ef)) res' e1; @@ -903,23 +925,23 @@ Definition transfer_aux (f: RTL.function) (shape: block_shape) (e: eqs) : option assertion (forallb (fun l => loc_unconstrained l e2) res'); assertion (can_undef (destroyed_by_builtin ef) e2); do e3 <- add_equations_args args (sig_args (ef_sig ef)) args' e2; - track_moves mv1 e3 + track_moves env mv1 e3 | BSannot txt typ args res mv1 args' s => do e1 <- add_equations_args args (annot_args_typ typ) args' e; - track_moves mv1 e1 + track_moves env mv1 e1 | BScond cond args mv args' s1 s2 => assertion (can_undef (destroyed_by_cond cond) e); do e1 <- add_equations args args' e; - track_moves mv e1 + track_moves env mv e1 | BSjumptable arg mv arg' tbl => assertion (can_undef destroyed_by_jumptable e); - track_moves mv (add_equation (Eq Full arg (R arg')) e) + track_moves env mv (add_equation (Eq Full arg (R arg')) e) | BSreturn None mv => - track_moves mv empty_eqs + track_moves env mv empty_eqs | BSreturn (Some arg) mv => let arg' := map R (loc_result (RTL.fn_sig f)) in do e1 <- add_equations_res arg (sig_res (RTL.fn_sig f)) arg' empty_eqs; - track_moves mv e1 + track_moves env mv e1 end. (** The main transfer function for the dataflow analysis. Like [transfer_aux], @@ -927,7 +949,7 @@ Definition transfer_aux (f: RTL.function) (shape: block_shape) (e: eqs) : option equations that must hold "after". It also handles error propagation and reporting. *) -Definition transfer (f: RTL.function) (shapes: PTree.t block_shape) +Definition transfer (f: RTL.function) (env: regenv) (shapes: PTree.t block_shape) (pc: node) (after: res eqs) : res eqs := match after with | Error _ => after @@ -935,7 +957,7 @@ Definition transfer (f: RTL.function) (shapes: PTree.t block_shape) match shapes!pc with | None => Error(MSG "At PC " :: POS pc :: MSG ": unmatched block" :: nil) | Some shape => - match transfer_aux f shape e with + match transfer_aux f env shape e with | None => Error(MSG "At PC " :: POS pc :: MSG ": invalid register allocation" :: nil) | Some e' => OK e' end @@ -1083,8 +1105,8 @@ Definition successors_block_shape (bsh: block_shape) : list node := | BSreturn optarg mv => nil end. -Definition analyze (f: RTL.function) (bsh: PTree.t block_shape) := - DS.fixpoint_allnodes bsh successors_block_shape (transfer f bsh). +Definition analyze (f: RTL.function) (env: regenv) (bsh: PTree.t block_shape) := + DS.fixpoint_allnodes bsh successors_block_shape (transfer f env bsh). (** * Validating and translating functions and programs *) @@ -1118,9 +1140,9 @@ Function compat_entry (rparams: list reg) (tys: list typ) (lparams: list loc) (e and stack size. *) Definition check_entrypoints_aux (rtl: RTL.function) (ltl: LTL.function) - (e1: eqs) : option unit := + (env: regenv) (e1: eqs) : option unit := do mv <- pair_entrypoints rtl ltl; - do e2 <- track_moves mv e1; + do e2 <- track_moves env mv e1; assertion (compat_entry (RTL.fn_params rtl) (sig_args (RTL.fn_sig rtl)) (loc_parameters (RTL.fn_sig rtl)) e2); @@ -1133,10 +1155,10 @@ Local Close Scope option_monad_scope. Local Open Scope error_monad_scope. Definition check_entrypoints (rtl: RTL.function) (ltl: LTL.function) - (bsh: PTree.t block_shape) + (env: regenv) (bsh: PTree.t block_shape) (a: PMap.t LEq.t): res unit := - do e1 <- transfer rtl bsh (RTL.fn_entrypoint rtl) a!!(RTL.fn_entrypoint rtl); - match check_entrypoints_aux rtl ltl e1 with + do e1 <- transfer rtl env bsh (RTL.fn_entrypoint rtl) a!!(RTL.fn_entrypoint rtl); + match check_entrypoints_aux rtl ltl env e1 with | None => Error (msg "invalid register allocation at entry point") | Some _ => OK tt end. @@ -1145,11 +1167,11 @@ Definition check_entrypoints (rtl: RTL.function) (ltl: LTL.function) a source RTL function and an LTL function generated by the external register allocator. *) -Definition check_function (rtl: RTL.function) (ltl: LTL.function) : res unit := +Definition check_function (rtl: RTL.function) (ltl: LTL.function) (env: regenv): res unit := let bsh := pair_codes rtl ltl in - match analyze rtl bsh with + match analyze rtl env bsh with | None => Error (msg "allocation analysis diverges") - | Some a => check_entrypoints rtl ltl bsh a + | Some a => check_entrypoints rtl ltl env bsh a end. (** [regalloc] is the external register allocator. It is written in OCaml @@ -1165,7 +1187,7 @@ Definition transf_function (f: RTL.function) : res LTL.function := | OK env => match regalloc f with | Error m => Error m - | OK tf => do x <- check_function f tf; OK tf + | OK tf => do x <- check_function f tf env; OK tf end end. diff --git a/backend/Allocproof.v b/backend/Allocproof.v index 93038783..588a674e 100644 --- a/backend/Allocproof.v +++ b/backend/Allocproof.v @@ -451,7 +451,7 @@ Lemma add_equations_args_lessdef: forall rs ls rl tyl ll e e', add_equations_args rl tyl ll e = Some e' -> satisf rs ls e' -> - Val.has_type_list (rs##rl) (normalize_list tyl) -> + Val.has_type_list (rs##rl) tyl -> Val.lessdef_list (rs##rl) (decode_longs tyl (map ls ll)). Proof. intros until e'. functional induction (add_equations_args rl tyl ll e); simpl; intros. @@ -969,7 +969,6 @@ Proof. split. apply eqs_same; auto. auto. Qed. -(* Lemma loc_type_compat_charact: forall env l e q, loc_type_compat env l e = true -> @@ -1009,8 +1008,7 @@ Proof. destruct (rs#r); exact I. eelim Loc.diff_not_eq. eexact A. auto. Qed. -*) -(* + Remark val_lessdef_normalize: forall v v' ty, Val.has_type v ty -> Val.lessdef v v' -> @@ -1018,19 +1016,23 @@ Remark val_lessdef_normalize: Proof. intros. inv H0. rewrite Val.load_result_same; auto. auto. Qed. -*) Lemma subst_loc_satisf: forall env src dst rs ls e e', subst_loc dst src e = Some e' -> - (*well_typed_move env dst e = true ->*) + well_typed_move env dst e = true -> wt_regset env rs -> satisf rs ls e' -> satisf rs (Locmap.set dst (ls src) ls) e. Proof. intros; red; intros. exploit in_subst_loc; eauto. intros [[A B] | [A B]]. - subst dst. rewrite Locmap.gss. apply (H1 _ B). + subst dst. rewrite Locmap.gss. + destruct q as [k r l]; simpl in *. + exploit well_typed_move_charact; eauto. + destruct l as [mr | sl ofs ty]; intros. + apply (H2 _ B). + apply val_lessdef_normalize; auto. apply (H2 _ B). rewrite Locmap.gso; auto. Qed. @@ -1077,18 +1079,22 @@ Proof. Qed. Lemma subst_loc_undef_satisf: - forall src dst rs ls ml e e', + forall env src dst rs ls ml e e', subst_loc dst src e = Some e' -> - (*well_typed_move env dst e = true ->*) + well_typed_move env dst e = true -> can_undef_except dst ml e = true -> - (*wt_regset env rs ->*) + wt_regset env rs -> satisf rs ls e' -> satisf rs (Locmap.set dst (ls src) (undef_regs ml ls)) e. Proof. intros; red; intros. exploit in_subst_loc; eauto. intros [[A B] | [A B]]. subst dst. rewrite Locmap.gss. - destruct q as [k r l]; simpl in *. apply (H1 _ B). + destruct q as [k r l]; simpl in *. + exploit well_typed_move_charact; eauto. + destruct l as [mr | sl ofs ty]; intros. + apply (H3 _ B). + apply val_lessdef_normalize; auto. apply (H3 _ B). rewrite Locmap.gso; auto. rewrite undef_regs_outside. eauto. eapply can_undef_except_sound; eauto. apply Loc.diff_sym; auto. Qed. @@ -1327,30 +1333,45 @@ Proof. auto. congruence. Qed. +Lemma loadv_int64_split: + forall m a v, + Mem.loadv Mint64 m a = Some v -> + exists v1 v2, + Mem.loadv Mint32 m a = Some (if Archi.big_endian then v1 else v2) + /\ Mem.loadv Mint32 m (Val.add a (Vint (Int.repr 4))) = Some (if Archi.big_endian then v2 else v1) + /\ Val.lessdef (Val.hiword v) v1 + /\ Val.lessdef (Val.loword v) v2. +Proof. + intros. exploit Mem.loadv_int64_split; eauto. intros (v1 & v2 & A & B & C). + exists v1, v2. split; auto. split; auto. + inv C; auto. destruct v1, v2; simpl; auto. + rewrite Int64.hi_ofwords, Int64.lo_ofwords; auto. +Qed. + (** * Properties of the dataflow analysis *) Lemma analyze_successors: - forall f bsh an pc bs s e, - analyze f bsh = Some an -> + forall f env bsh an pc bs s e, + analyze f env bsh = Some an -> bsh!pc = Some bs -> In s (successors_block_shape bs) -> an!!pc = OK e -> - exists e', transfer f bsh s an!!s = OK e' /\ EqSet.Subset e' e. + exists e', transfer f env bsh s an!!s = OK e' /\ EqSet.Subset e' e. Proof. unfold analyze; intros. exploit DS.fixpoint_allnodes_solution; eauto. - rewrite H2. unfold DS.L.ge. destruct (transfer f bsh s an#s); intros. + rewrite H2. unfold DS.L.ge. destruct (transfer f env bsh s an#s); intros. exists e0; auto. contradiction. Qed. Lemma satisf_successors: - forall f bsh an pc bs s e rs ls, - analyze f bsh = Some an -> + forall f env bsh an pc bs s e rs ls, + analyze f env bsh = Some an -> bsh!pc = Some bs -> In s (successors_block_shape bs) -> an!!pc = OK e -> satisf rs ls e -> - exists e', transfer f bsh s an!!s = OK e' /\ satisf rs ls e'. + exists e', transfer f env bsh s an!!s = OK e' /\ satisf rs ls e'. Proof. intros. exploit analyze_successors; eauto. intros [e' [A B]]. exists e'; split; auto. eapply satisf_incr; eauto. @@ -1360,12 +1381,13 @@ Qed. Inductive transf_function_spec (f: RTL.function) (tf: LTL.function) : Prop := | transf_function_spec_intro: - forall an mv k e1 e2, - analyze f (pair_codes f tf) = Some an -> + forall env an mv k e1 e2, + wt_function f env -> + analyze f env (pair_codes f tf) = Some an -> (LTL.fn_code tf)!(LTL.fn_entrypoint tf) = Some(expand_moves mv (Lbranch (RTL.fn_entrypoint f) :: k)) -> wf_moves mv -> - transfer f (pair_codes f tf) (RTL.fn_entrypoint f) an!!(RTL.fn_entrypoint f) = OK e1 -> - track_moves mv e1 = Some e2 -> + transfer f env (pair_codes f tf) (RTL.fn_entrypoint f) an!!(RTL.fn_entrypoint f) = OK e1 -> + track_moves env mv e1 = Some e2 -> compat_entry (RTL.fn_params f) (sig_args (RTL.fn_sig f)) (loc_parameters (fn_sig tf)) e2 = true -> can_undef destroyed_at_function_entry e2 = true -> RTL.fn_stacksize f = LTL.fn_stacksize tf -> @@ -1380,33 +1402,36 @@ Proof. unfold transf_function; intros. destruct (type_function f) as [env|] eqn:TY; try discriminate. destruct (regalloc f); try discriminate. - destruct (check_function f f0) as [] eqn:?; inv H. + destruct (check_function f f0 env) as [] eqn:?; inv H. unfold check_function in Heqr. - destruct (analyze f (pair_codes f tf)) as [an|] eqn:?; try discriminate. + destruct (analyze f env (pair_codes f tf)) as [an|] eqn:?; try discriminate. monadInv Heqr. - destruct (check_entrypoints_aux f tf x) as [y|] eqn:?; try discriminate. + destruct (check_entrypoints_aux f tf env x) as [y|] eqn:?; try discriminate. unfold check_entrypoints_aux, pair_entrypoints in Heqo0. MonadInv. exploit extract_moves_sound; eauto. intros [A B]. subst b. exploit check_succ_sound; eauto. intros [k EQ1]. subst b0. - econstructor; eauto. congruence. + econstructor; eauto. eapply type_function_correct; eauto. congruence. Qed. Lemma invert_code: - forall f tf pc i opte e, + forall f env tf pc i opte e, + wt_function f env -> (RTL.fn_code f)!pc = Some i -> - transfer f (pair_codes f tf) pc opte = OK e -> + transfer f env (pair_codes f tf) pc opte = OK e -> exists eafter, exists bsh, exists bb, opte = OK eafter /\ (pair_codes f tf)!pc = Some bsh /\ (LTL.fn_code tf)!pc = Some bb /\ expand_block_shape bsh i bb /\ - transfer_aux f bsh eafter = Some e. + transfer_aux f env bsh eafter = Some e /\ + wt_instr f env i. Proof. - intros. destruct opte as [eafter|]; simpl in H0; try discriminate. exists eafter. + intros. destruct opte as [eafter|]; simpl in H1; try discriminate. exists eafter. destruct (pair_codes f tf)!pc as [bsh|] eqn:?; try discriminate. exists bsh. exploit matching_instr_block; eauto. intros [bb [A B]]. - destruct (transfer_aux f bsh eafter) as [e1|] eqn:?; inv H0. - exists bb. tauto. + destruct (transfer_aux f env bsh eafter) as [e1|] eqn:?; inv H1. + exists bb. exploit wt_instr_at; eauto. + tauto. Qed. (** * Semantic preservation *) @@ -1480,10 +1505,11 @@ Proof. Qed. Lemma exec_moves: - forall mv rs s f sp bb m e e' ls, - track_moves mv e = Some e' -> + forall mv env rs s f sp bb m e e' ls, + track_moves env mv e = Some e' -> wf_moves mv -> satisf rs ls e' -> + wt_regset env rs -> exists ls', star step tge (Block s f sp (expand_moves mv bb) ls m) E0 (Block s f sp bb ls' m) @@ -1495,7 +1521,7 @@ Opaque destroyed_by_op. - unfold expand_moves; simpl. inv H. exists ls; split. apply star_refl. auto. (* step *) - destruct a as [src dst]. unfold expand_moves. simpl. - destruct (track_moves mv e) as [e1|] eqn:?; MonadInv. + destruct (track_moves env mv e) as [e1|] eqn:?; MonadInv. assert (wf_moves mv). red; intros. apply H0; auto with coqlib. destruct src as [rsrc | ssrc]; destruct dst as [rdst | sdst]. (* reg-reg *) @@ -1521,13 +1547,17 @@ Inductive match_stackframes: list RTL.stackframe -> list LTL.stackframe -> signa sg.(sig_res) = Some Tint -> match_stackframes nil nil sg | match_stackframes_cons: - forall res f sp pc rs s tf bb ls ts sg an e + forall res f sp pc rs s tf bb ls ts sg an e env (STACKS: match_stackframes s ts (fn_sig tf)) (FUN: transf_function f = OK tf) - (ANL: analyze f (pair_codes f tf) = Some an) - (EQ: transfer f (pair_codes f tf) pc an!!pc = OK e) + (ANL: analyze f env (pair_codes f tf) = Some an) + (EQ: transfer f env (pair_codes f tf) pc an!!pc = OK e) + (WTF: wt_function f env) + (WTRS: wt_regset env rs) + (WTRES: env res = proj_sig_res sg) (STEPS: forall v ls1 m, Val.lessdef_list (encode_long (sig_res sg) v) (map ls1 (map R (loc_result sg))) -> + Val.has_type v (env res) -> agree_callee_save ls ls1 -> exists ls2, star LTL.step tge (Block ts tf sp bb ls1 m) @@ -1540,13 +1570,15 @@ Inductive match_stackframes: list RTL.stackframe -> list LTL.stackframe -> signa Inductive match_states: RTL.state -> LTL.state -> Prop := | match_states_intro: - forall s f sp pc rs m ts tf ls m' an e + forall s f sp pc rs m ts tf ls m' an e env (STACKS: match_stackframes s ts (fn_sig tf)) (FUN: transf_function f = OK tf) - (ANL: analyze f (pair_codes f tf) = Some an) - (EQ: transfer f (pair_codes f tf) pc an!!pc = OK e) + (ANL: analyze f env (pair_codes f tf) = Some an) + (EQ: transfer f env (pair_codes f tf) pc an!!pc = OK e) (SAT: satisf rs ls e) - (MEM: Mem.extends m m'), + (MEM: Mem.extends m m') + (WTF: wt_function f env) + (WTRS: wt_regset env rs), match_states (RTL.State s f sp pc rs m) (LTL.State ts tf sp pc ls m') | match_states_call: @@ -1555,7 +1587,8 @@ Inductive match_states: RTL.state -> LTL.state -> Prop := (FUN: transf_fundef f = OK tf) (ARGS: Val.lessdef_list args (decode_longs (sig_args (funsig tf)) (map ls (loc_arguments (funsig tf))))) (AG: agree_callee_save (parent_locset ts) ls) - (MEM: Mem.extends m m'), + (MEM: Mem.extends m m') + (WTARGS: Val.has_type_list args (sig_args (funsig tf))), match_states (RTL.Callstate s f args m) (LTL.Callstate ts tf ls m') | match_states_return: @@ -1563,7 +1596,8 @@ Inductive match_states: RTL.state -> LTL.state -> Prop := (STACKS: match_stackframes s ts sg) (RES: Val.lessdef_list (encode_long (sig_res sg) res) (map ls (map R (loc_result sg)))) (AG: agree_callee_save (parent_locset ts) ls) - (MEM: Mem.extends m m'), + (MEM: Mem.extends m m') + (WTRES: Val.has_type res (proj_sig_res sg)), match_states (RTL.Returnstate s res m) (LTL.Returnstate ts ls m'). @@ -1576,13 +1610,14 @@ Proof. intros. inv H. constructor. congruence. econstructor; eauto. + unfold proj_sig_res in *. rewrite H0; auto. intros. unfold loc_result in H; rewrite H0 in H; eauto. Qed. Ltac UseShape := match goal with - | [ CODE: (RTL.fn_code _)!_ = Some _, EQ: transfer _ _ _ _ = OK _ |- _ ] => - destruct (invert_code _ _ _ _ _ _ CODE EQ) as (eafter & bsh & bb & AFTER & BSH & TCODE & EBS & TR); + | [ WT: wt_function _ _, CODE: (RTL.fn_code _)!_ = Some _, EQ: transfer _ _ _ _ _ = OK _ |- _ ] => + destruct (invert_code _ _ _ _ _ _ _ WT CODE EQ) as (eafter & bsh & bb & AFTER & BSH & TCODE & EBS & TR & WTI); inv EBS; unfold transfer_aux in TR; MonadInv end. @@ -1626,7 +1661,8 @@ Proof. econstructor; eauto. (* op move *) -- simpl in H0. inv H0. +- generalize (wt_exec_Iop _ _ _ _ _ _ _ _ _ _ _ WTI H0 WTRS). intros WTRS'. + simpl in H0. inv H0. exploit (exec_moves mv); eauto. intros [ls1 [X Y]]. econstructor; split. eapply plus_left. econstructor; eauto. @@ -1637,7 +1673,8 @@ Proof. econstructor; eauto. (* op makelong *) -- simpl in H0. inv H0. +- generalize (wt_exec_Iop _ _ _ _ _ _ _ _ _ _ _ WTI H0 WTRS). intros WTRS'. + simpl in H0. inv H0. exploit (exec_moves mv); eauto. intros [ls1 [X Y]]. econstructor; split. eapply plus_left. econstructor; eauto. @@ -1649,7 +1686,8 @@ Proof. econstructor; eauto. (* op lowlong *) -- simpl in H0. inv H0. +- generalize (wt_exec_Iop _ _ _ _ _ _ _ _ _ _ _ WTI H0 WTRS). intros WTRS'. + simpl in H0. inv H0. exploit (exec_moves mv); eauto. intros [ls1 [X Y]]. econstructor; split. eapply plus_left. econstructor; eauto. @@ -1661,7 +1699,8 @@ Proof. econstructor; eauto. (* op highlong *) -- simpl in H0. inv H0. +- generalize (wt_exec_Iop _ _ _ _ _ _ _ _ _ _ _ WTI H0 WTRS). intros WTRS'. + simpl in H0. inv H0. exploit (exec_moves mv); eauto. intros [ls1 [X Y]]. econstructor; split. eapply plus_left. econstructor; eauto. @@ -1673,7 +1712,8 @@ Proof. econstructor; eauto. (* op regular *) -- exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. +- generalize (wt_exec_Iop _ _ _ _ _ _ _ _ _ _ _ WTI H0 WTRS). intros WTRS'. + exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. exploit transfer_use_def_satisf; eauto. intros [X Y]. exploit eval_operation_lessdef; eauto. intros [v' [F G]]. exploit (exec_moves mv2); eauto. intros [ls2 [A2 B2]]. @@ -1697,9 +1737,11 @@ Proof. eapply reg_unconstrained_satisf; eauto. intros [enext [U V]]. econstructor; eauto. + eapply wt_exec_Iop; eauto. (* load regular *) -- exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. +- generalize (wt_exec_Iload _ _ _ _ _ _ _ _ _ _ _ WTI H1 WTRS). intros WTRS'. + exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. exploit transfer_use_def_satisf; eauto. intros [X Y]. exploit eval_addressing_lessdef; eauto. intros [a' [F G]]. exploit Mem.loadv_extends; eauto. intros [v' [P Q]]. @@ -1715,7 +1757,8 @@ Proof. econstructor; eauto. (* load pair *) -- exploit Mem.loadv_int64_split; eauto. intros (v1 & v2 & LOAD1 & LOAD2 & V12). +- generalize (wt_exec_Iload _ _ _ _ _ _ _ _ _ _ _ WTI H1 WTRS). intros WTRS'. + exploit loadv_int64_split; eauto. intros (v1 & v2 & LOAD1 & LOAD2 & V1 & V2). set (v2' := if Archi.big_endian then v2 else v1) in *. set (v1' := if Archi.big_endian then v1 else v2) in *. exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. @@ -1729,15 +1772,12 @@ Proof. eapply reg_unconstrained_satisf. eauto. eapply add_equations_satisf; eauto. assumption. rewrite Regmap.gss. apply Val.lessdef_trans with v1'; auto. - subst v. unfold v1', kind_first_word. - destruct Archi.big_endian; apply val_hiword_longofwords || apply val_loword_longofwords. } exploit (exec_moves mv2); eauto. intros [ls3 [A3 B3]]. assert (LD3: Val.lessdef_list rs##args (reglist ls3 args2')). { replace (rs##args) with ((rs#dst<-v)##args). eapply add_equations_lessdef; eauto. apply list_map_exten; intros. rewrite Regmap.gso; auto. - exploit wt_instr_inv; eauto. intros (env & WTI & WTRS). eapply addressing_not_long; eauto. } exploit eval_addressing_lessdef. eexact LD3. @@ -1747,9 +1787,7 @@ Proof. assert (SAT4: satisf (rs#dst <- v) ls4 e0). { eapply loc_unconstrained_satisf. eapply can_undef_satisf; eauto. eapply add_equations_satisf; eauto. assumption. - apply Val.lessdef_trans with v2'; auto. - rewrite Regmap.gss. subst v. unfold v2', kind_second_word. - destruct Archi.big_endian; apply val_hiword_longofwords || apply val_loword_longofwords. + rewrite Regmap.gss. apply Val.lessdef_trans with v2'; auto. } exploit (exec_moves mv3); eauto. intros [ls5 [A5 B5]]. econstructor; split. @@ -1769,7 +1807,8 @@ Proof. econstructor; eauto. (* load first word of a pair *) -- exploit Mem.loadv_int64_split; eauto. intros (v1 & v2 & LOAD1 & LOAD2 & V12). +- generalize (wt_exec_Iload _ _ _ _ _ _ _ _ _ _ _ WTI H1 WTRS). intros WTRS'. + exploit loadv_int64_split; eauto. intros (v1 & v2 & LOAD1 & LOAD2 & V1 & V2). set (v2' := if Archi.big_endian then v2 else v1) in *. set (v1' := if Archi.big_endian then v1 else v2) in *. exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. @@ -1781,8 +1820,6 @@ Proof. assert (SAT2: satisf (rs#dst <- v) ls2 e0). { eapply parallel_assignment_satisf; eauto. apply Val.lessdef_trans with v1'; auto. - subst v. unfold v1', kind_first_word. - destruct Archi.big_endian; apply val_hiword_longofwords || apply val_loword_longofwords. eapply can_undef_satisf. eauto. eapply add_equations_satisf; eauto. } exploit (exec_moves mv2); eauto. intros [ls3 [A3 B3]]. @@ -1799,7 +1836,8 @@ Proof. econstructor; eauto. (* load second word of a pair *) -- exploit Mem.loadv_int64_split; eauto. intros (v1 & v2 & LOAD1 & LOAD2 & V12). +- generalize (wt_exec_Iload _ _ _ _ _ _ _ _ _ _ _ WTI H1 WTRS). intros WTRS'. + exploit loadv_int64_split; eauto. intros (v1 & v2 & LOAD1 & LOAD2 & V1 & V2). set (v2' := if Archi.big_endian then v2 else v1) in *. set (v1' := if Archi.big_endian then v1 else v2) in *. exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. @@ -1812,8 +1850,6 @@ Proof. assert (SAT2: satisf (rs#dst <- v) ls2 e0). { eapply parallel_assignment_satisf; eauto. apply Val.lessdef_trans with v2'; auto. - subst v. unfold v2', kind_second_word. - destruct Archi.big_endian; apply val_hiword_longofwords || apply val_loword_longofwords. eapply can_undef_satisf. eauto. eapply add_equations_satisf; eauto. } exploit (exec_moves mv2); eauto. intros [ls3 [A3 B3]]. @@ -1839,6 +1875,7 @@ Proof. eapply reg_unconstrained_satisf; eauto. intros [enext [U V]]. econstructor; eauto. + eapply wt_exec_Iload; eauto. (* store *) - exploit exec_moves; eauto. intros [ls1 [X Y]]. @@ -1918,19 +1955,21 @@ Proof. exploit analyze_successors; eauto. simpl. left; eauto. intros [enext [U V]]. econstructor; eauto. econstructor; eauto. + inv WTI. congruence. intros. exploit (exec_moves mv2). eauto. eauto. eapply function_return_satisf with (v := v) (ls_before := ls1) (ls_after := ls0); eauto. eapply add_equation_ros_satisf; eauto. eapply add_equations_args_satisf; eauto. congruence. + apply wt_regset_assign; auto. intros [ls2 [A2 B2]]. exists ls2; split. eapply star_right. eexact A2. constructor. traceEq. apply satisf_incr with eafter; auto. rewrite SIG. eapply add_equations_args_lessdef; eauto. - exploit wt_instr_inv; eauto. intros (env & WTI & WTRS). inv WTI. rewrite <- H7. apply wt_regset_list; auto. - simpl. red; auto. + simpl. red; auto. + inv WTI. rewrite SIG. rewrite <- H7. apply wt_regset_list; auto. (* tailcall *) - set (sg := RTL.funsig fd) in *. @@ -1951,15 +1990,15 @@ Proof. eapply match_stackframes_change_sig; eauto. rewrite SIG. rewrite e0. decEq. destruct (transf_function_inv _ _ FUN); auto. rewrite SIG. rewrite return_regs_arg_values; auto. eapply add_equations_args_lessdef; eauto. - exploit wt_instr_inv; eauto. intros (env & WTI & WTRS). inv WTI. rewrite <- H6. apply wt_regset_list; auto. apply return_regs_agree_callee_save. + rewrite SIG. inv WTI. rewrite <- H6. apply wt_regset_list; auto. (* builtin *) -- exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. +- assert (WTRS': wt_regset env (rs#res <- v)) by (eapply wt_exec_Ibuiltin; eauto). + exploit (exec_moves mv1); eauto. intros [ls1 [A1 B1]]. exploit external_call_mem_extends; eauto. eapply add_equations_args_lessdef; eauto. - exploit wt_instr_inv; eauto. intros (env & WTI & WTRS). inv WTI. rewrite <- H4. apply wt_regset_list; auto. intros [v' [m'' [F [G [J K]]]]]. assert (E: map ls1 (map R args') = reglist ls1 args'). @@ -1990,7 +2029,6 @@ Proof. (* annot *) - exploit (exec_moves mv); eauto. intros [ls1 [A1 B1]]. exploit external_call_mem_extends; eauto. eapply add_equations_args_lessdef; eauto. - exploit wt_instr_inv; eauto. intros (env & WTI & WTRS). inv WTI. simpl in H4. rewrite <- H4. apply wt_regset_list; auto. intros [v' [m'' [F [G [J K]]]]]. assert (v = Vundef). red in H0; inv H0. auto. @@ -2009,6 +2047,7 @@ Proof. econstructor; eauto. change (destroyed_by_builtin (EF_annot txt typ)) with (@nil mreg). simpl. subst v. assumption. + apply wt_regset_assign; auto. subst v. constructor. (* cond *) - exploit (exec_moves mv); eauto. intros [ls1 [A1 B1]]. @@ -2040,43 +2079,44 @@ Proof. (* return *) - destruct (transf_function_inv _ _ FUN). - exploit Mem.free_parallel_extends; eauto. rewrite H9. intros [m'' [P Q]]. - destruct or as [r|]; MonadInv. - (* with an argument *) -+ exploit (exec_moves mv); eauto. intros [ls1 [A1 B1]]. + exploit Mem.free_parallel_extends; eauto. rewrite H10. intros [m'' [P Q]]. + inv WTI; MonadInv. ++ (* without an argument *) + exploit (exec_moves mv); eauto. intros [ls1 [A1 B1]]. econstructor; split. eapply plus_left. econstructor; eauto. eapply star_right. eexact A1. econstructor. eauto. eauto. traceEq. - simpl. econstructor; eauto. rewrite <- H10. - replace (map (return_regs (parent_locset ts) ls1) (map R (loc_result (RTL.fn_sig f)))) - with (map ls1 (map R (loc_result (RTL.fn_sig f)))). - eapply add_equations_res_lessdef; eauto. - rewrite !list_map_compose. apply list_map_exten; intros. - unfold return_regs. apply pred_dec_true. eapply loc_result_caller_save; eauto. + simpl. econstructor; eauto. + unfold encode_long, loc_result. rewrite <- H11; rewrite H13. simpl; auto. apply return_regs_agree_callee_save. - - (* without an argument *) -+ assert (SG: f.(RTL.fn_sig).(sig_res) = None). - { exploit wt_instr_inv; eauto. intros (env & WTI & WTRS). - inv WTI; auto. } + constructor. ++ (* with an argument *) exploit (exec_moves mv); eauto. intros [ls1 [A1 B1]]. econstructor; split. eapply plus_left. econstructor; eauto. eapply star_right. eexact A1. econstructor. eauto. eauto. traceEq. - simpl. econstructor; eauto. - unfold encode_long, loc_result. rewrite <- H10. rewrite SG. simpl; auto. + simpl. econstructor; eauto. rewrite <- H11. + replace (map (return_regs (parent_locset ts) ls1) (map R (loc_result (RTL.fn_sig f)))) + with (map ls1 (map R (loc_result (RTL.fn_sig f)))). + eapply add_equations_res_lessdef; eauto. + rewrite !list_map_compose. apply list_map_exten; intros. + unfold return_regs. apply pred_dec_true. eapply loc_result_caller_save; eauto. apply return_regs_agree_callee_save. + unfold proj_sig_res. rewrite <- H11; rewrite H13. apply WTRS. (* internal function *) - monadInv FUN. simpl in *. destruct (transf_function_inv _ _ EQ). - exploit Mem.alloc_extends; eauto. apply Zle_refl. rewrite H7; apply Zle_refl. + exploit Mem.alloc_extends; eauto. apply Zle_refl. rewrite H8; apply Zle_refl. intros [m'' [U V]]. + assert (WTRS: wt_regset env (init_regs args (fn_params f))). + { apply wt_init_regs. inv H0. rewrite wt_params. rewrite H9. auto. } exploit (exec_moves mv). eauto. eauto. eapply can_undef_satisf; eauto. eapply compat_entry_satisf; eauto. - rewrite call_regs_param_values. rewrite H8. eexact ARGS. + rewrite call_regs_param_values. rewrite H9. eexact ARGS. + exact WTRS. intros [ls1 [A B]]. econstructor; split. eapply plus_left. econstructor; eauto. @@ -2108,13 +2148,15 @@ Proof. exploit list_in_map_inv; eauto. intros [r [A B]]; subst l'. destruct l; simpl; auto. red; intros; subst r0; elim H0. eapply loc_result_caller_save; eauto. - + simpl. eapply external_call_well_typed; eauto. + (* return *) - inv STACKS. - exploit STEPS; eauto. intros [ls2 [A B]]. + exploit STEPS; eauto. rewrite WTRES0; auto. intros [ls2 [A B]]. econstructor; split. eapply plus_left. constructor. eexact A. traceEq. econstructor; eauto. + apply wt_regset_assign; auto. rewrite WTRES0; auto. Qed. Lemma initial_states_simulation: @@ -2135,6 +2177,7 @@ Proof. rewrite SIG; rewrite H3; simpl; auto. red; auto. apply Mem.extends_refl. + rewrite SIG, H3. constructor. Qed. Lemma final_states_simulation: diff --git a/backend/CMtypecheck.ml b/backend/CMtypecheck.ml index 5e46d769..02c3f210 100644 --- a/backend/CMtypecheck.ml +++ b/backend/CMtypecheck.ml @@ -34,12 +34,16 @@ let tint = Base Tint let tfloat = Base Tfloat let tlong = Base Tlong let tsingle = Base Tsingle +let tany32 = Base Tany32 +let tany64 = Base Tany64 let ty_of_typ = function | Tint -> tint | Tfloat -> tfloat | Tlong -> tlong - | Tsingle -> tfloat (* should be tsingle when supported *) + | Tsingle -> tsingle + | Tany32 -> tany32 + | Tany64 -> tany64 let ty_of_sig_args tyl = List.map ty_of_typ tyl @@ -88,6 +92,7 @@ let name_of_comparison = function let type_constant = function | Ointconst _ -> tint | Ofloatconst _ -> tfloat + | Osingleconst _ -> tsingle | Olongconst _ -> tlong | Oaddrsymbol _ -> tint | Oaddrstack _ -> tint @@ -101,11 +106,18 @@ let type_unary_operation = function | Onotint -> tint, tint | Onegf -> tfloat, tfloat | Oabsf -> tfloat, tfloat - | Osingleoffloat -> tfloat, tfloat + | Onegfs -> tsingle, tsingle + | Oabsfs -> tsingle, tsingle + | Osingleoffloat -> tfloat, tsingle + | Ofloatofsingle -> tsingle, tfloat | Ointoffloat -> tfloat, tint | Ointuoffloat -> tfloat, tint | Ofloatofint -> tint, tfloat | Ofloatofintu -> tint, tfloat + | Ointofsingle -> tsingle, tint + | Ointuofsingle -> tsingle, tint + | Osingleofint -> tint, tsingle + | Osingleofintu -> tint, tsingle | Onegl -> tlong, tlong | Onotl -> tlong, tlong | Ointoflong -> tlong, tint @@ -115,6 +127,8 @@ let type_unary_operation = function | Olonguoffloat -> tfloat, tlong | Ofloatoflong -> tlong, tfloat | Ofloatoflongu -> tlong, tfloat + | Olongofsingle -> tsingle, tlong + | Olonguofsingle -> tsingle, tlong | Osingleoflong -> tlong, tfloat | Osingleoflongu -> tlong, tfloat @@ -136,6 +150,10 @@ let type_binary_operation = function | Osubf -> tfloat, tfloat, tfloat | Omulf -> tfloat, tfloat, tfloat | Odivf -> tfloat, tfloat, tfloat + | Oaddfs -> tsingle, tsingle, tsingle + | Osubfs -> tsingle, tsingle, tsingle + | Omulfs -> tsingle, tsingle, tsingle + | Odivfs -> tsingle, tsingle, tsingle | Oaddl -> tlong, tlong, tlong | Osubl -> tlong, tlong, tlong | Omull -> tlong, tlong, tlong @@ -152,6 +170,7 @@ let type_binary_operation = function | Ocmp _ -> tint, tint, tint | Ocmpu _ -> tint, tint, tint | Ocmpf _ -> tfloat, tfloat, tint + | Ocmpfs _ -> tsingle, tsingle, tint | Ocmpl _ -> tlong, tlong, tint | Ocmplu _ -> tlong, tlong, tint @@ -166,8 +185,10 @@ let type_chunk = function | Mint16unsigned -> tint | Mint32 -> tint | Mint64 -> tlong - | Mfloat32 -> tfloat + | Mfloat32 -> tsingle | Mfloat64 -> tfloat + | Many32 -> tany32 + | Many64 -> tany64 let name_of_chunk = PrintAST.name_of_chunk diff --git a/backend/CSE.v b/backend/CSE.v index 373acce8..fa229871 100644 --- a/backend/CSE.v +++ b/backend/CSE.v @@ -293,7 +293,6 @@ Definition store_normalized_range (chunk: memory_chunk) : aval := | Mint8unsigned => Uns 8 | Mint16signed => Sgn 16 | Mint16unsigned => Uns 16 - | Mfloat32 => Fsingle | _ => Vtop end. @@ -380,7 +379,7 @@ Variable n: numbering. Fixpoint reduce_rec (niter: nat) (op: A) (args: list valnum) : option(A * list reg) := match niter with | O => None - | S niter' => + | Datatypes.S niter' => match f (fun v => find_valnum_num v n.(num_eqs)) op args with | None => None | Some(op', args') => diff --git a/backend/CSEproof.v b/backend/CSEproof.v index 317fb82f..af138f83 100644 --- a/backend/CSEproof.v +++ b/backend/CSEproof.v @@ -497,7 +497,6 @@ Proof. - inv H. rewrite is_uns_zero_ext in H3 by omega. rewrite H3; auto. - inv H. rewrite is_sgn_sign_ext in H3 by omega. rewrite H3; auto. - inv H. rewrite is_uns_zero_ext in H3 by omega. rewrite H3; auto. -- inv H. rewrite Float.singleoffloat_of_single by auto. auto. Qed. Lemma add_store_result_hold: diff --git a/backend/Cminor.v b/backend/Cminor.v index aaf75107..bf20de29 100644 --- a/backend/Cminor.v +++ b/backend/Cminor.v @@ -35,7 +35,8 @@ Require Import Switch. Inductive constant : Type := | Ointconst: int -> constant (**r integer constant *) - | Ofloatconst: float -> constant (**r floating-point constant *) + | Ofloatconst: float -> constant (**r double-precision floating-point constant *) + | Osingleconst: float32 -> constant (**r single-precision floating-point constant *) | Olongconst: int64 -> constant (**r long integer constant *) | Oaddrsymbol: ident -> int -> constant (**r address of the symbol plus the offset *) | Oaddrstack: int -> constant. (**r stack pointer plus the given offset *) @@ -47,22 +48,31 @@ Inductive unary_operation : Type := | Ocast16signed: unary_operation (**r 16-bit sign extension *) | Onegint: unary_operation (**r integer opposite *) | Onotint: unary_operation (**r bitwise complement *) - | Onegf: unary_operation (**r float opposite *) - | Oabsf: unary_operation (**r float absolute value *) + | Onegf: unary_operation (**r float64 opposite *) + | Oabsf: unary_operation (**r float64 absolute value *) + | Onegfs: unary_operation (**r float32 opposite *) + | Oabsfs: unary_operation (**r float32 absolute value *) | Osingleoffloat: unary_operation (**r float truncation to float32 *) - | Ointoffloat: unary_operation (**r signed integer to float *) - | Ointuoffloat: unary_operation (**r unsigned integer to float *) - | Ofloatofint: unary_operation (**r float to signed integer *) - | Ofloatofintu: unary_operation (**r float to unsigned integer *) + | Ofloatofsingle: unary_operation (**r float extension to float64 *) + | Ointoffloat: unary_operation (**r signed integer to float64 *) + | Ointuoffloat: unary_operation (**r unsigned integer to float64 *) + | Ofloatofint: unary_operation (**r float64 to signed integer *) + | Ofloatofintu: unary_operation (**r float64 to unsigned integer *) + | Ointofsingle: unary_operation (**r signed integer to float32 *) + | Ointuofsingle: unary_operation (**r unsigned integer to float32 *) + | Osingleofint: unary_operation (**r float32 to signed integer *) + | Osingleofintu: unary_operation (**r float32 to unsigned integer *) | Onegl: unary_operation (**r long integer opposite *) | Onotl: unary_operation (**r long bitwise complement *) | Ointoflong: unary_operation (**r long to int *) | Olongofint: unary_operation (**r signed int to long *) | Olongofintu: unary_operation (**r unsigned int to long *) - | Olongoffloat: unary_operation (**r float to signed long *) - | Olonguoffloat: unary_operation (**r float to unsigned long *) - | Ofloatoflong: unary_operation (**r signed long to float *) - | Ofloatoflongu: unary_operation (**r unsigned long to float *) + | Olongoffloat: unary_operation (**r float64 to signed long *) + | Olonguoffloat: unary_operation (**r float64 to unsigned long *) + | Ofloatoflong: unary_operation (**r signed long to float64 *) + | Ofloatoflongu: unary_operation (**r unsigned long to float64 *) + | Olongofsingle: unary_operation (**r float32 to signed long *) + | Olonguofsingle: unary_operation (**r float32 to unsigned long *) | Osingleoflong: unary_operation (**r signed long to float32 *) | Osingleoflongu: unary_operation. (**r unsigned long to float32 *) @@ -80,10 +90,14 @@ Inductive binary_operation : Type := | Oshl: binary_operation (**r integer left shift *) | Oshr: binary_operation (**r integer right signed shift *) | Oshru: binary_operation (**r integer right unsigned shift *) - | Oaddf: binary_operation (**r float addition *) - | Osubf: binary_operation (**r float subtraction *) - | Omulf: binary_operation (**r float multiplication *) - | Odivf: binary_operation (**r float division *) + | Oaddf: binary_operation (**r float64 addition *) + | Osubf: binary_operation (**r float64 subtraction *) + | Omulf: binary_operation (**r float64 multiplication *) + | Odivf: binary_operation (**r float64 division *) + | Oaddfs: binary_operation (**r float32 addition *) + | Osubfs: binary_operation (**r float32 subtraction *) + | Omulfs: binary_operation (**r float32 multiplication *) + | Odivfs: binary_operation (**r float32 division *) | Oaddl: binary_operation (**r long addition *) | Osubl: binary_operation (**r long subtraction *) | Omull: binary_operation (**r long multiplication *) @@ -99,7 +113,8 @@ Inductive binary_operation : Type := | Oshrlu: binary_operation (**r long right unsigned shift *) | Ocmp: comparison -> binary_operation (**r integer signed comparison *) | Ocmpu: comparison -> binary_operation (**r integer unsigned comparison *) - | Ocmpf: comparison -> binary_operation (**r float comparison *) + | Ocmpf: comparison -> binary_operation (**r float64 comparison *) + | Ocmpfs: comparison -> binary_operation (**r float32 comparison *) | Ocmpl: comparison -> binary_operation (**r long signed comparison *) | Ocmplu: comparison -> binary_operation. (**r long unsigned comparison *) @@ -240,6 +255,7 @@ Definition eval_constant (sp: val) (cst: constant) : option val := match cst with | Ointconst n => Some (Vint n) | Ofloatconst n => Some (Vfloat n) + | Osingleconst n => Some (Vsingle n) | Olongconst n => Some (Vlong n) | Oaddrsymbol s ofs => Some(match Genv.find_symbol ge s with @@ -258,11 +274,18 @@ Definition eval_unop (op: unary_operation) (arg: val) : option val := | Onotint => Some (Val.notint arg) | Onegf => Some (Val.negf arg) | Oabsf => Some (Val.absf arg) + | Onegfs => Some (Val.negfs arg) + | Oabsfs => Some (Val.absfs arg) | Osingleoffloat => Some (Val.singleoffloat arg) + | Ofloatofsingle => Some (Val.floatofsingle arg) | Ointoffloat => Val.intoffloat arg | Ointuoffloat => Val.intuoffloat arg | Ofloatofint => Val.floatofint arg | Ofloatofintu => Val.floatofintu arg + | Ointofsingle => Val.intofsingle arg + | Ointuofsingle => Val.intuofsingle arg + | Osingleofint => Val.singleofint arg + | Osingleofintu => Val.singleofintu arg | Onegl => Some (Val.negl arg) | Onotl => Some (Val.notl arg) | Ointoflong => Some (Val.loword arg) @@ -272,6 +295,8 @@ Definition eval_unop (op: unary_operation) (arg: val) : option val := | Olonguoffloat => Val.longuoffloat arg | Ofloatoflong => Val.floatoflong arg | Ofloatoflongu => Val.floatoflongu arg + | Olongofsingle => Val.longofsingle arg + | Olonguofsingle => Val.longuofsingle arg | Osingleoflong => Val.singleoflong arg | Osingleoflongu => Val.singleoflongu arg end. @@ -296,6 +321,10 @@ Definition eval_binop | Osubf => Some (Val.subf arg1 arg2) | Omulf => Some (Val.mulf arg1 arg2) | Odivf => Some (Val.divf arg1 arg2) + | Oaddfs => Some (Val.addfs arg1 arg2) + | Osubfs => Some (Val.subfs arg1 arg2) + | Omulfs => Some (Val.mulfs arg1 arg2) + | Odivfs => Some (Val.divfs arg1 arg2) | Oaddl => Some (Val.addl arg1 arg2) | Osubl => Some (Val.subl arg1 arg2) | Omull => Some (Val.mull arg1 arg2) @@ -312,6 +341,7 @@ Definition eval_binop | Ocmp c => Some (Val.cmp c arg1 arg2) | Ocmpu c => Some (Val.cmpu (Mem.valid_pointer m) c arg1 arg2) | Ocmpf c => Some (Val.cmpf c arg1 arg2) + | Ocmpfs c => Some (Val.cmpfs c arg1 arg2) | Ocmpl c => Val.cmpl c arg1 arg2 | Ocmplu c => Val.cmplu c arg1 arg2 end. diff --git a/backend/Constprop.v b/backend/Constprop.v index d2c4d447..bdfc4b17 100644 --- a/backend/Constprop.v +++ b/backend/Constprop.v @@ -75,6 +75,7 @@ Definition const_for_result (a: aval) : option operation := match a with | I n => Some(Ointconst n) | F n => if Compopts.generate_float_constants tt then Some(Ofloatconst n) else None + | FS n => if Compopts.generate_float_constants tt then Some(Osingleconst n) else None | Ptr(Gl symb ofs) => Some(Oaddrsymbol symb ofs) | Ptr(Stk ofs) => Some(Oaddrstack ofs) | _ => None diff --git a/backend/Constpropproof.v b/backend/Constpropproof.v index ecae5dc6..b79c721e 100644 --- a/backend/Constpropproof.v +++ b/backend/Constpropproof.v @@ -159,6 +159,8 @@ Proof. inv H. inv H0. exists (Vint n); auto. - (* float *) destruct (Compopts.generate_float_constants tt); inv H. inv H0. exists (Vfloat f); auto. +- (* single *) + destruct (Compopts.generate_float_constants tt); inv H. inv H0. exists (Vsingle f); auto. - (* pointer *) destruct p; try discriminate. + (* global *) diff --git a/backend/IRC.ml b/backend/IRC.ml index 6cb17e39..dcd8624a 100644 --- a/backend/IRC.ml +++ b/backend/IRC.ml @@ -35,7 +35,7 @@ type node = { ident: int; (*r unique identifier *) typ: typ; (*r its type *) var: var; (*r the XTL variable it comes from *) - regclass: int; (*r identifier of register class *) + mutable regclass: int; (*r identifier of register class *) mutable accesses: int; (*r number of defs and uses *) mutable spillcost: float; (*r estimated cost of spilling *) mutable adjlist: node list; (*r all nodes it interferes with *) @@ -239,15 +239,20 @@ type graph = { (* Register classes and reserved registers *) -let num_register_classes = 2 +(* We have two main register classes: + 0 for integer registers + 1 for floating-point registers + plus a third pseudo-class 2 that has no registers and forces + stack allocation. XTL variables are mapped to classes 0 and 1 + according to their types. A variable can be forced into class 2 + by giving it a negative spill cost. *) let class_of_type = function - | Tint -> 0 - | Tfloat | Tsingle -> 1 + | Tint | Tany32 -> 0 + | Tfloat | Tsingle | Tany64 -> 1 | Tlong -> assert false -let type_of_class c = - if c = 0 then Tint else Tfloat +let no_spill_class = 2 let reserved_registers = ref ([]: mreg list) @@ -267,14 +272,19 @@ let init costs = and float_callee_save = remove_reserved float_callee_save_regs in { caller_save_registers = - [| Array.of_list int_caller_save; Array.of_list float_caller_save |]; + [| Array.of_list int_caller_save; + Array.of_list float_caller_save; + [||] |]; callee_save_registers = - [| Array.of_list int_callee_save; Array.of_list float_callee_save |]; + [| Array.of_list int_callee_save; + Array.of_list float_callee_save; + [||] |]; num_available_registers = [| List.length int_caller_save + List.length int_callee_save; - List.length float_caller_save + List.length float_callee_save |]; + List.length float_caller_save + List.length float_callee_save; + 0 |]; start_points = - [| 0; 0 |]; + [| 0; 0; 0 |]; allocatable_registers = int_caller_save @ int_callee_save @ float_caller_save @ float_callee_save; stats_of_reg = costs; @@ -303,10 +313,13 @@ let newNodeOfReg g r ty = let st = g.stats_of_reg r in g.nextIdent <- g.nextIdent + 1; { ident = g.nextIdent; typ = ty; - var = V(r, ty); regclass = class_of_type ty; + var = V(r, ty); + regclass = if st.cost >= 0 then class_of_type ty else no_spill_class; accesses = st.usedefs; spillcost = weightedSpillCost st; - adjlist = []; degree = 0; movelist = []; extra_adj = []; extra_pref = []; + adjlist = []; + degree = if st.cost >= 0 then 0 else 1; + movelist = []; extra_adj = []; extra_pref = []; alias = None; color = None; nstate = Initial; @@ -382,11 +395,19 @@ let recordExtraPref n1 n2 = mnext = DLinkMove.dummy; mprev = DLinkMove.dummy } in n1.extra_pref <- m :: n1.extra_pref +let recordExtraPref2 n1 n2 = + let m = + { src = n1; dst = n2; mstate = FrozenMoves; + mnext = DLinkMove.dummy; mprev = DLinkMove.dummy } in + n1.extra_pref <- m :: n1.extra_pref; + n2.extra_pref <- m :: n2.extra_pref + let addMovePref g n1 n2 = - assert (n1.regclass = n2.regclass); match n1.color, n2.color with | None, None -> - recordMove g n1 n2 + if n1.regclass = n2.regclass + then recordMove g n1 n2 + else recordExtraPref2 n1 n2 | Some (R mr1), None -> if List.mem mr1 g.allocatable_registers then recordMove g n1 n2 | None, Some (R mr2) -> @@ -866,7 +887,7 @@ let assign_color g n = n.color <- Some loc | None -> (* Last, pick a Local stack slot *) - n.color <- Some (find_slot slot_conflicts (type_of_class n.regclass)) + n.color <- Some (find_slot slot_conflicts n.typ) (* Extract the location of a variable *) @@ -884,7 +905,7 @@ let location_of_var g v = match ty with | Tint -> R dummy_int_reg | Tfloat | Tsingle -> R dummy_float_reg - | Tlong -> assert false + | Tlong | Tany32 | Tany64 -> assert false (* The exported interface *) diff --git a/backend/Inliningproof.v b/backend/Inliningproof.v index 8e20442d..2564a736 100644 --- a/backend/Inliningproof.v +++ b/backend/Inliningproof.v @@ -691,6 +691,7 @@ Proof. apply Zone_divide. apply Zone_divide. apply H2; omega. + apply H2; omega. Qed. (** Preservation by external calls *) diff --git a/backend/Lineartyping.v b/backend/Lineartyping.v index b08fe878..ccf17e1e 100644 --- a/backend/Lineartyping.v +++ b/backend/Lineartyping.v @@ -12,15 +12,12 @@ (** Type-checking Linear code. *) -Require Import FSets. -Require FSetAVL. Require Import Coqlib. -Require Import Ordered. -Require Import Maps. -Require Import Iteration. Require Import AST. Require Import Integers. Require Import Values. +Require Import Globalenvs. +Require Import Memory. Require Import Events. Require Import Op. Require Import Machregs. @@ -29,172 +26,20 @@ Require Import Conventions. Require Import LTL. Require Import Linear. -(** The typechecker for Linear enforce several properties useful for - the proof of the [Stacking] pass: -- for each instruction, the type of the result register or slot - agrees with the type of values produced by the instruction; -- correctness conditions on the offsets of stack slots - accessed through [Lgetstack] and [Lsetstack] Linear instructions. -*) - -(** * Tracking the flow of single-precision floats *) - -Module Locset := FSetAVL.Make(OrderedLoc). - -(** At each program point, we infer a set of locations that are - guaranteed to contain single-precision floats. [None] means - that the program point is not reachable. *) - -Definition singlefloats := option Locset.t. - -Definition FSbot : singlefloats := None. -Definition FStop : singlefloats := Some Locset.empty. - -Definition setreg (fs: singlefloats) (r: mreg) (t: typ) := - match fs with - | None => None - | Some s => - Some(if typ_eq t Tsingle then Locset.add (R r) s else Locset.remove (R r) s) - end. - -Fixpoint setregs (fs: singlefloats) (rl: list mreg) (tl: list typ) := - match rl, tl with - | nil, nil => fs - | r1 :: rs, t1 :: ts => setregs (setreg fs r1 t1) rs ts - | _, _ => fs - end. - -Definition copyloc (fs: singlefloats) (dst src: loc) := - match fs with - | None => None - | Some s => - Some(if Locset.mem src s then Locset.add dst s else Locset.remove dst s) - end. - -Definition destroyed_at_call_regs := - List.fold_right (fun r fs => Locset.add (R r) fs) Locset.empty destroyed_at_call. - -Definition callregs (fs: singlefloats) := - match fs with - | None => None - | Some s => Some (Locset.diff s destroyed_at_call_regs) - end. - -(** The forward dataflow analysis below records [singlefloats] sets - at every label. Sets at other program points are recomputed when - needed. *) - -Definition labelmap := PTree.t Locset.t. - -(** [update_label lbl fs lm] updates the label map [lm] to reflect the - fact that the [singlefloats] set [fs] can flow into label [lbl]. - It returns the set after the label, an updated label map, and a - boolean indicating whether the label map changed. *) - -Definition update_label (lbl: label) (fs: singlefloats) (lm: labelmap) : - singlefloats * labelmap * bool := - match fs, PTree.get lbl lm with - | None, None => (None, lm, false) - | None, Some s => (Some s, lm, false) - | Some s, None => (Some s, PTree.set lbl s lm, true) - | Some s1, Some s2 => - if Locset.subset s2 s1 - then (Some s2, lm, false) - else let s := Locset.inter s1 s2 in (Some s, PTree.set lbl s lm, true) - end. - -(** [update_labels] is similar, for a list of labels (coming from a - [Ljumptable] instruction). *) - -Fixpoint update_labels (lbls: list label) (fs: singlefloats) (lm: labelmap): labelmap * bool := - match lbls with - | nil => (lm, false) - | lbl1 :: lbls => - let '(_, lm1, changed1) := update_label lbl1 fs lm in - let '(lm2, changed2) := update_labels lbls fs lm1 in - (lm2, changed1 || changed2) - end. - -(** One pass of forward analysis over the code [c]. Returns an updated - label map and a boolean indicating whether the label map changed. *) - -Fixpoint ana_code (lm: labelmap) (ch: bool) (fs: singlefloats) (c: code) : labelmap * bool := - match c with - | nil => (lm, ch) - | Lgetstack sl ofs ty rd :: c => - ana_code lm ch (copyloc fs (R rd) (S sl ofs ty)) c - | Lsetstack rs sl ofs ty :: c => - ana_code lm ch (copyloc fs (S sl ofs ty) (R rs)) c - | Lop op args dst :: c => - match is_move_operation op args with - | Some src => ana_code lm ch (copyloc fs (R dst) (R src)) c - | None => ana_code lm ch (setreg fs dst (snd (type_of_operation op))) c - end - | Lload chunk addr args dst :: c => - ana_code lm ch (setreg fs dst (type_of_chunk chunk)) c - | Lstore chunk addr args src :: c => - ana_code lm ch fs c - | Lcall sg ros :: c => - ana_code lm ch (callregs fs) c - | Ltailcall sg ros :: c => - ana_code lm ch None c - | Lbuiltin ef args res :: c => - ana_code lm ch (setregs fs res (proj_sig_res' (ef_sig ef))) c - | Lannot ef args :: c => - ana_code lm ch fs c - | Llabel lbl :: c => - let '(fs1, lm1, ch1) := update_label lbl fs lm in - ana_code lm1 (ch || ch1) fs1 c - | Lgoto lbl :: c => - let '(_, lm1, ch1) := update_label lbl fs lm in - ana_code lm1 (ch || ch1) None c - | Lcond cond args lbl :: c => - let '(_, lm1, ch1) := update_label lbl fs lm in - ana_code lm1 (ch || ch1) fs c - | Ljumptable r lbls :: c => - let '(lm1, ch1) := update_labels lbls fs lm in - ana_code lm1 (ch || ch1) None c - | Lreturn :: c => - ana_code lm ch None c - end. - -(** Iterating [ana_code] until the label map is stable. *) - -Definition ana_iter (c: code) (lm: labelmap) : labelmap + labelmap := - let '(lm1, ch) := ana_code lm false FStop c in - if ch then inr _ lm1 else inl _ lm. - -Definition ana_function (f: function): option labelmap := - PrimIter.iterate _ _ (ana_iter f.(fn_code)) (PTree.empty _). - -(** * The type-checker *) - -(** The typing rules are presented as boolean-valued functions so that we +(** The rules are presented as boolean-valued functions so that we get an executable type-checker for free. *) Section WT_INSTR. Variable funct: function. -Variable lm: labelmap. - -Definition FSmem (l: loc) (fs: singlefloats) : bool := - match fs with None => true | Some s => Locset.mem l s end. - -Definition loc_type (fs: singlefloats) (l: loc) := - let ty := Loc.type l in - if typ_eq ty Tfloat && FSmem l fs then Tsingle else ty. - Definition slot_valid (sl: slot) (ofs: Z) (ty: typ): bool := match sl with | Local => zle 0 ofs | Outgoing => zle 0 ofs | Incoming => In_dec Loc.eq (S Incoming ofs ty) (loc_parameters funct.(fn_sig)) end && - match ty with - | Tint | Tfloat | Tsingle => true - | Tlong => false - end. + match ty with Tlong => false | _ => true end. Definition slot_writable (sl: slot) : bool := match sl with @@ -210,475 +55,169 @@ Definition loc_valid (l: loc) : bool := | S _ _ _ => false end. -Fixpoint wt_code (fs: singlefloats) (c: code) : bool := - match c with - | nil => true - | Lgetstack sl ofs ty rd :: c => - subtype ty (mreg_type rd) && - slot_valid sl ofs ty && - wt_code (copyloc fs (R rd) (S sl ofs ty)) c - | Lsetstack rs sl ofs ty :: c => - subtype (loc_type fs (R rs)) ty && - slot_valid sl ofs ty && slot_writable sl && - wt_code (copyloc fs (S sl ofs ty) (R rs)) c - | Lop op args dst :: c => +Definition wt_instr (i: instruction) : bool := + match i with + | Lgetstack sl ofs ty r => + subtype ty (mreg_type r) && slot_valid sl ofs ty + | Lsetstack r sl ofs ty => + slot_valid sl ofs ty && slot_writable sl + | Lop op args res => match is_move_operation op args with - | Some src => - typ_eq (mreg_type src) (mreg_type dst) && - wt_code (copyloc fs (R dst) (R src)) c - | None => - let (ty_args, ty_res) := type_of_operation op in - subtype ty_res (mreg_type dst) && - wt_code (setreg fs dst ty_res) c + | Some arg => + subtype (mreg_type arg) (mreg_type res) + | None => + let (targs, tres) := type_of_operation op in + subtype tres (mreg_type res) end - | Lload chunk addr args dst :: c => - subtype (type_of_chunk chunk) (mreg_type dst) && - wt_code (setreg fs dst (type_of_chunk chunk)) c - | Lstore chunk addr args src :: c => - wt_code fs c - | Lcall sg ros :: c => - wt_code (callregs fs) c - | Ltailcall sg ros :: c => - zeq (size_arguments sg) 0 && - wt_code None c - | Lbuiltin ef args res :: c => - let ty_res := proj_sig_res' (ef_sig ef) in - subtype_list ty_res (map mreg_type res) && - wt_code (setregs fs res ty_res) c - | Lannot ef args :: c => - forallb loc_valid args && - wt_code fs c - | Llabel lbl :: c => - wt_code lm!lbl c - | Lgoto lbl :: c => - wt_code None c - | Lcond cond args lbl :: c => - wt_code fs c - | Ljumptable r lbls :: c => - wt_code None c - | Lreturn :: c => - wt_code None c + | Lload chunk addr args dst => + subtype (type_of_chunk chunk) (mreg_type dst) + | Ltailcall sg ros => + zeq (size_arguments sg) 0 + | Lbuiltin ef args res => + subtype_list (proj_sig_res' (ef_sig ef)) (map mreg_type res) + | Lannot ef args => + forallb loc_valid args + | _ => + true end. End WT_INSTR. -Definition wt_funcode (f: function) (lm: labelmap) : bool := - wt_code f lm FStop f.(fn_code). +Definition wt_code (f: function) (c: code) : bool := + forallb (wt_instr f) c. Definition wt_function (f: function) : bool := - match ana_function f with - | None => false - | Some lm => wt_funcode f lm - end. - -(** * Properties of the static analysis *) - -Inductive FSincl: singlefloats -> singlefloats -> Prop := - | FSincl_none: forall fs, - FSincl None fs - | FSincl_subset: forall s1 s2, - Locset.Subset s2 s1 -> FSincl (Some s1) (Some s2). - -Lemma update_label_false: - forall lbl fs lm fs' lm', - update_label lbl fs lm = (fs', lm', false) -> - FSincl fs lm!lbl /\ fs' = lm!lbl /\ lm' = lm. -Proof. - unfold update_label; intros. - destruct fs as [s1|]; destruct lm!lbl as [s2|]. -- destruct (Locset.subset s2 s1) eqn:S; inv H. - intuition. constructor. apply Locset.subset_2; auto. -- inv H. -- inv H. intuition. constructor. -- inv H. intuition. constructor. -Qed. - -Lemma update_labels_false: - forall fs lbls lm lm', - update_labels lbls fs lm = (lm', false) -> - (forall lbl, In lbl lbls -> FSincl fs lm!lbl) /\ lm' = lm. -Proof. - induction lbls; simpl; intros. -- inv H. tauto. -- destruct (update_label a fs lm) as [[fs1 lm1] changed1] eqn:UL. - destruct (update_labels lbls fs lm1) as [lm2 changed2] eqn:ULS. - inv H. apply orb_false_iff in H2. destruct H2; subst. - exploit update_label_false; eauto. intros (A & B & C). - exploit IHlbls; eauto. intros (D & E). subst. - split. intros. destruct H. congruence. auto. - auto. -Qed. - -Lemma ana_code_false: - forall lm' c lm ch fs, ana_code lm ch fs c = (lm', false) -> ch = false /\ lm' = lm. -Proof. - induction c; simpl; intros. -- inv H; auto. -- destruct a; try (eapply IHc; eauto; fail). - + destruct (is_move_operation o l); eapply IHc; eauto. - + destruct (update_label l fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit IHc; eauto. intros [A B]. apply orb_false_iff in A; destruct A; subst. - exploit update_label_false; eauto. intros (C & D & E). - auto. - + destruct (update_label l fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit IHc; eauto. intros [A B]. apply orb_false_iff in A; destruct A; subst. - exploit update_label_false; eauto. intros (C & D & E). - auto. - + destruct (update_label l0 fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit IHc; eauto. intros [A B]. apply orb_false_iff in A; destruct A; subst. - exploit update_label_false; eauto. intros (C & D & E). - auto. - + destruct (update_labels l fs lm) as [lm1 ch1] eqn:UL. - exploit IHc; eauto. intros [A B]. apply orb_false_iff in A; destruct A; subst. - exploit update_labels_false; eauto. intros (C & D); subst. - auto. -Qed. - -Lemma ana_function_inv: - forall f lm, - ana_function f = Some lm -> ana_code lm false FStop f.(fn_code) = (lm, false). -Proof. - intros. unfold ana_function in H. - eapply PrimIter.iterate_prop with - (Q := fun lm => ana_code lm false FStop (fn_code f) = (lm, false)) - (P := fun (lm: labelmap) => True); eauto. - intros. unfold ana_iter. - destruct (ana_code a false FStop (fn_code f)) as (lm1, ch1) eqn:ANA. - destruct ch1. auto. exploit ana_code_false; eauto. intros [A B]. congruence. -Qed. - -Remark wt_ana_code_cons: - forall f lm fs i c, - ana_code lm false fs (i :: c) = (lm, false) -> - wt_code f lm fs (i :: c) = true -> - exists fs', ana_code lm false fs' c = (lm, false) /\ wt_code f lm fs' c = true. -Proof. - simpl; intros; destruct i; InvBooleans; try (econstructor; split; eassumption). -- destruct (is_move_operation o l). - InvBooleans; econstructor; eauto. - destruct (type_of_operation o); InvBooleans; econstructor; eauto. -- destruct (update_label l fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - exploit update_label_false; eauto. intros (C & D & E); subst. - econstructor; eauto. -- destruct (update_label l fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - econstructor; eauto. -- destruct (update_label l0 fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - econstructor; eauto. -- destruct (update_labels l fs lm) as [lm1 ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - econstructor; eauto. -Qed. - -Lemma wt_find_label_rec: - forall f lm lbl c' c fs, - find_label lbl c = Some c' -> - ana_code lm false fs c = (lm, false) -> - wt_code f lm fs c = true -> - ana_code lm false (PTree.get lbl lm) c' = (lm, false) /\ wt_code f lm (PTree.get lbl lm) c' = true. -Proof. - induction c; intros. -- discriminate. -- simpl in H. specialize (is_label_correct lbl a). destruct (is_label lbl a); intros IL. - + subst a. inv H. simpl in *. - destruct (update_label lbl fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - exploit update_label_false; eauto. intros (C & D & E). subst. - InvBooleans. auto. - + exploit wt_ana_code_cons; eauto. intros (fs' & A & B). - eapply IHc; eauto. -Qed. + wt_code f f.(fn_code). -Lemma wt_find_label: - forall f lm lbl c, - ana_function f = Some lm -> - wt_funcode f lm = true -> - find_label lbl f.(fn_code) = Some c -> - ana_code lm false (PTree.get lbl lm) c = (lm, false) /\ wt_code f lm (PTree.get lbl lm) c = true. -Proof. - intros. eapply wt_find_label_rec; eauto. apply ana_function_inv; auto. -Qed. +(** Typing the run-time state. *) -(** * Soundness of the type system *) - -Require Import Values. -Require Import Globalenvs. -Require Import Memory. - -Module LSF := FSetFacts.Facts(Locset). - -(** ** Typing the run-time state *) - -Inductive wt_locset: singlefloats -> locset -> Prop := - | wt_locset_intro: forall s ls - (TY: forall l, Val.has_type (ls l) (Loc.type l)) - (SINGLE: forall l, Locset.In l s -> Val.has_type (ls l) Tsingle), - wt_locset (Some s) ls. - -Lemma wt_mreg: - forall fs ls r, wt_locset fs ls -> Val.has_type (ls (R r)) (mreg_type r). -Proof. - intros. inv H. apply (TY (R r)). -Qed. - -Lemma wt_loc_type: - forall fs ls l, wt_locset fs ls -> Val.has_type (ls l) (loc_type fs l). -Proof. - intros. inv H. unfold loc_type, FSmem. - destruct (typ_eq (Loc.type l) Tfloat); simpl; auto. - destruct (Locset.mem l s) eqn:MEM; auto. - apply SINGLE. apply Locset.mem_2; auto. -Qed. - -Lemma loc_type_subtype: - forall fs l, subtype (loc_type fs l) (Loc.type l) = true. -Proof. - unfold loc_type; intros. destruct (typ_eq (Loc.type l) Tfloat); simpl. - rewrite e. destruct (FSmem l fs); auto. - destruct (Loc.type l); auto. -Qed. - -Lemma wt_locset_top: - forall ls, - (forall l, Val.has_type (ls l) (Loc.type l)) -> - wt_locset FStop ls. -Proof. - intros; constructor; intros. - auto. - eelim Locset.empty_1; eauto. -Qed. - -Lemma wt_locset_mon: - forall fs1 fs2 ls, - FSincl fs1 fs2 -> wt_locset fs1 ls -> wt_locset fs2 ls. -Proof. - intros. inv H0; inv H. constructor; intros; auto. -Qed. +Definition wt_locset (ls: locset) : Prop := + forall l, Val.has_type (ls l) (Loc.type l). Lemma wt_setreg: - forall fs ls r v ty, - Val.has_type v ty -> subtype ty (mreg_type r) = true -> wt_locset fs ls -> - wt_locset (setreg fs r ty) (Locmap.set (R r) v ls). -Proof. - intros. inv H1. constructor; intros. -- unfold Locmap.set. destruct (Loc.eq (R r) l). - subst l; simpl. eapply Val.has_subtype; eauto. - destruct (Loc.diff_dec (R r) l); simpl; auto. -- unfold Locmap.set. destruct (Loc.eq (R r) l). - destruct (typ_eq ty Tsingle). congruence. - subst l. rewrite LSF.remove_iff in H1. intuition. - destruct (Loc.diff_dec (R r) l); simpl; auto. - apply SINGLE. - destruct (typ_eq ty Tsingle). - rewrite LSF.add_iff in H1; intuition. - rewrite LSF.remove_iff in H1; intuition. -Qed. - -Lemma wt_setregs: - forall vl tyl rl fs rs, - Val.has_type_list vl tyl -> - subtype_list tyl (map mreg_type rl) = true -> - wt_locset fs rs -> - wt_locset (setregs fs rl tyl) (Locmap.setlist (map R rl) vl rs). -Proof. - induction vl; simpl; intros. -- destruct tyl; try contradiction. destruct rl; try discriminate. - simpl. auto. -- destruct tyl as [ | ty tyl]; try contradiction. destruct H. - destruct rl as [ | r rl]; simpl in H0; try discriminate. InvBooleans. - simpl. eapply IHvl; eauto. eapply wt_setreg; eauto. -Qed. - -Lemma undef_regs_type: - forall ty l rl ls, - Val.has_type (ls l) ty -> Val.has_type (undef_regs rl ls l) ty. + forall ls r v, + Val.has_type v (mreg_type r) -> wt_locset ls -> wt_locset (Locmap.set (R r) v ls). Proof. - induction rl; simpl; intros. -- auto. -- unfold Locmap.set. destruct (Loc.eq (R a) l). red; auto. - destruct (Loc.diff_dec (R a) l); auto. red; auto. + intros; red; intros. + unfold Locmap.set. + destruct (Loc.eq (R r) l). + subst l; auto. + destruct (Loc.diff_dec (R r) l). auto. red. auto. Qed. -Lemma wt_copyloc_gen: - forall fs ls src dst temps, - Val.has_type (ls src) (Loc.type dst) -> - wt_locset fs ls -> - wt_locset (copyloc fs dst src) (Locmap.set dst (ls src) (undef_regs temps ls)). +Lemma wt_setstack: + forall ls sl ofs ty v, + wt_locset ls -> wt_locset (Locmap.set (S sl ofs ty) v ls). Proof. - intros. inversion H0; subst. constructor; intros. -- unfold Locmap.set. destruct (Loc.eq dst l). - subst l. auto. - destruct (Loc.diff_dec dst l); simpl; auto. - apply undef_regs_type; auto. -- unfold Locmap.set. destruct (Loc.eq dst l). - subst l. destruct (Locset.mem src s) eqn:E. - apply SINGLE. apply Locset.mem_2; auto. - rewrite LSF.remove_iff in H1. intuition. - destruct (Loc.diff_dec dst l); simpl; auto. - apply undef_regs_type; auto. - apply SINGLE. - destruct (Locset.mem src s). - rewrite LSF.add_iff in H1. intuition. - rewrite LSF.remove_iff in H1. intuition. -Qed. - -Lemma wt_copyloc: - forall fs ls src dst temps, - subtype (Loc.type src) (Loc.type dst) = true -> - wt_locset fs ls -> - wt_locset (copyloc fs dst src) (Locmap.set dst (ls src) (undef_regs temps ls)). -Proof. - intros. eapply wt_copyloc_gen; eauto. - eapply Val.has_subtype; eauto. inv H0; auto. + intros; red; intros. + unfold Locmap.set. + destruct (Loc.eq (S sl ofs ty) l). + subst l. simpl. + generalize (Val.load_result_type (chunk_of_type ty) v). + replace (type_of_chunk (chunk_of_type ty)) with ty. auto. + destruct ty; reflexivity. + destruct (Loc.diff_dec (S sl ofs ty) l). auto. red. auto. Qed. Lemma wt_undef_regs: - forall fs ls temps, wt_locset fs ls -> wt_locset fs (undef_regs temps ls). + forall rs ls, wt_locset ls -> wt_locset (undef_regs rs ls). Proof. - intros. inv H; constructor; intros. -- apply undef_regs_type; auto. -- apply undef_regs_type; auto. + induction rs; simpl; intros. auto. apply wt_setreg; auto. red; auto. Qed. Lemma wt_call_regs: - forall fs ls, wt_locset fs ls -> wt_locset FStop (call_regs ls). -Proof. - intros. inv H. apply wt_locset_top; intros. - unfold call_regs. - destruct l; auto. - destruct sl; try exact I. - change (Loc.type (S Incoming pos ty)) with (Loc.type (S Outgoing pos ty)); auto. -Qed. - -Remark destroyed_at_call_regs_charact: - forall l, - Locset.In l destroyed_at_call_regs <-> - match l with R r => In r destroyed_at_call | S _ _ _ => False end. + forall ls, wt_locset ls -> wt_locset (call_regs ls). Proof. - intros. unfold destroyed_at_call_regs. generalize destroyed_at_call. - induction l0; simpl. -- rewrite LSF.empty_iff. destruct l; tauto. -- rewrite LSF.add_iff. rewrite IHl0. destruct l; intuition congruence. + intros; red; intros. unfold call_regs. destruct l. auto. + destruct sl. + red; auto. + change (Loc.type (S Incoming pos ty)) with (Loc.type (S Outgoing pos ty)). auto. + red; auto. Qed. Lemma wt_return_regs: - forall fs caller fs' callee, - wt_locset fs caller -> wt_locset fs' callee -> - wt_locset (callregs fs) (return_regs caller callee). + forall caller callee, + wt_locset caller -> wt_locset callee -> wt_locset (return_regs caller callee). Proof. - intros. inv H; inv H0; constructor; intros. -- unfold return_regs. destruct l; auto. + intros; red; intros. + unfold return_regs. destruct l; auto. destruct (in_dec mreg_eq r destroyed_at_call); auto. -- unfold callregs. - rewrite LSF.diff_iff in H. rewrite destroyed_at_call_regs_charact in H. destruct H. - unfold return_regs. destruct l. -+ destruct (in_dec mreg_eq r destroyed_at_call). tauto. auto. -+ auto. Qed. Lemma wt_init: - forall s, wt_locset (Some s) (Locmap.init Vundef). + wt_locset (Locmap.init Vundef). Proof. - intros; constructor; intros; simpl; auto. + red; intros. unfold Locmap.init. red; auto. Qed. -Lemma callregs_setregs_result: - forall sg fs, - FSincl (setregs fs (loc_result sg) (proj_sig_res' sg)) (callregs fs). +Lemma wt_setlist: + forall vl rl rs, + Val.has_type_list vl (map mreg_type rl) -> + wt_locset rs -> + wt_locset (Locmap.setlist (map R rl) vl rs). Proof. - assert (X: forall rl tyl, setregs None rl tyl = None). - { - induction rl; destruct tyl; simpl; auto. - } - assert (Y: forall rl s tyl, - exists s', setregs (Some s) rl tyl = Some s' - /\ forall l, Locset.In l s -> ~In l (map R rl) -> Locset.In l s'). - { - induction rl; simpl; intros. - - exists s; split. destruct tyl; auto. tauto. - - destruct tyl. exists s; tauto. - destruct (IHrl (if typ_eq t Tsingle - then Locset.add (R a) s - else Locset.remove (R a) s) tyl) as (s1 & A & B). - exists s1; split; auto. intros. apply B. - destruct (typ_eq t Tsingle). - rewrite LSF.add_iff. tauto. - rewrite LSF.remove_iff. tauto. - tauto. - } - intros. destruct fs as [s|]; simpl. - - destruct (Y (loc_result sg) s (proj_sig_res' sg)) as (s' & A & B). - rewrite A. constructor. red; intros. - rewrite LSF.diff_iff in H. destruct H. apply B. auto. - red; intros. exploit list_in_map_inv; eauto. intros (r & U & V). - subst a. elim H0. rewrite destroyed_at_call_regs_charact. - eapply loc_result_caller_save; eauto. - - rewrite X. constructor. + induction vl; destruct rl; simpl; intros; try contradiction. + auto. + destruct H. apply IHvl; auto. apply wt_setreg; auto. +Qed. + +Lemma wt_find_label: + forall f lbl c, + wt_function f = true -> + find_label lbl f.(fn_code) = Some c -> + wt_code f c = true. +Proof. + unfold wt_function; intros until c. generalize (fn_code f). induction c0; simpl; intros. + discriminate. + InvBooleans. destruct (is_label lbl a). + congruence. + auto. Qed. +(** Soundness of the type system *) + Definition wt_fundef (fd: fundef) := match fd with | Internal f => wt_function f = true | External ef => True end. -Inductive wt_callstack: list stackframe -> singlefloats -> Prop := - | wt_callstack_nil: forall s, - wt_callstack nil (Some s) - | wt_callstack_cons: forall f sp rs c s fs lm fs0 fs1 - (WTSTK: wt_callstack s fs0) - (ANF: ana_function f = Some lm) - (WTF: wt_funcode f lm = true) - (ANC: ana_code lm false (callregs fs1) c = (lm, false)) - (WTC: wt_code f lm (callregs fs1) c = true) - (WTRS: wt_locset fs rs) - (INCL: FSincl (callregs fs) (callregs fs1)), - wt_callstack (Stackframe f sp rs c :: s) fs. +Inductive wt_callstack: list stackframe -> Prop := + | wt_callstack_nil: + wt_callstack nil + | wt_callstack_cons: forall f sp rs c s + (WTSTK: wt_callstack s) + (WTF: wt_function f = true) + (WTC: wt_code f c = true) + (WTRS: wt_locset rs), + wt_callstack (Stackframe f sp rs c :: s). Lemma wt_parent_locset: - forall s fs, wt_callstack s fs -> wt_locset fs (parent_locset s). + forall s, wt_callstack s -> wt_locset (parent_locset s). Proof. induction 1; simpl. - apply wt_init. - auto. Qed. -Lemma wt_callstack_change_fs: - forall s fs, wt_callstack s fs -> wt_callstack s (callregs fs). -Proof. - induction 1. -- constructor. -- econstructor; eauto. - apply wt_locset_mon with fs; auto. - + destruct fs; simpl; constructor. - red; intros. eapply Locset.diff_1; eauto. - + inv INCL; simpl; constructor. - destruct fs; simpl in H0; inv H0. - red; intros. exploit H2; eauto. rewrite ! LSF.diff_iff. tauto. -Qed. - Inductive wt_state: state -> Prop := - | wt_regular_state: forall s f sp c rs m lm fs fs0 - (WTSTK: wt_callstack s fs0) - (ANF: ana_function f = Some lm) - (WTF: wt_funcode f lm = true) - (ANC: ana_code lm false fs c = (lm, false)) - (WTC: wt_code f lm fs c = true) - (WTRS: wt_locset fs rs), + | wt_regular_state: forall s f sp c rs m + (WTSTK: wt_callstack s ) + (WTF: wt_function f = true) + (WTC: wt_code f c = true) + (WTRS: wt_locset rs), wt_state (State s f sp c rs m) - | wt_call_state: forall s fd rs m fs - (WTSTK: wt_callstack s fs) + | wt_call_state: forall s fd rs m + (WTSTK: wt_callstack s) (WTFD: wt_fundef fd) - (WTRS: wt_locset fs rs), + (WTRS: wt_locset rs), wt_state (Callstate s fd rs m) - | wt_return_state: forall s rs m fs - (WTSTK: wt_callstack s fs) - (WTRS: wt_locset (callregs fs) rs), + | wt_return_state: forall s rs m + (WTSTK: wt_callstack s) + (WTRS: wt_locset rs), wt_state (Returnstate s rs m). -(** ** Preservation of state typing by transitions *) +(** Preservation of state typing by transitions *) Section SOUNDNESS. @@ -709,129 +248,89 @@ Proof. - (* getstack *) simpl in *; InvBooleans. econstructor; eauto. - apply wt_copyloc; auto. + eapply wt_setreg; eauto. eapply Val.has_subtype; eauto. apply WTRS. + apply wt_undef_regs; auto. - (* setstack *) simpl in *; InvBooleans. econstructor; eauto. - apply wt_copyloc_gen; auto. - eapply Val.has_subtype; eauto. eapply wt_loc_type; eauto. + apply wt_setstack. apply wt_undef_regs; auto. - (* op *) simpl in *. destruct (is_move_operation op args) as [src | ] eqn:ISMOVE. + (* move *) InvBooleans. exploit is_move_operation_correct; eauto. intros [EQ1 EQ2]; subst. simpl in H. inv H. - econstructor; eauto. - apply wt_copyloc; auto. simpl. rewrite H0. - destruct (mreg_type res); auto. + econstructor; eauto. apply wt_setreg. eapply Val.has_subtype; eauto. apply WTRS. + apply wt_undef_regs; auto. + (* other ops *) destruct (type_of_operation op) as [ty_args ty_res] eqn:TYOP. InvBooleans. econstructor; eauto. - apply wt_setreg; auto. - change ty_res with (snd (ty_args, ty_res)). rewrite <- TYOP. - eapply type_of_operation_sound; eauto. + apply wt_setreg; auto. eapply Val.has_subtype; eauto. + change ty_res with (snd (ty_args, ty_res)). rewrite <- TYOP. eapply type_of_operation_sound; eauto. red; intros; subst op. simpl in ISMOVE. destruct args; try discriminate. destruct args; discriminate. apply wt_undef_regs; auto. - (* load *) simpl in *; InvBooleans. econstructor; eauto. - apply wt_setreg. + apply wt_setreg. eapply Val.has_subtype; eauto. destruct a; simpl in H0; try discriminate. eapply Mem.load_type; eauto. - auto. apply wt_undef_regs; auto. - (* store *) simpl in *; InvBooleans. - econstructor; eauto. + econstructor. eauto. eauto. eauto. apply wt_undef_regs; auto. - (* call *) simpl in *; InvBooleans. econstructor; eauto. econstructor; eauto. - destruct (callregs fs); constructor. red; auto. eapply wt_find_function; eauto. - (* tailcall *) simpl in *; InvBooleans. - econstructor. apply wt_callstack_change_fs; eauto. + econstructor; eauto. eapply wt_find_function; eauto. - eapply wt_return_regs. apply wt_parent_locset; auto. eauto. + apply wt_return_regs; auto. apply wt_parent_locset; auto. - (* builtin *) simpl in *; InvBooleans. econstructor; eauto. - apply wt_setregs. - eapply external_call_well_typed'; eauto. - auto. + apply wt_setlist. + eapply Val.has_subtype_list; eauto. eapply external_call_well_typed'; eauto. apply wt_undef_regs; auto. - (* annot *) simpl in *; InvBooleans. econstructor; eauto. - (* label *) - simpl in *. - destruct (update_label lbl fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - exploit update_label_false; eauto. intros (A & B & C); subst. - econstructor; eauto. - eapply wt_locset_mon; eauto. + simpl in *. econstructor; eauto. - (* goto *) - simpl in *. - destruct (update_label lbl fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - exploit update_label_false; eauto. intros (A & B & C); subst. - exploit wt_find_label; eauto. intros [P Q]. - econstructor; eauto. - eapply wt_locset_mon; eauto. + simpl in *. econstructor; eauto. eapply wt_find_label; eauto. - (* cond branch, taken *) - simpl in *. - destruct (update_label lbl fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - exploit update_label_false; eauto. intros (A & B & C); subst. - exploit wt_find_label; eauto. intros [P Q]. - econstructor; eauto. - eapply wt_locset_mon. eauto. + simpl in *. econstructor. auto. auto. eapply wt_find_label; eauto. apply wt_undef_regs; auto. - (* cond branch, not taken *) - simpl in *. - destruct (update_label lbl fs lm) as [[fs1 lm1] ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - exploit update_label_false; eauto. intros (A & B & C); subst. - econstructor. eauto. eauto. eauto. eauto. eauto. + simpl in *. econstructor. auto. auto. auto. apply wt_undef_regs; auto. - (* jumptable *) - simpl in *. - destruct (update_labels tbl fs lm) as [lm1 ch1] eqn:UL. - exploit ana_code_false; eauto. intros [A B]; subst. - exploit update_labels_false; eauto. intros (A & B); subst. - exploit wt_find_label; eauto. intros [P Q]. - econstructor; eauto. - apply wt_undef_regs. eapply wt_locset_mon; eauto. apply A. eapply list_nth_z_in; eauto. + simpl in *. econstructor. auto. auto. eapply wt_find_label; eauto. + apply wt_undef_regs; auto. - (* return *) - econstructor. eauto. - eapply wt_return_regs. - apply wt_parent_locset; auto. - eauto. + simpl in *. InvBooleans. + econstructor; eauto. + apply wt_return_regs; auto. apply wt_parent_locset; auto. - (* internal function *) - simpl in WTFD. unfold wt_function in WTFD. - destruct (ana_function f) as [lm|] eqn:ANF; try discriminate. - econstructor. eauto. eauto. eauto. apply ana_function_inv; auto. exact WTFD. - apply wt_undef_regs. eapply wt_call_regs; eauto. + simpl in WTFD. + econstructor. eauto. eauto. eauto. + apply wt_undef_regs. apply wt_call_regs. auto. - (* external function *) - econstructor. eauto. - eapply wt_locset_mon. - eapply callregs_setregs_result. - eapply wt_setregs. - eapply external_call_well_typed'; eauto. - unfold proj_sig_res', loc_result. destruct (sig_res (ef_sig ef) )as [[] | ]; auto. - auto. + econstructor. auto. apply wt_setlist; auto. + eapply Val.has_subtype_list. apply loc_result_type. eapply external_call_well_typed'; eauto. - (* return *) - inv WTSTK. econstructor; eauto. - apply wt_locset_mon with (callregs fs); auto. + inv WTSTK. econstructor; eauto. Qed. Theorem wt_initial_state: forall S, initial_state prog S -> wt_state S. Proof. - induction 1. econstructor. - apply wt_callstack_nil with (s := Locset.empty). + induction 1. econstructor. constructor. unfold ge0 in H1. exploit Genv.find_funct_ptr_inversion; eauto. - intros [id IN]. eapply wt_prog; eauto. + intros [id IN]. eapply wt_prog; eauto. apply wt_init. Qed. @@ -850,10 +349,9 @@ Qed. Lemma wt_state_setstack: forall s f sp sl ofs ty r c rs m, wt_state (State s f sp (Lsetstack r sl ofs ty :: c) rs m) -> - Val.has_type (rs (R r)) ty /\ slot_valid f sl ofs ty = true /\ slot_writable sl = true. + slot_valid f sl ofs ty = true /\ slot_writable sl = true. Proof. - intros. inv H. simpl in WTC; InvBooleans. intuition. - eapply Val.has_subtype; eauto. eapply wt_loc_type; eauto. + intros. inv H. simpl in WTC; InvBooleans. intuition. Qed. Lemma wt_state_tailcall: @@ -877,5 +375,5 @@ Lemma wt_callstate_wt_regs: wt_state (Callstate s f rs m) -> forall r, Val.has_type (rs (R r)) (mreg_type r). Proof. - intros. inv H. eapply wt_mreg; eauto. + intros. inv H. apply WTRS. Qed. diff --git a/backend/Locations.v b/backend/Locations.v index 96f1eba1..5674b93a 100644 --- a/backend/Locations.v +++ b/backend/Locations.v @@ -69,7 +69,14 @@ Defined. Open Scope Z_scope. Definition typesize (ty: typ) : Z := - match ty with Tint => 1 | Tlong => 2 | Tfloat => 2 | Tsingle => 1 end. + match ty with + | Tint => 1 + | Tlong => 2 + | Tfloat => 2 + | Tsingle => 1 + | Tany32 => 1 + | Tany64 => 2 + end. Lemma typesize_pos: forall (ty: typ), typesize ty > 0. @@ -301,16 +308,29 @@ Module Locmap. Definition set (l: loc) (v: val) (m: t) : t := fun (p: loc) => - if Loc.eq l p then v - else if Loc.diff_dec l p then m p + if Loc.eq l p then + match l with R r => v | S sl ofs ty => Val.load_result (chunk_of_type ty) v end + else if Loc.diff_dec l p then + m p else Vundef. Lemma gss: forall l v m, - (set l v m) l = v. + (set l v m) l = + match l with R r => v | S sl ofs ty => Val.load_result (chunk_of_type ty) v end. Proof. intros. unfold set. apply dec_eq_true. Qed. + Lemma gss_reg: forall r v m, (set (R r) v m) (R r) = v. + Proof. + intros. unfold set. rewrite dec_eq_true. auto. + Qed. + + Lemma gss_typed: forall l v m, Val.has_type v (Loc.type l) -> (set l v m) l = v. + Proof. + intros. rewrite gss. destruct l. auto. apply Val.load_result_same; auto. + Qed. + Lemma gso: forall l v m p, Loc.diff l p -> (set l v m) p = m p. Proof. intros. unfold set. destruct (Loc.eq l p). @@ -336,10 +356,11 @@ Module Locmap. Proof. assert (P: forall ll l m, m l = Vundef -> (undef ll m) l = Vundef). induction ll; simpl; intros. auto. apply IHll. - unfold set. destruct (Loc.eq a l). auto. + unfold set. destruct (Loc.eq a l). + destruct a. auto. destruct ty; reflexivity. destruct (Loc.diff_dec a l); auto. induction ll; simpl; intros. contradiction. - destruct H. apply P. subst a. apply gss. + destruct H. apply P. subst a. apply gss_typed. exact I. auto. Qed. @@ -364,10 +385,12 @@ Module IndexedTyp <: INDEXED_TYPE. Definition t := typ. Definition index (x: t) := match x with - | Tint => 1%positive - | Tsingle => 2%positive - | Tfloat => 3%positive - | Tlong => 4%positive + | Tany32 => 1%positive + | Tint => 2%positive + | Tsingle => 3%positive + | Tany64 => 4%positive + | Tfloat => 5%positive + | Tlong => 6%positive end. Lemma index_inj: forall x y, index x = index y -> x = y. Proof. destruct x; destruct y; simpl; congruence. Qed. @@ -456,7 +479,7 @@ Module OrderedLoc <: OrderedType. Definition diff_low_bound (l: loc) : loc := match l with | R mr => l - | S sl ofs ty => S sl (ofs - 1) Tfloat + | S sl ofs ty => S sl (ofs - 1) Tany64 end. Definition diff_high_bound (l: loc) : loc := @@ -480,9 +503,9 @@ Module OrderedLoc <: OrderedType. destruct H. right. destruct H0. right. generalize (RANGE ty'); omega. destruct H0. - assert (ty' = Tint \/ ty' = Tsingle). + assert (ty' = Tint \/ ty' = Tsingle \/ ty' = Tany32). { unfold OrderedTyp.lt in H1. destruct ty'; auto; compute in H1; congruence. } - right. destruct H2; subst ty'; simpl typesize; omega. + right. destruct H2 as [E|[E|E]]; subst ty'; simpl typesize; omega. + destruct H. left. apply OrderedSlot.lt_not_eq; auto. destruct H. right. destruct H0. left; omega. @@ -502,13 +525,13 @@ Module OrderedLoc <: OrderedType. destruct H. contradiction. destruct H. right; right; split; auto. left; omega. - left; right; split; auto. destruct ty'; simpl in *. - destruct (zlt ofs' (ofs - 1)). left; auto. - right; split. omega. compute. auto. + left; right; split; auto. + assert (EITHER: typesize ty' = 1 /\ OrderedTyp.lt ty' Tany64 \/ typesize ty' = 2). + { destruct ty'; compute; auto. } + destruct (zlt ofs' (ofs - 1)). left; auto. + destruct EITHER as [[P Q] | P]. + right; split; auto. omega. left; omega. - left; omega. - destruct (zlt ofs' (ofs - 1)). left; auto. - right; split. omega. compute. auto. Qed. End OrderedLoc. diff --git a/backend/NeedDomain.v b/backend/NeedDomain.v index f050c726..73b6831a 100644 --- a/backend/NeedDomain.v +++ b/backend/NeedDomain.v @@ -33,7 +33,6 @@ Require Import RTL. Inductive nval : Type := | Nothing (**r value is entirely unused *) | I (m: int) (**r only need the bits that are 1 in [m] *) - | Fsingle (**r only need the value after conversion to single float *) | All. (**r every bit of the value is used *) Definition eq_nval (x y: nval) : {x=y} + {x<>y}. @@ -56,12 +55,6 @@ Fixpoint vagree (v w: val) (x: nval) {struct x}: Prop := | Vint p, _ => False | _, _ => True end - | Fsingle => - match v, w with - | Vfloat f, Vfloat g => Float.singleoffloat f = Float.singleoffloat g - | Vfloat _, _ => False - | _, _ => True - end | All => Val.lessdef v w end. @@ -115,9 +108,7 @@ Inductive nge: nval -> nval -> Prop := | nge_all: forall x, nge x Nothing | nge_int: forall m1 m2, (forall i, 0 <= i < Int.zwordsize -> Int.testbit m2 i = true -> Int.testbit m1 i = true) -> - nge (I m1) (I m2) - | nge_single: - nge Fsingle Fsingle. + nge (I m1) (I m2). Lemma nge_refl: forall x, nge x x. Proof. @@ -145,7 +136,6 @@ Definition nlub (x y: nval) : nval := | Nothing, _ => y | _, Nothing => x | I m1, I m2 => I (Int.or m1 m2) - | Fsingle, Fsingle => Fsingle | _, _ => All end. @@ -388,14 +378,14 @@ Ltac InvAgree := auto || exact Logic.I || match goal with | [ H: False |- _ ] => contradiction - | [ H: match ?v with Vundef => _ | Vint _ => _ | Vlong _ => _ | Vfloat _ => _ | Vptr _ _ => _ end |- _ ] => destruct v + | [ H: match ?v with Vundef => _ | Vint _ => _ | Vlong _ => _ | Vfloat _ => _ | Vsingle _ => _ | Vptr _ _ => _ end |- _ ] => destruct v end). (** And immediate, or immediate *) Definition andimm (x: nval) (n: int) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.and m n) | All => I n end. @@ -408,13 +398,12 @@ Proof. unfold andimm; intros; destruct x; simpl in *; unfold Val.and. - auto. - InvAgree. apply iagree_and; auto. -- destruct v; destruct w; tauto. - InvAgree. rewrite iagree_and_eq in H. rewrite H; auto. Qed. Definition orimm (x: nval) (n: int) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.and m (Int.not n)) | _ => I (Int.not n) end. @@ -427,21 +416,16 @@ Proof. unfold orimm; intros; destruct x; simpl in *. - auto. - unfold Val.or; InvAgree. apply iagree_or; auto. -- destruct v; destruct w; tauto. - InvAgree. simpl. apply Val.lessdef_same. f_equal. apply iagree_mone. apply iagree_or. rewrite Int.and_commut. rewrite Int.and_mone. auto. Qed. (** Bitwise operations: and, or, xor, not *) -Definition bitwise (x: nval) := - match x with - | Fsingle => Nothing - | _ => x - end. +Definition bitwise (x: nval) := x. Remark bitwise_idem: forall nv, bitwise (bitwise nv) = bitwise nv. -Proof. destruct nv; auto. Qed. +Proof. auto. Qed. Lemma vagree_bitwise_binop: forall f, @@ -456,7 +440,6 @@ Proof. unfold bitwise; intros. destruct x; simpl in *. - auto. - InvAgree. -- destruct v1; auto. destruct v2; auto. - inv H0; auto. inv H1; auto. destruct w1; auto. Qed. @@ -489,7 +472,7 @@ Qed. Definition shlimm (x: nval) (n: int) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.shru m n) | All => I (Int.shru Int.mone n) end. @@ -504,14 +487,13 @@ Proof. destruct x; simpl in *. - auto. - InvAgree. apply iagree_shl; auto. -- destruct v; destruct w; auto. - InvAgree. apply Val.lessdef_same. f_equal. apply iagree_mone. apply iagree_shl; auto. - destruct v; auto with na. Qed. Definition shruimm (x: nval) (n: int) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.shl m n) | All => I (Int.shl Int.mone n) end. @@ -526,14 +508,13 @@ Proof. destruct x; simpl in *. - auto. - InvAgree. apply iagree_shru; auto. -- destruct v; destruct w; auto. - InvAgree. apply Val.lessdef_same. f_equal. apply iagree_mone. apply iagree_shru; auto. - destruct v; auto with na. Qed. Definition shrimm (x: nval) (n: int) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (let m' := Int.shl m n in if Int.eq_dec (Int.shru m' n) m then m' @@ -554,14 +535,13 @@ Proof. destruct (Int.eq_dec (Int.shru (Int.shl m n) n) m). apply iagree_shr_1; auto. apply iagree_shr; auto. -- destruct v; destruct w; auto. - InvAgree. apply Val.lessdef_same. f_equal. apply iagree_mone. apply iagree_shr. auto. - destruct v; auto with na. Qed. Definition rolm (x: nval) (amount mask: int) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.ror (Int.and m mask) amount) | _ => I (Int.ror mask amount) end. @@ -575,7 +555,6 @@ Proof. - auto. - unfold Val.rolm; InvAgree. unfold Int.rolm. apply iagree_and. apply iagree_rol. auto. -- unfold Val.rolm; destruct v, w; auto. - unfold Val.rolm; InvAgree. apply Val.lessdef_same. f_equal. apply iagree_mone. unfold Int.rolm. apply iagree_and. apply iagree_rol. rewrite Int.and_commut. rewrite Int.and_mone. auto. @@ -583,7 +562,7 @@ Qed. Definition ror (x: nval) (amount: int) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.rol m amount) | All => All end. @@ -598,7 +577,6 @@ Proof. destruct x; simpl in *. - auto. - InvAgree. apply iagree_ror; auto. -- destruct v, w; auto. - inv H; auto. - destruct v; auto with na. Qed. @@ -608,7 +586,7 @@ Qed. Definition modarith (x: nval) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (complete_mask m) | All => All end. @@ -621,7 +599,6 @@ Proof. unfold modarith; intros. destruct x; simpl in *. - auto. - unfold Val.add; InvAgree. apply eqmod_iagree. apply Int.eqmod_add; apply iagree_eqmod; auto. -- unfold Val.add; destruct v1, w1; auto; destruct v2, w2; auto. - inv H; auto. inv H0; auto. destruct w1; auto. Qed. @@ -638,7 +615,6 @@ Proof. unfold mul, add; intros. destruct x; simpl in *. - auto. - unfold Val.mul; InvAgree. apply eqmod_iagree. apply Int.eqmod_mult; apply iagree_eqmod; auto. -- unfold Val.mul; destruct v1, w1; auto; destruct v2, w2; auto. - inv H; auto. inv H0; auto. destruct w1; auto. Qed. @@ -651,7 +627,6 @@ Proof. - auto. - unfold Val.neg; InvAgree. apply eqmod_iagree. apply Int.eqmod_neg. apply iagree_eqmod; auto. -- destruct v, w; simpl; auto. - inv H; simpl; auto. Qed. @@ -659,7 +634,7 @@ Qed. Definition zero_ext (n: Z) (x: nval) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.zero_ext n m) | All => I (Int.zero_ext n Int.mone) end. @@ -677,7 +652,6 @@ Proof. red; intros. autorewrite with ints; try omega. destruct (zlt i1 n); auto. apply H; auto. autorewrite with ints; try omega. rewrite zlt_true; auto. -- unfold Val.zero_ext; destruct v; destruct w; auto. - unfold Val.zero_ext; InvAgree; auto. apply Val.lessdef_same. f_equal. Int.bit_solve; try omega. destruct (zlt i1 n); auto. apply H; auto. autorewrite with ints; try omega. apply zlt_true; auto. @@ -685,7 +659,7 @@ Qed. Definition sign_ext (n: Z) (x: nval) := match x with - | Nothing | Fsingle => Nothing + | Nothing => Nothing | I m => I (Int.or (Int.zero_ext n m) (Int.shl Int.one (Int.repr (n - 1)))) | All => I (Int.zero_ext n Int.mone) end. @@ -710,7 +684,6 @@ Proof. right. rewrite Int.unsigned_repr. rewrite zlt_false by omega. replace (n - 1 - (n - 1)) with 0 by omega. reflexivity. generalize Int.wordsize_max_unsigned; omega. -- unfold Val.sign_ext; destruct v; destruct w; auto. - unfold Val.sign_ext; InvAgree; auto. apply Val.lessdef_same. f_equal. Int.bit_solve; try omega. set (j := if zlt i1 n then i1 else n - 1). @@ -721,31 +694,12 @@ Proof. unfold j. destruct (zlt i1 n); omega. Qed. -Definition singleoffloat (x: nval) := - match x with - | Nothing | I _ => Nothing - | Fsingle | All => Fsingle - end. - -Lemma singleoffloat_sound: - forall v w x, - vagree v w (singleoffloat x) -> - vagree (Val.singleoffloat v) (Val.singleoffloat w) x. -Proof. - unfold singleoffloat; intros. destruct x; simpl in *. -- auto. -- unfold Val.singleoffloat; destruct v, w; auto. -- unfold Val.singleoffloat; InvAgree. congruence. -- unfold Val.singleoffloat; InvAgree; auto. rewrite H; auto. -Qed. - (** The needs of a memory store concerning the value being stored. *) Definition store_argument (chunk: memory_chunk) := match chunk with | Mint8signed | Mint8unsigned => I (Int.repr 255) | Mint16signed | Mint16unsigned => I (Int.repr 65535) - | Mfloat32 => Fsingle | _ => All end. @@ -781,10 +735,9 @@ Proof. change 16 with (Int.size (Int.repr 65535)). apply iagree_eqmod; auto. - apply encode_val_inject. rewrite val_inject_id; auto. - apply encode_val_inject. rewrite val_inject_id; auto. -- InvAgree. apply SAME. simpl. - rewrite <- (Float.bits_of_singleoffloat f). - rewrite <- (Float.bits_of_singleoffloat f0). - congruence. +- apply encode_val_inject. rewrite val_inject_id; auto. +- apply encode_val_inject. rewrite val_inject_id; auto. +- apply encode_val_inject. rewrite val_inject_id; auto. - apply encode_val_inject. rewrite val_inject_id; auto. Qed. @@ -803,8 +756,6 @@ Proof. auto. compute; auto. - apply zero_ext_sound with (v := Vint i) (w := Vint i0) (x := All) (n := 16). auto. omega. -- apply singleoffloat_sound with (v := Vfloat f) (w := Vfloat f0) (x := All). - auto. Qed. (** The needs of a comparison *) @@ -1034,24 +985,6 @@ Proof. rewrite Int.bits_zero_ext in H3 by omega. rewrite zlt_false in H3 by auto. discriminate. Qed. -Definition singleoffloat_redundant (x: nval) := - match x with - | Nothing => true - | Fsingle => true - | _ => false - end. - -Lemma singleoffloat_redundant_sound: - forall v w x, - singleoffloat_redundant x = true -> - vagree v w (singleoffloat x) -> - vagree (Val.singleoffloat v) w x. -Proof. - unfold singleoffloat; intros. destruct x; try discriminate. -- auto. -- simpl in *; InvAgree. simpl. rewrite Float.singleoffloat_idem; auto. -Qed. - (** * Neededness for register environments *) Module NVal <: SEMILATTICE. diff --git a/backend/PrintCminor.ml b/backend/PrintCminor.ml index 4c53c5ea..51a45b27 100644 --- a/backend/PrintCminor.ml +++ b/backend/PrintCminor.ml @@ -31,10 +31,10 @@ let rec precedence = function | Evar _ -> (16, NA) | Econst _ -> (16, NA) | Eunop _ -> (15, RtoL) - | Ebinop((Omul|Odiv|Odivu|Omod|Omodu|Omulf|Odivf|Omull|Odivl|Odivlu|Omodl|Omodlu), _, _) -> (13, LtoR) - | Ebinop((Oadd|Osub|Oaddf|Osubf|Oaddl|Osubl), _, _) -> (12, LtoR) + | Ebinop((Omul|Odiv|Odivu|Omod|Omodu|Omulf|Odivf|Omulfs|Odivfs|Omull|Odivl|Odivlu|Omodl|Omodlu), _, _) -> (13, LtoR) + | Ebinop((Oadd|Osub|Oaddf|Osubf|Oaddfs|Osubfs|Oaddl|Osubl), _, _) -> (12, LtoR) | Ebinop((Oshl|Oshr|Oshru|Oshll|Oshrl|Oshrlu), _, _) -> (11, LtoR) - | Ebinop((Ocmp _|Ocmpu _|Ocmpf _|Ocmpl _|Ocmplu _), _, _) -> (10, LtoR) + | Ebinop((Ocmp _|Ocmpu _|Ocmpf _|Ocmpfs _|Ocmpl _|Ocmplu _), _, _) -> (10, LtoR) | Ebinop((Oand|Oandl), _, _) -> (8, LtoR) | Ebinop((Oxor|Oxorl), _, _) -> (7, LtoR) | Ebinop((Oor|Oorl), _, _) -> (6, LtoR) @@ -55,11 +55,18 @@ let name_of_unop = function | Onotint -> "~" | Onegf -> "-f" | Oabsf -> "absf" + | Onegfs -> "-s" + | Oabsfs -> "abss" | Osingleoffloat -> "float32" + | Ofloatofsingle -> "float64" | Ointoffloat -> "intoffloat" | Ointuoffloat -> "intuoffloat" | Ofloatofint -> "floatofint" | Ofloatofintu -> "floatofintu" + | Ointofsingle -> "intofsingle" + | Ointuofsingle -> "intuofsingle" + | Osingleofint -> "singleofint" + | Osingleofintu -> "singleofintu" | Onegl -> "-l" | Onotl -> "~l" | Ointoflong -> "intoflong" @@ -69,6 +76,8 @@ let name_of_unop = function | Olonguoffloat -> "longuoffloat" | Ofloatoflong -> "floatoflong" | Ofloatoflongu -> "floatoflongu" + | Olongofsingle -> "longofsingle" + | Olonguofsingle -> "longuofsingle" | Osingleoflong -> "singleoflong" | Osingleoflongu -> "singleoflongu" @@ -98,6 +107,10 @@ let name_of_binop = function | Osubf -> "-f" | Omulf -> "*f" | Odivf -> "/f" + | Oaddfs -> "+s" + | Osubfs -> "-s" + | Omulfs -> "*s" + | Odivfs -> "/s" | Oaddl -> "+l" | Osubl -> "-l" | Omull -> "*l" @@ -114,6 +127,7 @@ let name_of_binop = function | Ocmp c -> comparison_name c | Ocmpu c -> comparison_name c ^ "u" | Ocmpf c -> comparison_name c ^ "f" + | Ocmpfs c -> comparison_name c ^ "s" | Ocmpl c -> comparison_name c ^ "l" | Ocmplu c -> comparison_name c ^ "lu" @@ -135,6 +149,8 @@ let rec expr p (prec, e) = fprintf p "%ld" (camlint_of_coqint n) | Econst(Ofloatconst f) -> fprintf p "%F" (camlfloat_of_coqfloat f) + | Econst(Osingleconst f) -> + fprintf p "%Ff" (camlfloat_of_coqfloat32 f) | Econst(Olongconst n) -> fprintf p "%LdLL" (camlint64_of_coqint n) | Econst(Oaddrsymbol(id, ofs)) -> @@ -171,6 +187,8 @@ let name_of_type = function | Tfloat -> "float" | Tlong -> "long" | Tsingle -> "single" + | Tany32 -> "any32" + | Tany64 -> "any64" let print_sig p sg = List.iter diff --git a/backend/PrintXTL.ml b/backend/PrintXTL.ml index ff458091..08ccf129 100644 --- a/backend/PrintXTL.ml +++ b/backend/PrintXTL.ml @@ -34,6 +34,8 @@ let short_name_of_type = function | Tfloat -> 'f' | Tlong -> 'l' | Tsingle -> 's' + | Tany32 -> 'w' + | Tany64 -> 'd' let loc pp = function | R r -> mreg pp r diff --git a/backend/RTLtyping.v b/backend/RTLtyping.v index e8ae7ae0..5042c775 100644 --- a/backend/RTLtyping.v +++ b/backend/RTLtyping.v @@ -36,8 +36,6 @@ Require Import Conventions. is very simple, consisting of the four types [Tint] (for integers and pointers), [Tfloat] (for double-precision floats), [Tlong] (for 64-bit integers) and [Tsingle] (for single-precision floats). - At the RTL level, we simplify things further by equating [Tsingle] - with [Tfloat]. Additionally, we impose that each pseudo-register has the same type throughout the function. This requirement helps with register allocation, @@ -57,11 +55,6 @@ Require Import Conventions. which can work over both integers and floats. *) -Definition normalize (ty: typ) : typ := - match ty with Tsingle => Tfloat | _ => ty end. - -Definition normalize_list (tyl: list typ) : list typ := map normalize tyl. - Definition regenv := reg -> typ. Section WT_INSTR. @@ -86,39 +79,39 @@ Inductive wt_instr : instruction -> Prop := forall op args res s, op <> Omove -> map env args = fst (type_of_operation op) -> - env res = normalize (snd (type_of_operation op)) -> + env res = snd (type_of_operation op) -> valid_successor s -> wt_instr (Iop op args res s) | wt_Iload: forall chunk addr args dst s, map env args = type_of_addressing addr -> - env dst = normalize (type_of_chunk chunk) -> + env dst = type_of_chunk chunk -> valid_successor s -> wt_instr (Iload chunk addr args dst s) | wt_Istore: forall chunk addr args src s, map env args = type_of_addressing addr -> - env src = type_of_chunk_use chunk -> + env src = type_of_chunk chunk -> valid_successor s -> wt_instr (Istore chunk addr args src s) | wt_Icall: forall sig ros args res s, match ros with inl r => env r = Tint | inr s => True end -> - map env args = normalize_list sig.(sig_args) -> - env res = normalize (proj_sig_res sig) -> + map env args = sig.(sig_args) -> + env res = proj_sig_res sig -> valid_successor s -> wt_instr (Icall sig ros args res s) | wt_Itailcall: forall sig ros args, match ros with inl r => env r = Tint | inr s => True end -> - map env args = normalize_list sig.(sig_args) -> + map env args = sig.(sig_args) -> sig.(sig_res) = funct.(fn_sig).(sig_res) -> tailcall_possible sig -> wt_instr (Itailcall sig ros args) | wt_Ibuiltin: forall ef args res s, - map env args = normalize_list (ef_sig ef).(sig_args) -> - env res = normalize (proj_sig_res (ef_sig ef)) -> + map env args = (ef_sig ef).(sig_args) -> + env res = proj_sig_res (ef_sig ef) -> valid_successor s -> wt_instr (Ibuiltin ef args res s) | wt_Icond: @@ -139,7 +132,7 @@ Inductive wt_instr : instruction -> Prop := | wt_Ireturn_some: forall arg ty, funct.(fn_sig).(sig_res) = Some ty -> - env arg = normalize ty -> + env arg = ty -> wt_instr (Ireturn (Some arg)). End WT_INSTR. @@ -152,7 +145,7 @@ End WT_INSTR. Record wt_function (f: function) (env: regenv): Prop := mk_wt_function { wt_params: - map env f.(fn_params) = normalize_list f.(fn_sig).(sig_args); + map env f.(fn_params) = f.(fn_sig).(sig_args); wt_norepet: list_norepet f.(fn_params); wt_instrs: @@ -231,23 +224,23 @@ Definition type_instr (e: S.typenv) (i: instruction) : res S.typenv := end else (let (targs, tres) := type_of_operation op in - do e1 <- S.set_list e args targs; S.set e1 res (normalize tres)) + do e1 <- S.set_list e args targs; S.set e1 res tres) | Iload chunk addr args dst s => do x <- check_successor s; do e1 <- S.set_list e args (type_of_addressing addr); - S.set e1 dst (normalize (type_of_chunk chunk)) + S.set e1 dst (type_of_chunk chunk) | Istore chunk addr args src s => do x <- check_successor s; do e1 <- S.set_list e args (type_of_addressing addr); - S.set e1 src (type_of_chunk_use chunk) + S.set e1 src (type_of_chunk chunk) | Icall sig ros args res s => do x <- check_successor s; do e1 <- type_ros e ros; - do e2 <- S.set_list e1 args (normalize_list sig.(sig_args)); - S.set e2 res (normalize (proj_sig_res sig)) + do e2 <- S.set_list e1 args sig.(sig_args); + S.set e2 res (proj_sig_res sig) | Itailcall sig ros args => do e1 <- type_ros e ros; - do e2 <- S.set_list e1 args (normalize_list sig.(sig_args)); + do e2 <- S.set_list e1 args sig.(sig_args); if opt_typ_eq sig.(sig_res) f.(fn_sig).(sig_res) then if tailcall_is_possible sig then OK e2 @@ -256,8 +249,8 @@ Definition type_instr (e: S.typenv) (i: instruction) : res S.typenv := | Ibuiltin ef args res s => let sig := ef_sig ef in do x <- check_successor s; - do e1 <- S.set_list e args (normalize_list sig.(sig_args)); - S.set e1 res (normalize (proj_sig_res sig)) + do e1 <- S.set_list e args sig.(sig_args); + S.set e1 res (proj_sig_res sig) | Icond cond args s1 s2 => do x1 <- check_successor s1; do x2 <- check_successor s2; @@ -271,7 +264,7 @@ Definition type_instr (e: S.typenv) (i: instruction) : res S.typenv := | Ireturn optres => match optres, f.(fn_sig).(sig_res) with | None, None => OK e - | Some r, Some t => S.set e r (normalize t) + | Some r, Some t => S.set e r t | _, _ => Error(msg "bad return") end end. @@ -297,7 +290,7 @@ Definition check_params_norepet (params: list reg): res unit := Definition type_function : res regenv := do e1 <- type_code S.initial; - do e2 <- S.set_list e1 f.(fn_params) (normalize_list f.(fn_sig).(sig_args)); + do e2 <- S.set_list e1 f.(fn_params) f.(fn_sig).(sig_args); do te <- S.solve e2; do x1 <- check_params_norepet f.(fn_params); do x2 <- check_successor f.(fn_entrypoint); @@ -597,7 +590,7 @@ Proof. intros. destruct H. destruct (type_code_complete te S.initial) as (e1 & A & B). auto. apply S.satisf_initial. - destruct (S.set_list_complete te f.(fn_params) (normalize_list f.(fn_sig).(sig_args)) e1) as (e2 & C & D); auto. + destruct (S.set_list_complete te f.(fn_params) f.(fn_sig).(sig_args) e1) as (e2 & C & D); auto. destruct (S.solve_complete te e2) as (te' & E); auto. exists te'; unfold type_function. rewrite A; simpl. rewrite C; simpl. rewrite E; simpl. @@ -637,23 +630,6 @@ Proof. apply H. Qed. -Lemma normalize_subtype: - forall ty, subtype ty (normalize ty) = true. -Proof. - intros. destruct ty; reflexivity. -Qed. - -Lemma wt_regset_assign2: - forall env rs v r ty, - wt_regset env rs -> - Val.has_type v ty -> - env r = normalize ty -> - wt_regset env (rs#r <- v). -Proof. - intros. eapply wt_regset_assign; eauto. - rewrite H1. eapply Val.has_subtype; eauto. apply normalize_subtype. -Qed. - Lemma wt_regset_list: forall env rs, wt_regset env rs -> @@ -684,9 +660,8 @@ Proof. intros. inv H. simpl in H0. inv H0. apply wt_regset_assign; auto. rewrite H4; auto. - eapply wt_regset_assign2; auto. - eapply type_of_operation_sound; eauto. - auto. + eapply wt_regset_assign; auto. + rewrite H8. eapply type_of_operation_sound; eauto. Qed. Lemma wt_exec_Iload: @@ -697,8 +672,7 @@ Lemma wt_exec_Iload: wt_regset env (rs#dst <- v). Proof. intros. destruct a; simpl in H0; try discriminate. inv H. - eapply wt_regset_assign2; eauto. - eapply Mem.load_type; eauto. + eapply wt_regset_assign; eauto. rewrite H8; eapply Mem.load_type; eauto. Qed. Lemma wt_exec_Ibuiltin: @@ -709,8 +683,8 @@ Lemma wt_exec_Ibuiltin: wt_regset env (rs#res <- vres). Proof. intros. inv H. - eapply wt_regset_assign2; eauto. - eapply external_call_well_typed; eauto. + eapply wt_regset_assign; eauto. + rewrite H7; eapply external_call_well_typed; eauto. Qed. Lemma wt_instr_at: @@ -728,7 +702,7 @@ Inductive wt_stackframes: list stackframe -> signature -> Prop := forall s res f sp pc rs env sg, wt_function f env -> wt_regset env rs -> - env res = normalize (proj_sig_res sg) -> + env res = proj_sig_res sg -> wt_stackframes s (fn_sig f) -> wt_stackframes (Stackframe res f sp pc rs :: s) sg. @@ -743,12 +717,12 @@ Inductive wt_state: state -> Prop := forall s f args m, wt_stackframes s (funsig f) -> wt_fundef f -> - Val.has_type_list args (normalize_list (sig_args (funsig f))) -> + Val.has_type_list args (sig_args (funsig f)) -> wt_state (Callstate s f args m) | wt_state_return: forall s v m sg, wt_stackframes s sg -> - Val.has_type v (normalize (proj_sig_res sg)) -> + Val.has_type v (proj_sig_res sg) -> wt_state (Returnstate s v m). Remark wt_stackframes_change_sig: @@ -814,15 +788,13 @@ Proof. econstructor; eauto. (* Ireturn *) econstructor; eauto. - inv WTI; simpl. auto. unfold proj_sig_res; rewrite H2. rewrite <- H3. auto. + inv WTI; simpl. auto. unfold proj_sig_res; rewrite H2. auto. (* internal function *) simpl in *. inv H5. econstructor; eauto. inv H1. apply wt_init_regs; auto. rewrite wt_params0. auto. (* external function *) econstructor; eauto. simpl. - change (Val.has_type res (normalize (proj_sig_res (ef_sig ef)))). - eapply Val.has_subtype. apply normalize_subtype. eapply external_call_well_typed; eauto. (* return *) inv H1. econstructor; eauto. diff --git a/backend/Regalloc.ml b/backend/Regalloc.ml index b736f291..687cbbd3 100644 --- a/backend/Regalloc.ml +++ b/backend/Regalloc.ml @@ -92,7 +92,14 @@ let rec constrain_regs vl cl = | v1 :: vl', None :: cl' -> v1 :: constrain_regs vl' cl' let move v1 v2 k = - if v1 = v2 then k else Xmove(v1, v2) :: k + if v1 = v2 then + k + else if is_stack_reg v1 then begin + let t = new_temp (typeof v2) in Xmove(v1, t) :: Xmove(t, v2) :: k + end else if is_stack_reg v2 then begin + let t = new_temp (typeof v1) in Xmove(v1, t) :: Xmove(t, v2) :: k + end else + Xmove(v1, v2) :: k let rec movelist vl1 vl2 k = match vl1, vl2 with @@ -104,7 +111,7 @@ let xparmove srcs dsts k = assert (List.length srcs = List.length dsts); match srcs, dsts with | [], [] -> k - | [src], [dst] -> Xmove(src, dst) :: k + | [src], [dst] -> move src dst k | _, _ -> Xparmove(srcs, dsts, new_temp Tint, new_temp Tfloat) :: k (* Return the XTL basic block corresponding to the given RTL instruction. @@ -365,9 +372,15 @@ let spill_costs f = | L l -> () | V(r, ty) -> let st = get_stats r in - let c1 = st.cost + amount in - let c2 = if c1 >= 0 then c1 else max_int (* overflow *) in - st.cost <- c2; + if st.cost < 0 then + (* the variable must be spilled, don't change its cost *) + assert (amount < max_int) + else begin + (* saturating addition *) + let c1 = st.cost + amount in + let c2 = if c1 >= 0 then c1 else max_int (* overflow *) in + st.cost <- c2 + end; st.usedefs <- st.usedefs + uses in let charge_list amount uses vl = @@ -376,9 +389,23 @@ let spill_costs f = let charge_ros amount ros = match ros with Coq_inl v -> charge amount 1 v | Coq_inr id -> () in + let force_stack_allocation v = + match v with + | L l -> () + | V(r, ty) -> + let st = get_stats r in + assert (st.cost < max_int); + st.cost <- (-1) in + let charge_instr = function | Xmove(src, dst) -> - charge 1 1 src; charge 1 1 dst + if is_stack_reg src then + force_stack_allocation dst + else if is_stack_reg dst then + force_stack_allocation src + else begin + charge 1 1 src; charge 1 1 dst + end | Xreload(src, dst) -> charge 1 1 src; charge max_int 1 dst (* dest must not be spilled! *) @@ -491,10 +518,13 @@ let add_interfs_instr g instr live = add_interfs_list g (vmreg mr) srcs; IRC.add_interf g (vmreg mr) ftmp) (destroyed_by_setstack Tsingle) + | Xop(Ofloatofsingle, arg1::_, res) when Configuration.arch = "powerpc" -> + add_interfs_def g res live; + IRC.add_pref g arg1 res | Xop(op, args, res) -> begin match is_two_address op args with | None -> - add_interfs_def g res live; + add_interfs_def g res live | Some(arg1, argl) -> (* Treat as "res := arg1; res := op(res, argl)" *) add_interfs_def g res live; @@ -502,7 +532,7 @@ let add_interfs_instr g instr live = add_interfs_move g arg1 res (vset_addlist (res :: argl) (VSet.remove res live)) end; - add_interfs_destroyed g (VSet.remove res live) (destroyed_by_op op); + add_interfs_destroyed g (VSet.remove res live) (destroyed_by_op op) | Xload(chunk, addr, args, dst) -> add_interfs_def g dst live; add_interfs_destroyed g (VSet.remove dst live) @@ -576,10 +606,16 @@ let tospill_instr alloc instr ts = then ts else VSet.add src (VSet.add dst ts) | Xreload(src, dst) -> - assert (is_reg alloc dst); + if not (is_reg alloc dst) then begin + printf "Error: %a was spilled\n" PrintXTL.var dst; + assert false + end; ts | Xspill(src, dst) -> - assert (is_reg alloc src); + if not (is_reg alloc src) then begin + printf "Error: %a was spilled\n" PrintXTL.var src; + assert false + end; ts | Xparmove(srcs, dsts, itmp, ftmp) -> assert (is_reg alloc itmp && is_reg alloc ftmp); @@ -834,7 +870,9 @@ let make_parmove srcs dsts itmp ftmp k = assert (Array.length dst = n); let status = Array.make n To_move in let temp_for = - function Tint -> itmp | (Tfloat|Tsingle) -> ftmp | Tlong -> assert false in + function (Tint|Tany32) -> itmp + | (Tfloat|Tsingle|Tany64) -> ftmp + | Tlong -> assert false in let code = ref [] in let add_move s d = match s, d with diff --git a/backend/SelectDiv.vp b/backend/SelectDiv.vp index 938ce5d8..a275a850 100644 --- a/backend/SelectDiv.vp +++ b/backend/SelectDiv.vp @@ -182,4 +182,16 @@ Nondetfunction divf (e1: expr) (e2: expr) := | _ => Eop Odivf (e1 ::: e2 ::: Enil) end. +Definition divfsimm (e: expr) (n: float32) := + match Float32.exact_inverse n with + | Some n' => Eop Omulfs (e ::: Eop (Osingleconst n') Enil ::: Enil) + | None => Eop Odivfs (e ::: Eop (Osingleconst n) Enil ::: Enil) + end. + +Nondetfunction divfs (e1: expr) (e2: expr) := + match e2 with + | Eop (Osingleconst n2) Enil => divfsimm e1 n2 + | _ => Eop Odivfs (e1 ::: e2 ::: Enil) + end. + diff --git a/backend/SelectDivproof.v b/backend/SelectDivproof.v index 9228cde4..d4bd4f5c 100644 --- a/backend/SelectDivproof.v +++ b/backend/SelectDivproof.v @@ -570,4 +570,20 @@ Proof. - TrivialExists. Qed. +Theorem eval_divfs: + forall le a b x y, + eval_expr ge sp e m le a x -> + eval_expr ge sp e m le b y -> + exists v, eval_expr ge sp e m le (divfs a b) v /\ Val.lessdef (Val.divfs x y) v. +Proof. + intros until y. unfold divfs. destruct (divfs_match b); intros. +- unfold divfsimm. destruct (Float32.exact_inverse n2) as [n2' | ] eqn:EINV. + + inv H0. inv H4. simpl in H6. inv H6. econstructor; split. + EvalOp. constructor. eauto. constructor. EvalOp. simpl; eauto. constructor. + simpl; eauto. + destruct x; simpl; auto. erewrite Float32.div_mul_inverse; eauto. + + TrivialExists. +- TrivialExists. +Qed. + End CMCONSTRS. diff --git a/backend/SelectLong.vp b/backend/SelectLong.vp index 0c1cbb3a..ab4ab8c4 100644 --- a/backend/SelectLong.vp +++ b/backend/SelectLong.vp @@ -138,6 +138,10 @@ Definition floatoflong (arg: expr) := Eexternal hf.(i64_stod) sig_l_f (arg ::: Enil). Definition floatoflongu (arg: expr) := Eexternal hf.(i64_utod) sig_l_f (arg ::: Enil). +Definition longofsingle (arg: expr) := + longoffloat (floatofsingle arg). +Definition longuofsingle (arg: expr) := + longuoffloat (floatofsingle arg). Definition singleoflong (arg: expr) := Eexternal hf.(i64_stof) sig_l_s (arg ::: Enil). Definition singleoflongu (arg: expr) := diff --git a/backend/SelectLongproof.v b/backend/SelectLongproof.v index ec0dd2cd..c7c7ab2d 100644 --- a/backend/SelectLongproof.v +++ b/backend/SelectLongproof.v @@ -435,6 +435,34 @@ Proof. auto. Qed. +Theorem eval_longofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.longofsingle x = Some y -> + exists v, eval_expr ge sp e m le (longofsingle hf a) v /\ Val.lessdef y v. +Proof. + intros; unfold longofsingle. + destruct x; simpl in H0; inv H0. destruct (Float32.to_long f) as [n|] eqn:EQ; simpl in H2; inv H2. + exploit eval_floatofsingle; eauto. intros (v & A & B). simpl in B. inv B. + apply Float32.to_long_double in EQ. + eapply eval_longoffloat; eauto. simpl. + change (Float.of_single f) with (Float32.to_double f); rewrite EQ; auto. +Qed. + +Theorem eval_longuofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.longuofsingle x = Some y -> + exists v, eval_expr ge sp e m le (longuofsingle hf a) v /\ Val.lessdef y v. +Proof. + intros; unfold longuofsingle. + destruct x; simpl in H0; inv H0. destruct (Float32.to_longu f) as [n|] eqn:EQ; simpl in H2; inv H2. + exploit eval_floatofsingle; eauto. intros (v & A & B). simpl in B. inv B. + apply Float32.to_longu_double in EQ. + eapply eval_longuoffloat; eauto. simpl. + change (Float.of_single f) with (Float32.to_double f); rewrite EQ; auto. +Qed. + Theorem eval_singleoflong: forall le a x y, eval_expr ge sp e m le a x -> diff --git a/backend/Selection.v b/backend/Selection.v index f62a888e..9bd37ef5 100644 --- a/backend/Selection.v +++ b/backend/Selection.v @@ -70,6 +70,7 @@ Definition sel_constant (cst: Cminor.constant) : expr := match cst with | Cminor.Ointconst n => Eop (Ointconst n) Enil | Cminor.Ofloatconst f => Eop (Ofloatconst f) Enil + | Cminor.Osingleconst f => Eop (Osingleconst f) Enil | Cminor.Olongconst n => longconst n | Cminor.Oaddrsymbol id ofs => addrsymbol id ofs | Cminor.Oaddrstack ofs => addrstack ofs @@ -85,11 +86,18 @@ Definition sel_unop (op: Cminor.unary_operation) (arg: expr) : expr := | Cminor.Onotint => notint arg | Cminor.Onegf => negf arg | Cminor.Oabsf => absf arg + | Cminor.Onegfs => negfs arg + | Cminor.Oabsfs => absfs arg | Cminor.Osingleoffloat => singleoffloat arg + | Cminor.Ofloatofsingle => floatofsingle arg | Cminor.Ointoffloat => intoffloat arg | Cminor.Ointuoffloat => intuoffloat arg | Cminor.Ofloatofint => floatofint arg | Cminor.Ofloatofintu => floatofintu arg + | Cminor.Ointofsingle => intofsingle arg + | Cminor.Ointuofsingle => intuofsingle arg + | Cminor.Osingleofint => singleofint arg + | Cminor.Osingleofintu => singleofintu arg | Cminor.Onegl => negl hf arg | Cminor.Onotl => notl arg | Cminor.Ointoflong => intoflong arg @@ -99,6 +107,8 @@ Definition sel_unop (op: Cminor.unary_operation) (arg: expr) : expr := | Cminor.Olonguoffloat => longuoffloat hf arg | Cminor.Ofloatoflong => floatoflong hf arg | Cminor.Ofloatoflongu => floatoflongu hf arg + | Cminor.Olongofsingle => longofsingle hf arg + | Cminor.Olonguofsingle => longuofsingle hf arg | Cminor.Osingleoflong => singleoflong hf arg | Cminor.Osingleoflongu => singleoflongu hf arg end. @@ -122,6 +132,10 @@ Definition sel_binop (op: Cminor.binary_operation) (arg1 arg2: expr) : expr := | Cminor.Osubf => subf arg1 arg2 | Cminor.Omulf => mulf arg1 arg2 | Cminor.Odivf => divf arg1 arg2 + | Cminor.Oaddfs => addfs arg1 arg2 + | Cminor.Osubfs => subfs arg1 arg2 + | Cminor.Omulfs => mulfs arg1 arg2 + | Cminor.Odivfs => divfs arg1 arg2 | Cminor.Oaddl => addl hf arg1 arg2 | Cminor.Osubl => subl hf arg1 arg2 | Cminor.Omull => mull hf arg1 arg2 @@ -138,6 +152,7 @@ Definition sel_binop (op: Cminor.binary_operation) (arg1 arg2: expr) : expr := | Cminor.Ocmp c => comp c arg1 arg2 | Cminor.Ocmpu c => compu c arg1 arg2 | Cminor.Ocmpf c => compf c arg1 arg2 + | Cminor.Ocmpfs c => compfs c arg1 arg2 | Cminor.Ocmpl c => cmpl c arg1 arg2 | Cminor.Ocmplu c => cmplu c arg1 arg2 end. diff --git a/backend/Selectionproof.v b/backend/Selectionproof.v index 175b025c..55977b47 100644 --- a/backend/Selectionproof.v +++ b/backend/Selectionproof.v @@ -191,11 +191,18 @@ Proof. apply eval_notint; auto. apply eval_negf; auto. apply eval_absf; auto. + apply eval_negfs; auto. + apply eval_absfs; auto. apply eval_singleoffloat; auto. + apply eval_floatofsingle; auto. eapply eval_intoffloat; eauto. eapply eval_intuoffloat; eauto. eapply eval_floatofint; eauto. eapply eval_floatofintu; eauto. + eapply eval_intofsingle; eauto. + eapply eval_intuofsingle; eauto. + eapply eval_singleofint; eauto. + eapply eval_singleofintu; eauto. eapply eval_negl; eauto. eapply eval_notl; eauto. eapply eval_intoflong; eauto. @@ -205,6 +212,8 @@ Proof. eapply eval_longuoffloat; eauto. eapply eval_floatoflong; eauto. eapply eval_floatoflongu; eauto. + eapply eval_longofsingle; eauto. + eapply eval_longuofsingle; eauto. eapply eval_singleoflong; eauto. eapply eval_singleoflongu; eauto. Qed. @@ -234,6 +243,10 @@ Proof. apply eval_subf; auto. apply eval_mulf; auto. apply eval_divf; auto. + apply eval_addfs; auto. + apply eval_subfs; auto. + apply eval_mulfs; auto. + apply eval_divfs; auto. eapply eval_addl; eauto. eapply eval_subl; eauto. eapply eval_mull; eauto. @@ -250,6 +263,7 @@ Proof. apply eval_comp; auto. apply eval_compu; auto. apply eval_compf; auto. + apply eval_compfs; auto. exists v; split; auto. eapply eval_cmpl; eauto. exists v; split; auto. eapply eval_cmplu; eauto. Qed. @@ -377,6 +391,7 @@ Proof. destruct cst; simpl in *; inv H. exists (Vint i); split; auto. econstructor. constructor. auto. exists (Vfloat f); split; auto. econstructor. constructor. auto. + exists (Vsingle f); split; auto. econstructor. constructor. auto. exists (Val.longofwords (Vint (Int64.hiword i)) (Vint (Int64.loword i))); split. eapply eval_Eop. constructor. EvalOp. simpl; eauto. constructor. EvalOp. simpl; eauto. constructor. auto. simpl. rewrite Int64.ofwords_recompose. auto. diff --git a/backend/Stacking.v b/backend/Stacking.v index f7c16d14..4ee43bb1 100644 --- a/backend/Stacking.v +++ b/backend/Stacking.v @@ -73,12 +73,12 @@ Definition save_callee_save_regs Definition save_callee_save_int (fe: frame_env) := save_callee_save_regs fe_num_int_callee_save index_int_callee_save FI_saved_int - Tint fe int_callee_save_regs. + Tany32 fe int_callee_save_regs. Definition save_callee_save_float (fe: frame_env) := save_callee_save_regs fe_num_float_callee_save index_float_callee_save FI_saved_float - Tfloat fe float_callee_save_regs. + Tany64 fe float_callee_save_regs. Definition save_callee_save (fe: frame_env) (k: Mach.code) := save_callee_save_int fe (save_callee_save_float fe k). @@ -103,12 +103,12 @@ Definition restore_callee_save_regs Definition restore_callee_save_int (fe: frame_env) := restore_callee_save_regs fe_num_int_callee_save index_int_callee_save FI_saved_int - Tint fe int_callee_save_regs. + Tany32 fe int_callee_save_regs. Definition restore_callee_save_float (fe: frame_env) := restore_callee_save_regs fe_num_float_callee_save index_float_callee_save FI_saved_float - Tfloat fe float_callee_save_regs. + Tany64 fe float_callee_save_regs. Definition restore_callee_save (fe: frame_env) (k: Mach.code) := restore_callee_save_int fe (restore_callee_save_float fe k). diff --git a/backend/Stackingproof.v b/backend/Stackingproof.v index b69f1ec2..28b155ab 100644 --- a/backend/Stackingproof.v +++ b/backend/Stackingproof.v @@ -132,8 +132,8 @@ Definition type_of_index (idx: frame_index) := | FI_retaddr => Tint | FI_local x ty => ty | FI_arg x ty => ty - | FI_saved_int x => Tint - | FI_saved_float x => Tfloat + | FI_saved_int x => Tany32 + | FI_saved_float x => Tany64 end. (** Non-overlap between the memory areas corresponding to two @@ -194,8 +194,8 @@ Proof. destruct idx1; destruct idx2; simpl in V1; simpl in V2; repeat InvIndexValid; simpl in DIFF; unfold offset_of_index, type_of_index; + change (AST.typesize Tany32) with 4; change (AST.typesize Tany64) with 8; change (AST.typesize Tint) with 4; - change (AST.typesize Tfloat) with 8; omega. Qed. @@ -211,8 +211,8 @@ Proof. destruct idx; simpl in V; repeat InvIndexValid; unfold offset_of_index, type_of_index; + change (AST.typesize Tany32) with 4; change (AST.typesize Tany64) with 8; change (AST.typesize Tint) with 4; - change (AST.typesize Tfloat) with 8; omega. Qed. @@ -254,6 +254,17 @@ Proof. auto with align_4. Qed. +Lemma offset_of_index_aligned_2: + forall idx, index_valid idx -> + (align_chunk (chunk_of_type (type_of_index idx)) | offset_of_index fe idx). +Proof. + intros. replace (align_chunk (chunk_of_type (type_of_index idx))) with 4. + apply offset_of_index_aligned. + assert (type_of_index idx <> Tlong) by + (destruct idx; simpl; simpl in H; intuition congruence). + destruct (type_of_index idx); auto; congruence. +Qed. + Lemma fe_stack_data_aligned: (8 | fe_stack_data fe). Proof. @@ -271,7 +282,7 @@ Lemma index_local_valid: Proof. unfold slot_within_bounds, slot_valid, index_valid; intros. InvBooleans. - split. destruct ty; congruence. auto. + split. destruct ty; auto || discriminate. auto. Qed. Lemma index_arg_valid: @@ -281,7 +292,7 @@ Lemma index_arg_valid: Proof. unfold slot_within_bounds, slot_valid, index_valid; intros. InvBooleans. - split. destruct ty; congruence. auto. + split. destruct ty; auto || discriminate. auto. Qed. Lemma index_saved_int_valid: @@ -322,8 +333,8 @@ Proof. AddPosProps. destruct idx; simpl in V; repeat InvIndexValid; unfold offset_of_index, type_of_index; + change (AST.typesize Tany32) with 4; change (AST.typesize Tany64) with 8; change (AST.typesize Tint) with 4; - change (AST.typesize Tfloat) with 8; omega. Qed. @@ -414,7 +425,7 @@ Proof. intros. exploit gss_index_contains_base; eauto. intros [v' [A B]]. assert (v' = v). destruct v; destruct (type_of_index idx); simpl in *; - try contradiction; auto. rewrite Floats.Float.singleoffloat_of_single in B; auto. + try contradiction; auto. subst v'. auto. Qed. @@ -479,11 +490,7 @@ Proof. rewrite size_type_chunk. apply Mem.range_perm_implies with Freeable; auto with mem. apply offset_of_index_perm; auto. - replace (align_chunk (chunk_of_type (type_of_index idx))) with 4. - apply offset_of_index_aligned; auto. - assert (type_of_index idx <> Tlong). - destruct idx; simpl in *; tauto || congruence. - destruct (type_of_index idx); reflexivity || congruence. + apply offset_of_index_aligned_2; auto. exists m'; auto. Qed. @@ -514,7 +521,8 @@ Proof. intros. exploit gss_index_contains_base; eauto. intros [v'' [A B]]. exists v''; split; auto. inv H2; destruct (type_of_index idx); simpl in *; try contradiction; subst; auto. - rewrite Floats.Float.singleoffloat_of_single; auto. + econstructor; eauto. + econstructor; eauto. econstructor; eauto. Qed. @@ -529,6 +537,8 @@ Proof. exists v''; split; auto. inv H1; destruct (type_of_index idx); simpl in *; try contradiction; subst; auto. econstructor; eauto. + econstructor; eauto. + econstructor; eauto. Qed. Lemma gso_index_contains_inj: @@ -576,11 +586,7 @@ Proof. rewrite size_type_chunk. apply Mem.range_perm_implies with Freeable; auto with mem. apply offset_of_index_perm; auto. - replace (align_chunk (chunk_of_type (type_of_index idx))) with 4. - apply offset_of_index_aligned. - assert (type_of_index idx <> Tlong). - destruct idx; simpl in *; tauto || congruence. - destruct (type_of_index idx); reflexivity || congruence. + apply offset_of_index_aligned_2; auto. intros [v C]. exists v; split; auto. constructor; auto. Qed. @@ -844,18 +850,17 @@ Lemma agree_frame_set_local: forall j ls ls0 m sp m' sp' parent retaddr ofs ty v v' m'', agree_frame j ls ls0 m sp m' sp' parent retaddr -> slot_within_bounds b Local ofs ty -> slot_valid f Local ofs ty = true -> - Val.has_type v ty -> val_inject j v v' -> Mem.store (chunk_of_type ty) m' sp' (offset_of_index fe (FI_local ofs ty)) v' = Some m'' -> agree_frame j (Locmap.set (S Local ofs ty) v ls) ls0 m sp m'' sp' parent retaddr. Proof. intros. inv H. - change (chunk_of_type ty) with (chunk_of_type (type_of_index (FI_local ofs ty))) in H4. + change (chunk_of_type ty) with (chunk_of_type (type_of_index (FI_local ofs ty))) in H3. constructor; auto; intros. (* local *) unfold Locmap.set. destruct (Loc.eq (S Local ofs ty) (S Local ofs0 ty0)). - inv e. eapply gss_index_contains_inj; eauto with stacking. + inv e. eapply gss_index_contains_inj'; eauto with stacking. destruct (Loc.diff_dec (S Local ofs ty) (S Local ofs0 ty0)). eapply gso_index_contains_inj. eauto. eauto with stacking. eauto. simpl. simpl in d. intuition. @@ -884,20 +889,19 @@ Lemma agree_frame_set_outgoing: forall j ls ls0 m sp m' sp' parent retaddr ofs ty v v' m'', agree_frame j ls ls0 m sp m' sp' parent retaddr -> slot_within_bounds b Outgoing ofs ty -> slot_valid f Outgoing ofs ty = true -> - Val.has_type v ty -> val_inject j v v' -> Mem.store (chunk_of_type ty) m' sp' (offset_of_index fe (FI_arg ofs ty)) v' = Some m'' -> agree_frame j (Locmap.set (S Outgoing ofs ty) v ls) ls0 m sp m'' sp' parent retaddr. Proof. intros. inv H. - change (chunk_of_type ty) with (chunk_of_type (type_of_index (FI_arg ofs ty))) in H4. + change (chunk_of_type ty) with (chunk_of_type (type_of_index (FI_arg ofs ty))) in H3. constructor; auto; intros. (* local *) rewrite Locmap.gso. eapply gso_index_contains_inj; eauto with stacking. red; auto. red; left; congruence. (* outgoing *) unfold Locmap.set. destruct (Loc.eq (S Outgoing ofs ty) (S Outgoing ofs0 ty0)). - inv e. eapply gss_index_contains_inj; eauto with stacking. + inv e. eapply gss_index_contains_inj'; eauto with stacking. destruct (Loc.diff_dec (S Outgoing ofs ty) (S Outgoing ofs0 ty0)). eapply gso_index_contains_inj; eauto with stacking. red. red in d. intuition. @@ -1145,12 +1149,6 @@ Proof. apply check_mreg_list_incl; compute; auto. Qed. -Remark destroyed_by_move_at_function_entry: - incl (destroyed_by_op Omove) destroyed_at_function_entry. -Proof. - apply check_mreg_list_incl; compute; auto. -Qed. - Remark temp_for_parent_frame_caller_save: In temp_for_parent_frame destroyed_at_call. Proof. @@ -1164,6 +1162,12 @@ Hint Resolve destroyed_by_op_caller_save destroyed_by_load_caller_save destroyed_by_cond_caller_save destroyed_by_jumptable_caller_save destroyed_at_function_entry_caller_save: stacking. +Remark destroyed_by_setstack_function_entry: + forall ty, incl (destroyed_by_setstack ty) destroyed_at_function_entry. +Proof. + destruct ty; apply check_mreg_list_incl; compute; auto. +Qed. + Remark transl_destroyed_by_op: forall op e, destroyed_by_op (transl_op e op) = destroyed_by_op op. Proof. @@ -1367,12 +1371,12 @@ Lemma save_callee_save_correct: /\ agree_regs j ls rs'. Proof. intros. - assert (UNDEF: forall r, In r (destroyed_by_op Omove) -> ls (R r) = Vundef). - intros. unfold ls. apply LTL_undef_regs_same. apply destroyed_by_move_at_function_entry; auto. + assert (UNDEF: forall r ty, In r (destroyed_by_setstack ty) -> ls (R r) = Vundef). + intros. unfold ls. apply LTL_undef_regs_same. eapply destroyed_by_setstack_function_entry; eauto. exploit (save_callee_save_regs_correct fe_num_int_callee_save index_int_callee_save - FI_saved_int Tint + FI_saved_int Tany32 j cs fb sp int_callee_save_regs ls). intros. apply index_int_callee_save_inj; auto. intros. simpl. split. apply Zge_le. apply index_int_callee_save_pos; auto. assumption. @@ -1380,8 +1384,7 @@ Proof. intros; congruence. intros; simpl. destruct idx; auto. congruence. intros. apply int_callee_save_type. auto. -Local Transparent destroyed_by_setstack. - simpl. tauto. + eauto. auto. apply incl_refl. apply int_callee_save_norepet. @@ -1391,7 +1394,7 @@ Local Transparent destroyed_by_setstack. exploit (save_callee_save_regs_correct fe_num_float_callee_save index_float_callee_save - FI_saved_float Tfloat + FI_saved_float Tany64 j cs fb sp float_callee_save_regs ls). intros. apply index_float_callee_save_inj; auto. intros. simpl. split. apply Zge_le. apply index_float_callee_save_pos; auto. assumption. @@ -1399,7 +1402,7 @@ Local Transparent destroyed_by_setstack. intros; congruence. intros; simpl. destruct idx; auto. congruence. intros. apply float_callee_save_type. auto. - simpl. tauto. + eauto. auto. apply incl_refl. apply float_callee_save_norepet. @@ -1462,6 +1465,16 @@ Proof. left. apply Plt_ne; auto. Qed. +Lemma undef_regs_type: + forall ty l rl ls, + Val.has_type (ls l) ty -> Val.has_type (LTL.undef_regs rl ls l) ty. +Proof. + induction rl; simpl; intros. +- auto. +- unfold Locmap.set. destruct (Loc.eq (R a) l). red; auto. + destruct (Loc.diff_dec (R a) l); auto. red; auto. +Qed. + (** As a corollary of the previous lemmas, we obtain the following correctness theorem for the execution of a function prologue (allocation of the frame + saving of the link and return address + @@ -1732,7 +1745,7 @@ Proof. fe_num_int_callee_save index_int_callee_save FI_saved_int - Tint + Tany32 int_callee_save_regs j cs fb sp' ls0 m'); auto. intros. unfold mreg_within_bounds. split; intros. @@ -1750,7 +1763,7 @@ Proof. fe_num_float_callee_save index_float_callee_save FI_saved_float - Tfloat + Tany64 float_callee_save_regs j cs fb sp' ls0 m'); auto. intros. unfold mreg_within_bounds. split; intros. @@ -2318,8 +2331,8 @@ Proof. unfold parent_sp. assert (slot_valid f Outgoing pos ty = true). exploit loc_arguments_acceptable; eauto. intros [A B]. - unfold slot_valid. unfold proj_sumbool. rewrite zle_true. - destruct ty; reflexivity || congruence. omega. + unfold slot_valid. unfold proj_sumbool. rewrite zle_true by omega. + destruct ty; auto; congruence. assert (slot_within_bounds (function_bounds f) Outgoing pos ty). eauto. exploit agree_outgoing; eauto. intros [v [A B]]. @@ -2525,7 +2538,7 @@ Proof. apply agree_frame_set_reg; auto. - (* Lsetstack *) - exploit wt_state_setstack; eauto. intros (VTY & SV & SW). + exploit wt_state_setstack; eauto. intros (SV & SW). set (idx := match sl with | Local => FI_local ofs ty | Incoming => FI_link (*dummy*) @@ -2552,15 +2565,15 @@ Proof. omega. apply match_stacks_change_mach_mem with m'; auto. eauto with mem. eauto with mem. intros. rewrite <- H1; eapply Mem.load_store_other; eauto. left; apply Plt_ne; auto. - eauto. eauto. auto. + eauto. eauto. apply agree_regs_set_slot. apply agree_regs_undef_regs; auto. destruct sl. + eapply agree_frame_set_local. eapply agree_frame_undef_locs; eauto. - apply destroyed_by_setstack_caller_save. auto. auto. auto. apply AGREGS. + apply destroyed_by_setstack_caller_save. auto. auto. auto. assumption. + simpl in SW; discriminate. + eapply agree_frame_set_outgoing. eapply agree_frame_undef_locs; eauto. - apply destroyed_by_setstack_caller_save. auto. auto. auto. apply AGREGS. + apply destroyed_by_setstack_caller_save. auto. auto. auto. assumption. + eauto with coqlib. @@ -2613,13 +2626,15 @@ Proof. apply plus_one. econstructor. instantiate (1 := a'). rewrite <- A. apply eval_addressing_preserved. exact symbols_preserved. eexact C. eauto. - econstructor; eauto with coqlib. + econstructor. eauto. eapply match_stacks_parallel_stores. eexact MINJ. eexact B. eauto. eauto. auto. + eauto. eauto. rewrite transl_destroyed_by_store. apply agree_regs_undef_regs; auto. apply agree_frame_undef_locs; auto. eapply agree_frame_parallel_stores; eauto. - apply destroyed_by_store_caller_save. + apply destroyed_by_store_caller_save. + eauto with coqlib. - (* Lcall *) exploit find_function_translated; eauto. intros [bf [tf' [A [B C]]]]. diff --git a/backend/ValueAnalysis.v b/backend/ValueAnalysis.v index e293028d..813944d6 100644 --- a/backend/ValueAnalysis.v +++ b/backend/ValueAnalysis.v @@ -166,7 +166,7 @@ Definition store_init_data (ab: ablock) (p: Z) (id: init_data) : ablock := | Init_int32 n => ablock_store Mint32 ab p (I n) | Init_int64 n => ablock_store Mint64 ab p (L n) | Init_float32 n => ablock_store Mfloat32 ab p - (if propagate_float_constants tt then F n else ftop) + (if propagate_float_constants tt then FS n else ftop) | Init_float64 n => ablock_store Mfloat64 ab p (if propagate_float_constants tt then F n else ftop) | Init_addrof symb ofs => ablock_store Mint32 ab p (Ptr (Gl symb ofs)) @@ -922,7 +922,7 @@ Proof. eapply VMTOP; eauto. - exploit BC'INV; eauto. intros (b'' & delta & J'). exploit Mem.loadbytes_inject. eexact IMEM. eauto. eauto. intros (bytes & A & B). - inv B. inv H3. eapply PMTOP; eauto. + inv B. inv H3. inv H7. eapply PMTOP; eauto. } (* Conclusions *) exists bc'; splitall. diff --git a/backend/ValueDomain.v b/backend/ValueDomain.v index c8431bb7..75dd9d20 100644 --- a/backend/ValueDomain.v +++ b/backend/ValueDomain.v @@ -509,13 +509,13 @@ Inductive aval : Type := | Sgn (n: Z) | L (n: int64) | F (f: float) - | Fsingle + | FS (f: float32) | Ptr (p: aptr) | Ifptr (p: aptr). Definition eq_aval: forall (v1 v2: aval), {v1=v2} + {v1<>v2}. Proof. - intros. generalize zeq Int.eq_dec Int64.eq_dec Float.eq_dec eq_aptr; intros. + intros. generalize zeq Int.eq_dec Int64.eq_dec Float.eq_dec Float32.eq_dec eq_aptr; intros. decide equality. Defined. @@ -537,14 +537,14 @@ Inductive vmatch : val -> aval -> Prop := | vmatch_Sgn_undef: forall n, vmatch Vundef (Sgn n) | vmatch_l: forall i, vmatch (Vlong i) (L i) | vmatch_f: forall f, vmatch (Vfloat f) (F f) - | vmatch_single: forall f, Float.is_single f -> vmatch (Vfloat f) Fsingle - | vmatch_single_undef: vmatch Vundef Fsingle + | vmatch_s: forall f, vmatch (Vsingle f) (FS f) | vmatch_ptr: forall b ofs p, pmatch b ofs p -> vmatch (Vptr b ofs) (Ptr p) | vmatch_ptr_undef: forall p, vmatch Vundef (Ptr p) | vmatch_ifptr_undef: forall p, vmatch Vundef (Ifptr p) | vmatch_ifptr_i: forall i p, vmatch (Vint i) (Ifptr p) | vmatch_ifptr_l: forall i p, vmatch (Vlong i) (Ifptr p) | vmatch_ifptr_f: forall f p, vmatch (Vfloat f) (Ifptr p) + | vmatch_ifptr_s: forall f p, vmatch (Vsingle f) (Ifptr p) | vmatch_ifptr_p: forall b ofs p, pmatch b ofs p -> vmatch (Vptr b ofs) (Ifptr p). Lemma vmatch_ifptr: @@ -569,11 +569,14 @@ Proof. constructor. Qed. Lemma vmatch_ftop: forall f, vmatch (Vfloat f) ftop. Proof. intros; constructor. Qed. +Lemma vmatch_ftop_single: forall f, vmatch (Vsingle f) ftop. +Proof. intros; constructor. Qed. + Lemma vmatch_undef_ftop: vmatch Vundef ftop. Proof. constructor. Qed. Hint Constructors vmatch : va. -Hint Resolve vmatch_itop vmatch_undef_itop vmatch_ftop vmatch_undef_ftop : va. +Hint Resolve vmatch_itop vmatch_undef_itop vmatch_ftop vmatch_ftop_single vmatch_undef_ftop : va. (* Some properties about [is_uns] and [is_sgn]. *) @@ -770,22 +773,21 @@ Inductive vge: aval -> aval -> Prop := | vge_i: forall i, vge (I i) (I i) | vge_l: forall i, vge (L i) (L i) | vge_f: forall f, vge (F f) (F f) + | vge_s: forall f, vge (FS f) (FS f) | vge_uns_i: forall n i, 0 <= n -> is_uns n i -> vge (Uns n) (I i) | vge_uns_uns: forall n1 n2, n1 >= n2 -> vge (Uns n1) (Uns n2) | vge_sgn_i: forall n i, 0 < n -> is_sgn n i -> vge (Sgn n) (I i) | vge_sgn_sgn: forall n1 n2, n1 >= n2 -> vge (Sgn n1) (Sgn n2) | vge_sgn_uns: forall n1 n2, n1 > n2 -> vge (Sgn n1) (Uns n2) - | vge_single_f: forall f, Float.is_single f -> vge Fsingle (F f) - | vge_single: vge Fsingle Fsingle | vge_p_p: forall p q, pge p q -> vge (Ptr p) (Ptr q) | vge_ip_p: forall p q, pge p q -> vge (Ifptr p) (Ptr q) | vge_ip_ip: forall p q, pge p q -> vge (Ifptr p) (Ifptr q) | vge_ip_i: forall p i, vge (Ifptr p) (I i) | vge_ip_l: forall p i, vge (Ifptr p) (L i) | vge_ip_f: forall p f, vge (Ifptr p) (F f) + | vge_ip_s: forall p f, vge (Ifptr p) (FS f) | vge_ip_uns: forall p n, vge (Ifptr p) (Uns n) - | vge_ip_sgn: forall p n, vge (Ifptr p) (Sgn n) - | vge_ip_single: forall p, vge (Ifptr p) Fsingle. + | vge_ip_sgn: forall p n, vge (Ifptr p) (Sgn n). Hint Constructors vge : va. @@ -836,12 +838,9 @@ Definition vlub (v w: aval) : aval := | Sgn n1, Uns n2 => sgn (Z.max n1 (n2 + 1)) | Sgn n1, Sgn n2 => sgn (Z.max n1 n2) | F f1, F f2 => - if Float.eq_dec f1 f2 then v else - if Float.is_single_dec f1 && Float.is_single_dec f2 then Fsingle else ftop - | F f, Fsingle | Fsingle, F f => - if Float.is_single_dec f then Fsingle else ftop - | Fsingle, Fsingle => - Fsingle + if Float.eq_dec f1 f2 then v else ftop + | FS f1, FS f2 => + if Float32.eq_dec f1 f2 then v else ftop | L i1, L i2 => if Int64.eq i1 i2 then v else ltop | Ptr p1, Ptr p2 => Ptr(plub p1 p2) @@ -865,8 +864,8 @@ Proof. - f_equal; apply Z.max_comm. - f_equal; apply Z.max_comm. - rewrite Int64.eq_sym. predSpec Int64.eq Int64.eq_spec n0 n; congruence. -- rewrite dec_eq_sym. destruct (Float.eq_dec f0 f). congruence. - rewrite andb_comm. auto. +- rewrite dec_eq_sym. destruct (Float.eq_dec f0 f). congruence. auto. +- rewrite dec_eq_sym. destruct (Float32.eq_dec f0 f). congruence. auto. - f_equal; apply plub_comm. - f_equal; apply plub_comm. - f_equal; apply plub_comm. @@ -937,12 +936,8 @@ Proof. - eapply vge_trans. apply vge_sgn_sgn'. eauto with va. - eapply vge_trans. apply vge_sgn_sgn'. eauto with va. - destruct (Int64.eq n n0); constructor. -- destruct (Float.eq_dec f f0). constructor. - destruct (Float.is_single_dec f && Float.is_single_dec f0) eqn:FS. - InvBooleans. auto with va. - constructor. -- destruct (Float.is_single_dec f); constructor; auto. -- destruct (Float.is_single_dec f); constructor; auto. +- destruct (Float.eq_dec f f0); constructor. +- destruct (Float32.eq_dec f f0); constructor. - constructor; apply pge_lub_l; auto. - constructor; apply pge_lub_l; auto. - constructor; apply pge_lub_l; auto. @@ -1043,8 +1038,7 @@ Definition vincl (v w: aval) : bool := | Sgn n, Sgn m => zle n m | L i, L j => Int64.eq_dec i j | F i, F j => Float.eq_dec i j - | F i, Fsingle => Float.is_single_dec i - | Fsingle, Fsingle => true + | FS i, FS j => Float32.eq_dec i j | Ptr p, Ptr q => pincl p q | Ptr p, Ifptr q => pincl p q | Ifptr p, Ifptr q => pincl p q @@ -1063,7 +1057,7 @@ Proof. InvBooleans. constructor; auto with va. InvBooleans. subst; auto with va. InvBooleans. subst; auto with va. - InvBooleans. auto with va. + InvBooleans. subst; auto with va. constructor; apply pincl_ge; auto. constructor; apply pincl_ge; auto. constructor; apply pincl_ge; auto. @@ -1154,6 +1148,28 @@ Proof. intros. unfold binop_float; inv H; auto with va; inv H0; auto with va. Qed. +Definition unop_single (sem: float32 -> float32) (x: aval) := + match x with FS n => FS (sem n) | _ => ftop end. + +Lemma unop_single_sound: + forall sem v x, + vmatch v x -> + vmatch (match v with Vsingle i => Vsingle(sem i) | _ => Vundef end) (unop_single sem x). +Proof. + intros. unfold unop_single; inv H; auto with va. +Qed. + +Definition binop_single (sem: float32 -> float32 -> float32) (x y: aval) := + match x, y with FS n, FS m => FS (sem n m) | _, _ => ftop end. + +Lemma binop_single_sound: + forall sem v x w y, + vmatch v x -> vmatch w y -> + vmatch (match v, w with Vsingle i, Vsingle j => Vsingle(sem i j) | _, _ => Vundef end) (binop_single sem x y). +Proof. + intros. unfold binop_single; inv H; auto with va; inv H0; auto with va. +Qed. + (** Logical operations *) Definition shl (v w: aval) := @@ -1636,6 +1652,42 @@ Lemma divf_sound: forall v x w y, vmatch v x -> vmatch w y -> vmatch (Val.divf v w) (divf x y). Proof (binop_float_sound Float.div). +Definition negfs := unop_single Float32.neg. + +Lemma negfs_sound: + forall v x, vmatch v x -> vmatch (Val.negfs v) (negfs x). +Proof (unop_single_sound Float32.neg). + +Definition absfs := unop_single Float32.abs. + +Lemma absfs_sound: + forall v x, vmatch v x -> vmatch (Val.absfs v) (absfs x). +Proof (unop_single_sound Float32.abs). + +Definition addfs := binop_single Float32.add. + +Lemma addfs_sound: + forall v x w y, vmatch v x -> vmatch w y -> vmatch (Val.addfs v w) (addfs x y). +Proof (binop_single_sound Float32.add). + +Definition subfs := binop_single Float32.sub. + +Lemma subfs_sound: + forall v x w y, vmatch v x -> vmatch w y -> vmatch (Val.subfs v w) (subfs x y). +Proof (binop_single_sound Float32.sub). + +Definition mulfs := binop_single Float32.mul. + +Lemma mulfs_sound: + forall v x w y, vmatch v x -> vmatch w y -> vmatch (Val.mulfs v w) (mulfs x y). +Proof (binop_single_sound Float32.mul). + +Definition divfs := binop_single Float32.div. + +Lemma divfs_sound: + forall v x w y, vmatch v x -> vmatch w y -> vmatch (Val.divfs v w) (divfs x y). +Proof (binop_single_sound Float32.div). + (** Conversions *) Definition zero_ext (nbits: Z) (v: aval) := @@ -1683,23 +1735,38 @@ Qed. Definition singleoffloat (v: aval) := match v with - | F f => F (Float.singleoffloat f) - | _ => Fsingle + | F f => FS (Float.to_single f) + | _ => ftop end. Lemma singleoffloat_sound: forall v x, vmatch v x -> vmatch (Val.singleoffloat v) (singleoffloat x). Proof. - intros. - assert (DEFAULT: vmatch (Val.singleoffloat v) Fsingle). - { destruct v; constructor. apply Float.singleoffloat_is_single. } + intros. + assert (DEFAULT: vmatch (Val.singleoffloat v) ftop). + { destruct v; constructor. } + destruct x; auto. inv H. constructor. +Qed. + +Definition floatofsingle (v: aval) := + match v with + | FS f => F (Float.of_single f) + | _ => ftop + end. + +Lemma floatofsingle_sound: + forall v x, vmatch v x -> vmatch (Val.floatofsingle v) (floatofsingle x). +Proof. + intros. + assert (DEFAULT: vmatch (Val.floatofsingle v) ftop). + { destruct v; constructor. } destruct x; auto. inv H. constructor. Qed. Definition intoffloat (x: aval) := match x with | F f => - match Float.intoffloat f with + match Float.to_int f with | Some i => I i | None => if va_strict tt then Vbot else itop end @@ -1710,14 +1777,14 @@ Lemma intoffloat_sound: forall v x w, vmatch v x -> Val.intoffloat v = Some w -> vmatch w (intoffloat x). Proof. unfold Val.intoffloat; intros. destruct v; try discriminate. - destruct (Float.intoffloat f) as [i|] eqn:E; simpl in H0; inv H0. + destruct (Float.to_int f) as [i|] eqn:E; simpl in H0; inv H0. inv H; simpl; auto with va. rewrite E; constructor. Qed. Definition intuoffloat (x: aval) := match x with | F f => - match Float.intuoffloat f with + match Float.to_intu f with | Some i => I i | None => if va_strict tt then Vbot else itop end @@ -1728,13 +1795,13 @@ Lemma intuoffloat_sound: forall v x w, vmatch v x -> Val.intuoffloat v = Some w -> vmatch w (intuoffloat x). Proof. unfold Val.intuoffloat; intros. destruct v; try discriminate. - destruct (Float.intuoffloat f) as [i|] eqn:E; simpl in H0; inv H0. + destruct (Float.to_intu f) as [i|] eqn:E; simpl in H0; inv H0. inv H; simpl; auto with va. rewrite E; constructor. Qed. Definition floatofint (x: aval) := match x with - | I i => F(Float.floatofint i) + | I i => F(Float.of_int i) | _ => ftop end. @@ -1747,7 +1814,7 @@ Qed. Definition floatofintu (x: aval) := match x with - | I i => F(Float.floatofintu i) + | I i => F(Float.of_intu i) | _ => ftop end. @@ -1758,6 +1825,68 @@ Proof. inv H; simpl; auto with va. Qed. +Definition intofsingle (x: aval) := + match x with + | FS f => + match Float32.to_int f with + | Some i => I i + | None => if va_strict tt then Vbot else itop + end + | _ => itop + end. + +Lemma intofsingle_sound: + forall v x w, vmatch v x -> Val.intofsingle v = Some w -> vmatch w (intofsingle x). +Proof. + unfold Val.intofsingle; intros. destruct v; try discriminate. + destruct (Float32.to_int f) as [i|] eqn:E; simpl in H0; inv H0. + inv H; simpl; auto with va. rewrite E; constructor. +Qed. + +Definition intuofsingle (x: aval) := + match x with + | FS f => + match Float32.to_intu f with + | Some i => I i + | None => if va_strict tt then Vbot else itop + end + | _ => itop + end. + +Lemma intuofsingle_sound: + forall v x w, vmatch v x -> Val.intuofsingle v = Some w -> vmatch w (intuofsingle x). +Proof. + unfold Val.intuofsingle; intros. destruct v; try discriminate. + destruct (Float32.to_intu f) as [i|] eqn:E; simpl in H0; inv H0. + inv H; simpl; auto with va. rewrite E; constructor. +Qed. + +Definition singleofint (x: aval) := + match x with + | I i => FS(Float32.of_int i) + | _ => ftop + end. + +Lemma singleofint_sound: + forall v x w, vmatch v x -> Val.singleofint v = Some w -> vmatch w (singleofint x). +Proof. + unfold Val.singleofint; intros. destruct v; inv H0. + inv H; simpl; auto with va. +Qed. + +Definition singleofintu (x: aval) := + match x with + | I i => FS(Float32.of_intu i) + | _ => ftop + end. + +Lemma singleofintu_sound: + forall v x w, vmatch v x -> Val.singleofintu v = Some w -> vmatch w (singleofintu x). +Proof. + unfold Val.singleofintu; intros. destruct v; inv H0. + inv H; simpl; auto with va. +Qed. + Definition floatofwords (x y: aval) := match x, y with | I i, I j => F(Float.from_words i j) @@ -1862,6 +1991,18 @@ Proof. intros. inv H; try constructor; inv H0; constructor. Qed. +Definition cmpfs_bool (c: comparison) (v w: aval) : abool := + match v, w with + | FS f1, FS f2 => Just (Float32.cmp c f1 f2) + | _, _ => Btop + end. + +Lemma cmpfs_bool_sound: + forall c v w x y, vmatch v x -> vmatch w y -> cmatch (Val.cmpfs_bool c v w) (cmpfs_bool c x y). +Proof. + intros. inv H; try constructor; inv H0; constructor. +Qed. + Definition maskzero (x: aval) (mask: int) : abool := match x with | I i => Just (Int.eq (Int.and i mask) Int.zero) @@ -1939,9 +2080,10 @@ Definition vnormalize (chunk: memory_chunk) (v: aval) := | Mint16unsigned, _ => Uns 16 | Mint32, (I _ | Ptr _ | Ifptr _) => v | Mint64, L _ => v - | Mfloat32, F f => F (Float.singleoffloat f) - | Mfloat32, _ => Fsingle + | Mfloat32, FS f => v | Mfloat64, F f => v + | Many32, (I _ | Ptr _ | Ifptr _ | FS _) => v + | Many64, _ => v | _, _ => Ifptr Pbot end. @@ -1963,12 +2105,10 @@ Proof. - constructor. omega. apply is_zero_ext_uns; auto with va. - constructor. xomega. apply is_sign_ext_sgn; auto with va. apply Z.min_case; auto with va. - constructor. omega. apply is_zero_ext_uns; auto with va. -- constructor. apply Float.singleoffloat_is_single. - constructor. omega. apply is_sign_ext_sgn; auto with va. - constructor. omega. apply is_zero_ext_uns; auto with va. - constructor. omega. apply is_sign_ext_sgn; auto with va. - constructor. omega. apply is_zero_ext_uns; auto with va. -- constructor. apply Float.singleoffloat_is_single. Qed. Lemma vnormalize_cast: @@ -1992,9 +2132,13 @@ Proof. - (* int64 *) destruct v; try contradiction; constructor. - (* float32 *) - rewrite H2. destruct v; simpl; constructor. apply Float.singleoffloat_is_single. + destruct v; try contradiction; constructor. - (* float64 *) destruct v; try contradiction; constructor. +- (* any32 *) + auto. +- (* any64 *) + auto. Qed. Lemma vnormalize_monotone: @@ -2024,12 +2168,10 @@ Proof. - constructor; auto with va. apply is_zero_ext_uns; auto with va. - destruct (zlt n2 8); constructor; auto with va. - destruct (zlt n2 16); constructor; auto with va. -- constructor. apply Float.singleoffloat_is_single. - constructor; auto with va. apply is_sign_ext_sgn; auto with va. - constructor; auto with va. apply is_zero_ext_uns; auto with va. - constructor; auto with va. apply is_sign_ext_sgn; auto with va. - constructor; auto with va. apply is_zero_ext_uns; auto with va. -- constructor. apply Float.singleoffloat_is_single. - destruct (zlt n 8); constructor; auto with va. - destruct (zlt n 16); constructor; auto with va. Qed. @@ -2064,6 +2206,8 @@ Definition chunk_compat (chunk chunk': memory_chunk) : bool := | Mfloat32, Mfloat32 => true | Mint64, Mint64 => true | Mfloat64, Mfloat64 => true + | Many32, Many32 => true + | Many64, Many64 => true | _, _ => false end. @@ -2148,7 +2292,7 @@ Definition ablock_storebytes_anywhere (ab: ablock) (p: aptr) := Definition smatch (m: mem) (b: block) (p: aptr) : Prop := (forall chunk ofs v, Mem.load chunk m b ofs = Some v -> vmatch v (Ifptr p)) -/\(forall ofs b' ofs' i, Mem.loadbytes m b ofs 1 = Some (Pointer b' ofs' i :: nil) -> pmatch b' ofs' p). +/\(forall ofs b' ofs' q i, Mem.loadbytes m b ofs 1 = Some (Fragment (Vptr b' ofs') q i :: nil) -> pmatch b' ofs' p). Remark loadbytes_load_ext: forall b m m', @@ -2219,10 +2363,10 @@ Proof. Qed. Lemma smatch_loadbytes: - forall m b p b' ofs' i n ofs bytes, + forall m b p b' ofs' q i n ofs bytes, Mem.loadbytes m b ofs n = Some bytes -> smatch m b p -> - In (Pointer b' ofs' i) bytes -> + In (Fragment (Vptr b' ofs') q i) bytes -> pmatch b' ofs' p. Proof. intros. exploit In_loadbytes; eauto. intros (ofs1 & A & B). @@ -2251,11 +2395,11 @@ Proof. Qed. Lemma storebytes_provenance: - forall m b ofs bytes m' b' ofs' b'' ofs'' i, + forall m b ofs bytes m' b' ofs' b'' ofs'' q i, Mem.storebytes m b ofs bytes = Some m' -> - Mem.loadbytes m' b' ofs' 1 = Some (Pointer b'' ofs'' i :: nil) -> - In (Pointer b'' ofs'' i) bytes - \/ Mem.loadbytes m b' ofs' 1 = Some (Pointer b'' ofs'' i :: nil). + Mem.loadbytes m' b' ofs' 1 = Some (Fragment (Vptr b'' ofs'') q i :: nil) -> + In (Fragment (Vptr b'' ofs'') q i) bytes + \/ Mem.loadbytes m b' ofs' 1 = Some (Fragment (Vptr b'' ofs'') q i :: nil). Proof. intros. assert (EITHER: @@ -2275,26 +2419,24 @@ Proof. Qed. Lemma store_provenance: - forall chunk m b ofs v m' b' ofs' b'' ofs'' i, + forall chunk m b ofs v m' b' ofs' b'' ofs'' q i, Mem.store chunk m b ofs v = Some m' -> - Mem.loadbytes m' b' ofs' 1 = Some (Pointer b'' ofs'' i :: nil) -> - v = Vptr b'' ofs'' /\ chunk = Mint32 - \/ Mem.loadbytes m b' ofs' 1 = Some (Pointer b'' ofs'' i :: nil). + Mem.loadbytes m' b' ofs' 1 = Some (Fragment (Vptr b'' ofs'') q i :: nil) -> + v = Vptr b'' ofs'' /\ (chunk = Mint32 \/ chunk = Many32 \/ chunk = Many64) + \/ Mem.loadbytes m b' ofs' 1 = Some (Fragment (Vptr b'' ofs'') q i :: nil). Proof. intros. exploit storebytes_provenance; eauto. eapply Mem.store_storebytes; eauto. intros [A|A]; auto. left. - assert (IN_ENC_BYTES: forall bl, ~In (Pointer b'' ofs'' i) (inj_bytes bl)). - { - induction bl; simpl. tauto. red; intros; elim IHbl. destruct H1. congruence. auto. - } - assert (IN_REP_UNDEF: forall n, ~In (Pointer b'' ofs'' i) (list_repeat n Undef)). - { - intros; red; intros. exploit in_list_repeat; eauto. congruence. - } - unfold encode_val in A; destruct chunk, v; - try (eelim IN_ENC_BYTES; eassumption); - try (eelim IN_REP_UNDEF; eassumption). - simpl in A. split; auto. intuition congruence. + generalize (encode_val_shape chunk v). intros ENC; inv ENC. +- split; auto. rewrite <- H1 in A; destruct A. + + congruence. + + exploit H5; eauto. intros (j & P & Q); congruence. +- rewrite <- H1 in A; destruct A. + + congruence. + + exploit H3; eauto. intros [byte P]; congruence. +- rewrite <- H1 in A; destruct A. + + congruence. + + exploit H2; eauto. congruence. Qed. Lemma smatch_store: @@ -2307,7 +2449,7 @@ Proof. intros. destruct H0 as [A B]. split. - intros chunk' ofs' v' LOAD. destruct v'; auto with va. exploit Mem.load_pointer_store; eauto. - intros [(P & Q & R & S & T) | DISJ]. + intros [(P & Q & R & S) | DISJ]. + subst. apply vmatch_vplub_l. auto. + apply vmatch_vplub_r. apply A with (chunk := chunk') (ofs := ofs'). rewrite <- LOAD. symmetry. eapply Mem.load_store_other; eauto. @@ -2325,17 +2467,20 @@ Lemma smatch_storebytes: forall m b ofs bytes m' b' p p', Mem.storebytes m b ofs bytes = Some m' -> smatch m b' p -> - (forall b' ofs' i, In (Pointer b' ofs' i) bytes -> pmatch b' ofs' p') -> + (forall b' ofs' q i, In (Fragment (Vptr b' ofs') q i) bytes -> pmatch b' ofs' p') -> smatch m' b' (plub p' p). Proof. intros. destruct H0 as [A B]. split. - intros. apply vmatch_ifptr. intros bx ofsx EQ; subst v. exploit Mem.load_loadbytes; eauto. intros (bytes' & P & Q). - exploit decode_val_pointer_inv; eauto. intros [U V]. - subst chunk bytes'. - exploit In_loadbytes; eauto. - instantiate (1 := Pointer bx ofsx 3%nat). simpl; auto. - intros (ofs' & X & Y). + destruct bytes' as [ | byte1' bytes']. + exploit Mem.loadbytes_length; eauto. intros. destruct chunk; discriminate. + generalize (decode_val_shape chunk byte1' bytes'). rewrite <- Q. + intros DEC; inv DEC; try contradiction. + assert (v = Vptr bx ofsx). + { destruct H5 as [E|[E|E]]; rewrite E in H4; destruct v; simpl in H4; congruence. } + exploit In_loadbytes; eauto. eauto with coqlib. + intros (ofs' & X & Y). subst v. exploit storebytes_provenance; eauto. intros [Z | Z]. apply pmatch_lub_l. eauto. apply pmatch_lub_r. eauto. @@ -2530,10 +2675,10 @@ Proof. Qed. Lemma ablock_loadbytes_sound: - forall m b ab b' ofs' i n ofs bytes, + forall m b ab b' ofs' q i n ofs bytes, Mem.loadbytes m b ofs n = Some bytes -> bmatch m b ab -> - In (Pointer b' ofs' i) bytes -> + In (Fragment (Vptr b' ofs') q i) bytes -> pmatch b' ofs' (ablock_loadbytes ab). Proof. intros. destruct H0. eapply smatch_loadbytes; eauto. @@ -2542,7 +2687,7 @@ Qed. Lemma ablock_storebytes_anywhere_sound: forall m b ofs bytes p m' b' ab, Mem.storebytes m b ofs bytes = Some m' -> - (forall b' ofs' i, In (Pointer b' ofs' i) bytes -> pmatch b' ofs' p) -> + (forall b' ofs' q i, In (Fragment (Vptr b' ofs') q i) bytes -> pmatch b' ofs' p) -> bmatch m b' ab -> bmatch m' b' (ablock_storebytes_anywhere ab p). Proof. @@ -2566,7 +2711,7 @@ Lemma ablock_storebytes_sound: forall m b ofs bytes m' p ab sz, Mem.storebytes m b ofs bytes = Some m' -> length bytes = nat_of_Z sz -> - (forall b' ofs' i, In (Pointer b' ofs' i) bytes -> pmatch b' ofs' p) -> + (forall b' ofs' q i, In (Fragment (Vptr b' ofs') q i) bytes -> pmatch b' ofs' p) -> bmatch m b ab -> bmatch m' b (ablock_storebytes ab p ofs sz). Proof. @@ -3060,7 +3205,7 @@ Theorem loadbytes_sound: romatch m rm -> mmatch m am -> pmatch b ofs p -> - forall b' ofs' i, In (Pointer b' ofs' i) bytes -> pmatch b' ofs' (loadbytes am rm p). + forall b' ofs' q i, In (Fragment (Vptr b' ofs') q i) bytes -> pmatch b' ofs' (loadbytes am rm p). Proof. intros. unfold loadbytes; inv H2. - (* Gl id ofs *) @@ -3093,7 +3238,7 @@ Theorem storebytes_sound: mmatch m am -> pmatch b ofs p -> length bytes = nat_of_Z sz -> - (forall b' ofs' i, In (Pointer b' ofs' i) bytes -> pmatch b' ofs' q) -> + (forall b' ofs' qt i, In (Fragment (Vptr b' ofs') qt i) bytes -> pmatch b' ofs' q) -> mmatch m' (storebytes am p sz q). Proof. intros until q; intros STORE MM PM LENGTH BYTES. @@ -3403,7 +3548,7 @@ Proof. unfold Mem.loadbytes. rewrite pred_dec_true. reflexivity. red; intros. replace ofs0 with ofs by omega. auto. } - destruct mv; econstructor. + destruct mv; econstructor. destruct v; econstructor. apply inj_of_bc_valid. assert (PM: pmatch bc b i Ptop). { exploit mmatch_top; eauto. intros [P Q]. @@ -3456,7 +3601,7 @@ Proof. - exploit Mem.load_inject. eauto. eauto. apply inj_of_bc_valid; auto. intros (v' & A & B). eapply vmatch_inj_top; eauto. - exploit Mem.loadbytes_inject. eauto. eauto. apply inj_of_bc_valid; auto. - intros (bytes' & A & B). inv B. inv H4. eapply pmatch_inj_top; eauto. + intros (bytes' & A & B). inv B. inv H4. inv H8. eapply pmatch_inj_top; eauto. } constructor; simpl; intros. - apply ablock_init_sound. apply SM. congruence. @@ -3701,7 +3846,11 @@ Hint Resolve cnot_sound symbol_address_sound divs_sound divu_sound mods_sound modu_sound shrx_sound negf_sound absf_sound addf_sound subf_sound mulf_sound divf_sound - zero_ext_sound sign_ext_sound singleoffloat_sound + negfs_sound absfs_sound + addfs_sound subfs_sound mulfs_sound divfs_sound + zero_ext_sound sign_ext_sound singleoffloat_sound floatofsingle_sound intoffloat_sound intuoffloat_sound floatofint_sound floatofintu_sound + intofsingle_sound intuofsingle_sound singleofint_sound singleofintu_sound longofwords_sound loword_sound hiword_sound - cmpu_bool_sound cmp_bool_sound cmpf_bool_sound maskzero_sound : va. + cmpu_bool_sound cmp_bool_sound cmpf_bool_sound cmpfs_bool_sound + maskzero_sound : va. diff --git a/cfrontend/C2C.ml b/cfrontend/C2C.ml index d389d0a8..e7d83377 100644 --- a/cfrontend/C2C.ml +++ b/cfrontend/C2C.ml @@ -441,7 +441,13 @@ let z_of_str hex str fst = let convertFloat f kind = let mant = z_of_str f.C.hex (f.C.intPart ^ f.C.fracPart) 0 in match mant with - | Z.Z0 -> Float.zero + | Z.Z0 -> + begin match kind with + | FFloat -> + Vsingle (Float.to_single Float.zero) + | FDouble | FLongDouble -> + Vfloat Float.zero + end | Z.Zpos mant -> let sgExp = match f.C.exp.[0] with '+' | '-' -> true | _ -> false in @@ -454,10 +460,10 @@ let convertFloat f kind = let base = P.of_int (if f.C.hex then 2 else 10) in begin match kind with - | FFloat -> - Float.build_from_parsed32 base mant exp - | FDouble | FLongDouble -> - Float.build_from_parsed64 base mant exp + | FFloat -> + Vsingle (Float32.from_parsed base mant exp) + | FDouble | FLongDouble -> + Vfloat (Float.from_parsed base mant exp) end | Z.Zneg _ -> assert false @@ -482,7 +488,7 @@ let rec convertExpr env e = | C.EConst(C.CFloat(f, k)) -> if k = C.FLongDouble && not !Clflags.option_flongdouble then unsupported "'long double' floating-point literal"; - Eval(Vfloat(convertFloat f k), ty) + Eval(convertFloat f k, ty) | C.EConst(C.CStr s) -> let ty = typeStringLiteral s in Evalof(Evar(name_for_string_literal env s, ty), ty) diff --git a/cfrontend/Cexec.v b/cfrontend/Cexec.v index b41902c9..eea1997e 100644 --- a/cfrontend/Cexec.v +++ b/cfrontend/Cexec.v @@ -104,7 +104,7 @@ Definition eventval_of_val (v: val) (t: typ) : option eventval := match v, t with | Vint i, AST.Tint => Some (EVint i) | Vfloat f, AST.Tfloat => Some (EVfloat f) - | Vfloat f, AST.Tsingle => if Float.is_single_dec f then Some (EVfloatsingle f) else None + | Vsingle f, AST.Tsingle => Some (EVsingle f) | Vlong n, AST.Tlong => Some (EVlong n) | Vptr b ofs, AST.Tint => do id <- Genv.invert_symbol ge b; Some (EVptr_global id ofs) | _, _ => None @@ -124,7 +124,7 @@ Definition val_of_eventval (ev: eventval) (t: typ) : option val := match ev, t with | EVint i, AST.Tint => Some (Vint i) | EVfloat f, AST.Tfloat => Some (Vfloat f) - | EVfloatsingle f, AST.Tsingle => if Float.is_single_dec f then Some (Vfloat f) else None + | EVsingle f, AST.Tsingle => Some (Vsingle f) | EVlong n, AST.Tlong => Some (Vlong n) | EVptr_global id ofs, AST.Tint => do b <- Genv.find_symbol ge id; Some (Vptr b ofs) | _, _ => None @@ -133,11 +133,7 @@ Definition val_of_eventval (ev: eventval) (t: typ) : option val := Lemma eventval_of_val_sound: forall v t ev, eventval_of_val v t = Some ev -> eventval_match ge ev t v. Proof. - intros. destruct v; destruct t; simpl in H; inv H. - constructor. - constructor. - constructor. - destruct (Float.is_single_dec f); inv H1. constructor; auto. + intros. destruct v; destruct t; simpl in H; inv H; try constructor. destruct (Genv.invert_symbol ge b) as [id|] eqn:?; inv H1. constructor. apply Genv.invert_find_symbol; auto. Qed. @@ -146,7 +142,6 @@ Lemma eventval_of_val_complete: forall ev t v, eventval_match ge ev t v -> eventval_of_val v t = Some ev. Proof. induction 1; simpl; auto. - rewrite pred_dec_true; auto. rewrite (Genv.find_invert_symbol _ _ H). auto. Qed. @@ -170,11 +165,7 @@ Qed. Lemma val_of_eventval_sound: forall ev t v, val_of_eventval ev t = Some v -> eventval_match ge ev t v. Proof. - intros. destruct ev; destruct t; simpl in H; inv H. - constructor. - constructor. - constructor. - destruct (Float.is_single_dec f); inv H1. constructor; auto. + intros. destruct ev; destruct t; simpl in H; inv H; try constructor. destruct (Genv.find_symbol ge i) as [b|] eqn:?; inv H1. constructor. auto. Qed. @@ -182,7 +173,7 @@ Qed. Lemma val_of_eventval_complete: forall ev t v, eventval_match ge ev t v -> val_of_eventval ev t = Some v. Proof. - induction 1; simpl; auto. rewrite pred_dec_true; auto. rewrite H; auto. + induction 1; simpl; auto. rewrite H; auto. Qed. (** Volatile memory accesses. *) diff --git a/cfrontend/Clight.v b/cfrontend/Clight.v index f2ba2405..d9fb6501 100644 --- a/cfrontend/Clight.v +++ b/cfrontend/Clight.v @@ -47,7 +47,8 @@ Require Import Cop. Inductive expr : Type := | Econst_int: int -> type -> expr (**r integer literal *) - | Econst_float: float -> type -> expr (**r float literal *) + | Econst_float: float -> type -> expr (**r double float literal *) + | Econst_single: float32 -> type -> expr (**r single float literal *) | Econst_long: int64 -> type -> expr (**r long integer literal *) | Evar: ident -> type -> expr (**r variable *) | Etempvar: ident -> type -> expr (**r temporary variable *) @@ -69,6 +70,7 @@ Definition typeof (e: expr) : type := match e with | Econst_int _ ty => ty | Econst_float _ ty => ty + | Econst_single _ ty => ty | Econst_long _ ty => ty | Evar _ ty => ty | Etempvar _ ty => ty @@ -352,6 +354,8 @@ Inductive eval_expr: expr -> val -> Prop := eval_expr (Econst_int i ty) (Vint i) | eval_Econst_float: forall f ty, eval_expr (Econst_float f ty) (Vfloat f) + | eval_Econst_single: forall f ty, + eval_expr (Econst_single f ty) (Vsingle f) | eval_Econst_long: forall i ty, eval_expr (Econst_long i ty) (Vlong i) | eval_Etempvar: forall id ty v, diff --git a/cfrontend/Cminorgen.v b/cfrontend/Cminorgen.v index 8ecf4989..ae6ec56b 100644 --- a/cfrontend/Cminorgen.v +++ b/cfrontend/Cminorgen.v @@ -70,6 +70,8 @@ Definition transl_constant (cst: Csharpminor.constant): constant := Ointconst n | Csharpminor.Ofloatconst n => Ofloatconst n + | Csharpminor.Osingleconst n => + Osingleconst n | Csharpminor.Olongconst n => Olongconst n end. diff --git a/cfrontend/Cminorgenproof.v b/cfrontend/Cminorgenproof.v index dba445df..3bf790ca 100644 --- a/cfrontend/Cminorgenproof.v +++ b/cfrontend/Cminorgenproof.v @@ -1299,8 +1299,15 @@ Proof. inv H; inv H0; simpl; TrivialExists. inv H; inv H0; simpl; TrivialExists. inv H; inv H0; simpl; TrivialExists. - inv H0; simpl in H; inv H. simpl. destruct (Float.intoffloat f0); simpl in *; inv H1. TrivialExists. - inv H0; simpl in H; inv H. simpl. destruct (Float.intuoffloat f0); simpl in *; inv H1. TrivialExists. + inv H; inv H0; simpl; TrivialExists. + inv H; inv H0; simpl; TrivialExists. + inv H; inv H0; simpl; TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float.to_int f0); simpl in *; inv H1. TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float.to_intu f0); simpl in *; inv H1. TrivialExists. + inv H0; simpl in H; inv H. simpl. TrivialExists. + inv H0; simpl in H; inv H. simpl. TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float32.to_int f0); simpl in *; inv H1. TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float32.to_intu f0); simpl in *; inv H1. TrivialExists. inv H0; simpl in H; inv H. simpl. TrivialExists. inv H0; simpl in H; inv H. simpl. TrivialExists. inv H; inv H0; simpl; TrivialExists. @@ -1308,10 +1315,12 @@ Proof. inv H; inv H0; simpl; TrivialExists. inv H; inv H0; simpl; TrivialExists. inv H; inv H0; simpl; TrivialExists. - inv H0; simpl in H; inv H. simpl. destruct (Float.longoffloat f0); simpl in *; inv H1. TrivialExists. - inv H0; simpl in H; inv H. simpl. destruct (Float.longuoffloat f0); simpl in *; inv H1. TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float.to_long f0); simpl in *; inv H1. TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float.to_longu f0); simpl in *; inv H1. TrivialExists. inv H0; simpl in H; inv H. simpl. TrivialExists. inv H0; simpl in H; inv H. simpl. TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float32.to_long f0); simpl in *; inv H1. TrivialExists. + inv H0; simpl in H; inv H. simpl. destruct (Float32.to_longu f0); simpl in *; inv H1. TrivialExists. inv H0; simpl in H; inv H. simpl. TrivialExists. inv H0; simpl in H; inv H. simpl. TrivialExists. Qed. @@ -1361,6 +1370,10 @@ Proof. inv H; inv H0; inv H1; TrivialExists. inv H; inv H0; inv H1; TrivialExists. inv H; inv H0; inv H1; TrivialExists. + inv H; inv H0; inv H1; TrivialExists. + inv H; inv H0; inv H1; TrivialExists. + inv H; inv H0; inv H1; TrivialExists. + inv H; inv H0; inv H1; TrivialExists. inv H0; try discriminate; inv H1; try discriminate. simpl in *. destruct (Int64.eq i0 Int64.zero || Int64.eq i (Int64.repr Int64.min_signed) && Int64.eq i0 Int64.mone); inv H; TrivialExists. @@ -1392,6 +1405,8 @@ Proof. simpl; auto. (* cmpf *) inv H; inv H0; inv H1; TrivialExists. apply val_inject_val_of_optbool. +(* cmpfs *) + inv H; inv H0; inv H1; TrivialExists. apply val_inject_val_of_optbool. (* cmpl *) unfold Val.cmpl in *. inv H0; inv H1; simpl in H; inv H. econstructor; split. simpl; eauto. apply val_inject_val_of_bool. @@ -1470,6 +1485,7 @@ Proof. destruct cst; simpl; intros; inv H. exists (Vint i); auto. exists (Vfloat f0); auto. + exists (Vsingle f0); auto. exists (Vlong i); auto. Qed. diff --git a/cfrontend/Cop.v b/cfrontend/Cop.v index 83fe7726..ff43cbd5 100644 --- a/cfrontend/Cop.v +++ b/cfrontend/Cop.v @@ -74,17 +74,25 @@ Inductive incr_or_decr : Type := Incr | Decr. Inductive classify_cast_cases : Type := | cast_case_neutral (**r int|pointer -> int32|pointer *) | cast_case_i2i (sz2:intsize) (si2:signedness) (**r int -> int *) - | cast_case_f2f (sz2:floatsize) (**r float -> float *) - | cast_case_i2f (si1:signedness) (sz2:floatsize) (**r int -> float *) - | cast_case_f2i (sz2:intsize) (si2:signedness) (**r float -> int *) + | cast_case_f2f (**r double -> double *) + | cast_case_s2s (**r single -> single *) + | cast_case_f2s (**r double -> single *) + | cast_case_s2f (**r single -> double *) + | cast_case_i2f (si1: signedness) (**r int -> double *) + | cast_case_i2s (si1: signedness) (**r int -> single *) + | cast_case_f2i (sz2:intsize) (si2:signedness) (**r double -> int *) + | cast_case_s2i (sz2:intsize) (si2:signedness) (**r single -> int *) | cast_case_l2l (**r long -> long *) | cast_case_i2l (si1: signedness) (**r int -> long *) | cast_case_l2i (sz2: intsize) (si2: signedness) (**r long -> int *) - | cast_case_l2f (si1: signedness) (sz2: floatsize) (**r long -> float *) - | cast_case_f2l (si2: signedness) (**r float -> long *) - | cast_case_f2bool (**r float -> bool *) - | cast_case_l2bool (**r long -> bool *) - | cast_case_p2bool (**r pointer -> bool *) + | cast_case_l2f (si1: signedness) (**r long -> double *) + | cast_case_l2s (si1: signedness) (**r long -> single *) + | cast_case_f2l (si2:signedness) (**r double -> long *) + | cast_case_s2l (si2:signedness) (**r single -> long *) + | cast_case_f2bool (**r double -> bool *) + | cast_case_s2bool (**r single -> bool *) + | cast_case_l2bool (**r long -> bool *) + | cast_case_p2bool (**r pointer -> bool *) | cast_case_struct (id1: ident) (fld1: fieldlist) (id2: ident) (fld2: fieldlist) (**r struct -> struct *) | cast_case_union (id1: ident) (fld1: fieldlist) (id2: ident) (fld2: fieldlist) (**r union -> union *) | cast_case_void (**r any -> void *) @@ -93,19 +101,27 @@ Inductive classify_cast_cases : Type := Definition classify_cast (tfrom tto: type) : classify_cast_cases := match tto, tfrom with | Tint I32 si2 _, (Tint _ _ _ | Tpointer _ _ | Tcomp_ptr _ _ | Tarray _ _ _ | Tfunction _ _ _) => cast_case_neutral - | Tint IBool _ _, Tfloat _ _ => cast_case_f2bool + | Tint IBool _ _, Tfloat F64 _ => cast_case_f2bool + | Tint IBool _ _, Tfloat F32 _ => cast_case_s2bool | Tint IBool _ _, (Tpointer _ _ | Tcomp_ptr _ _ | Tarray _ _ _ | Tfunction _ _ _) => cast_case_p2bool | Tint sz2 si2 _, Tint sz1 si1 _ => cast_case_i2i sz2 si2 - | Tint sz2 si2 _, Tfloat sz1 _ => cast_case_f2i sz2 si2 - | Tfloat sz2 _, Tfloat sz1 _ => cast_case_f2f sz2 - | Tfloat sz2 _, Tint sz1 si1 _ => cast_case_i2f si1 sz2 + | Tint sz2 si2 _, Tfloat F64 _ => cast_case_f2i sz2 si2 + | Tint sz2 si2 _, Tfloat F32 _ => cast_case_s2i sz2 si2 + | Tfloat F64 _, Tfloat F64 _ => cast_case_f2f + | Tfloat F32 _, Tfloat F32 _ => cast_case_s2s + | Tfloat F64 _, Tfloat F32 _ => cast_case_s2f + | Tfloat F32 _, Tfloat F64 _ => cast_case_f2s + | Tfloat F64 _, Tint sz1 si1 _ => cast_case_i2f si1 + | Tfloat F32 _, Tint sz1 si1 _ => cast_case_i2s si1 | (Tpointer _ _ | Tcomp_ptr _ _), (Tint _ _ _ | Tpointer _ _ | Tcomp_ptr _ _ | Tarray _ _ _ | Tfunction _ _ _) => cast_case_neutral | Tlong _ _, Tlong _ _ => cast_case_l2l | Tlong _ _, Tint sz1 si1 _ => cast_case_i2l si1 | Tint IBool _ _, Tlong _ _ => cast_case_l2bool | Tint sz2 si2 _, Tlong _ _ => cast_case_l2i sz2 si2 - | Tlong si2 _, Tfloat sz1 _ => cast_case_f2l si2 - | Tfloat sz2 _, Tlong si1 _ => cast_case_l2f si1 sz2 + | Tlong si2 _, Tfloat F64 _ => cast_case_f2l si2 + | Tlong si2 _, Tfloat F32 _ => cast_case_s2l si2 + | Tfloat F64 _, Tlong si1 _ => cast_case_l2f si1 + | Tfloat F32 _, Tlong si1 _ => cast_case_l2s si1 | (Tpointer _ _ | Tcomp_ptr _ _), Tlong _ _ => cast_case_l2i I32 Unsigned | Tlong si2 _, (Tpointer _ _ | Tcomp_ptr _ _ | Tarray _ _ _ | Tfunction _ _ _) => cast_case_i2l si2 | Tstruct id2 fld2 _, Tstruct id1 fld1 _ => cast_case_struct id1 fld1 id2 fld2 @@ -128,24 +144,28 @@ Definition cast_int_int (sz: intsize) (sg: signedness) (i: int) : int := | IBool, _ => if Int.eq i Int.zero then Int.zero else Int.one end. -Definition cast_int_float (si: signedness) (sz: floatsize) (i: int) : float := - match si, sz with - | Signed, F64 => Float.floatofint i - | Unsigned, F64 => Float.floatofintu i - | Signed, F32 => Float.singleofint i - | Unsigned, F32 => Float.singleofintu i +Definition cast_int_float (si: signedness) (i: int) : float := + match si with + | Signed => Float.of_int i + | Unsigned => Float.of_intu i end. Definition cast_float_int (si : signedness) (f: float) : option int := match si with - | Signed => Float.intoffloat f - | Unsigned => Float.intuoffloat f + | Signed => Float.to_int f + | Unsigned => Float.to_intu f end. -Definition cast_float_float (sz: floatsize) (f: float) : float := - match sz with - | F32 => Float.singleoffloat f - | F64 => f +Definition cast_int_single (si: signedness) (i: int) : float32 := + match si with + | Signed => Float32.of_int i + | Unsigned => Float32.of_intu i + end. + +Definition cast_single_int (si : signedness) (f: float32) : option int := + match si with + | Signed => Float32.to_int f + | Unsigned => Float32.to_intu f end. Definition cast_int_long (si: signedness) (i: int) : int64 := @@ -154,18 +174,28 @@ Definition cast_int_long (si: signedness) (i: int) : int64 := | Unsigned => Int64.repr (Int.unsigned i) end. -Definition cast_long_float (si: signedness) (sz: floatsize) (i: int64) : float := - match si, sz with - | Signed, F64 => Float.floatoflong i - | Unsigned, F64 => Float.floatoflongu i - | Signed, F32 => Float.singleoflong i - | Unsigned, F32 => Float.singleoflongu i +Definition cast_long_float (si: signedness) (i: int64) : float := + match si with + | Signed => Float.of_long i + | Unsigned => Float.of_longu i + end. + +Definition cast_long_single (si: signedness) (i: int64) : float32 := + match si with + | Signed => Float32.of_long i + | Unsigned => Float32.of_longu i end. Definition cast_float_long (si : signedness) (f: float) : option int64 := match si with - | Signed => Float.longoffloat f - | Unsigned => Float.longuoffloat f + | Signed => Float.to_long f + | Unsigned => Float.to_longu f + end. + +Definition cast_single_long (si : signedness) (f: float32) : option int64 := + match si with + | Signed => Float32.to_long f + | Unsigned => Float32.to_longu f end. Definition sem_cast (v: val) (t1 t2: type) : option val := @@ -180,14 +210,34 @@ Definition sem_cast (v: val) (t1 t2: type) : option val := | Vint i => Some (Vint (cast_int_int sz2 si2 i)) | _ => None end - | cast_case_f2f sz2 => + | cast_case_f2f => + match v with + | Vfloat f => Some (Vfloat f) + | _ => None + end + | cast_case_s2s => + match v with + | Vsingle f => Some (Vsingle f) + | _ => None + end + | cast_case_s2f => match v with - | Vfloat f => Some (Vfloat (cast_float_float sz2 f)) + | Vsingle f => Some (Vfloat (Float.of_single f)) | _ => None end - | cast_case_i2f si1 sz2 => + | cast_case_f2s => match v with - | Vint i => Some (Vfloat (cast_int_float si1 sz2 i)) + | Vfloat f => Some (Vsingle (Float.to_single f)) + | _ => None + end + | cast_case_i2f si1 => + match v with + | Vint i => Some (Vfloat (cast_int_float si1 i)) + | _ => None + end + | cast_case_i2s si1 => + match v with + | Vint i => Some (Vsingle (cast_int_single si1 i)) | _ => None end | cast_case_f2i sz2 si2 => @@ -199,12 +249,27 @@ Definition sem_cast (v: val) (t1 t2: type) : option val := end | _ => None end + | cast_case_s2i sz2 si2 => + match v with + | Vsingle f => + match cast_single_int si2 f with + | Some i => Some (Vint (cast_int_int sz2 si2 i)) + | None => None + end + | _ => None + end | cast_case_f2bool => match v with | Vfloat f => Some(Vint(if Float.cmp Ceq f Float.zero then Int.zero else Int.one)) | _ => None end + | cast_case_s2bool => + match v with + | Vsingle f => + Some(Vint(if Float32.cmp Ceq f Float32.zero then Int.zero else Int.one)) + | _ => None + end | cast_case_p2bool => match v with | Vint i => Some (Vint (cast_int_int IBool Signed i)) @@ -232,9 +297,14 @@ Definition sem_cast (v: val) (t1 t2: type) : option val := Some(Vint(if Int64.eq n Int64.zero then Int.zero else Int.one)) | _ => None end - | cast_case_l2f si1 sz2 => + | cast_case_l2f si1 => + match v with + | Vlong i => Some (Vfloat (cast_long_float si1 i)) + | _ => None + end + | cast_case_l2s si1 => match v with - | Vlong i => Some (Vfloat (cast_long_float si1 sz2 i)) + | Vlong i => Some (Vsingle (cast_long_single si1 i)) | _ => None end | cast_case_f2l si2 => @@ -246,6 +316,15 @@ Definition sem_cast (v: val) (t1 t2: type) : option val := end | _ => None end + | cast_case_s2l si2 => + match v with + | Vsingle f => + match cast_single_long si2 f with + | Some i => Some (Vlong i) + | None => None + end + | _ => None + end | cast_case_struct id1 fld1 id2 fld2 => match v with | Vptr b ofs => @@ -271,7 +350,8 @@ Definition sem_cast (v: val) (t1 t2: type) : option val := Inductive classify_bool_cases : Type := | bool_case_i (**r integer *) - | bool_case_f (**r float *) + | bool_case_f (**r double float *) + | bool_case_s (**r single float *) | bool_case_p (**r pointer *) | bool_case_l (**r long *) | bool_default. @@ -280,7 +360,8 @@ Definition classify_bool (ty: type) : classify_bool_cases := match typeconv ty with | Tint _ _ _ => bool_case_i | Tpointer _ _ | Tcomp_ptr _ _ => bool_case_p - | Tfloat _ _ => bool_case_f + | Tfloat F64 _ => bool_case_f + | Tfloat F32 _ => bool_case_s | Tlong _ _ => bool_case_l | _ => bool_default end. @@ -302,6 +383,11 @@ Definition bool_val (v: val) (t: type) : option bool := | Vfloat f => Some (negb (Float.cmp Ceq f Float.zero)) | _ => None end + | bool_case_s => + match v with + | Vsingle f => Some (negb (Float32.cmp Ceq f Float32.zero)) + | _ => None + end | bool_case_p => match v with | Vint n => Some (negb (Int.eq n Int.zero)) @@ -333,6 +419,11 @@ Definition sem_notbool (v: val) (ty: type) : option val := | Vfloat f => Some (Val.of_bool (Float.cmp Ceq f Float.zero)) | _ => None end + | bool_case_s => + match v with + | Vsingle f => Some (Val.of_bool (Float32.cmp Ceq f Float32.zero)) + | _ => None + end | bool_case_p => match v with | Vint n => Some (Val.of_bool (Int.eq n Int.zero)) @@ -351,7 +442,8 @@ Definition sem_notbool (v: val) (ty: type) : option val := Inductive classify_neg_cases : Type := | neg_case_i(s: signedness) (**r int *) - | neg_case_f(sz: floatsize) (**r float *) + | neg_case_f (**r double float *) + | neg_case_s (**r single float *) | neg_case_l(s: signedness) (**r long *) | neg_default. @@ -359,7 +451,8 @@ Definition classify_neg (ty: type) : classify_neg_cases := match ty with | Tint I32 Unsigned _ => neg_case_i Unsigned | Tint _ _ _ => neg_case_i Signed - | Tfloat sz _ => neg_case_f sz + | Tfloat F64 _ => neg_case_f + | Tfloat F32 _ => neg_case_s | Tlong si _ => neg_case_l si | _ => neg_default end. @@ -371,11 +464,16 @@ Definition sem_neg (v: val) (ty: type) : option val := | Vint n => Some (Vint (Int.neg n)) | _ => None end - | neg_case_f sz => + | neg_case_f => match v with | Vfloat f => Some (Vfloat (Float.neg f)) | _ => None end + | neg_case_s => + match v with + | Vsingle f => Some (Vsingle (Float32.neg f)) + | _ => None + end | neg_case_l sg => match v with | Vlong n => Some (Vlong (Int64.neg n)) @@ -388,17 +486,22 @@ Definition sem_absfloat (v: val) (ty: type) : option val := match classify_neg ty with | neg_case_i sg => match v with - | Vint n => Some (Vfloat (Float.abs (cast_int_float sg F64 n))) + | Vint n => Some (Vfloat (Float.abs (cast_int_float sg n))) | _ => None end - | neg_case_f sz => + | neg_case_f => match v with | Vfloat f => Some (Vfloat (Float.abs f)) | _ => None end + | neg_case_s => + match v with + | Vsingle f => Some (Vfloat (Float.abs (Float.of_single f))) + | _ => None + end | neg_case_l sg => match v with - | Vlong n => Some (Vfloat (Float.abs (cast_long_float sg F64 n))) + | Vlong n => Some (Vfloat (Float.abs (cast_long_float sg n))) | _ => None end | neg_default => None @@ -446,7 +549,8 @@ Definition sem_notint (v: val) (ty: type): option val := Inductive binarith_cases: Type := | bin_case_i (s: signedness) (**r at int type *) | bin_case_l (s: signedness) (**r at long int type *) - | bin_case_f (sz: floatsize) (**r at float type *) + | bin_case_f (**r at double float type *) + | bin_case_s (**r at single float type *) | bin_default. (**r error *) Definition classify_binarith (ty1: type) (ty2: type) : binarith_cases := @@ -458,31 +562,24 @@ Definition classify_binarith (ty1: type) (ty2: type) : binarith_cases := | Tlong _ _, Tlong _ _ => bin_case_l Unsigned | Tlong sg _, Tint _ _ _ => bin_case_l sg | Tint _ _ _, Tlong sg _ => bin_case_l sg - | Tfloat F32 _, Tfloat F32 _ => bin_case_f F32 - | Tfloat _ _, Tfloat _ _ => bin_case_f F64 - | Tfloat sz _, (Tint _ _ _ | Tlong _ _) => bin_case_f sz - | (Tint _ _ _ | Tlong _ _), Tfloat sz _ => bin_case_f sz + | Tfloat F32 _, Tfloat F32 _ => bin_case_s + | Tfloat _ _, Tfloat _ _ => bin_case_f + | Tfloat F64 _, (Tint _ _ _ | Tlong _ _) => bin_case_f + | (Tint _ _ _ | Tlong _ _), Tfloat F64 _ => bin_case_f + | Tfloat F32 _, (Tint _ _ _ | Tlong _ _) => bin_case_s + | (Tint _ _ _ | Tlong _ _), Tfloat F32 _ => bin_case_s | _, _ => bin_default end. -(** The static type of the result. *) - -Definition binarith_result_type (c: binarith_cases) : option type := - match c with - | bin_case_i sg => Some(Tint I32 sg noattr) - | bin_case_l sg => Some(Tlong sg noattr) - | bin_case_f sz => Some(Tfloat sz noattr) - | bin_default => None - end. - -(** The type at which the computation is done. Both arguments are - converted to this type before the actual computation. *) +(** The static type of the result. Both arguments are converted to this type + before the actual computation. *) Definition binarith_type (c: binarith_cases) : type := match c with | bin_case_i sg => Tint I32 sg noattr | bin_case_l sg => Tlong sg noattr - | bin_case_f sz => Tfloat F64 noattr + | bin_case_f => Tfloat F64 noattr + | bin_case_s => Tfloat F32 noattr | bin_default => Tvoid end. @@ -490,6 +587,7 @@ Definition sem_binarith (sem_int: signedness -> int -> int -> option val) (sem_long: signedness -> int64 -> int64 -> option val) (sem_float: float -> float -> option val) + (sem_single: float32 -> float32 -> option val) (v1: val) (t1: type) (v2: val) (t2: type) : option val := let c := classify_binarith t1 t2 in let t := binarith_type c in @@ -505,11 +603,16 @@ Definition sem_binarith | Vint n1, Vint n2 => sem_int sg n1 n2 | _, _ => None end - | bin_case_f sz => + | bin_case_f => match v1', v2' with | Vfloat n1, Vfloat n2 => sem_float n1 n2 | _, _ => None end + | bin_case_s => + match v1', v2' with + | Vsingle n1, Vsingle n2 => sem_single n1 n2 + | _, _ => None + end | bin_case_l sg => match v1', v2' with | Vlong n1, Vlong n2 => sem_long sg n1 n2 @@ -569,6 +672,7 @@ Definition sem_add (v1:val) (t1:type) (v2: val) (t2:type) : option val := (fun sg n1 n2 => Some(Vint(Int.add n1 n2))) (fun sg n1 n2 => Some(Vlong(Int64.add n1 n2))) (fun n1 n2 => Some(Vfloat(Float.add n1 n2))) + (fun n1 n2 => Some(Vsingle(Float32.add n1 n2))) v1 t1 v2 t2 end. @@ -617,6 +721,7 @@ Definition sem_sub (v1:val) (t1:type) (v2: val) (t2:type) : option val := (fun sg n1 n2 => Some(Vint(Int.sub n1 n2))) (fun sg n1 n2 => Some(Vlong(Int64.sub n1 n2))) (fun n1 n2 => Some(Vfloat(Float.sub n1 n2))) + (fun n1 n2 => Some(Vsingle(Float32.sub n1 n2))) v1 t1 v2 t2 end. @@ -627,6 +732,7 @@ Definition sem_mul (v1:val) (t1:type) (v2: val) (t2:type) : option val := (fun sg n1 n2 => Some(Vint(Int.mul n1 n2))) (fun sg n1 n2 => Some(Vlong(Int64.mul n1 n2))) (fun n1 n2 => Some(Vfloat(Float.mul n1 n2))) + (fun n1 n2 => Some(Vsingle(Float32.mul n1 n2))) v1 t1 v2 t2. Definition sem_div (v1:val) (t1:type) (v2: val) (t2:type) : option val := @@ -652,6 +758,7 @@ Definition sem_div (v1:val) (t1:type) (v2: val) (t2:type) : option val := then None else Some(Vlong(Int64.divu n1 n2)) end) (fun n1 n2 => Some(Vfloat(Float.div n1 n2))) + (fun n1 n2 => Some(Vsingle(Float32.div n1 n2))) v1 t1 v2 t2. Definition sem_mod (v1:val) (t1:type) (v2: val) (t2:type) : option val := @@ -677,6 +784,7 @@ Definition sem_mod (v1:val) (t1:type) (v2: val) (t2:type) : option val := then None else Some(Vlong(Int64.modu n1 n2)) end) (fun n1 n2 => None) + (fun n1 n2 => None) v1 t1 v2 t2. Definition sem_and (v1:val) (t1:type) (v2: val) (t2:type) : option val := @@ -684,6 +792,7 @@ Definition sem_and (v1:val) (t1:type) (v2: val) (t2:type) : option val := (fun sg n1 n2 => Some(Vint(Int.and n1 n2))) (fun sg n1 n2 => Some(Vlong(Int64.and n1 n2))) (fun n1 n2 => None) + (fun n1 n2 => None) v1 t1 v2 t2. Definition sem_or (v1:val) (t1:type) (v2: val) (t2:type) : option val := @@ -691,6 +800,7 @@ Definition sem_or (v1:val) (t1:type) (v2: val) (t2:type) : option val := (fun sg n1 n2 => Some(Vint(Int.or n1 n2))) (fun sg n1 n2 => Some(Vlong(Int64.or n1 n2))) (fun n1 n2 => None) + (fun n1 n2 => None) v1 t1 v2 t2. Definition sem_xor (v1:val) (t1:type) (v2: val) (t2:type) : option val := @@ -698,6 +808,7 @@ Definition sem_xor (v1:val) (t1:type) (v2: val) (t2:type) : option val := (fun sg n1 n2 => Some(Vint(Int.xor n1 n2))) (fun sg n1 n2 => Some(Vlong(Int64.xor n1 n2))) (fun n1 n2 => None) + (fun n1 n2 => None) v1 t1 v2 t2. (** *** Shifts *) @@ -818,6 +929,8 @@ Definition sem_cmp (c:comparison) Some(Val.of_bool(match sg with Signed => Int64.cmp c n1 n2 | Unsigned => Int64.cmpu c n1 n2 end))) (fun n1 n2 => Some(Val.of_bool(Float.cmp c n1 n2))) + (fun n1 n2 => + Some(Val.of_bool(Float32.cmp c n1 n2))) v1 t1 v2 t2 end. @@ -946,7 +1059,9 @@ Proof. inv H0; inv H; TrivialInject. - econstructor; eauto. - destruct (cast_float_int si2 f0); inv H1; TrivialInject. +- destruct (cast_single_int si2 f0); inv H1; TrivialInject. - destruct (cast_float_long si2 f0); inv H1; TrivialInject. +- destruct (cast_single_long si2 f0); inv H1; TrivialInject. - destruct (ident_eq id1 id2 && fieldlist_eq fld1 fld2); inv H2; TrivialInject. econstructor; eauto. - destruct (ident_eq id1 id2 && fieldlist_eq fld1 fld2); inv H2; TrivialInject. econstructor; eauto. - econstructor; eauto. @@ -976,18 +1091,19 @@ Definition optval_self_injects (ov: option val) : Prop := end. Remark sem_binarith_inject: - forall sem_int sem_long sem_float v1 t1 v2 t2 v v1' v2', - sem_binarith sem_int sem_long sem_float v1 t1 v2 t2 = Some v -> + forall sem_int sem_long sem_float sem_single v1 t1 v2 t2 v v1' v2', + sem_binarith sem_int sem_long sem_float sem_single v1 t1 v2 t2 = Some v -> val_inject f v1 v1' -> val_inject f v2 v2' -> (forall sg n1 n2, optval_self_injects (sem_int sg n1 n2)) -> (forall sg n1 n2, optval_self_injects (sem_long sg n1 n2)) -> (forall n1 n2, optval_self_injects (sem_float n1 n2)) -> - exists v', sem_binarith sem_int sem_long sem_float v1' t1 v2' t2 = Some v' /\ val_inject f v v'. + (forall n1 n2, optval_self_injects (sem_single n1 n2)) -> + exists v', sem_binarith sem_int sem_long sem_float sem_single v1' t1 v2' t2 = Some v' /\ val_inject f v v'. Proof. intros. assert (SELF: forall ov v, ov = Some v -> optval_self_injects ov -> val_inject f v v). { - intros. subst ov; simpl in H6. destruct v0; contradiction || constructor. + intros. subst ov; simpl in H7. destruct v0; contradiction || constructor. } unfold sem_binarith in *. set (c := classify_binarith t1 t2) in *. @@ -1093,6 +1209,7 @@ Proof. || Int64.eq n1 (Int64.repr Int64.min_signed) && Int64.eq n2 Int64.mone); exact I. destruct (Int64.eq n2 Int64.zero); exact I. exact I. + exact I. - (* mod *) eapply sem_binarith_inject; eauto; intros. destruct sg. @@ -1104,6 +1221,7 @@ Proof. || Int64.eq n1 (Int64.repr Int64.min_signed) && Int64.eq n2 Int64.mone); exact I. destruct (Int64.eq n2 Int64.zero); exact I. exact I. + exact I. - (* and *) eapply sem_binarith_inject; eauto; intros; exact I. - (* or *) @@ -1168,7 +1286,8 @@ Proof. match t with | Tint _ _ _ => bool_case_i | Tpointer _ _ | Tcomp_ptr _ _ | Tarray _ _ _ | Tfunction _ _ _ => bool_case_p - | Tfloat _ _ => bool_case_f + | Tfloat F64 _ => bool_case_f + | Tfloat F32 _ => bool_case_s | Tlong _ _ => bool_case_l | _ => bool_default end). @@ -1178,7 +1297,14 @@ Proof. unfold bool_val. rewrite A. unfold sem_cast. destruct t; simpl; auto; destruct v; auto. destruct (Int.eq i0 Int.zero); auto. destruct (Int64.eq i Int64.zero); auto. + destruct f; auto. + destruct f; auto. + destruct f; auto. + destruct f; auto. destruct (Float.cmp Ceq f0 Float.zero); auto. + destruct f; auto. + destruct (Float32.cmp Ceq f0 Float32.zero); auto. + destruct f; auto. destruct (Int.eq i Int.zero); auto. destruct (Int.eq i Int.zero); auto. destruct (Int.eq i Int.zero); auto. @@ -1360,8 +1486,8 @@ Qed. Lemma classify_binarith_arithmetic_conversion: forall t1 t2, - binarith_result_type (classify_binarith (proj_type t1) (proj_type t2)) = - Some (proj_type (usual_arithmetic_conversion t1 t2)). + binarith_type (classify_binarith (proj_type t1) (proj_type t2)) = + proj_type (usual_arithmetic_conversion t1 t2). Proof. destruct t1; destruct t2; try reflexivity. - destruct it; destruct it0; reflexivity. diff --git a/cfrontend/Csharpminor.v b/cfrontend/Csharpminor.v index 4b2e915c..d37fa819 100644 --- a/cfrontend/Csharpminor.v +++ b/cfrontend/Csharpminor.v @@ -33,7 +33,8 @@ Require Import Smallstep. Inductive constant : Type := | Ointconst: int -> constant (**r integer constant *) - | Ofloatconst: float -> constant (**r floating-point constant *) + | Ofloatconst: float -> constant (**r double-precision floating-point constant *) + | Osingleconst: float32 -> constant (**r single-precision floating-point constant *) | Olongconst: int64 -> constant. (**r long integer constant *) Definition unary_operation : Type := Cminor.unary_operation. @@ -253,6 +254,7 @@ Definition eval_constant (cst: constant) : option val := match cst with | Ointconst n => Some (Vint n) | Ofloatconst n => Some (Vfloat n) + | Osingleconst n => Some (Vsingle n) | Olongconst n => Some (Vlong n) end. diff --git a/cfrontend/Cshmgen.v b/cfrontend/Cshmgen.v index 685fa71b..cedfd8bf 100644 --- a/cfrontend/Cshmgen.v +++ b/cfrontend/Cshmgen.v @@ -50,16 +50,25 @@ Open Local Scope error_monad_scope. Definition make_intconst (n: int) := Econst (Ointconst n). +Definition make_longconst (f: int64) := Econst (Olongconst f). + Definition make_floatconst (f: float) := Econst (Ofloatconst f). -Definition make_longconst (f: int64) := Econst (Olongconst f). +Definition make_singleconst (f: float32) := Econst (Osingleconst f). -Definition make_floatofint (e: expr) (sg: signedness) (sz: floatsize) := - match sg, sz with - | Signed, F64 => Eunop Ofloatofint e - | Unsigned, F64 => Eunop Ofloatofintu e - | Signed, F32 => Eunop Osingleoffloat (Eunop Ofloatofint e) - | Unsigned, F32 => Eunop Osingleoffloat (Eunop Ofloatofintu e) +Definition make_singleoffloat (e: expr) := Eunop Osingleoffloat e. +Definition make_floatofsingle (e: expr) := Eunop Ofloatofsingle e. + +Definition make_floatofint (e: expr) (sg: signedness) := + match sg with + | Signed => Eunop Ofloatofint e + | Unsigned => Eunop Ofloatofintu e + end. + +Definition make_singleofint (e: expr) (sg: signedness) := + match sg with + | Signed => Eunop Osingleofint e + | Unsigned => Eunop Osingleofintu e end. Definition make_intoffloat (e: expr) (sg: signedness) := @@ -68,18 +77,28 @@ Definition make_intoffloat (e: expr) (sg: signedness) := | Unsigned => Eunop Ointuoffloat e end. +Definition make_intofsingle (e: expr) (sg: signedness) := + match sg with + | Signed => Eunop Ointofsingle e + | Unsigned => Eunop Ointuofsingle e + end. + Definition make_longofint (e: expr) (sg: signedness) := match sg with | Signed => Eunop Olongofint e | Unsigned => Eunop Olongofintu e end. -Definition make_floatoflong (e: expr) (sg: signedness) (sz: floatsize) := - match sg, sz with - | Signed, F64 => Eunop Ofloatoflong e - | Unsigned, F64 => Eunop Ofloatoflongu e - | Signed, F32 => Eunop Osingleoflong e - | Unsigned, F32 => Eunop Osingleoflongu e +Definition make_floatoflong (e: expr) (sg: signedness) := + match sg with + | Signed => Eunop Ofloatoflong e + | Unsigned => Eunop Ofloatoflongu e + end. + +Definition make_singleoflong (e: expr) (sg: signedness) := + match sg with + | Signed => Eunop Osingleoflong e + | Unsigned => Eunop Osingleoflongu e end. Definition make_longoffloat (e: expr) (sg: signedness) := @@ -88,11 +107,18 @@ Definition make_longoffloat (e: expr) (sg: signedness) := | Unsigned => Eunop Olonguoffloat e end. +Definition make_longofsingle (e: expr) (sg: signedness) := + match sg with + | Signed => Eunop Olongofsingle e + | Unsigned => Eunop Olonguofsingle e + end. + Definition make_cmp_ne_zero (e: expr) := match e with | Ebinop (Ocmp c) e1 e2 => e | Ebinop (Ocmpu c) e1 e2 => e | Ebinop (Ocmpf c) e1 e2 => e + | Ebinop (Ocmpfs c) e1 e2 => e | Ebinop (Ocmpl c) e1 e2 => e | Ebinop (Ocmplu c) e1 e2 => e | _ => Ebinop (Ocmp Cne) e (make_intconst Int.zero) @@ -111,25 +137,27 @@ Definition make_cast_int (e: expr) (sz: intsize) (si: signedness) := | IBool, _ => make_cmp_ne_zero e end. -Definition make_cast_float (e: expr) (sz: floatsize) := - match sz with - | F32 => Eunop Osingleoffloat e - | F64 => e - end. - Definition make_cast (from to: type) (e: expr) := match classify_cast from to with | cast_case_neutral => OK e | cast_case_i2i sz2 si2 => OK (make_cast_int e sz2 si2) - | cast_case_f2f sz2 => OK (make_cast_float e sz2) - | cast_case_i2f si1 sz2 => OK (make_floatofint e si1 sz2) + | cast_case_f2f => OK e + | cast_case_s2s => OK e + | cast_case_f2s => OK (make_singleoffloat e) + | cast_case_s2f => OK (make_floatofsingle e) + | cast_case_i2f si1 => OK (make_floatofint e si1) + | cast_case_i2s si1 => OK (make_singleofint e si1) | cast_case_f2i sz2 si2 => OK (make_cast_int (make_intoffloat e si2) sz2 si2) + | cast_case_s2i sz2 si2 => OK (make_cast_int (make_intofsingle e si2) sz2 si2) | cast_case_l2l => OK e | cast_case_i2l si1 => OK (make_longofint e si1) | cast_case_l2i sz2 si2 => OK (make_cast_int (Eunop Ointoflong e) sz2 si2) - | cast_case_l2f si1 sz2 => OK (make_floatoflong e si1 sz2) + | cast_case_l2f si1 => OK (make_floatoflong e si1) + | cast_case_l2s si1 => OK (make_singleoflong e si1) | cast_case_f2l si2 => OK (make_longoffloat e si2) + | cast_case_s2l si2 => OK (make_longofsingle e si2) | cast_case_f2bool => OK (Ebinop (Ocmpf Cne) e (make_floatconst Float.zero)) + | cast_case_s2bool => OK (Ebinop (Ocmpfs Cne) e (make_singleconst Float32.zero)) | cast_case_l2bool => OK (Ebinop (Ocmpl Cne) e (make_longconst Int64.zero)) | cast_case_p2bool => OK (Ebinop (Ocmpu Cne) e (make_intconst Int.zero)) | cast_case_struct id1 fld1 id2 fld2 => OK e @@ -145,6 +173,7 @@ Definition make_boolean (e: expr) (ty: type) := match classify_bool ty with | bool_case_i => make_cmp_ne_zero e | bool_case_f => Ebinop (Ocmpf Cne) e (make_floatconst Float.zero) + | bool_case_s => Ebinop (Ocmpfs Cne) e (make_singleconst Float32.zero) | bool_case_p => Ebinop (Ocmpu Cne) e (make_intconst Int.zero) | bool_case_l => Ebinop (Ocmpl Cne) e (make_longconst Int64.zero) | bool_default => e (**r should not happen *) @@ -156,6 +185,7 @@ Definition make_notbool (e: expr) (ty: type) := match classify_bool ty with | bool_case_i => OK (Ebinop (Ocmp Ceq) e (make_intconst Int.zero)) | bool_case_f => OK (Ebinop (Ocmpf Ceq) e (make_floatconst Float.zero)) + | bool_case_s => OK (Ebinop (Ocmpfs Ceq) e (make_singleconst Float32.zero)) | bool_case_p => OK (Ebinop (Ocmpu Ceq) e (make_intconst Int.zero)) | bool_case_l => OK (Ebinop (Ocmpl Ceq) e (make_longconst Int64.zero)) | _ => Error (msg "Cshmgen.make_notbool") @@ -164,16 +194,18 @@ Definition make_notbool (e: expr) (ty: type) := Definition make_neg (e: expr) (ty: type) := match classify_neg ty with | neg_case_i _ => OK (Eunop Onegint e) - | neg_case_f _ => OK (Eunop Onegf e) + | neg_case_f => OK (Eunop Onegf e) + | neg_case_s => OK (Eunop Onegfs e) | neg_case_l _ => OK (Eunop Onegl e) | _ => Error (msg "Cshmgen.make_neg") end. Definition make_absfloat (e: expr) (ty: type) := match classify_neg ty with - | neg_case_i sg => OK (Eunop Oabsf (make_floatofint e sg F64)) - | neg_case_f _ => OK (Eunop Oabsf e) - | neg_case_l sg => OK (Eunop Oabsf (make_floatoflong e sg F64)) + | neg_case_i sg => OK (Eunop Oabsf (make_floatofint e sg)) + | neg_case_f => OK (Eunop Oabsf e) + | neg_case_s => OK (Eunop Oabsf (make_floatofsingle e)) + | neg_case_l sg => OK (Eunop Oabsf (make_floatoflong e sg)) | _ => Error (msg "Cshmgen.make_absfloat") end. @@ -186,7 +218,7 @@ Definition make_notint (e: expr) (ty: type) := (** Binary operators *) -Definition make_binarith (iop iopu fop lop lopu: binary_operation) +Definition make_binarith (iop iopu fop sop lop lopu: binary_operation) (e1: expr) (ty1: type) (e2: expr) (ty2: type) := let c := classify_binarith ty1 ty2 in let ty := binarith_type c in @@ -195,7 +227,8 @@ Definition make_binarith (iop iopu fop lop lopu: binary_operation) match c with | bin_case_i Signed => OK (Ebinop iop e1' e2') | bin_case_i Unsigned => OK (Ebinop iopu e1' e2') - | bin_case_f _ => OK (Ebinop fop e1' e2') + | bin_case_f => OK (Ebinop fop e1' e2') + | bin_case_s => OK (Ebinop sop e1' e2') | bin_case_l Signed => OK (Ebinop lop e1' e2') | bin_case_l Unsigned => OK (Ebinop lopu e1' e2') | bin_default => Error (msg "Cshmgen.make_binarith") @@ -216,7 +249,7 @@ Definition make_add (e1: expr) (ty1: type) (e2: expr) (ty2: type) := let n := make_intconst (Int.repr (Ctypes.sizeof ty)) in OK (Ebinop Oadd e2 (Ebinop Omul n (Eunop Ointoflong e1))) | add_default => - make_binarith Oadd Oadd Oaddf Oaddl Oaddl e1 ty1 e2 ty2 + make_binarith Oadd Oadd Oaddf Oaddfs Oaddl Oaddl e1 ty1 e2 ty2 end. Definition make_sub (e1: expr) (ty1: type) (e2: expr) (ty2: type) := @@ -231,14 +264,14 @@ Definition make_sub (e1: expr) (ty1: type) (e2: expr) (ty2: type) := let n := make_intconst (Int.repr (Ctypes.sizeof ty)) in OK (Ebinop Osub e1 (Ebinop Omul n (Eunop Ointoflong e2))) | sub_default => - make_binarith Osub Osub Osubf Osubl Osubl e1 ty1 e2 ty2 + make_binarith Osub Osub Osubf Osubfs Osubl Osubl e1 ty1 e2 ty2 end. Definition make_mul (e1: expr) (ty1: type) (e2: expr) (ty2: type) := - make_binarith Omul Omul Omulf Omull Omull e1 ty1 e2 ty2. + make_binarith Omul Omul Omulf Omulfs Omull Omull e1 ty1 e2 ty2. Definition make_div (e1: expr) (ty1: type) (e2: expr) (ty2: type) := - make_binarith Odiv Odivu Odivf Odivl Odivlu e1 ty1 e2 ty2. + make_binarith Odiv Odivu Odivf Odivfs Odivl Odivlu e1 ty1 e2 ty2. Definition make_binarith_int (iop iopu lop lopu: binary_operation) (e1: expr) (ty1: type) (e2: expr) (ty2: type) := @@ -251,7 +284,7 @@ Definition make_binarith_int (iop iopu lop lopu: binary_operation) | bin_case_i Unsigned => OK (Ebinop iopu e1' e2') | bin_case_l Signed => OK (Ebinop lop e1' e2') | bin_case_l Unsigned => OK (Ebinop lopu e1' e2') - | bin_case_f _ | bin_default => Error (msg "Cshmgen.make_binarith_int") + | bin_case_f | bin_case_s | bin_default => Error (msg "Cshmgen.make_binarith_int") end. Definition make_mod (e1: expr) (ty1: type) (e2: expr) (ty2: type) := @@ -294,7 +327,9 @@ Definition make_cmp (c: comparison) (e1: expr) (ty1: type) (e2: expr) (ty2: type | cmp_case_pl => OK (Ebinop (Ocmpu c) e1 (Eunop Ointoflong e2)) | cmp_case_lp => OK (Ebinop (Ocmpu c) (Eunop Ointoflong e1) e2) | cmp_default => - make_binarith (Ocmp c) (Ocmpu c) (Ocmpf c) (Ocmpl c) (Ocmplu c) e1 ty1 e2 ty2 + make_binarith + (Ocmp c) (Ocmpu c) (Ocmpf c) (Ocmpfs c) (Ocmpl c) (Ocmplu c) + e1 ty1 e2 ty2 end. (** [make_load addr ty_res] loads a value of type [ty_res] from @@ -374,6 +409,8 @@ Fixpoint transl_expr (a: Clight.expr) {struct a} : res expr := OK(make_intconst n) | Clight.Econst_float n _ => OK(make_floatconst n) + | Clight.Econst_single n _ => + OK(make_singleconst n) | Clight.Econst_long n _ => OK(make_longconst n) | Clight.Evar id ty => @@ -454,8 +491,9 @@ Fixpoint transl_arglist (al: list Clight.expr) (tyl: typelist) | a1 :: a2, Tnil => (* Tolerance for calls to K&R or variadic functions *) do ta1 <- transl_expr a1; + do ta1' <- make_cast (typeof a1) (default_argument_conversion (typeof a1)) ta1; do ta2 <- transl_arglist a2 Tnil; - OK (ta1 :: ta2) + OK (ta1' :: ta2) | _, _ => Error(msg "Cshmgen.transl_arglist: arity mismatch") end. @@ -470,7 +508,7 @@ Fixpoint typlist_of_arglist (al: list Clight.expr) (tyl: typelist) typ_of_type ty1 :: typlist_of_arglist a2 ty2 | a1 :: a2, Tnil => (* Tolerance for calls to K&R or variadic functions *) - typ_of_type_default (typeof a1) :: typlist_of_arglist a2 Tnil + typ_of_type (default_argument_conversion (typeof a1)) :: typlist_of_arglist a2 Tnil end. (** * Translation of statements *) diff --git a/cfrontend/Cshmgenproof.v b/cfrontend/Cshmgenproof.v index 15c4e4cd..a977e0f4 100644 --- a/cfrontend/Cshmgenproof.v +++ b/cfrontend/Cshmgenproof.v @@ -159,6 +159,13 @@ Proof. intros. unfold make_floatconst. econstructor. reflexivity. Qed. +Lemma make_singleconst_correct: + forall n e le m, + eval_expr ge e le m (make_singleconst n) (Vsingle n). +Proof. + intros. unfold make_singleconst. econstructor. reflexivity. +Qed. + Lemma make_longconst_correct: forall n e le m, eval_expr ge e le m (make_longconst n) (Vlong n). @@ -166,62 +173,35 @@ Proof. intros. unfold make_floatconst. econstructor. reflexivity. Qed. -Lemma make_floatofint_correct: - forall a n sg sz e le m, - eval_expr ge e le m a (Vint n) -> - eval_expr ge e le m (make_floatofint a sg sz) (Vfloat(cast_int_float sg sz n)). +Lemma make_singleoffloat_correct: + forall a n e le m, + eval_expr ge e le m a (Vfloat n) -> + eval_expr ge e le m (make_singleoffloat a) (Vsingle (Float.to_single n)). Proof. - intros. unfold make_floatofint, cast_int_float. - destruct sz. - destruct sg. - rewrite Float.singleofint_floatofint. econstructor. econstructor; eauto. simpl; reflexivity. auto. - rewrite Float.singleofintu_floatofintu. econstructor. econstructor; eauto. simpl; reflexivity. auto. - destruct sg; econstructor; eauto. + intros. econstructor. eauto. auto. Qed. -Lemma make_intoffloat_correct: - forall e le m a sg f i, - eval_expr ge e le m a (Vfloat f) -> - cast_float_int sg f = Some i -> - eval_expr ge e le m (make_intoffloat a sg) (Vint i). +Lemma make_floatofsingle_correct: + forall a n e le m, + eval_expr ge e le m a (Vsingle n) -> + eval_expr ge e le m (make_floatofsingle a) (Vfloat (Float.of_single n)). Proof. - unfold cast_float_int, make_intoffloat; intros. - destruct sg; econstructor; eauto; simpl; rewrite H0; auto. + intros. econstructor. eauto. auto. Qed. -Lemma make_longofint_correct: +Lemma make_floatofint_correct: forall a n sg e le m, eval_expr ge e le m a (Vint n) -> - eval_expr ge e le m (make_longofint a sg) (Vlong(cast_int_long sg n)). + eval_expr ge e le m (make_floatofint a sg) (Vfloat(cast_int_float sg n)). Proof. - intros. unfold make_longofint, cast_int_long. + intros. unfold make_floatofint, cast_int_float. destruct sg; econstructor; eauto. Qed. -Lemma make_floatoflong_correct: - forall a n sg sz e le m, - eval_expr ge e le m a (Vlong n) -> - eval_expr ge e le m (make_floatoflong a sg sz) (Vfloat(cast_long_float sg sz n)). -Proof. - intros. unfold make_floatoflong, cast_int_long. - destruct sg; destruct sz; econstructor; eauto. -Qed. - -Lemma make_longoffloat_correct: - forall e le m a sg f i, - eval_expr ge e le m a (Vfloat f) -> - cast_float_long sg f = Some i -> - eval_expr ge e le m (make_longoffloat a sg) (Vlong i). -Proof. - unfold cast_float_long, make_longoffloat; intros. - destruct sg; econstructor; eauto; simpl; rewrite H0; auto. -Qed. - Hint Resolve make_intconst_correct make_floatconst_correct make_longconst_correct - make_floatofint_correct make_intoffloat_correct - make_longofint_correct - make_floatoflong_correct make_longoffloat_correct - eval_Eunop eval_Ebinop: cshm. + make_singleconst_correct make_singleoffloat_correct make_floatofsingle_correct + make_floatofint_correct: cshm. +Hint Constructors eval_expr eval_exprlist: cshm. Hint Extern 2 (@eq trace _ _) => traceEq: cshm. Lemma make_cmp_ne_zero_correct: @@ -248,6 +228,9 @@ Proof. inv H. econstructor; eauto. rewrite H6. decEq. decEq. simpl in H6. inv H6. unfold Val.cmp in H0. eauto. inv H. econstructor; eauto. rewrite H6. decEq. decEq. + simpl in H6. unfold Val.cmpfs in H6. + destruct (Val.cmpfs_bool c v1 v2) as [[]|]; inv H6; reflexivity. + inv H. econstructor; eauto. rewrite H6. decEq. decEq. simpl in H6. unfold Val.cmpl in H6. destruct (Val.cmpl_bool c v1 v2) as [[]|]; inv H6; reflexivity. inv H. econstructor; eauto. rewrite H6. decEq. decEq. @@ -268,16 +251,7 @@ Proof. apply make_cmp_ne_zero_correct; auto. Qed. -Lemma make_cast_float_correct: - forall e le m a n sz, - eval_expr ge e le m a (Vfloat n) -> - eval_expr ge e le m (make_cast_float a sz) (Vfloat (cast_float_float sz n)). -Proof. - intros. unfold make_cast_float, cast_float_float. - destruct sz. eauto with cshm. auto. -Qed. - -Hint Resolve make_cast_int_correct make_cast_float_correct: cshm. +Hint Resolve make_cast_int_correct: cshm. Lemma make_cast_correct: forall e le m a b v ty1 ty2 v', @@ -288,14 +262,40 @@ Lemma make_cast_correct: Proof. intros. unfold make_cast, sem_cast in *; destruct (classify_cast ty1 ty2); inv H; destruct v; inv H1; eauto with cshm. + (* single -> int *) + unfold make_singleofint, cast_int_float. destruct si1; eauto with cshm. (* float -> int *) - destruct (cast_float_int si2 f) as [i|] eqn:E; inv H2. eauto with cshm. + destruct (cast_float_int si2 f) as [i|] eqn:E; inv H2. + apply make_cast_int_correct. + unfold cast_float_int in E. unfold make_intoffloat. + destruct si2; econstructor; eauto; simpl; rewrite E; auto. + (* single -> int *) + destruct (cast_single_int si2 f) as [i|] eqn:E; inv H2. + apply make_cast_int_correct. + unfold cast_single_int in E. unfold make_intofsingle. + destruct si2; econstructor; eauto with cshm; simpl; rewrite E; auto. + (* long -> int *) + unfold make_longofint, cast_int_long. destruct si1; eauto with cshm. + (* long -> float *) + unfold make_floatoflong, cast_long_float. destruct si1; eauto with cshm. + (* long -> single *) + unfold make_singleoflong, cast_long_single. destruct si1; eauto with cshm. (* float -> long *) - destruct (cast_float_long si2 f) as [i|] eqn:E; inv H2. eauto with cshm. + destruct (cast_float_long si2 f) as [i|] eqn:E; inv H2. + unfold cast_float_long in E. unfold make_longoffloat. + destruct si2; econstructor; eauto; simpl; rewrite E; auto. + (* single -> long *) + destruct (cast_single_long si2 f) as [i|] eqn:E; inv H2. + unfold cast_single_long in E. unfold make_longofsingle. + destruct si2; econstructor; eauto with cshm; simpl; rewrite E; auto. (* float -> bool *) econstructor; eauto with cshm. simpl. unfold Val.cmpf, Val.cmpf_bool. rewrite Float.cmp_ne_eq. destruct (Float.cmp Ceq f Float.zero); auto. + (* single -> bool *) + econstructor; eauto with cshm. + simpl. unfold Val.cmpfs, Val.cmpfs_bool. rewrite Float32.cmp_ne_eq. + destruct (Float32.cmp Ceq f Float32.zero); auto. (* long -> bool *) econstructor; eauto with cshm. simpl. unfold Val.cmpl, Val.cmpl_bool, Int64.cmp. @@ -327,6 +327,10 @@ Proof. econstructor; split. econstructor; eauto with cshm. simpl. eauto. unfold Val.cmpf, Val.cmpf_bool. simpl. rewrite <- Float.cmp_ne_eq. destruct (Float.cmp Cne f Float.zero); constructor. +(* single *) + econstructor; split. econstructor; eauto with cshm. simpl. eauto. + unfold Val.cmpfs, Val.cmpfs_bool. simpl. rewrite <- Float32.cmp_ne_eq. + destruct (Float32.cmp Cne f Float32.zero); constructor. (* pointer *) econstructor; split. econstructor; eauto with cshm. simpl. eauto. unfold Val.cmpu, Val.cmpu_bool. simpl. @@ -357,6 +361,9 @@ Lemma make_absfloat_correct: Proof. unfold sem_absfloat, make_absfloat; intros until m; intros SEM MAKE EV1; destruct (classify_neg tya); inv MAKE; destruct va; inv SEM; eauto with cshm. + unfold make_floatoflong, cast_long_float. destruct s. + econstructor. econstructor; simpl; eauto. simpl; eauto. simpl; eauto. + econstructor. econstructor; simpl; eauto. simpl; eauto. simpl; eauto. Qed. Lemma make_notbool_correct: @@ -396,7 +403,8 @@ Section MAKE_BIN. Variable sem_int: signedness -> int -> int -> option val. Variable sem_long: signedness -> int64 -> int64 -> option val. Variable sem_float: float -> float -> option val. -Variables iop iopu fop lop lopu: binary_operation. +Variable sem_single: float32 -> float32 -> option val. +Variables iop iopu fop sop lop lopu: binary_operation. Hypothesis iop_ok: forall x y m, eval_binop iop (Vint x) (Vint y) m = sem_int Signed x y. @@ -408,11 +416,13 @@ Hypothesis lopu_ok: forall x y m, eval_binop lopu (Vlong x) (Vlong y) m = sem_long Unsigned x y. Hypothesis fop_ok: forall x y m, eval_binop fop (Vfloat x) (Vfloat y) m = sem_float x y. +Hypothesis sop_ok: + forall x y m, eval_binop sop (Vsingle x) (Vsingle y) m = sem_single x y. Lemma make_binarith_correct: binary_constructor_correct - (make_binarith iop iopu fop lop lopu) - (sem_binarith sem_int sem_long sem_float). + (make_binarith iop iopu fop sop lop lopu) + (sem_binarith sem_int sem_long sem_float sem_single). Proof. red; unfold make_binarith, sem_binarith; intros until m; intros SEM MAKE EV1 EV2. @@ -429,12 +439,13 @@ Proof. - destruct s; inv H0; econstructor; eauto with cshm. rewrite lop_ok; auto. rewrite lopu_ok; auto. - erewrite <- fop_ok in SEM; eauto with cshm. +- erewrite <- sop_ok in SEM; eauto with cshm. Qed. Lemma make_binarith_int_correct: binary_constructor_correct (make_binarith_int iop iopu lop lopu) - (sem_binarith sem_int sem_long (fun x y => None)). + (sem_binarith sem_int sem_long (fun x y => None) (fun x y => None)). Proof. red; unfold make_binarith_int, sem_binarith; intros until m; intros SEM MAKE EV1 EV2. @@ -930,6 +941,8 @@ Proof. apply make_intconst_correct. (* const float *) apply make_floatconst_correct. +(* const single *) + apply make_singleconst_correct. (* const long *) apply make_longconst_correct. (* temp var *) diff --git a/cfrontend/Ctypes.v b/cfrontend/Ctypes.v index 41d0dcbf..c437a6bc 100644 --- a/cfrontend/Ctypes.v +++ b/cfrontend/Ctypes.v @@ -196,6 +196,18 @@ Definition typeconv (ty: type) : type := | _ => remove_attributes ty end. +(** Default conversion for arguments to an unprototyped or variadic function. + Like [typeconv] but also converts single floats to double floats. *) + +Definition default_argument_conversion (ty: type) : type := + match ty with + | Tint (I8 | I16 | IBool) _ _ => Tint I32 Signed noattr + | Tfloat _ _ => Tfloat F64 noattr + | Tarray t sz a => Tpointer t noattr + | Tfunction _ _ _ => Tpointer ty noattr + | _ => remove_attributes ty + end. + (** * Operations over types *) (** Alignment of a type, in bytes. *) @@ -723,12 +735,3 @@ Fixpoint typlist_of_typelist (tl: typelist) : list AST.typ := Definition signature_of_type (args: typelist) (res: type) (cc: calling_convention): signature := mksignature (typlist_of_typelist args) (opttyp_of_type res) cc. -(** Like [typ_of_type], but apply default argument promotion. *) - -Definition typ_of_type_default (t: type) : AST.typ := - match t with - | Tfloat _ _ => AST.Tfloat - | Tlong _ _ => AST.Tlong - | _ => AST.Tint - end. - diff --git a/cfrontend/Initializers.v b/cfrontend/Initializers.v index 4054f6e1..3454ddcf 100644 --- a/cfrontend/Initializers.v +++ b/cfrontend/Initializers.v @@ -55,7 +55,7 @@ Fixpoint constval (a: expr) : res val := match a with | Eval v ty => match v with - | Vint _ | Vfloat _ | Vlong _ => OK v + | Vint _ | Vfloat _ | Vsingle _ | Vlong _ => OK v | Vptr _ _ | Vundef => Error(msg "illegal constant") end | Evalof l ty => @@ -156,7 +156,7 @@ Definition transl_init_single (ty: type) (a: expr) : res init_data := | Vint n, Tpointer _ _ => OK(Init_int32 n) | Vint n, Tcomp_ptr _ _ => OK(Init_int32 n) | Vlong n, Tlong _ _ => OK(Init_int64 n) - | Vfloat f, Tfloat F32 _ => OK(Init_float32 f) + | Vsingle f, Tfloat F32 _ => OK(Init_float32 f) | Vfloat f, Tfloat F64 _ => OK(Init_float64 f) | Vptr id ofs, Tint I32 sg _ => OK(Init_addrof id ofs) | Vptr id ofs, Tpointer _ _ => OK(Init_addrof id ofs) diff --git a/cfrontend/Initializersproof.v b/cfrontend/Initializersproof.v index f11901d7..16004fc0 100644 --- a/cfrontend/Initializersproof.v +++ b/cfrontend/Initializersproof.v @@ -506,6 +506,9 @@ Proof. (* float *) destruct ty; try discriminate. destruct f1; inv EQ2; simpl in H2; inv H2; assumption. + (* single *) + destruct ty; try discriminate. + destruct f1; inv EQ2; simpl in H2; inv H2; assumption. (* pointer *) unfold inj in H. assert (data = Init_addrof b1 ofs1 /\ chunk = Mint32). @@ -525,20 +528,18 @@ Lemma transl_init_single_size: Genv.init_data_size data = sizeof ty. Proof. intros. monadInv H. destruct x0. - monadInv EQ2. - destruct ty; try discriminate. +- monadInv EQ2. +- destruct ty; try discriminate. destruct i0; inv EQ2; auto. inv EQ2; auto. inv EQ2; auto. - destruct ty; inv EQ2; auto. - destruct ty; try discriminate. +- destruct ty; inv EQ2; auto. +- destruct ty; try discriminate. destruct f0; inv EQ2; auto. - destruct ty; try discriminate. - destruct i0; auto. - inv EQ2. - inv EQ2. - inv EQ2; auto. - inv EQ2. +- destruct ty; try discriminate. + destruct f0; inv EQ2; auto. +- destruct ty; try discriminate. + destruct i0; inv EQ2; auto. inv EQ2; auto. inv EQ2; auto. Qed. diff --git a/cfrontend/PrintClight.ml b/cfrontend/PrintClight.ml index 49705c46..c5a6e6e1 100644 --- a/cfrontend/PrintClight.ml +++ b/cfrontend/PrintClight.ml @@ -43,6 +43,7 @@ let rec precedence = function | Efield _ -> (16, LtoR) | Econst_int _ -> (16, NA) | Econst_float _ -> (16, NA) + | Econst_single _ -> (16, NA) | Econst_long _ -> (16, NA) | Eunop _ -> (15, RtoL) | Eaddrof _ -> (15, RtoL) @@ -82,6 +83,8 @@ let rec expr p (prec, e) = fprintf p "%ld" (camlint_of_coqint n) | Econst_float(f, _) -> fprintf p "%F" (camlfloat_of_coqfloat f) + | Econst_single(f, _) -> + fprintf p "%Ff" (camlfloat_of_coqfloat32 f) | Econst_long(n, Tlong(Unsigned, _)) -> fprintf p "%LuLLU" (camlint64_of_coqint n) | Econst_long(n, _) -> @@ -269,6 +272,7 @@ let rec collect_expr e = match e with | Econst_int _ -> () | Econst_float _ -> () + | Econst_single _ -> () | Econst_long _ -> () | Evar _ -> () | Etempvar _ -> () diff --git a/cfrontend/PrintCsyntax.ml b/cfrontend/PrintCsyntax.ml index 9441d71e..4e0ee7f3 100644 --- a/cfrontend/PrintCsyntax.ml +++ b/cfrontend/PrintCsyntax.ml @@ -191,6 +191,8 @@ let print_typed_value p v ty = fprintf p "%ld" (camlint_of_coqint n) | Vfloat f, _ -> fprintf p "%F" (camlfloat_of_coqfloat f) + | Vsingle f, _ -> + fprintf p "%Ff" (camlfloat_of_coqfloat32 f) | Vlong n, Tlong(Unsigned, _) -> fprintf p "%LuLLU" (camlint64_of_coqint n) | Vlong n, _ -> diff --git a/cfrontend/SimplExpr.v b/cfrontend/SimplExpr.v index f9aa8dbf..05d9c860 100644 --- a/cfrontend/SimplExpr.v +++ b/cfrontend/SimplExpr.v @@ -116,6 +116,7 @@ Function eval_simpl_expr (a: expr) : option val := match a with | Econst_int n _ => Some(Vint n) | Econst_float n _ => Some(Vfloat n) + | Econst_single n _ => Some(Vsingle n) | Econst_long n _ => Some(Vlong n) | Ecast b ty => match eval_simpl_expr b with @@ -242,6 +243,8 @@ Fixpoint transl_expr (dst: destination) (a: Csyntax.expr) : mon (list statement ret (finish dst nil (Econst_int n ty)) | Csyntax.Eval (Vfloat n) ty => ret (finish dst nil (Econst_float n ty)) + | Csyntax.Eval (Vsingle n) ty => + ret (finish dst nil (Econst_single n ty)) | Csyntax.Eval (Vlong n) ty => ret (finish dst nil (Econst_long n ty)) | Csyntax.Eval _ ty => diff --git a/cfrontend/SimplExprspec.v b/cfrontend/SimplExprspec.v index 3dae7ab1..1ef2e76d 100644 --- a/cfrontend/SimplExprspec.v +++ b/cfrontend/SimplExprspec.v @@ -799,6 +799,10 @@ Opaque makeif. constructor. auto. intros; constructor. constructor. constructor. auto. intros; constructor. + intros. destruct dst; simpl in *; inv H2. + constructor. auto. intros; constructor. + constructor. + constructor. auto. intros; constructor. (* var *) monadInv H; econstructor; split; auto with gensym. UseFinish. constructor. (* field *) diff --git a/cfrontend/SimplLocals.v b/cfrontend/SimplLocals.v index 29565089..9c529280 100644 --- a/cfrontend/SimplLocals.v +++ b/cfrontend/SimplLocals.v @@ -45,7 +45,8 @@ Definition make_cast (a: expr) (tto: type) : expr := match classify_cast (typeof a) tto with | cast_case_neutral => a | cast_case_i2i I32 _ => a - | cast_case_f2f F64 => a + | cast_case_f2f => a + | cast_case_s2s => a | cast_case_l2l => a | cast_case_struct _ _ _ _ => a | cast_case_union _ _ _ _ => a @@ -59,6 +60,7 @@ Fixpoint simpl_expr (cenv: compilenv) (a: expr) : expr := match a with | Econst_int _ _ => a | Econst_float _ _ => a + | Econst_single _ _ => a | Econst_long _ _ => a | Evar id ty => if VSet.mem id cenv then Etempvar id ty else Evar id ty | Etempvar id ty => Etempvar id ty @@ -157,6 +159,7 @@ Fixpoint addr_taken_expr (a: expr): VSet.t := match a with | Econst_int _ _ => VSet.empty | Econst_float _ _ => VSet.empty + | Econst_single _ _ => VSet.empty | Econst_long _ _ => VSet.empty | Evar id ty => VSet.empty | Etempvar id ty => VSet.empty diff --git a/cfrontend/SimplLocalsproof.v b/cfrontend/SimplLocalsproof.v index 6eec8cce..6024de4d 100644 --- a/cfrontend/SimplLocalsproof.v +++ b/cfrontend/SimplLocalsproof.v @@ -189,9 +189,10 @@ Inductive val_casted: val -> type -> Prop := | val_casted_int: forall sz si attr n, cast_int_int sz si n = n -> val_casted (Vint n) (Tint sz si attr) - | val_casted_float: forall sz attr n, - cast_float_float sz n = n -> - val_casted (Vfloat n) (Tfloat sz attr) + | val_casted_float: forall attr n, + val_casted (Vfloat n) (Tfloat F64 attr) + | val_casted_single: forall attr n, + val_casted (Vsingle n) (Tfloat F32 attr) | val_casted_long: forall si attr n, val_casted (Vlong n) (Tlong si attr) | val_casted_ptr_ptr: forall b ofs ty attr, @@ -220,14 +221,6 @@ Proof. destruct (Int.eq i Int.zero); auto. Qed. -Remark cast_float_float_idem: - forall sz f, cast_float_float sz (cast_float_float sz f) = cast_float_float sz f. -Proof. - intros; destruct sz; simpl. - apply Float.singleoffloat_idem; auto. - auto. -Qed. - Lemma cast_val_is_casted: forall v ty ty' v', sem_cast v ty ty' = Some v' -> val_casted v' ty'. Proof. @@ -235,28 +228,32 @@ Proof. (* void *) constructor. (* int *) - destruct i; destruct ty; simpl in H; try discriminate; destruct v; inv H. + destruct i; destruct ty; simpl in H; try (destruct f); try discriminate; destruct v; inv H. constructor. apply (cast_int_int_idem I8 s). constructor. apply (cast_int_int_idem I8 s). - destruct (cast_float_int s f0); inv H1. constructor. apply (cast_int_int_idem I8 s). + destruct (cast_single_int s f); inv H1. constructor. apply (cast_int_int_idem I8 s). + destruct (cast_float_int s f); inv H1. constructor. apply (cast_int_int_idem I8 s). constructor. apply (cast_int_int_idem I16 s). constructor. apply (cast_int_int_idem I16 s). - destruct (cast_float_int s f0); inv H1. constructor. apply (cast_int_int_idem I16 s). + destruct (cast_single_int s f); inv H1. constructor. apply (cast_int_int_idem I16 s). + destruct (cast_float_int s f); inv H1. constructor. apply (cast_int_int_idem I16 s). constructor. auto. constructor. constructor. auto. - destruct (cast_float_int s f0); inv H1. constructor. auto. - constructor. auto. - constructor. + destruct (cast_single_int s f); inv H1. constructor. auto. + destruct (cast_float_int s f); inv H1. constructor; auto. constructor; auto. constructor. constructor; auto. + constructor. constructor; auto. + constructor. constructor; auto. - constructor; auto. + constructor. constructor. simpl. destruct (Int.eq i0 Int.zero); auto. constructor. simpl. destruct (Int64.eq i Int64.zero); auto. - constructor. simpl. destruct (Float.cmp Ceq f0 Float.zero); auto. + constructor. simpl. destruct (Float32.cmp Ceq f Float32.zero); auto. + constructor. simpl. destruct (Float.cmp Ceq f Float.zero); auto. constructor. simpl. destruct (Int.eq i Int.zero); auto. constructor; auto. constructor. simpl. destruct (Int.eq i Int.zero); auto. @@ -266,23 +263,17 @@ Proof. constructor. simpl. destruct (Int.eq i0 Int.zero); auto. constructor; auto. (* long *) - destruct ty; try discriminate. + destruct ty; try (destruct f); try discriminate. destruct v; inv H. constructor. destruct v; inv H. constructor. - destruct v; try discriminate. destruct (cast_float_long s f0); inv H. constructor. + destruct v; try discriminate. destruct (cast_single_long s f); inv H. constructor. + destruct v; try discriminate. destruct (cast_float_long s f); inv H. constructor. destruct v; inv H. constructor. destruct v; inv H. constructor. destruct v; inv H. constructor. destruct v; inv H. constructor. (* float *) - destruct ty; simpl in H; try discriminate; destruct v; inv H. - constructor. unfold cast_float_float, cast_int_float. - destruct f; destruct s; auto. - rewrite Float.singleofint_floatofint. apply Float.singleoffloat_idem. - rewrite Float.singleofintu_floatofintu. apply Float.singleoffloat_idem. - constructor. unfold cast_float_float, cast_long_float. - destruct f; destruct s; auto. apply Float.singleoflong_idem. apply Float.singleoflongu_idem. - constructor. apply cast_float_float_idem. + destruct f; destruct ty; simpl in H; try (destruct f); try discriminate; destruct v; inv H; constructor. (* pointer *) destruct ty; simpl in H; try discriminate; destruct v; inv H; try constructor. (* impossible cases *) @@ -310,7 +301,8 @@ Proof. clear H1. inv H0. auto. inversion H0; clear H0; subst chunk. simpl in *. destruct (Int.eq n Int.zero); subst n; reflexivity. - destruct sz; inversion H0; clear H0; subst chunk; simpl in *; congruence. + inv H0; auto. + inv H0; auto. inv H0; auto. inv H0; auto. inv H0; auto. @@ -327,7 +319,6 @@ Lemma cast_val_casted: Proof. intros. inversion H; clear H; subst v ty; unfold sem_cast; simpl; auto. destruct sz; congruence. - congruence. unfold proj_sumbool; repeat rewrite dec_eq_true; auto. unfold proj_sumbool; repeat rewrite dec_eq_true; auto. Qed. @@ -373,7 +364,8 @@ Proof. destruct (classify_cast (typeof a) tto); auto. destruct v1; inv H0; auto. destruct sz2; auto. destruct v1; inv H0; auto. - destruct sz2; auto. destruct v1; inv H0; auto. + destruct v1; inv H0; auto. + destruct v1; inv H0; auto. destruct v1; inv H0; auto. destruct v1; try discriminate. destruct (ident_eq id1 id2 && fieldlist_eq fld1 fld2); inv H0; auto. @@ -1416,6 +1408,7 @@ Proof. (* const *) exists (Vint i); split; auto. constructor. exists (Vfloat f0); split; auto. constructor. + exists (Vsingle f0); split; auto. constructor. exists (Vlong i); split; auto. constructor. (* tempvar *) exploit me_temps; eauto. intros [[tv [A B]] C]. diff --git a/common/AST.v b/common/AST.v index 29d14524..1c46389e 100644 --- a/common/AST.v +++ b/common/AST.v @@ -35,22 +35,15 @@ Definition ident_eq := peq. Parameter ident_of_string : String.string -> ident. -(** The intermediate languages are weakly typed, using only four - types: [Tint] for 32-bit integers and pointers, [Tfloat] for 64-bit - floating-point numbers, [Tlong] for 64-bit integers, [Tsingle] - for 32-bit floating-point numbers. *) +(** The intermediate languages are weakly typed, using the following types: *) Inductive typ : Type := - | Tint - | Tfloat - | Tlong - | Tsingle. - -Definition typesize (ty: typ) : Z := - match ty with Tint => 4 | Tfloat => 8 | Tlong => 8 | Tsingle => 4 end. - -Lemma typesize_pos: forall ty, typesize ty > 0. -Proof. destruct ty; simpl; omega. Qed. + | Tint (**r 32-bit integers or pointers *) + | Tfloat (**r 64-bit double-precision floats *) + | Tlong (**r 64-bit integers *) + | Tsingle (**r 32-bit single-precision floats *) + | Tany32 (**r any 32-bit value *) + | Tany64. (**r any 64-bit value, i.e. any value *) Lemma typ_eq: forall (t1 t2: typ), {t1=t2} + {t1<>t2}. Proof. decide equality. Defined. @@ -62,8 +55,22 @@ Definition opt_typ_eq: forall (t1 t2: option typ), {t1=t2} + {t1<>t2} Definition list_typ_eq: forall (l1 l2: list typ), {l1=l2} + {l1<>l2} := list_eq_dec typ_eq. -(** All values of type [Tsingle] are also of type [Tfloat]. This - corresponds to the following subtyping relation over types. *) +Definition typesize (ty: typ) : Z := + match ty with + | Tint => 4 + | Tfloat => 8 + | Tlong => 8 + | Tsingle => 4 + | Tany32 => 4 + | Tany64 => 8 + end. + +Lemma typesize_pos: forall ty, typesize ty > 0. +Proof. destruct ty; simpl; omega. Qed. + +(** All values of size 32 bits are also of type [Tany32]. All values + are of type [Tany64]. This corresponds to the following subtyping + relation over types. *) Definition subtype (ty1 ty2: typ) : bool := match ty1, ty2 with @@ -71,7 +78,8 @@ Definition subtype (ty1 ty2: typ) : bool := | Tlong, Tlong => true | Tfloat, Tfloat => true | Tsingle, Tsingle => true - | Tsingle, Tfloat => true + | (Tint | Tsingle | Tany32), Tany32 => true + | _, Tany64 => true | _, _ => false end. @@ -133,7 +141,9 @@ Inductive memory_chunk : Type := | Mint32 (**r 32-bit integer, or pointer *) | Mint64 (**r 64-bit integer *) | Mfloat32 (**r 32-bit single-precision float *) - | Mfloat64. (**r 64-bit double-precision float *) + | Mfloat64 (**r 64-bit double-precision float *) + | Many32 (**r any value that fits in 32 bits *) + | Many64. (**r any value *) Definition chunk_eq: forall (c1 c2: memory_chunk), {c1=c2} + {c1<>c2}. Proof. decide equality. Defined. @@ -151,18 +161,8 @@ Definition type_of_chunk (c: memory_chunk) : typ := | Mint64 => Tlong | Mfloat32 => Tsingle | Mfloat64 => Tfloat - end. - -Definition type_of_chunk_use (c: memory_chunk) : typ := - match c with - | Mint8signed => Tint - | Mint8unsigned => Tint - | Mint16signed => Tint - | Mint16unsigned => Tint - | Mint32 => Tint - | Mint64 => Tlong - | Mfloat32 => Tfloat - | Mfloat64 => Tfloat + | Many32 => Tany32 + | Many64 => Tany64 end. (** The chunk that is appropriate to store and reload a value of @@ -174,6 +174,8 @@ Definition chunk_of_type (ty: typ) := | Tfloat => Mfloat64 | Tlong => Mint64 | Tsingle => Mfloat32 + | Tany32 => Many32 + | Tany64 => Many64 end. (** Initialization data for global variables. *) @@ -183,7 +185,7 @@ Inductive init_data: Type := | Init_int16: int -> init_data | Init_int32: int -> init_data | Init_int64: int64 -> init_data - | Init_float32: float -> init_data + | Init_float32: float32 -> init_data | Init_float64: float -> init_data | Init_space: Z -> init_data | Init_addrof: ident -> int -> init_data. (**r address of symbol + offset *) @@ -586,9 +588,9 @@ Definition ef_sig (ef: external_function): signature := | EF_external name sg => sg | EF_builtin name sg => sg | EF_vload chunk => mksignature (Tint :: nil) (Some (type_of_chunk chunk)) cc_default - | EF_vstore chunk => mksignature (Tint :: type_of_chunk_use chunk :: nil) None cc_default + | EF_vstore chunk => mksignature (Tint :: type_of_chunk chunk :: nil) None cc_default | EF_vload_global chunk _ _ => mksignature nil (Some (type_of_chunk chunk)) cc_default - | EF_vstore_global chunk _ _ => mksignature (type_of_chunk_use chunk :: nil) None cc_default + | EF_vstore_global chunk _ _ => mksignature (type_of_chunk chunk :: nil) None cc_default | EF_malloc => mksignature (Tint :: nil) (Some Tint) cc_default | EF_free => mksignature (Tint :: nil) None cc_default | EF_memcpy sz al => mksignature (Tint :: Tint :: nil) None cc_default diff --git a/common/Events.v b/common/Events.v index 48cd91e4..5eee93a7 100644 --- a/common/Events.v +++ b/common/Events.v @@ -57,7 +57,7 @@ Inductive eventval: Type := | EVint: int -> eventval | EVlong: int64 -> eventval | EVfloat: float -> eventval - | EVfloatsingle: float -> eventval + | EVsingle: float32 -> eventval | EVptr_global: ident -> int -> eventval. Inductive event: Type := @@ -274,8 +274,7 @@ Inductive eventval_match: eventval -> typ -> val -> Prop := | ev_match_float: forall f, eventval_match (EVfloat f) Tfloat (Vfloat f) | ev_match_single: forall f, - Float.is_single f -> - eventval_match (EVfloatsingle f) Tsingle (Vfloat f) + eventval_match (EVsingle f) Tsingle (Vsingle f) | ev_match_ptr: forall id b ofs, Genv.find_symbol ge id = Some b -> eventval_match (EVptr_global id ofs) Tint (Vptr b ofs). @@ -388,7 +387,7 @@ Definition eventval_valid (ev: eventval) : Prop := | EVint _ => True | EVlong _ => True | EVfloat _ => True - | EVfloatsingle f => Float.is_single f + | EVsingle _ => True | EVptr_global id ofs => exists b, Genv.find_symbol ge id = Some b end. @@ -397,7 +396,7 @@ Definition eventval_type (ev: eventval) : typ := | EVint _ => Tint | EVlong _ => Tlong | EVfloat _ => Tfloat - | EVfloatsingle _ => Tsingle + | EVsingle _ => Tsingle | EVptr_global id ofs => Tint end. @@ -412,7 +411,7 @@ Proof. destruct H1 as [b EQ]. exists (Vptr b i1); constructor; auto. exists (Vlong i0); constructor. exists (Vfloat f1); constructor. - exists (Vfloat f1); constructor; auto. + exists (Vsingle f1); constructor; auto. exists (Vint i); constructor. destruct H1 as [b' EQ]. exists (Vptr b' i0); constructor; auto. Qed. @@ -967,7 +966,7 @@ Qed. Lemma volatile_store_ok: forall chunk, extcall_properties (volatile_store_sem chunk) - (mksignature (Tint :: type_of_chunk_use chunk :: nil) None cc_default). + (mksignature (Tint :: type_of_chunk chunk :: nil) None cc_default). Proof. intros; constructor; intros. (* well typed *) @@ -1022,7 +1021,7 @@ Qed. Lemma volatile_store_global_ok: forall chunk id ofs, extcall_properties (volatile_store_global_sem chunk id ofs) - (mksignature (type_of_chunk_use chunk :: nil) None cc_default). + (mksignature (type_of_chunk chunk :: nil) None cc_default). Proof. intros; constructor; intros. (* well typed *) diff --git a/common/Globalenvs.v b/common/Globalenvs.v index bd38fa3d..a34a05dc 100644 --- a/common/Globalenvs.v +++ b/common/Globalenvs.v @@ -516,7 +516,7 @@ Definition store_init_data (m: mem) (b: block) (p: Z) (id: init_data) : option m | Init_int16 n => Mem.store Mint16unsigned m b p (Vint n) | Init_int32 n => Mem.store Mint32 m b p (Vint n) | Init_int64 n => Mem.store Mint64 m b p (Vlong n) - | Init_float32 n => Mem.store Mfloat32 m b p (Vfloat n) + | Init_float32 n => Mem.store Mfloat32 m b p (Vsingle n) | Init_float64 n => Mem.store Mfloat64 m b p (Vfloat n) | Init_addrof symb ofs => match find_symbol ge symb with @@ -848,7 +848,7 @@ Fixpoint load_store_init_data (m: mem) (b: block) (p: Z) (il: list init_data) {s Mem.load Mint64 m b p = Some(Vlong n) /\ load_store_init_data m b (p + 8) il' | Init_float32 n :: il' => - Mem.load Mfloat32 m b p = Some(Vfloat(Float.singleoffloat n)) + Mem.load Mfloat32 m b p = Some(Vsingle n) /\ load_store_init_data m b (p + 4) il' | Init_float64 n :: il' => Mem.load Mfloat64 m b p = Some(Vfloat n) @@ -895,7 +895,7 @@ Proof. eapply (A Mint16unsigned (Vint i)); eauto. eapply (A Mint32 (Vint i)); eauto. eapply (A Mint64 (Vlong i)); eauto. - eapply (A Mfloat32 (Vfloat f)); eauto. + eapply (A Mfloat32 (Vsingle f)); eauto. eapply (A Mfloat64 (Vfloat f)); eauto. inv Heqo. red; intros. transitivity (Mem.load chunk m1 b p0). eapply store_init_data_list_outside; eauto. right. simpl. xomega. diff --git a/common/Memdata.v b/common/Memdata.v index 1b74705f..96278a29 100644 --- a/common/Memdata.v +++ b/common/Memdata.v @@ -39,6 +39,8 @@ Definition size_chunk (chunk: memory_chunk) : Z := | Mint64 => 8 | Mfloat32 => 4 | Mfloat64 => 8 + | Many32 => 4 + | Many64 => 8 end. Lemma size_chunk_pos: @@ -86,6 +88,8 @@ Definition align_chunk (chunk: memory_chunk) : Z := | Mint64 => 8 | Mfloat32 => 4 | Mfloat64 => 4 + | Many32 => 4 + | Many64 => 4 end. Lemma align_chunk_pos: @@ -112,12 +116,27 @@ Proof. | exists 8; reflexivity ]. Qed. +Inductive quantity : Type := Q32 | Q64. + +Definition quantity_eq (q1 q2: quantity) : {q1 = q2} + {q1 <> q2}. +Proof. decide equality. Defined. +Global Opaque quantity_eq. + +Definition size_quantity_nat (q: quantity) := + match q with Q32 => 4%nat | Q64 => 8%nat end. + +Lemma size_quantity_nat_pos: + forall q, exists n, size_quantity_nat q = S n. +Proof. + intros. destruct q; [exists 3%nat | exists 7%nat]; auto. +Qed. + (** * Memory values *) (** A ``memory value'' is a byte-sized quantity that describes the current content of a memory cell. It can be either: - a concrete 8-bit integer; -- a byte-sized fragment of an opaque pointer; +- a byte-sized fragment of an opaque value; - the special constant [Undef] that represents uninitialized memory. *) @@ -126,7 +145,7 @@ Qed. Inductive memval: Type := | Undef: memval | Byte: byte -> memval - | Pointer: block -> int -> nat -> memval. + | Fragment: val -> quantity -> nat -> memval. (** * Encoding and decoding integers *) @@ -311,25 +330,28 @@ Proof. simpl. decEq. auto. Qed. -Fixpoint inj_pointer (n: nat) (b: block) (ofs: int) {struct n}: list memval := +Fixpoint inj_value_rec (n: nat) (v: val) (q: quantity) {struct n}: list memval := match n with | O => nil - | S m => Pointer b ofs m :: inj_pointer m b ofs + | S m => Fragment v q m :: inj_value_rec m v q end. -Fixpoint check_pointer (n: nat) (b: block) (ofs: int) (vl: list memval) +Definition inj_value (q: quantity) (v: val): list memval := + inj_value_rec (size_quantity_nat q) v q. + +Fixpoint check_value (n: nat) (v: val) (q: quantity) (vl: list memval) {struct n} : bool := match n, vl with | O, nil => true - | S m, Pointer b' ofs' m' :: vl' => - eq_block b b' && Int.eq_dec ofs ofs' && beq_nat m m' && check_pointer m b ofs vl' + | S m, Fragment v' q' m' :: vl' => + Val.eq v v' && quantity_eq q q' && beq_nat m m' && check_value m v q vl' | _, _ => false end. -Definition proj_pointer (vl: list memval) : val := +Definition proj_value (q: quantity) (vl: list memval) : val := match vl with - | Pointer b ofs n :: vl' => - if check_pointer 4%nat b ofs vl then Vptr b ofs else Vundef + | Fragment v q' n :: vl' => + if check_value (size_quantity_nat q) v q vl then v else Vundef | _ => Vundef end. @@ -338,10 +360,12 @@ Definition encode_val (chunk: memory_chunk) (v: val) : list memval := | Vint n, (Mint8signed | Mint8unsigned) => inj_bytes (encode_int 1%nat (Int.unsigned n)) | Vint n, (Mint16signed | Mint16unsigned) => inj_bytes (encode_int 2%nat (Int.unsigned n)) | Vint n, Mint32 => inj_bytes (encode_int 4%nat (Int.unsigned n)) - | Vptr b ofs, Mint32 => inj_pointer 4%nat b ofs + | Vptr b ofs, Mint32 => inj_value Q32 v | Vlong n, Mint64 => inj_bytes (encode_int 8%nat (Int64.unsigned n)) - | Vfloat n, Mfloat32 => inj_bytes (encode_int 4%nat (Int.unsigned (Float.bits_of_single n))) - | Vfloat n, Mfloat64 => inj_bytes (encode_int 8%nat (Int64.unsigned (Float.bits_of_double n))) + | Vsingle n, Mfloat32 => inj_bytes (encode_int 4%nat (Int.unsigned (Float32.to_bits n))) + | Vfloat n, Mfloat64 => inj_bytes (encode_int 8%nat (Int64.unsigned (Float.to_bits n))) + | _, Many32 => inj_value Q32 v + | _, Many64 => inj_value Q64 v | _, _ => list_repeat (size_chunk_nat chunk) Undef end. @@ -355,12 +379,15 @@ Definition decode_val (chunk: memory_chunk) (vl: list memval) : val := | Mint16unsigned => Vint(Int.zero_ext 16 (Int.repr (decode_int bl))) | Mint32 => Vint(Int.repr(decode_int bl)) | Mint64 => Vlong(Int64.repr(decode_int bl)) - | Mfloat32 => Vfloat(Float.single_of_bits (Int.repr (decode_int bl))) - | Mfloat64 => Vfloat(Float.double_of_bits (Int64.repr (decode_int bl))) + | Mfloat32 => Vsingle(Float32.of_bits (Int.repr (decode_int bl))) + | Mfloat64 => Vfloat(Float.of_bits (Int64.repr (decode_int bl))) + | Many32 => Vundef + | Many64 => Vundef end | None => match chunk with - | Mint32 => proj_pointer vl + | Mint32 | Many32 => Val.load_result chunk (proj_value Q32 vl) + | Many64 => Val.load_result chunk (proj_value Q64 vl) | _ => Vundef end end. @@ -374,14 +401,40 @@ Proof. | rewrite length_inj_bytes; apply encode_int_length ]. Qed. -Lemma check_inj_pointer: - forall b ofs n, check_pointer n b ofs (inj_pointer n b ofs) = true. +Lemma check_inj_value: + forall v q n, check_value n v q (inj_value_rec n v q) = true. Proof. induction n; simpl. auto. unfold proj_sumbool. rewrite dec_eq_true. rewrite dec_eq_true. rewrite <- beq_nat_refl. simpl; auto. Qed. +Lemma proj_inj_value: + forall q v, proj_value q (inj_value q v) = v. +Proof. + intros. unfold proj_value, inj_value. destruct (size_quantity_nat_pos q) as [n EQ]. + rewrite EQ at 1. simpl. rewrite check_inj_value. auto. +Qed. + +Remark in_inj_value: + forall mv v q, In mv (inj_value q v) -> exists n, mv = Fragment v q n. +Proof. +Local Transparent inj_value. + unfold inj_value; intros until q. generalize (size_quantity_nat q). induction n; simpl; intros. + contradiction. + destruct H. exists n; auto. eauto. +Qed. + +Lemma proj_inj_value_mismatch: + forall q1 q2 v, q1 <> q2 -> proj_value q1 (inj_value q2 v) = Vundef. +Proof. + intros. unfold proj_value. destruct (inj_value q2 v) eqn:V. auto. destruct m; auto. + destruct (in_inj_value (Fragment v0 q n) v q2) as [n' EQ]. + rewrite V; auto with coqlib. inv EQ. + destruct (size_quantity_nat_pos q1) as [p EQ1]; rewrite EQ1; simpl. + unfold proj_sumbool. rewrite dec_eq_true. rewrite dec_eq_false by congruence. auto. +Qed. + Definition decode_encode_val (v1: val) (chunk1 chunk2: memory_chunk) (v2: val) : Prop := match v1, chunk1, chunk2 with | Vundef, _, _ => v2 = Vundef @@ -394,21 +447,30 @@ Definition decode_encode_val (v1: val) (chunk1 chunk2: memory_chunk) (v2: val) : | Vint n, Mint16signed, Mint16unsigned => v2 = Vint(Int.zero_ext 16 n) | Vint n, Mint16unsigned, Mint16unsigned => v2 = Vint(Int.zero_ext 16 n) | Vint n, Mint32, Mint32 => v2 = Vint n - | Vint n, Mint32, Mfloat32 => v2 = Vfloat(Float.single_of_bits n) - | Vint n, (Mint64 | Mfloat32 | Mfloat64), _ => v2 = Vundef + | Vint n, Many32, (Mint32 | Many32) => v2 = Vint n + | Vint n, Mint32, Mfloat32 => v2 = Vsingle(Float32.of_bits n) + | Vint n, Many64, Many64 => v2 = Vint n + | Vint n, (Mint64 | Mfloat32 | Mfloat64 | Many64), _ => v2 = Vundef | Vint n, _, _ => True (**r nothing meaningful to say about v2 *) - | Vptr b ofs, Mint32, Mint32 => v2 = Vptr b ofs + | Vptr b ofs, (Mint32 | Many32), (Mint32 | Many32) => v2 = Vptr b ofs + | Vptr b ofs, Many64, Many64 => v2 = Vptr b ofs | Vptr b ofs, _, _ => v2 = Vundef | Vlong n, Mint64, Mint64 => v2 = Vlong n - | Vlong n, Mint64, Mfloat64 => v2 = Vfloat(Float.double_of_bits n) - | Vlong n, (Mint8signed|Mint8unsigned|Mint16signed|Mint16unsigned|Mint32|Mfloat32|Mfloat64), _ => v2 = Vundef + | Vlong n, Mint64, Mfloat64 => v2 = Vfloat(Float.of_bits n) + | Vlong n, Many64, Many64 => v2 = Vlong n + | Vlong n, (Mint8signed|Mint8unsigned|Mint16signed|Mint16unsigned|Mint32|Mfloat32|Mfloat64|Many32), _ => v2 = Vundef | Vlong n, _, _ => True (**r nothing meaningful to say about v2 *) - | Vfloat f, Mfloat32, Mfloat32 => v2 = Vfloat(Float.singleoffloat f) - | Vfloat f, Mfloat32, Mint32 => v2 = Vint(Float.bits_of_single f) | Vfloat f, Mfloat64, Mfloat64 => v2 = Vfloat f - | Vfloat f, (Mint8signed|Mint8unsigned|Mint16signed|Mint16unsigned|Mint32|Mint64), _ => v2 = Vundef - | Vfloat f, Mfloat64, Mint64 => v2 = Vlong(Float.bits_of_double f) + | Vfloat f, Mfloat64, Mint64 => v2 = Vlong(Float.to_bits f) + | Vfloat f, Many64, Many64 => v2 = Vfloat f + | Vfloat f, (Mint8signed|Mint8unsigned|Mint16signed|Mint16unsigned|Mint32|Mfloat32|Mint64|Many32), _ => v2 = Vundef | Vfloat f, _, _ => True (* nothing interesting to say about v2 *) + | Vsingle f, Mfloat32, Mfloat32 => v2 = Vsingle f + | Vsingle f, Mfloat32, Mint32 => v2 = Vint(Float32.to_bits f) + | Vsingle f, Many32, Many32 => v2 = Vsingle f + | Vsingle f, Many64, Many64 => v2 = Vsingle f + | Vsingle f, (Mint8signed|Mint8unsigned|Mint16signed|Mint16unsigned|Mint32|Mint64|Mfloat64|Many64), _ => v2 = Vundef + | Vsingle f, _, _ => True (* nothing interesting to say about v2 *) end. Remark decode_val_undef: @@ -417,14 +479,23 @@ Proof. intros. unfold decode_val. simpl. destruct chunk; auto. Qed. +Remark proj_bytes_inj_value: + forall q v, proj_bytes (inj_value q v) = None. +Proof. + intros. destruct q; reflexivity. +Qed. + Lemma decode_encode_val_general: forall v chunk1 chunk2, decode_encode_val v chunk1 chunk2 (decode_val chunk2 (encode_val chunk1 v)). Proof. -Opaque inj_pointer. +Opaque inj_value. intros. - destruct v; destruct chunk1; simpl; try (apply decode_val_undef); - destruct chunk2; unfold decode_val; auto; try (rewrite proj_inj_bytes). + destruct v; destruct chunk1 eqn:C1; simpl; try (apply decode_val_undef); + destruct chunk2 eqn:C2; unfold decode_val; auto; + try (rewrite proj_inj_bytes); try (rewrite proj_bytes_inj_value); + try (rewrite proj_inj_value); try (rewrite proj_inj_value_mismatch by congruence); + auto. rewrite decode_encode_int_1. decEq. apply Int.sign_ext_zero_ext. omega. rewrite decode_encode_int_1. decEq. apply Int.zero_ext_idem. omega. rewrite decode_encode_int_1. decEq. apply Int.sign_ext_zero_ext. omega. @@ -437,14 +508,10 @@ Opaque inj_pointer. rewrite decode_encode_int_4. auto. rewrite decode_encode_int_8. auto. rewrite decode_encode_int_8. auto. - rewrite decode_encode_int_4. auto. - rewrite decode_encode_int_4. decEq. apply Float.single_of_bits_of_single. rewrite decode_encode_int_8. auto. - rewrite decode_encode_int_8. decEq. apply Float.double_of_bits_of_double. - change (proj_bytes (inj_pointer 4 b i)) with (@None (list byte)). simpl. - unfold proj_pointer. generalize (check_inj_pointer b i 4%nat). -Transparent inj_pointer. - simpl. intros EQ; rewrite EQ; auto. + rewrite decode_encode_int_8. decEq. apply Float.of_to_bits. + rewrite decode_encode_int_4. auto. + rewrite decode_encode_int_4. decEq. apply Float32.of_to_bits. Qed. Lemma decode_encode_val_similar: @@ -465,12 +532,8 @@ Lemma decode_val_type: Proof. intros. unfold decode_val. destruct (proj_bytes cl). - destruct chunk; simpl; auto. apply Float.single_of_bits_is_single. destruct chunk; simpl; auto. - unfold proj_pointer. destruct cl; try (exact I). - destruct m; try (exact I). - destruct (check_pointer 4%nat b i (Pointer b i n :: cl)); - exact I. + destruct chunk; exact I || apply Val.load_result_type. Qed. Lemma encode_val_int8_signed_unsigned: @@ -518,7 +581,6 @@ Lemma decode_val_cast: | Mint8unsigned => v = Val.zero_ext 8 v | Mint16signed => v = Val.sign_ext 16 v | Mint16unsigned => v = Val.zero_ext 16 v - | Mfloat32 => v = Val.singleoffloat v | _ => True end. Proof. @@ -527,145 +589,132 @@ Proof. unfold Val.zero_ext. rewrite Int.zero_ext_idem; auto. omega. unfold Val.sign_ext. rewrite Int.sign_ext_idem; auto. omega. unfold Val.zero_ext. rewrite Int.zero_ext_idem; auto. omega. - simpl. rewrite Float.singleoffloat_of_bits. auto. Qed. (** Pointers cannot be forged. *) -Definition memval_valid_first (mv: memval) : Prop := - match mv with - | Pointer b ofs n => n = 3%nat - | _ => True - end. - -Definition memval_valid_cont (mv: memval) : Prop := - match mv with - | Pointer b ofs n => n <> 3%nat - | _ => True +Definition quantity_chunk (chunk: memory_chunk) := + match chunk with + | Mint64 | Mfloat64 | Many64 => Q64 + | _ => Q32 end. -Inductive encoding_shape: list memval -> Prop := - | encoding_shape_intro: forall mv1 mvl, - memval_valid_first mv1 -> - (forall mv, In mv mvl -> memval_valid_cont mv) -> - encoding_shape (mv1 :: mvl). - -Lemma encode_val_shape: - forall chunk v, encoding_shape (encode_val chunk v). +Inductive shape_encoding (chunk: memory_chunk) (v: val): list memval -> Prop := + | shape_encoding_f: forall q i mvl, + (chunk = Mint32 \/ chunk = Many32 \/ chunk = Many64) -> + q = quantity_chunk chunk -> + S i = size_quantity_nat q -> + (forall mv, In mv mvl -> exists j, mv = Fragment v q j /\ S j <> size_quantity_nat q) -> + shape_encoding chunk v (Fragment v q i :: mvl) + | shape_encoding_b: forall b mvl, + match v with Vint _ => True | Vlong _ => True | Vfloat _ => True | Vsingle _ => True | _ => False end -> + (forall mv, In mv mvl -> exists b', mv = Byte b') -> + shape_encoding chunk v (Byte b :: mvl) + | shape_encoding_u: forall mvl, + (forall mv, In mv mvl -> mv = Undef) -> + shape_encoding chunk v (Undef :: mvl). + +Lemma encode_val_shape: forall chunk v, shape_encoding chunk v (encode_val chunk v). Proof. intros. - destruct (size_chunk_nat_pos chunk) as [sz1 EQ]. - assert (A: encoding_shape (list_repeat (size_chunk_nat chunk) Undef)). - rewrite EQ; simpl; constructor. exact I. - intros. replace mv with Undef. exact I. symmetry; eapply in_list_repeat; eauto. - assert (B: forall bl, length bl = size_chunk_nat chunk -> - encoding_shape (inj_bytes bl)). - intros. destruct bl; simpl in *. congruence. - constructor. exact I. unfold inj_bytes. intros. - exploit list_in_map_inv; eauto. intros [x [C D]]. subst mv. exact I. - destruct v; auto; destruct chunk; simpl; auto; try (apply B; apply encode_int_length). - constructor. red. auto. - simpl; intros. intuition; subst mv; red; simpl; congruence. -Qed. - -Lemma check_pointer_inv: - forall b ofs n mv, - check_pointer n b ofs mv = true -> mv = inj_pointer n b ofs. -Proof. - induction n; destruct mv; simpl. - auto. - congruence. - congruence. - destruct m; try congruence. intro. - destruct (andb_prop _ _ H). destruct (andb_prop _ _ H0). - destruct (andb_prop _ _ H2). - decEq. decEq. symmetry; eapply proj_sumbool_true; eauto. - symmetry; eapply proj_sumbool_true; eauto. - symmetry; apply beq_nat_true; auto. - auto. + destruct (size_chunk_nat_pos chunk) as [sz EQ]. + assert (A: forall mv q n, + (n < size_quantity_nat q)%nat -> + In mv (inj_value_rec n v q) -> + exists j, mv = Fragment v q j /\ S j <> size_quantity_nat q). + { + induction n; simpl; intros. contradiction. destruct H0. + exists n; split; auto. omega. apply IHn; auto. omega. + } + assert (B: forall q, + q = quantity_chunk chunk -> + (chunk = Mint32 \/ chunk = Many32 \/ chunk = Many64) -> + shape_encoding chunk v (inj_value q v)). + { +Local Transparent inj_value. + intros. unfold inj_value. destruct (size_quantity_nat_pos q) as [sz' EQ']. + rewrite EQ'. simpl. constructor; auto. + intros; eapply A; eauto. omega. + } + assert (C: forall bl, + match v with Vint _ => True | Vlong _ => True | Vfloat _ => True | Vsingle _ => True | _ => False end -> + length (inj_bytes bl) = size_chunk_nat chunk -> + shape_encoding chunk v (inj_bytes bl)). + { + intros. destruct bl as [|b1 bl]. simpl in H0; congruence. simpl. + constructor; auto. unfold inj_bytes; intros. exploit list_in_map_inv; eauto. + intros (b & P & Q); exists b; auto. + } + assert (D: shape_encoding chunk v (list_repeat (size_chunk_nat chunk) Undef)). + { + intros. rewrite EQ; simpl; constructor; auto. + intros. eapply in_list_repeat; eauto. + } + generalize (encode_val_length chunk v). intros LEN. + unfold encode_val; unfold encode_val in LEN; destruct v; destruct chunk; (apply B || apply C || apply D); auto; red; auto. Qed. -Inductive decoding_shape: list memval -> Prop := - | decoding_shape_intro: forall mv1 mvl, - memval_valid_first mv1 -> mv1 <> Undef -> - (forall mv, In mv mvl -> memval_valid_cont mv /\ mv <> Undef) -> - decoding_shape (mv1 :: mvl). - -Lemma decode_val_shape: - forall chunk mvl, - List.length mvl = size_chunk_nat chunk -> - decode_val chunk mvl = Vundef \/ decoding_shape mvl. +Inductive shape_decoding (chunk: memory_chunk): list memval -> val -> Prop := + | shape_decoding_f: forall v q i mvl, + (chunk = Mint32 \/ chunk = Many32 \/ chunk = Many64) -> + q = quantity_chunk chunk -> + S i = size_quantity_nat q -> + (forall mv, In mv mvl -> exists j, mv = Fragment v q j /\ S j <> size_quantity_nat q) -> + shape_decoding chunk (Fragment v q i :: mvl) (Val.load_result chunk v) + | shape_decoding_b: forall b mvl v, + match v with Vint _ => True | Vlong _ => True | Vfloat _ => True | Vsingle _ => True | _ => False end -> + (forall mv, In mv mvl -> exists b', mv = Byte b') -> + shape_decoding chunk (Byte b :: mvl) v + | shape_decoding_u: forall mvl, + shape_decoding chunk mvl Vundef. + +Lemma decode_val_shape: forall chunk mv1 mvl, + shape_decoding chunk (mv1 :: mvl) (decode_val chunk (mv1 :: mvl)). Proof. - intros. destruct (size_chunk_nat_pos chunk) as [sz EQ]. + intros. + assert (A: forall mv mvs bs, proj_bytes mvs = Some bs -> In mv mvs -> + exists b, mv = Byte b). + { + induction mvs; simpl; intros. + contradiction. + destruct a; try discriminate. destruct H0. exists i; auto. + destruct (proj_bytes mvs); try discriminate. eauto. + } + assert (B: forall v q mv n mvs, + check_value n v q mvs = true -> In mv mvs -> (n < size_quantity_nat q)%nat -> + exists j, mv = Fragment v q j /\ S j <> size_quantity_nat q). + { + induction n; destruct mvs; simpl; intros; try discriminate. + contradiction. + destruct m; try discriminate. InvBooleans. apply beq_nat_true in H4. subst. + destruct H0. subst mv. exists n0; split; auto. omega. + eapply IHn; eauto. omega. + } + assert (U: forall mvs, shape_decoding chunk mvs (Val.load_result chunk Vundef)). + { + intros. replace (Val.load_result chunk Vundef) with Vundef. constructor. + destruct chunk; auto. + } + assert (C: forall q, size_quantity_nat q = size_chunk_nat chunk -> + (chunk = Mint32 \/ chunk = Many32 \/ chunk = Many64) -> + shape_decoding chunk (mv1 :: mvl) (Val.load_result chunk (proj_value q (mv1 :: mvl)))). + { + intros. unfold proj_value. destruct mv1; auto. + destruct (size_quantity_nat_pos q) as [sz EQ]. rewrite EQ. + simpl. unfold proj_sumbool. rewrite dec_eq_true. + destruct (quantity_eq q q0); auto. + destruct (beq_nat sz n) eqn:EQN; auto. + destruct (check_value sz v q mvl) eqn:CHECK; auto. + simpl. apply beq_nat_true in EQN. subst n q0. constructor. auto. + destruct H0 as [E|[E|E]]; subst chunk; destruct q; auto || discriminate. + congruence. + intros. eapply B; eauto. omega. + } unfold decode_val. - caseEq (proj_bytes mvl). - intros bl PROJ. right. exploit inj_proj_bytes; eauto. intros. subst mvl. - destruct bl; simpl in H. congruence. simpl. constructor. - red; auto. congruence. - unfold inj_bytes; intros. exploit list_in_map_inv; eauto. intros [b [A B]]. - subst mv. split. red; auto. congruence. - intros. destruct chunk; auto. unfold proj_pointer. - destruct mvl; auto. destruct m; auto. - caseEq (check_pointer 4%nat b i (Pointer b i n :: mvl)); auto. - intros. right. exploit check_pointer_inv; eauto. simpl; intros; inv H2. - constructor. red. auto. congruence. - simpl; intros. intuition; subst mv; simpl; congruence. -Qed. - -Lemma encode_val_pointer_inv: - forall chunk v b ofs n mvl, - encode_val chunk v = Pointer b ofs n :: mvl -> - chunk = Mint32 /\ v = Vptr b ofs /\ mvl = inj_pointer 3%nat b ofs. -Proof. - intros until mvl. - assert (A: list_repeat (size_chunk_nat chunk) Undef = Pointer b ofs n :: mvl -> - chunk = Mint32 /\ v = Vptr b ofs /\ mvl = inj_pointer 3 b ofs). - intros. destruct (size_chunk_nat_pos chunk) as [sz SZ]. rewrite SZ in H. simpl in H. discriminate. - assert (B: forall bl, length bl <> 0%nat -> inj_bytes bl = Pointer b ofs n :: mvl -> - chunk = Mint32 /\ v = Vptr b ofs /\ mvl = inj_pointer 3 b ofs). - intros. destruct bl; simpl in *; congruence. - unfold encode_val; destruct v; destruct chunk; - (apply A; assumption) || - (apply B; rewrite encode_int_length; congruence) || idtac. - simpl. intros EQ; inv EQ; auto. -Qed. - -Lemma decode_val_pointer_inv: - forall chunk mvl b ofs, - decode_val chunk mvl = Vptr b ofs -> - chunk = Mint32 /\ mvl = inj_pointer 4%nat b ofs. -Proof. - intros until ofs; unfold decode_val. - destruct (proj_bytes mvl). - destruct chunk; congruence. - destruct chunk; try congruence. - unfold proj_pointer. destruct mvl. congruence. destruct m; try congruence. - case_eq (check_pointer 4%nat b0 i (Pointer b0 i n :: mvl)); intros. - inv H0. split; auto. apply check_pointer_inv; auto. - congruence. -Qed. - -Inductive pointer_encoding_shape: list memval -> Prop := - | pointer_encoding_shape_intro: forall mv1 mvl, - ~memval_valid_cont mv1 -> - (forall mv, In mv mvl -> ~memval_valid_first mv) -> - pointer_encoding_shape (mv1 :: mvl). - -Lemma encode_pointer_shape: - forall b ofs, pointer_encoding_shape (encode_val Mint32 (Vptr b ofs)). -Proof. - intros. simpl. constructor. - unfold memval_valid_cont. red; intro. elim H. auto. - unfold memval_valid_first. simpl; intros; intuition; subst mv; congruence. -Qed. - -Lemma decode_pointer_shape: - forall chunk mvl b ofs, - decode_val chunk mvl = Vptr b ofs -> - chunk = Mint32 /\ pointer_encoding_shape mvl. -Proof. - intros. exploit decode_val_pointer_inv; eauto. intros [A B]. - split; auto. subst mvl. apply encode_pointer_shape. + destruct (proj_bytes (mv1 :: mvl)) as [bl|] eqn:PB. + exploit (A mv1); eauto with coqlib. intros [b1 EQ1]; subst mv1. + destruct chunk; (apply shape_decoding_u || apply shape_decoding_b); eauto with coqlib. + destruct chunk; (apply shape_decoding_u || apply C); auto. Qed. (** * Compatibility with memory injections *) @@ -675,18 +724,17 @@ Qed. Inductive memval_inject (f: meminj): memval -> memval -> Prop := | memval_inject_byte: forall n, memval_inject f (Byte n) (Byte n) - | memval_inject_ptr: - forall b1 ofs1 b2 ofs2 delta n, - f b1 = Some (b2, delta) -> - ofs2 = Int.add ofs1 (Int.repr delta) -> - memval_inject f (Pointer b1 ofs1 n) (Pointer b2 ofs2 n) + | memval_inject_frag: + forall v1 v2 q n, + val_inject f v1 v2 -> + memval_inject f (Fragment v1 q n) (Fragment v2 q n) | memval_inject_undef: forall mv, memval_inject f Undef mv. Lemma memval_inject_incr: forall f f' v1 v2, memval_inject f v1 v2 -> inject_incr f f' -> memval_inject f' v1 v2. Proof. - intros. inv H; econstructor. rewrite (H0 _ _ _ H1). reflexivity. auto. + intros. inv H; econstructor. eapply val_inject_incr; eauto. Qed. (** [decode_val], applied to lists of memory values that are pairwise @@ -706,38 +754,33 @@ Proof. congruence. Qed. -Lemma check_pointer_inject: +Lemma check_value_inject: forall f vl vl', list_forall2 (memval_inject f) vl vl' -> - forall n b ofs b' delta, - check_pointer n b ofs vl = true -> - f b = Some(b', delta) -> - check_pointer n b' (Int.add ofs (Int.repr delta)) vl' = true. + forall v v' q n, + check_value n v q vl = true -> + val_inject f v v' -> v <> Vundef -> + check_value n v' q vl' = true. Proof. induction 1; intros; destruct n; simpl in *; auto. inv H; auto. - destruct (andb_prop _ _ H1). destruct (andb_prop _ _ H). - destruct (andb_prop _ _ H5). - assert (n = n0) by (apply beq_nat_true; auto). - assert (b = b0) by (eapply proj_sumbool_true; eauto). - assert (ofs = ofs1) by (eapply proj_sumbool_true; eauto). - subst. rewrite H3 in H2; inv H2. - unfold proj_sumbool. rewrite dec_eq_true. rewrite dec_eq_true. - rewrite <- beq_nat_refl. simpl. eauto. - congruence. + InvBooleans. assert (n = n0) by (apply beq_nat_true; auto). subst v1 q0 n0. + replace v2 with v'. + unfold proj_sumbool; rewrite ! dec_eq_true. rewrite <- beq_nat_refl. simpl; eauto. + inv H2; try discriminate; inv H4; congruence. + discriminate. Qed. -Lemma proj_pointer_inject: - forall f vl1 vl2, +Lemma proj_value_inject: + forall f q vl1 vl2, list_forall2 (memval_inject f) vl1 vl2 -> - val_inject f (proj_pointer vl1) (proj_pointer vl2). + val_inject f (proj_value q vl1) (proj_value q vl2). Proof. - intros. unfold proj_pointer. + intros. unfold proj_value. inversion H; subst. auto. inversion H0; subst; auto. - case_eq (check_pointer 4%nat b0 ofs1 (Pointer b0 ofs1 n :: al)); intros. - exploit check_pointer_inject. eexact H. eauto. eauto. - intro. rewrite H4. econstructor; eauto. - constructor. + destruct (check_value (size_quantity_nat q) v1 q (Fragment v1 q0 n :: al)) eqn:B; auto. + destruct (Val.eq v1 Vundef). subst; auto. + erewrite check_value_inject by eauto. auto. Qed. Lemma proj_bytes_not_inject: @@ -754,9 +797,9 @@ Proof. auto. Qed. -Lemma check_pointer_undef: - forall n b ofs vl, - In Undef vl -> check_pointer n b ofs vl = false. +Lemma check_value_undef: + forall n q v vl, + In Undef vl -> check_value n v q vl = false. Proof. induction n; intros; simpl. destruct vl. elim H. auto. @@ -765,12 +808,12 @@ Proof. rewrite IHn; auto. apply andb_false_r. Qed. -Lemma proj_pointer_undef: - forall vl, In Undef vl -> proj_pointer vl = Vundef. +Lemma proj_value_undef: + forall q vl, In Undef vl -> proj_value q vl = Vundef. Proof. - intros; unfold proj_pointer. + intros; unfold proj_value. destruct vl; auto. destruct m; auto. - rewrite check_pointer_undef. auto. auto. + rewrite check_value_undef. auto. auto. Qed. Theorem decode_val_inject: @@ -779,13 +822,20 @@ Theorem decode_val_inject: val_inject f (decode_val chunk vl1) (decode_val chunk vl2). Proof. intros. unfold decode_val. - case_eq (proj_bytes vl1); intros. - exploit proj_bytes_inject; eauto. intros. rewrite H1. + destruct (proj_bytes vl1) as [bl1|] eqn:PB1. + exploit proj_bytes_inject; eauto. intros PB2. rewrite PB2. destruct chunk; constructor. + assert (A: forall q fn, + val_inject f (Val.load_result chunk (proj_value q vl1)) + (match proj_bytes vl2 with + | Some bl => fn bl + | None => Val.load_result chunk (proj_value q vl2) + end)). + { intros. destruct (proj_bytes vl2) as [bl2|] eqn:PB2. + rewrite proj_value_undef. destruct chunk; auto. eapply proj_bytes_not_inject; eauto. congruence. + apply val_load_result_inject. apply proj_value_inject; auto. + } destruct chunk; auto. - case_eq (proj_bytes vl2); intros. - rewrite proj_pointer_undef. auto. eapply proj_bytes_not_inject; eauto. congruence. - apply proj_pointer_inject; auto. Qed. (** Symmetrically, [encode_val], applied to values related by [val_inject], @@ -805,6 +855,13 @@ Proof. induction vl; simpl; constructor; auto. constructor. Qed. +Lemma repeat_Undef_inject_encode_val: + forall f chunk v, + list_forall2 (memval_inject f) (list_repeat (size_chunk_nat chunk) Undef) (encode_val chunk v). +Proof. + intros. rewrite <- (encode_val_length chunk v). apply repeat_Undef_inject_any. +Qed. + Lemma repeat_Undef_inject_self: forall f n, list_forall2 (memval_inject f) (list_repeat n Undef) (list_repeat n Undef). @@ -812,19 +869,24 @@ Proof. induction n; simpl; constructor; auto. constructor. Qed. +Lemma inj_value_inject: + forall f v1 v2 q, val_inject f v1 v2 -> list_forall2 (memval_inject f) (inj_value q v1) (inj_value q v2). +Proof. + intros. +Local Transparent inj_value. + unfold inj_value. generalize (size_quantity_nat q). induction n; simpl; constructor; auto. + constructor; auto. +Qed. + Theorem encode_val_inject: forall f v1 v2 chunk, val_inject f v1 v2 -> list_forall2 (memval_inject f) (encode_val chunk v1) (encode_val chunk v2). Proof. - intros. inv H; simpl. - destruct chunk; apply inj_bytes_inject || apply repeat_Undef_inject_self. - destruct chunk; apply inj_bytes_inject || apply repeat_Undef_inject_self. - destruct chunk; apply inj_bytes_inject || apply repeat_Undef_inject_self. - destruct chunk; try (apply repeat_Undef_inject_self). - repeat econstructor; eauto. - replace (size_chunk_nat chunk) with (length (encode_val chunk v2)). - apply repeat_Undef_inject_any. apply encode_val_length. + intros. inversion H; subst; simpl; destruct chunk; + auto using inj_bytes_inject, inj_value_inject, repeat_Undef_inject_self, repeat_Undef_inject_encode_val. + unfold encode_val. destruct v2; apply inj_value_inject; auto. + unfold encode_val. destruct v2; apply inj_value_inject; auto. Qed. Definition memval_lessdef: memval -> memval -> Prop := memval_inject inject_id. @@ -832,8 +894,7 @@ Definition memval_lessdef: memval -> memval -> Prop := memval_inject inject_id. Lemma memval_lessdef_refl: forall mv, memval_lessdef mv mv. Proof. - red. destruct mv; econstructor. - unfold inject_id; reflexivity. rewrite Int.add_zero; auto. + red. destruct mv; econstructor. apply val_inject_id. auto. Qed. (** [memval_inject] and compositions *) @@ -845,9 +906,8 @@ Lemma memval_inject_compose: Proof. intros. inv H. inv H0. constructor. - inv H0. econstructor. - unfold compose_meminj; rewrite H1; rewrite H5; eauto. - rewrite Int.add_assoc. decEq. unfold Int.add. apply Int.eqm_samerepr. auto with ints. + inv H0. econstructor. + eapply val_inject_compose; eauto. constructor. Qed. @@ -905,28 +965,22 @@ Qed. Lemma decode_val_int64: forall l1 l2, length l1 = 4%nat -> length l2 = 4%nat -> - decode_val Mint64 (l1 ++ l2) = - Val.longofwords (decode_val Mint32 (if Archi.big_endian then l1 else l2)) - (decode_val Mint32 (if Archi.big_endian then l2 else l1)). + Val.lessdef + (decode_val Mint64 (l1 ++ l2)) + (Val.longofwords (decode_val Mint32 (if Archi.big_endian then l1 else l2)) + (decode_val Mint32 (if Archi.big_endian then l2 else l1))). Proof. intros. unfold decode_val. - assert (PP: forall vl, match proj_pointer vl with Vundef => True | Vptr _ _ => True | _ => False end). - intros. unfold proj_pointer. destruct vl; auto. destruct m; auto. - destruct (check_pointer 4 b i (Pointer b i n :: vl)); auto. - assert (PP1: forall vl v2, Val.longofwords (proj_pointer vl) v2 = Vundef). - intros. generalize (PP vl). destruct (proj_pointer vl); reflexivity || contradiction. - assert (PP2: forall v1 vl, Val.longofwords v1 (proj_pointer vl) = Vundef). - intros. destruct v1; simpl; auto. - generalize (PP vl). destruct (proj_pointer vl); reflexivity || contradiction. rewrite proj_bytes_append. - destruct (proj_bytes l1) as [b1|] eqn:B1; destruct (proj_bytes l2) as [b2|] eqn:B2. -- exploit length_proj_bytes. eexact B1. rewrite H; intro L1. + destruct (proj_bytes l1) as [b1|] eqn:B1; destruct (proj_bytes l2) as [b2|] eqn:B2; auto. + exploit length_proj_bytes. eexact B1. rewrite H; intro L1. exploit length_proj_bytes. eexact B2. rewrite H0; intro L2. assert (UR: forall l, length l = 4%nat -> Int.unsigned (Int.repr (int_of_bytes l)) = int_of_bytes l). intros. apply Int.unsigned_repr. generalize (int_of_bytes_range l). rewrite H1. change (two_p (Z.of_nat 4 * 8)) with (Int.max_unsigned + 1). omega. + apply Val.lessdef_same. unfold decode_int, rev_if_be. destruct Archi.big_endian; rewrite B1; rewrite B2. + rewrite <- (rev_length b1) in L1. rewrite <- (rev_length b2) in L2. @@ -938,9 +992,6 @@ Proof. + unfold Val.longofwords. f_equal. rewrite Int64.ofwords_add. f_equal. rewrite !UR by auto. rewrite int_of_bytes_append. rewrite L1. change (Z.of_nat 4 * 8) with 32. ring. -- destruct Archi.big_endian; rewrite B1; rewrite B2; auto. -- destruct Archi.big_endian; rewrite B1; rewrite B2; auto. -- destruct Archi.big_endian; rewrite B1; rewrite B2; auto. Qed. Lemma bytes_of_int_append: diff --git a/common/Memory.v b/common/Memory.v index 9afdfd39..e45df664 100644 --- a/common/Memory.v +++ b/common/Memory.v @@ -690,7 +690,6 @@ Theorem load_cast: | Mint8unsigned => v = Val.zero_ext 8 v | Mint16signed => v = Val.sign_ext 16 v | Mint16unsigned => v = Val.zero_ext 16 v - | Mfloat32 => v = Val.singleoffloat v | _ => True end. Proof. @@ -727,24 +726,6 @@ Proof. rewrite pred_dec_false; auto. Qed. -(* -Theorem load_float64al32: - forall m b ofs v, - load Mfloat64 m b ofs = Some v -> load Mfloat64al32 m b ofs = Some v. -Proof. - unfold load; intros. destruct (valid_access_dec m Mfloat64 b ofs Readable); try discriminate. - rewrite pred_dec_true. assumption. - apply valid_access_compat with Mfloat64; auto. simpl; omega. -Qed. - -Theorem loadv_float64al32: - forall m a v, - loadv Mfloat64 m a = Some v -> loadv Mfloat64al32 m a = Some v. -Proof. - unfold loadv; intros. destruct a; auto. apply load_float64al32; auto. -Qed. -*) - (** ** Properties related to [loadbytes] *) Theorem range_perm_loadbytes: @@ -896,7 +877,7 @@ Theorem load_int64_split: exists v1 v2, load Mint32 m b ofs = Some (if Archi.big_endian then v1 else v2) /\ load Mint32 m b (ofs + 4) = Some (if Archi.big_endian then v2 else v1) - /\ v = Val.longofwords v1 v2. + /\ Val.lessdef v (Val.longofwords v1 v2). Proof. intros. exploit load_valid_access; eauto. intros [A B]. simpl in *. @@ -927,7 +908,7 @@ Theorem loadv_int64_split: exists v1 v2, loadv Mint32 m a = Some (if Archi.big_endian then v1 else v2) /\ loadv Mint32 m (Val.add a (Vint (Int.repr 4))) = Some (if Archi.big_endian then v2 else v1) - /\ v = Val.longofwords v1 v2. + /\ Val.lessdef v (Val.longofwords v1 v2). Proof. intros. destruct a; simpl in H; try discriminate. exploit load_int64_split; eauto. intros (v1 & v2 & L1 & L2 & EQ). @@ -1138,18 +1119,17 @@ Proof. red; intro; elim n0; red; intros; eauto with mem. Qed. -Lemma setN_property: - forall (P: memval -> Prop) vl p q c, - (forall v, In v vl -> P v) -> +Lemma setN_in: + forall vl p q c, p <= q < p + Z_of_nat (length vl) -> - P(ZMap.get q (setN vl p c)). + In (ZMap.get q (setN vl p c)) vl. Proof. induction vl; intros. - simpl in H0. omegaContradiction. - simpl length in H0. rewrite inj_S in H0. simpl. + simpl in H. omegaContradiction. + simpl length in H. rewrite inj_S in H. simpl. destruct (zeq p q). subst q. rewrite setN_outside. rewrite ZMap.gss. auto with coqlib. omega. - apply IHvl. auto with coqlib. omega. + right. apply IHvl. omega. Qed. Lemma getN_in: @@ -1164,84 +1144,114 @@ Proof. right. apply IHn. omega. Qed. +End STORE. + +Local Hint Resolve perm_store_1 perm_store_2: mem. +Local Hint Resolve store_valid_block_1 store_valid_block_2: mem. +Local Hint Resolve store_valid_access_1 store_valid_access_2 + store_valid_access_3: mem. + +Lemma load_store_overlap: + forall chunk m1 b ofs v m2 chunk' ofs' v', + store chunk m1 b ofs v = Some m2 -> + load chunk' m2 b ofs' = Some v' -> + ofs' + size_chunk chunk' > ofs -> + ofs + size_chunk chunk > ofs' -> + exists mv1 mvl mv1' mvl', + shape_encoding chunk v (mv1 :: mvl) + /\ shape_decoding chunk' (mv1' :: mvl') v' + /\ ( (ofs' = ofs /\ mv1' = mv1) + \/ (ofs' > ofs /\ In mv1' mvl) + \/ (ofs' < ofs /\ In mv1 mvl')). +Proof. + intros. + exploit load_result; eauto. erewrite store_mem_contents by eauto; simpl. + rewrite PMap.gss. + set (c := (mem_contents m1)#b). intros V'. + destruct (size_chunk_nat_pos chunk) as [sz SIZE]. + destruct (size_chunk_nat_pos chunk') as [sz' SIZE']. + destruct (encode_val chunk v) as [ | mv1 mvl] eqn:ENC. + generalize (encode_val_length chunk v); rewrite ENC; simpl; congruence. + set (c' := setN (mv1::mvl) ofs c) in *. + exists mv1, mvl, (ZMap.get ofs' c'), (getN sz' (ofs' + 1) c'). + split. rewrite <- ENC. apply encode_val_shape. + split. rewrite V', SIZE'. apply decode_val_shape. + destruct (zeq ofs' ofs). +- subst ofs'. left; split. auto. unfold c'. simpl. + rewrite setN_outside by omega. apply ZMap.gss. +- right. destruct (zlt ofs ofs'). +(* If ofs < ofs': the load reads (at ofs') a continuation byte from the write. + ofs ofs' ofs+|chunk| + [-------------------] write + [-------------------] read +*) ++ left; split. omega. unfold c'. simpl. apply setN_in. + assert (Z.of_nat (length (mv1 :: mvl)) = size_chunk chunk). + { rewrite <- ENC; rewrite encode_val_length. rewrite size_chunk_conv; auto. } + simpl length in H3. rewrite inj_S in H3. omega. +(* If ofs > ofs': the load reads (at ofs) the first byte from the write. + ofs' ofs ofs'+|chunk'| + [-------------------] write + [----------------] read +*) ++ right; split. omega. replace mv1 with (ZMap.get ofs c'). + apply getN_in. + assert (size_chunk chunk' = Zsucc (Z.of_nat sz')). + { rewrite size_chunk_conv. rewrite SIZE'. rewrite inj_S; auto. } + omega. + unfold c'. simpl. rewrite setN_outside by omega. apply ZMap.gss. +Qed. + +Definition compat_pointer_chunks (chunk1 chunk2: memory_chunk) : Prop := + match chunk1, chunk2 with + | (Mint32 | Many32), (Mint32 | Many32) => True + | Many64, Many64 => True + | _, _ => False + end. + +Lemma compat_pointer_chunks_true: + forall chunk1 chunk2, + (chunk1 = Mint32 \/ chunk1 = Many32 \/ chunk1 = Many64) -> + (chunk2 = Mint32 \/ chunk2 = Many32 \/ chunk2 = Many64) -> + quantity_chunk chunk1 = quantity_chunk chunk2 -> + compat_pointer_chunks chunk1 chunk2. +Proof. + intros. destruct H as [P|[P|P]]; destruct H0 as [Q|[Q|Q]]; + subst; red; auto; discriminate. +Qed. + Theorem load_pointer_store: - forall chunk' b' ofs' v_b v_o, + forall chunk m1 b ofs v m2 chunk' b' ofs' v_b v_o, + store chunk m1 b ofs v = Some m2 -> load chunk' m2 b' ofs' = Some(Vptr v_b v_o) -> - (chunk = Mint32 /\ v = Vptr v_b v_o /\ chunk' = Mint32 /\ b' = b /\ ofs' = ofs) + (v = Vptr v_b v_o /\ compat_pointer_chunks chunk chunk' /\ b' = b /\ ofs' = ofs) \/ (b' <> b \/ ofs' + size_chunk chunk' <= ofs \/ ofs + size_chunk chunk <= ofs'). Proof. - intros. exploit load_result; eauto. rewrite store_mem_contents; simpl. - rewrite PMap.gsspec. destruct (peq b' b); auto. subst b'. intro DEC. + intros. + destruct (peq b' b); auto. subst b'. destruct (zle (ofs' + size_chunk chunk') ofs); auto. destruct (zle (ofs + size_chunk chunk) ofs'); auto. - destruct (size_chunk_nat_pos chunk) as [sz SZ]. - destruct (size_chunk_nat_pos chunk') as [sz' SZ']. - exploit decode_pointer_shape; eauto. intros [CHUNK' PSHAPE]. clear CHUNK'. - generalize (encode_val_shape chunk v). intro VSHAPE. - set (c := m1.(mem_contents)#b) in *. - set (c' := setN (encode_val chunk v) ofs c) in *. - destruct (zeq ofs ofs'). - -(* 1. ofs = ofs': must be same chunks and same value *) - subst ofs'. inv VSHAPE. - exploit decode_val_pointer_inv; eauto. intros [A B]. - subst chunk'. simpl in B. inv B. - generalize H4. unfold c'. rewrite <- H0. simpl. - rewrite setN_outside; try omega. rewrite ZMap.gss. intros. - exploit (encode_val_pointer_inv chunk v v_b v_o). - rewrite <- H0. subst mv1. eauto. intros [C [D E]]. - left; auto. - - destruct (zlt ofs ofs'). - -(* 2. ofs < ofs': - - ofs ofs' ofs+|chunk| - [-------------------] write - [-------------------] read - - The byte at ofs' satisfies memval_valid_cont (consequence of write). - For the read to return a pointer, it must satisfy ~memval_valid_cont. -*) - elimtype False. - assert (~memval_valid_cont (ZMap.get ofs' c')). - rewrite SZ' in PSHAPE. simpl in PSHAPE. inv PSHAPE. auto. - assert (memval_valid_cont (ZMap.get ofs' c')). - inv VSHAPE. unfold c'. rewrite <- H1. simpl. - apply setN_property. auto. - assert (length mvl = sz). - generalize (encode_val_length chunk v). rewrite <- H1. rewrite SZ. - simpl; congruence. - rewrite H4. rewrite size_chunk_conv in *. omega. - contradiction. - -(* 3. ofs > ofs': - - ofs' ofs ofs'+|chunk'| - [-------------------] write - [----------------] read - - The byte at ofs satisfies memval_valid_first (consequence of write). - For the read to return a pointer, it must satisfy ~memval_valid_first. -*) - elimtype False. - assert (memval_valid_first (ZMap.get ofs c')). - inv VSHAPE. unfold c'. rewrite <- H0. simpl. - rewrite setN_outside. rewrite ZMap.gss. auto. omega. - assert (~memval_valid_first (ZMap.get ofs c')). - rewrite SZ' in PSHAPE. simpl in PSHAPE. inv PSHAPE. - apply H4. apply getN_in. rewrite size_chunk_conv in *. - rewrite SZ' in *. rewrite inj_S in *. omega. - contradiction. + exploit load_store_overlap; eauto. + intros (mv1 & mvl & mv1' & mvl' & ENC & DEC & CASES). + inv DEC; try contradiction. + destruct CASES as [(A & B) | [(A & B) | (A & B)]]. +- (* Same offset *) + subst. inv ENC. + assert (chunk = Mint32 \/ chunk = Many32 \/ chunk = Many64) + by (destruct chunk; auto || contradiction). + left; split. rewrite H3. + destruct H4 as [P|[P|P]]; subst chunk'; destruct v0; simpl in H3; congruence. + split. apply compat_pointer_chunks_true; auto. + auto. +- (* ofs' > ofs *) + inv ENC. + + exploit H10; eauto. intros (j & P & Q). inv P. congruence. + + exploit H8; eauto. intros (n & P); congruence. + + exploit H2; eauto. congruence. +- (* ofs' < ofs *) + exploit H7; eauto. intros (j & P & Q). subst mv1. inv ENC. congruence. Qed. -End STORE. - -Local Hint Resolve perm_store_1 perm_store_2: mem. -Local Hint Resolve store_valid_block_1 store_valid_block_2: mem. -Local Hint Resolve store_valid_access_1 store_valid_access_2 - store_valid_access_3: mem. - Theorem load_store_pointer_overlap: forall chunk m1 b ofs v_b v_o m2 chunk' ofs' v, store chunk m1 b ofs (Vptr v_b v_o) = Some m2 -> @@ -1251,102 +1261,37 @@ Theorem load_store_pointer_overlap: ofs + size_chunk chunk > ofs' -> v = Vundef. Proof. - intros. - exploit store_mem_contents; eauto. intro ST. - exploit load_result; eauto. intro LD. - rewrite LD; clear LD. -Opaque encode_val. - rewrite ST; simpl. - rewrite PMap.gss. - set (c := m1.(mem_contents)#b). - set (c' := setN (encode_val chunk (Vptr v_b v_o)) ofs c). - destruct (decode_val_shape chunk' (getN (size_chunk_nat chunk') ofs' c')) - as [OK | VSHAPE]. - apply getN_length. - exact OK. - elimtype False. - destruct (size_chunk_nat_pos chunk) as [sz SZ]. - destruct (size_chunk_nat_pos chunk') as [sz' SZ']. - assert (ENC: encode_val chunk (Vptr v_b v_o) = list_repeat (size_chunk_nat chunk) Undef - \/ pointer_encoding_shape (encode_val chunk (Vptr v_b v_o))). - destruct chunk; try (left; reflexivity). - right. apply encode_pointer_shape. - assert (GET: getN (size_chunk_nat chunk) ofs c' = encode_val chunk (Vptr v_b v_o)). - unfold c'. rewrite <- (encode_val_length chunk (Vptr v_b v_o)). - apply getN_setN_same. - destruct (zlt ofs ofs'). - -(* ofs < ofs': - - ofs ofs' ofs+|chunk| - [-------------------] write - [-------------------] read - - The byte at ofs' is Undef or not memval_valid_first (because write of pointer). - The byte at ofs' must be memval_valid_first and not Undef (otherwise load returns Vundef). -*) - assert (memval_valid_first (ZMap.get ofs' c') /\ ZMap.get ofs' c' <> Undef). - rewrite SZ' in VSHAPE. simpl in VSHAPE. inv VSHAPE. auto. - assert (~memval_valid_first (ZMap.get ofs' c') \/ ZMap.get ofs' c' = Undef). - unfold c'. destruct ENC. - right. apply setN_property. rewrite H5. intros. eapply in_list_repeat; eauto. - rewrite encode_val_length. rewrite <- size_chunk_conv. omega. - left. revert H5. rewrite <- GET. rewrite SZ. simpl. intros. inv H5. - apply setN_property. apply H9. rewrite getN_length. - rewrite size_chunk_conv in H3. rewrite SZ in H3. rewrite inj_S in H3. omega. - intuition. - -(* ofs > ofs': - - ofs' ofs ofs'+|chunk'| - [-------------------] write - [----------------] read - - The byte at ofs is Undef or not memval_valid_cont (because write of pointer). - The byte at ofs must be memval_valid_cont and not Undef (otherwise load returns Vundef). -*) - assert (memval_valid_cont (ZMap.get ofs c') /\ ZMap.get ofs c' <> Undef). - rewrite SZ' in VSHAPE. simpl in VSHAPE. inv VSHAPE. - apply H8. apply getN_in. rewrite size_chunk_conv in H2. - rewrite SZ' in H2. rewrite inj_S in H2. omega. - assert (~memval_valid_cont (ZMap.get ofs c') \/ ZMap.get ofs c' = Undef). - elim ENC. - rewrite <- GET. rewrite SZ. simpl. intros. right; congruence. - rewrite <- GET. rewrite SZ. simpl. intros. inv H5. auto. - intuition. + intros. + exploit load_store_overlap; eauto. + intros (mv1 & mvl & mv1' & mvl' & ENC & DEC & CASES). + destruct CASES as [(A & B) | [(A & B) | (A & B)]]. +- congruence. +- inv ENC. + + exploit H9; eauto. intros (j & P & Q). subst mv1'. inv DEC. congruence. auto. + + contradiction. + + exploit H5; eauto. intros; subst. inv DEC; auto. +- inv DEC. + + exploit H10; eauto. intros (j & P & Q). subst mv1. inv ENC. congruence. + + exploit H8; eauto. intros (n & P). subst mv1. inv ENC. contradiction. + + auto. Qed. Theorem load_store_pointer_mismatch: forall chunk m1 b ofs v_b v_o m2 chunk' v, store chunk m1 b ofs (Vptr v_b v_o) = Some m2 -> load chunk' m2 b ofs = Some v -> - chunk <> Mint32 \/ chunk' <> Mint32 -> + ~compat_pointer_chunks chunk chunk' -> v = Vundef. Proof. intros. - exploit store_mem_contents; eauto. intro ST. - exploit load_result; eauto. intro LD. - rewrite LD; clear LD. -Opaque encode_val. - rewrite ST; simpl. - rewrite PMap.gss. - set (c1 := m1.(mem_contents)#b). - set (e := encode_val chunk (Vptr v_b v_o)). - destruct (size_chunk_nat_pos chunk) as [sz SZ]. - destruct (size_chunk_nat_pos chunk') as [sz' SZ']. - assert (match e with - | Undef :: _ => True - | Pointer _ _ _ :: _ => chunk = Mint32 - | _ => False - end). -Transparent encode_val. - unfold e, encode_val. rewrite SZ. destruct chunk; simpl; auto. - destruct e as [ | e1 el]. contradiction. - rewrite SZ'. simpl. rewrite setN_outside. rewrite ZMap.gss. - destruct e1; try contradiction. - destruct chunk'; auto. - destruct chunk'; auto. intuition. - omega. + exploit load_store_overlap; eauto. + generalize (size_chunk_pos chunk'); omega. + generalize (size_chunk_pos chunk); omega. + intros (mv1 & mvl & mv1' & mvl' & ENC & DEC & CASES). + destruct CASES as [(A & B) | [(A & B) | (A & B)]]; try omegaContradiction. + inv ENC; inv DEC; auto. +- elim H1. apply compat_pointer_chunks_true; auto. +- contradiction. Qed. Lemma store_similar_chunks: @@ -1403,16 +1348,6 @@ Theorem store_int16_sign_ext: store Mint16signed m b ofs (Vint n). Proof. intros. apply store_similar_chunks. apply encode_val_int16_sign_ext. auto. Qed. -Theorem store_float32_truncate: - forall m b ofs n, - store Mfloat32 m b ofs (Vfloat (Float.singleoffloat n)) = - store Mfloat32 m b ofs (Vfloat n). -Proof. - intros. apply store_similar_chunks. simpl. decEq. - repeat rewrite encode_float32_eq. rewrite Float.bits_of_singleoffloat. auto. - auto. -Qed. - (* Theorem store_float64al32: forall m b ofs v m', diff --git a/common/Memtype.v b/common/Memtype.v index 37ab33c2..1ab1cb73 100644 --- a/common/Memtype.v +++ b/common/Memtype.v @@ -310,7 +310,6 @@ Axiom load_cast: | Mint8unsigned => v = Val.zero_ext 8 v | Mint16signed => v = Val.sign_ext 16 v | Mint16unsigned => v = Val.zero_ext 16 v - | Mfloat32 => v = Val.singleoffloat v | _ => True end. @@ -443,6 +442,13 @@ Axiom load_store_other: (** Integrity of pointer values. *) +Definition compat_pointer_chunks (chunk1 chunk2: memory_chunk) : Prop := + match chunk1, chunk2 with + | (Mint32 | Many32), (Mint32 | Many32) => True + | Many64, Many64 => True + | _, _ => False + end. + Axiom load_store_pointer_overlap: forall chunk m1 b ofs v_b v_o m2 chunk' ofs' v, store chunk m1 b ofs (Vptr v_b v_o) = Some m2 -> @@ -455,13 +461,13 @@ Axiom load_store_pointer_mismatch: forall chunk m1 b ofs v_b v_o m2 chunk' v, store chunk m1 b ofs (Vptr v_b v_o) = Some m2 -> load chunk' m2 b ofs = Some v -> - chunk <> Mint32 \/ chunk' <> Mint32 -> + ~compat_pointer_chunks chunk chunk' -> v = Vundef. Axiom load_pointer_store: - forall chunk m1 b ofs v m2, store chunk m1 b ofs v = Some m2 -> - forall chunk' b' ofs' v_b v_o, + forall chunk m1 b ofs v m2 chunk' b' ofs' v_b v_o, + store chunk m1 b ofs v = Some m2 -> load chunk' m2 b' ofs' = Some(Vptr v_b v_o) -> - (chunk = Mint32 /\ v = Vptr v_b v_o /\ chunk' = Mint32 /\ b' = b /\ ofs' = ofs) + (v = Vptr v_b v_o /\ compat_pointer_chunks chunk chunk' /\ b' = b /\ ofs' = ofs) \/ (b' <> b \/ ofs' + size_chunk chunk' <= ofs \/ ofs + size_chunk chunk <= ofs'). (** Load-store properties for [loadbytes]. *) @@ -500,10 +506,6 @@ Axiom store_int16_sign_ext: forall m b ofs n, store Mint16signed m b ofs (Vint (Int.sign_ext 16 n)) = store Mint16signed m b ofs (Vint n). -Axiom store_float32_truncate: - forall m b ofs n, - store Mfloat32 m b ofs (Vfloat (Float.singleoffloat n)) = - store Mfloat32 m b ofs (Vfloat n). (** ** Properties of [storebytes]. *) diff --git a/common/PrintAST.ml b/common/PrintAST.ml index 329c0c79..c0eab04f 100644 --- a/common/PrintAST.ml +++ b/common/PrintAST.ml @@ -21,6 +21,8 @@ let name_of_type = function | Tfloat -> "float" | Tlong -> "long" | Tsingle -> "single" + | Tany32 -> "any32" + | Tany64 -> "any64" let name_of_chunk = function | Mint8signed -> "int8s" @@ -31,6 +33,8 @@ let name_of_chunk = function | Mint64 -> "int64" | Mfloat32 -> "float32" | Mfloat64 -> "float64" + | Many32 -> "any32" + | Many64 -> "any64" let name_of_external = function | EF_external(name, sg) -> sprintf "extern %S" (extern_atom name) diff --git a/common/Values.v b/common/Values.v index 5ac9f3ee..a12fb636 100644 --- a/common/Values.v +++ b/common/Values.v @@ -38,6 +38,7 @@ Inductive val: Type := | Vint: int -> val | Vlong: int64 -> val | Vfloat: float -> val + | Vsingle: float32 -> val | Vptr: block -> int -> val. Definition Vzero: val := Vint Int.zero. @@ -55,14 +56,28 @@ Definition Vfalse: val := Vint Int.zero. Module Val. +Definition eq (x y: val): {x=y} + {x<>y}. +Proof. + decide equality. + apply Int.eq_dec. + apply Int64.eq_dec. + apply Float.eq_dec. + apply Float32.eq_dec. + apply Int.eq_dec. + apply eq_block. +Defined. +Global Opaque eq. + Definition has_type (v: val) (t: typ) : Prop := match v, t with | Vundef, _ => True | Vint _, Tint => True | Vlong _, Tlong => True | Vfloat _, Tfloat => True - | Vfloat f, Tsingle => Float.is_single f + | Vsingle _, Tsingle => True | Vptr _ _, Tint => True + | (Vint _ | Vptr _ _ | Vsingle _), Tany32 => True + | _, Tany64 => True | _, _ => False end. @@ -83,8 +98,8 @@ Lemma has_subtype: forall ty1 ty2 v, subtype ty1 ty2 = true -> has_type v ty1 -> has_type v ty2. Proof. - intros. destruct ty1; destruct ty2; simpl in H; discriminate || assumption || idtac. - unfold has_type in *. destruct v; auto. + intros. destruct ty1; destruct ty2; simpl in H; discriminate || assumption || idtac; + unfold has_type in *; destruct v; auto. Qed. Lemma has_subtype_list: @@ -126,30 +141,66 @@ Definition absf (v: val) : val := | _ => Vundef end. +Definition negfs (v: val) : val := + match v with + | Vsingle f => Vsingle (Float32.neg f) + | _ => Vundef + end. + +Definition absfs (v: val) : val := + match v with + | Vsingle f => Vsingle (Float32.abs f) + | _ => Vundef + end. + Definition maketotal (ov: option val) : val := match ov with Some v => v | None => Vundef end. Definition intoffloat (v: val) : option val := match v with - | Vfloat f => option_map Vint (Float.intoffloat f) + | Vfloat f => option_map Vint (Float.to_int f) | _ => None end. Definition intuoffloat (v: val) : option val := match v with - | Vfloat f => option_map Vint (Float.intuoffloat f) + | Vfloat f => option_map Vint (Float.to_intu f) | _ => None end. Definition floatofint (v: val) : option val := match v with - | Vint n => Some (Vfloat (Float.floatofint n)) + | Vint n => Some (Vfloat (Float.of_int n)) | _ => None end. Definition floatofintu (v: val) : option val := match v with - | Vint n => Some (Vfloat (Float.floatofintu n)) + | Vint n => Some (Vfloat (Float.of_intu n)) + | _ => None + end. + +Definition intofsingle (v: val) : option val := + match v with + | Vsingle f => option_map Vint (Float32.to_int f) + | _ => None + end. + +Definition intuofsingle (v: val) : option val := + match v with + | Vsingle f => option_map Vint (Float32.to_intu f) + | _ => None + end. + +Definition singleofint (v: val) : option val := + match v with + | Vint n => Some (Vsingle (Float32.of_int n)) + | _ => None + end. + +Definition singleofintu (v: val) : option val := + match v with + | Vint n => Some (Vsingle (Float32.of_intu n)) | _ => None end. @@ -195,7 +246,13 @@ Definition sign_ext (nbits: Z) (v: val) : val := Definition singleoffloat (v: val) : val := match v with - | Vfloat f => Vfloat(Float.singleoffloat f) + | Vfloat f => Vsingle (Float.to_single f) + | _ => Vundef + end. + +Definition floatofsingle (v: val) : val := + match v with + | Vsingle f => Vfloat (Float.of_single f) | _ => Vundef end. @@ -394,6 +451,30 @@ Definition floatofwords (v1 v2: val) : val := | _, _ => Vundef end. +Definition addfs (v1 v2: val): val := + match v1, v2 with + | Vsingle f1, Vsingle f2 => Vsingle(Float32.add f1 f2) + | _, _ => Vundef + end. + +Definition subfs (v1 v2: val): val := + match v1, v2 with + | Vsingle f1, Vsingle f2 => Vsingle(Float32.sub f1 f2) + | _, _ => Vundef + end. + +Definition mulfs (v1 v2: val): val := + match v1, v2 with + | Vsingle f1, Vsingle f2 => Vsingle(Float32.mul f1 f2) + | _, _ => Vundef + end. + +Definition divfs (v1 v2: val): val := + match v1, v2 with + | Vsingle f1, Vsingle f2 => Vsingle(Float32.div f1 f2) + | _, _ => Vundef + end. + (** Operations on 64-bit integers *) Definition longofwords (v1 v2: val) : val := @@ -440,37 +521,49 @@ Definition longofintu (v: val) : val := Definition longoffloat (v: val) : option val := match v with - | Vfloat f => option_map Vlong (Float.longoffloat f) + | Vfloat f => option_map Vlong (Float.to_long f) | _ => None end. Definition longuoffloat (v: val) : option val := match v with - | Vfloat f => option_map Vlong (Float.longuoffloat f) + | Vfloat f => option_map Vlong (Float.to_longu f) + | _ => None + end. + +Definition longofsingle (v: val) : option val := + match v with + | Vsingle f => option_map Vlong (Float32.to_long f) + | _ => None + end. + +Definition longuofsingle (v: val) : option val := + match v with + | Vsingle f => option_map Vlong (Float32.to_longu f) | _ => None end. Definition floatoflong (v: val) : option val := match v with - | Vlong n => Some (Vfloat (Float.floatoflong n)) + | Vlong n => Some (Vfloat (Float.of_long n)) | _ => None end. Definition floatoflongu (v: val) : option val := match v with - | Vlong n => Some (Vfloat (Float.floatoflongu n)) + | Vlong n => Some (Vfloat (Float.of_longu n)) | _ => None end. Definition singleoflong (v: val) : option val := match v with - | Vlong n => Some (Vfloat (Float.singleoflong n)) + | Vlong n => Some (Vsingle (Float32.of_long n)) | _ => None end. Definition singleoflongu (v: val) : option val := match v with - | Vlong n => Some (Vfloat (Float.singleoflongu n)) + | Vlong n => Some (Vsingle (Float32.of_longu n)) | _ => None end. @@ -625,6 +718,12 @@ Definition cmpf_bool (c: comparison) (v1 v2: val): option bool := | _, _ => None end. +Definition cmpfs_bool (c: comparison) (v1 v2: val): option bool := + match v1, v2 with + | Vsingle f1, Vsingle f2 => Some (Float32.cmp c f1 f2) + | _, _ => None + end. + Definition cmpl_bool (c: comparison) (v1 v2: val): option bool := match v1, v2 with | Vlong n1, Vlong n2 => Some (Int64.cmp c n1 n2) @@ -649,6 +748,9 @@ Definition cmpu (c: comparison) (v1 v2: val): val := Definition cmpf (c: comparison) (v1 v2: val): val := of_optbool (cmpf_bool c v1 v2). +Definition cmpfs (c: comparison) (v1 v2: val): val := + of_optbool (cmpfs_bool c v1 v2). + Definition cmpl (c: comparison) (v1 v2: val): option val := option_map of_bool (cmpl_bool c v1 v2). @@ -681,22 +783,23 @@ Definition load_result (chunk: memory_chunk) (v: val) := | Mint32, Vint n => Vint n | Mint32, Vptr b ofs => Vptr b ofs | Mint64, Vlong n => Vlong n - | Mfloat32, Vfloat f => Vfloat(Float.singleoffloat f) + | Mfloat32, Vsingle f => Vsingle f | Mfloat64, Vfloat f => Vfloat f + | Many32, (Vint _ | Vptr _ _ | Vsingle _) => v + | Many64, _ => v | _, _ => Vundef end. Lemma load_result_type: forall chunk v, has_type (load_result chunk v) (type_of_chunk chunk). Proof. - intros. destruct chunk; destruct v; simpl; auto. apply Float.singleoffloat_is_single. + intros. destruct chunk; destruct v; simpl; auto. Qed. Lemma load_result_same: forall v ty, has_type v ty -> load_result (chunk_of_type ty) v = v. Proof. unfold has_type; intros. destruct v; destruct ty; try contradiction; auto. - simpl. rewrite Float.singleoffloat_of_single; auto. Qed. (** Theorems on arithmetic operations. *) @@ -1388,6 +1491,8 @@ Inductive val_inject (mi: meminj): val -> val -> Prop := forall i, val_inject mi (Vlong i) (Vlong i) | val_inject_float: forall f, val_inject mi (Vfloat f) (Vfloat f) + | val_inject_single: + forall f, val_inject mi (Vsingle f) (Vsingle f) | val_inject_ptr: forall b1 ofs1 b2 ofs2 delta, mi b1 = Some (b2, delta) -> diff --git a/driver/Interp.ml b/driver/Interp.ml index af657890..db4537cb 100644 --- a/driver/Interp.ml +++ b/driver/Interp.ml @@ -46,7 +46,7 @@ let print_id_ofs p (id, ofs) = let print_eventval p = function | EVint n -> fprintf p "%ld" (camlint_of_coqint n) | EVfloat f -> fprintf p "%F" (camlfloat_of_coqfloat f) - | EVfloatsingle f -> fprintf p "%F" (camlfloat_of_coqfloat f) + | EVsingle f -> fprintf p "%F" (camlfloat_of_coqfloat32 f) | EVlong n -> fprintf p "%LdLL" (camlint64_of_coqint n) | EVptr_global(id, ofs) -> fprintf p "&%a" print_id_ofs (id, ofs) diff --git a/exportclight/ExportClight.ml b/exportclight/ExportClight.ml index 409abcfa..02f1249c 100644 --- a/exportclight/ExportClight.ml +++ b/exportclight/ExportClight.ml @@ -121,10 +121,10 @@ let coqint p n = else fprintf p "(Int.repr (%ld))" n let coqfloat p n = - let n = camlint64_of_coqint(Floats.Float.bits_of_double n) in + let n = camlint64_of_coqint(Floats.Float.to_bits n) in if n >= 0L - then fprintf p "(Float.double_of_bits (Int64.repr %Ld))" n - else fprintf p "(Float.double_of_bits (Int64.repr (%Ld)))" n + then fprintf p "(Float.of_bits (Int64.repr %Ld))" n + else fprintf p "(Float.of_bits (Int64.repr (%Ld)))" n let coqint64 p n = let n = camlint64_of_coqint n in diff --git a/flocq/Appli/Fappli_IEEE_bits.v b/flocq/Appli/Fappli_IEEE_bits.v index a41fba98..937e8d43 100644 --- a/flocq/Appli/Fappli_IEEE_bits.v +++ b/flocq/Appli/Fappli_IEEE_bits.v @@ -248,6 +248,57 @@ discriminate. now apply Zlt_0_le_0_pred. Qed. +Theorem bits_of_binary_float_range: + forall x, (0 <= bits_of_binary_float x < 2^(mw+ew+1))%Z. +Proof. + intros. +Local Open Scope Z_scope. + assert (J: forall s m e, + 0 <= m < 2^mw -> 0 <= e < 2^ew -> + 0 <= join_bits s m e < 2^(mw+ew+1)). + { + intros. unfold join_bits. + set (se := (if s then 2 ^ ew else 0) + e). + assert (0 <= se < 2^(ew+1)). + { rewrite (Zpower_plus radix2) by omega. change (radix2^1) with 2. simpl. + unfold se. destruct s; omega. } + assert (0 <= se * 2^mw <= (2^(ew+1) - 1) * 2^mw). + { split. apply Zmult_gt_0_le_0_compat; omega. + apply Zmult_le_compat_r; omega. } + rewrite Z.mul_sub_distr_r in H2. + replace (mw + ew + 1) with ((ew + 1) + mw) by omega. + rewrite (Zpower_plus radix2) by omega. simpl. omega. + } + assert (D: forall p n, Z.of_nat (S (digits2_Pnat p)) <= n -> + 0 <= Z.pos p < 2^n). + { + intros. + generalize (digits2_Pnat_correct p). simpl. rewrite ! Zpower_nat_Z. intros [A B]. + split. zify; omega. eapply Zlt_le_trans. eassumption. + apply (Zpower_le radix2); auto. + } + destruct x; unfold bits_of_binary_float. +- apply J; omega. +- apply J; omega. +- destruct n as [pl pl_range]. apply Z.ltb_lt in pl_range. + apply J. apply D. unfold prec, Z_of_nat' in pl_range; omega. omega. +- unfold bounded in e0. apply Bool.andb_true_iff in e0; destruct e0 as [A B]. + apply Z.leb_le in B. + unfold canonic_mantissa, FLT_exp in A. apply Zeq_bool_eq in A. + assert (G: Z.of_nat (S (digits2_Pnat m)) <= prec) by (zify; omega). + assert (M: emin <= e) by (unfold emin; zify; omega). + generalize (Zle_bool_spec (2^mw) (Z.pos m)); intro SPEC; inversion SPEC. + + apply J. + * split. omega. generalize (D _ _ G). unfold prec. + rewrite (Zpower_plus radix2) by omega. + change (radix2^1) with 2. simpl radix_val. omega. + * split. omega. unfold emin. replace (2^ew) with (2 * emax). omega. + symmetry. replace ew with (1 + (ew - 1)) by omega. + apply (Zpower_plus radix2); omega. + + apply J. zify; omega. omega. +Local Close Scope Z_scope. +Qed. + Definition binary_float_of_bits_aux x := let '(sx, mx, ex) := split_bits x in if Zeq_bool ex 0 then diff --git a/ia32/Archi.v b/ia32/Archi.v index a2e136c3..fff25cfc 100644 --- a/ia32/Archi.v +++ b/ia32/Archi.v @@ -25,10 +25,19 @@ Definition big_endian := false. Notation align_int64 := 4%Z (only parsing). Notation align_float64 := 4%Z (only parsing). -Program Definition default_pl : bool * nan_pl 53 := +Program Definition default_pl_64 : bool * nan_pl 53 := (true, nat_iter 51 xO xH). -Definition choose_binop_pl (s1: bool) (pl1: nan_pl 53) (s2: bool) (pl2: nan_pl 53) := +Definition choose_binop_pl_64 (s1: bool) (pl1: nan_pl 53) (s2: bool) (pl2: nan_pl 53) := false. (**r always choose first NaN *) -Global Opaque big_endian default_pl choose_binop_pl. +Program Definition default_pl_32 : bool * nan_pl 24 := + (true, nat_iter 22 xO xH). + +Definition choose_binop_pl_32 (s1: bool) (pl1: nan_pl 24) (s2: bool) (pl2: nan_pl 24) := + false. (**r always choose first NaN *) + +Global Opaque big_endian + default_pl_64 choose_binop_pl_64 + default_pl_32 choose_binop_pl_32. + diff --git a/ia32/Asm.v b/ia32/Asm.v index 2e5b8b9d..15f80e42 100644 --- a/ia32/Asm.v +++ b/ia32/Asm.v @@ -118,10 +118,13 @@ Inductive instruction: Type := | Pmovsd_fi (rd: freg) (n: float) (**r (pseudo-instruction) *) | Pmovsd_fm (rd: freg) (a: addrmode) | Pmovsd_mf (a: addrmode) (r1: freg) - | Pfld_f (r1: freg) (**r [fld] from XMM register (pseudo) *) - | Pfld_m (a: addrmode) (**r [fld] from memory *) - | Pfstp_f (rd: freg) (**r [fstp] to XMM register (pseudo) *) - | Pfstp_m (a: addrmode) (**r [fstp] to memory *) + | Pmovss_fi (rd: freg) (n: float32) (**r [movss] (single 32-bit float) *) + | Pmovss_fm (rd: freg) (a: addrmode) + | Pmovss_mf (a: addrmode) (r1: freg) + | Pfldl_m (a: addrmode) (**r [fld] double precision *) + | Pfstpl_m (a: addrmode) (**r [fstp] double precision *) + | Pflds_m (a: addrmode) (**r [fld] simple precision *) + | Pfstps_m (a: addrmode) (**r [fstp] simple precision *) | Pxchg_rr (r1: ireg) (r2: ireg) (**r register-register exchange *) (** Moves with conversion *) | Pmovb_mr (a: addrmode) (rs: ireg) (**r [mov] (8-bit int) *) @@ -134,11 +137,12 @@ Inductive instruction: Type := | Pmovzw_rm (rd: ireg) (a: addrmode) | Pmovsw_rr (rd: ireg) (rs: ireg) (**r [movsw] (16-bit sign-extension) *) | Pmovsw_rm (rd: ireg) (a: addrmode) - | Pcvtss2sd_fm (rd: freg) (a: addrmode) (**r [cvtss2sd] (single float load) *) - | Pcvtsd2ss_ff (rd: freg) (r1: freg) (**r conversion to single (pseudo) *) - | Pcvtsd2ss_mf (a: addrmode) (r1: freg) (**r [cvtsd2ss] (single float store *) + | Pcvtsd2ss_ff (rd: freg) (r1: freg) (**r conversion to single float *) + | Pcvtss2sd_ff (rd: freg) (r1: freg) (**r conversion to double float *) | Pcvttsd2si_rf (rd: ireg) (r1: freg) (**r double to signed int *) | Pcvtsi2sd_fr (rd: freg) (r1: ireg) (**r signed int to double *) + | Pcvttss2si_rf (rd: ireg) (r1: freg) (**r single to signed int *) + | Pcvtsi2ss_fr (rd: freg) (r1: ireg) (**r signed int to single *) (** Integer arithmetic *) | Plea (rd: ireg) (a: addrmode) | Pneg (rd: ireg) @@ -180,6 +184,14 @@ Inductive instruction: Type := | Pabsd (rd: freg) | Pcomisd_ff (r1 r2: freg) | Pxorpd_f (rd: freg) (**r [xor] with self = set to zero *) + | Padds_ff (rd: freg) (r1: freg) + | Psubs_ff (rd: freg) (r1: freg) + | Pmuls_ff (rd: freg) (r1: freg) + | Pdivs_ff (rd: freg) (r1: freg) + | Pnegs (rd: freg) + | Pabss (rd: freg) + | Pcomiss_ff (r1 r2: freg) + | Pxorps_f (rd: freg) (**r [xor] with self = set to zero *) (** Branches and calls *) | Pjmp_l (l: label) | Pjmp_s (symb: ident) (sg: signature) @@ -190,6 +202,11 @@ Inductive instruction: Type := | Pcall_s (symb: ident) (sg: signature) | Pcall_r (r: ireg) (sg: signature) | Pret + (** Saving and restoring registers *) + | Pmov_rm_a (rd: ireg) (a: addrmode) (**r like [Pmov_rm], using [Many32] chunk *) + | Pmov_mr_a (a: addrmode) (rs: ireg) (**r like [Pmov_mr], using [Many32] chunk *) + | Pmovsd_fm_a (rd: freg) (a: addrmode) (**r like [Pmovsd_fm], using [Many64] chunk *) + | Pmovsd_mf_a (a: addrmode) (r1: freg) (**r like [Pmovsd_mf], using [Many64] chunk *) (** Pseudo-instructions *) | Plabel(l: label) | Pallocframe(sz: Z)(ofs_ra ofs_link: int) @@ -330,6 +347,18 @@ Definition compare_floats (vx vy: val) (rs: regset) : regset := undef_regs (CR ZF :: CR CF :: CR PF :: CR SF :: CR OF :: nil) rs end. +Definition compare_floats32 (vx vy: val) (rs: regset) : regset := + match vx, vy with + | Vsingle x, Vsingle y => + rs #ZF <- (Val.of_bool (negb (Float32.cmp Cne x y))) + #CF <- (Val.of_bool (negb (Float32.cmp Cge x y))) + #PF <- (Val.of_bool (negb (Float32.cmp Ceq x y || Float32.cmp Clt x y || Float32.cmp Cgt x y))) + #SF <- Vundef + #OF <- Vundef + | _, _ => + undef_regs (CR ZF :: CR CF :: CR PF :: CR SF :: CR OF :: nil) rs + end. + (** Testing a condition *) Definition eval_testcond (c: testcond) (rs: regset) : option bool := @@ -484,14 +513,20 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out exec_load Mfloat64 m a rs rd | Pmovsd_mf a r1 => exec_store Mfloat64 m a rs r1 nil - | Pfld_f r1 => - Next (nextinstr (rs#ST0 <- (rs r1))) m - | Pfld_m a => + | Pmovss_fi rd n => + Next (nextinstr (rs#rd <- (Vsingle n))) m + | Pmovss_fm rd a => + exec_load Mfloat32 m a rs rd + | Pmovss_mf a r1 => + exec_store Mfloat32 m a rs r1 nil + | Pfldl_m a => exec_load Mfloat64 m a rs ST0 - | Pfstp_f rd => - Next (nextinstr (rs#rd <- (rs ST0) #ST0 <- Vundef)) m - | Pfstp_m a => + | Pfstpl_m a => exec_store Mfloat64 m a rs ST0 (ST0 :: nil) + | Pflds_m a => + exec_load Mfloat32 m a rs ST0 + | Pfstps_m a => + exec_store Mfloat32 m a rs ST0 (ST0 :: nil) | Pxchg_rr r1 r2 => Next (nextinstr (rs#r1 <- (rs r2) #r2 <- (rs r1))) m (** Moves with conversion *) @@ -515,16 +550,18 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out Next (nextinstr (rs#rd <- (Val.sign_ext 16 rs#r1))) m | Pmovsw_rm rd a => exec_load Mint16signed m a rs rd - | Pcvtss2sd_fm rd a => - exec_load Mfloat32 m a rs rd | Pcvtsd2ss_ff rd r1 => Next (nextinstr (rs#rd <- (Val.singleoffloat rs#r1))) m - | Pcvtsd2ss_mf a r1 => - exec_store Mfloat32 m a rs r1 (FR XMM7 :: nil) + | Pcvtss2sd_ff rd r1 => + Next (nextinstr (rs#rd <- (Val.floatofsingle rs#r1))) m | Pcvttsd2si_rf rd r1 => Next (nextinstr (rs#rd <- (Val.maketotal (Val.intoffloat rs#r1)))) m | Pcvtsi2sd_fr rd r1 => Next (nextinstr (rs#rd <- (Val.maketotal (Val.floatofint rs#r1)))) m + | Pcvttss2si_rf rd r1 => + Next (nextinstr (rs#rd <- (Val.maketotal (Val.intofsingle rs#r1)))) m + | Pcvtsi2ss_fr rd r1 => + Next (nextinstr (rs#rd <- (Val.maketotal (Val.singleofint rs#r1)))) m (** Integer arithmetic *) | Plea rd a => Next (nextinstr (rs#rd <- (eval_addrmode a rs))) m @@ -604,7 +641,7 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out end | Psetcc c rd => Next (nextinstr (rs#rd <- (Val.of_optbool (eval_testcond c rs)))) m - (** Arithmetic operations over floats *) + (** Arithmetic operations over double-precision floats *) | Paddd_ff rd r1 => Next (nextinstr (rs#rd <- (Val.addf rs#rd rs#r1))) m | Psubd_ff rd r1 => @@ -621,6 +658,23 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out Next (nextinstr (compare_floats (rs r1) (rs r2) rs)) m | Pxorpd_f rd => Next (nextinstr_nf (rs#rd <- (Vfloat Float.zero))) m + (** Arithmetic operations over single-precision floats *) + | Padds_ff rd r1 => + Next (nextinstr (rs#rd <- (Val.addfs rs#rd rs#r1))) m + | Psubs_ff rd r1 => + Next (nextinstr (rs#rd <- (Val.subfs rs#rd rs#r1))) m + | Pmuls_ff rd r1 => + Next (nextinstr (rs#rd <- (Val.mulfs rs#rd rs#r1))) m + | Pdivs_ff rd r1 => + Next (nextinstr (rs#rd <- (Val.divfs rs#rd rs#r1))) m + | Pnegs rd => + Next (nextinstr (rs#rd <- (Val.negfs rs#rd))) m + | Pabss rd => + Next (nextinstr (rs#rd <- (Val.absfs rs#rd))) m + | Pcomiss_ff r1 r2 => + Next (nextinstr (compare_floats32 (rs r1) (rs r2) rs)) m + | Pxorps_f rd => + Next (nextinstr_nf (rs#rd <- (Vsingle Float32.zero))) m (** Branches and calls *) | Pjmp_l lbl => goto_label f lbl rs m @@ -655,6 +709,15 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out Next (rs#RA <- (Val.add rs#PC Vone) #PC <- (rs r)) m | Pret => Next (rs#PC <- (rs#RA)) m + (** Saving and restoring registers *) + | Pmov_rm_a rd a => + exec_load Many32 m a rs rd + | Pmov_mr_a a r1 => + exec_store Many32 m a rs r1 nil + | Pmovsd_fm_a rd a => + exec_load Many64 m a rs rd + | Pmovsd_mf_a a r1 => + exec_store Many64 m a rs r1 nil (** Pseudo-instructions *) | Plabel lbl => Next (nextinstr rs) m diff --git a/ia32/Asmgen.v b/ia32/Asmgen.v index f92d72c0..9c0a76e0 100644 --- a/ia32/Asmgen.v +++ b/ia32/Asmgen.v @@ -50,8 +50,6 @@ Definition mk_mov (rd rs: preg) (k: code) : res code := match rd, rs with | IR rd, IR rs => OK (Pmov_rr rd rs :: k) | FR rd, FR rs => OK (Pmovsd_ff rd rs :: k) - | ST0, FR rs => OK (Pfld_f rs :: k) - | FR rd, ST0 => OK (Pfstp_f rd :: k) | _, _ => Error(msg "Asmgen.mk_mov") end. @@ -90,42 +88,30 @@ Definition mk_smallstore (sto: addrmode -> ireg ->instruction) else OK (Pmov_rr EAX rs :: sto addr EAX :: k). -(** Accessing slots in the stack frame. *) +(** Accessing slots in the stack frame. *) Definition loadind (base: ireg) (ofs: int) (ty: typ) (dst: mreg) (k: code) := - match ty with - | Tint => - do r <- ireg_of dst; - OK (Pmov_rm r (Addrmode (Some base) None (inl _ ofs)) :: k) - | Tfloat => - match preg_of dst with - | FR r => OK (Pmovsd_fm r (Addrmode (Some base) None (inl _ ofs)) :: k) - | ST0 => OK (Pfld_m (Addrmode (Some base) None (inl _ ofs)) :: k) - | _ => Error (msg "Asmgen.loadind") - end - | Tsingle => - do r <- freg_of dst; - OK (Pcvtss2sd_fm r (Addrmode (Some base) None (inl _ ofs)) :: k) - | Tlong => - Error (msg "Asmgen.loadind") + match ty, preg_of dst with + | Tint, IR r => OK (Pmov_rm r (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tsingle, FR r => OK (Pmovss_fm r (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tsingle, ST0 => OK (Pflds_m (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tfloat, FR r => OK (Pmovsd_fm r (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tfloat, ST0 => OK (Pfldl_m (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tany32, IR r => OK (Pmov_rm_a r (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tany64, FR r => OK (Pmovsd_fm_a r (Addrmode (Some base) None (inl _ ofs)) :: k) + | _, _ => Error (msg "Asmgen.loadind") end. Definition storeind (src: mreg) (base: ireg) (ofs: int) (ty: typ) (k: code) := - match ty with - | Tint => - do r <- ireg_of src; - OK (Pmov_mr (Addrmode (Some base) None (inl _ ofs)) r :: k) - | Tfloat => - match preg_of src with - | FR r => OK (Pmovsd_mf (Addrmode (Some base) None (inl _ ofs)) r :: k) - | ST0 => OK (Pfstp_m (Addrmode (Some base) None (inl _ ofs)) :: k) - | _ => Error (msg "Asmgen.loadind") - end - | Tsingle => - do r <- freg_of src; - OK (Pcvtsd2ss_mf (Addrmode (Some base) None (inl _ ofs)) r :: k) - | Tlong => - Error (msg "Asmgen.storeind") + match ty, preg_of src with + | Tint, IR r => OK (Pmov_mr (Addrmode (Some base) None (inl _ ofs)) r :: k) + | Tsingle, FR r => OK (Pmovss_mf (Addrmode (Some base) None (inl _ ofs)) r :: k) + | Tsingle, ST0 => OK (Pfstps_m (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tfloat, FR r => OK (Pmovsd_mf (Addrmode (Some base) None (inl _ ofs)) r :: k) + | Tfloat, ST0 => OK (Pfstpl_m (Addrmode (Some base) None (inl _ ofs)) :: k) + | Tany32, IR r => OK (Pmov_mr_a (Addrmode (Some base) None (inl _ ofs)) r :: k) + | Tany64, FR r => OK (Pmovsd_mf_a (Addrmode (Some base) None (inl _ ofs)) r :: k) + | _, _ => Error (msg "Asmgen.storeind") end. (** Translation of addressing modes *) @@ -163,6 +149,12 @@ Definition floatcomp (cmp: comparison) (r1 r2: freg) : instruction := | Ceq | Cne | Cgt | Cge => Pcomisd_ff r1 r2 end. +Definition floatcomp32 (cmp: comparison) (r1 r2: freg) : instruction := + match cmp with + | Clt | Cle => Pcomiss_ff r2 r1 + | Ceq | Cne | Cgt | Cge => Pcomiss_ff r1 r2 + end. + (** Translation of a condition. Prepends to [k] the instructions that evaluate the condition and leave its boolean result in bits of the condition register. *) @@ -183,6 +175,10 @@ Definition transl_cond do r1 <- freg_of a1; do r2 <- freg_of a2; OK (floatcomp cmp r1 r2 :: k) | Cnotcompf cmp, a1 :: a2 :: nil => do r1 <- freg_of a1; do r2 <- freg_of a2; OK (floatcomp cmp r1 r2 :: k) + | Ccompfs cmp, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; OK (floatcomp32 cmp r1 r2 :: k) + | Cnotcompfs cmp, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; OK (floatcomp32 cmp r1 r2 :: k) | Cmaskzero n, a1 :: nil => do r1 <- ireg_of a1; OK (Ptest_ri r1 n :: k) | Cmasknotzero n, a1 :: nil => @@ -224,7 +220,7 @@ Definition testcond_for_condition (cond: condition) : extcond := | Ccompu c => Cond_base(testcond_for_unsigned_comparison c) | Ccompimm c n => Cond_base(testcond_for_signed_comparison c) | Ccompuimm c n => Cond_base(testcond_for_unsigned_comparison c) - | Ccompf c => + | Ccompf c | Ccompfs c => match c with | Ceq => Cond_and Cond_np Cond_e | Cne => Cond_or Cond_p Cond_ne @@ -233,7 +229,7 @@ Definition testcond_for_condition (cond: condition) : extcond := | Cgt => Cond_base Cond_a | Cge => Cond_base Cond_ae end - | Cnotcompf c => + | Cnotcompf c | Cnotcompfs c => match c with | Ceq => Cond_or Cond_p Cond_ne | Cne => Cond_and Cond_np Cond_e @@ -288,6 +284,9 @@ Definition transl_op | Ofloatconst f, nil => do r <- freg_of res; OK ((if Float.eq_dec f Float.zero then Pxorpd_f r else Pmovsd_fi r f) :: k) + | Osingleconst f, nil => + do r <- freg_of res; + OK ((if Float32.eq_dec f Float32.zero then Pxorps_f r else Pmovss_fi r f) :: k) | Oindirectsymbol id, nil => do r <- ireg_of res; OK (Pmov_ra r id :: k) @@ -412,12 +411,36 @@ Definition transl_op | Odivf, a1 :: a2 :: nil => assertion (mreg_eq a1 res); do r <- freg_of res; do r2 <- freg_of a2; OK (Pdivd_ff r r2 :: k) + | Onegfs, a1 :: nil => + assertion (mreg_eq a1 res); + do r <- freg_of res; OK (Pnegs r :: k) + | Oabsfs, a1 :: nil => + assertion (mreg_eq a1 res); + do r <- freg_of res; OK (Pabss r :: k) + | Oaddfs, a1 :: a2 :: nil => + assertion (mreg_eq a1 res); + do r <- freg_of res; do r2 <- freg_of a2; OK (Padds_ff r r2 :: k) + | Osubfs, a1 :: a2 :: nil => + assertion (mreg_eq a1 res); + do r <- freg_of res; do r2 <- freg_of a2; OK (Psubs_ff r r2 :: k) + | Omulfs, a1 :: a2 :: nil => + assertion (mreg_eq a1 res); + do r <- freg_of res; do r2 <- freg_of a2; OK (Pmuls_ff r r2 :: k) + | Odivfs, a1 :: a2 :: nil => + assertion (mreg_eq a1 res); + do r <- freg_of res; do r2 <- freg_of a2; OK (Pdivs_ff r r2 :: k) | Osingleoffloat, a1 :: nil => do r <- freg_of res; do r1 <- freg_of a1; OK (Pcvtsd2ss_ff r r1 :: k) + | Ofloatofsingle, a1 :: nil => + do r <- freg_of res; do r1 <- freg_of a1; OK (Pcvtss2sd_ff r r1 :: k) | Ointoffloat, a1 :: nil => do r <- ireg_of res; do r1 <- freg_of a1; OK (Pcvttsd2si_rf r r1 :: k) | Ofloatofint, a1 :: nil => do r <- freg_of res; do r1 <- ireg_of a1; OK (Pcvtsi2sd_fr r r1 :: k) + | Ointofsingle, a1 :: nil => + do r <- ireg_of res; do r1 <- freg_of a1; OK (Pcvttss2si_rf r r1 :: k) + | Osingleofint, a1 :: nil => + do r <- freg_of res; do r1 <- ireg_of a1; OK (Pcvtsi2ss_fr r r1 :: k) | Ocmp c, args => do r <- ireg_of res; transl_cond c args (mk_setcc (testcond_for_condition c) r k) @@ -443,10 +466,10 @@ Definition transl_load (chunk: memory_chunk) | Mint32 => do r <- ireg_of dest; OK(Pmov_rm r am :: k) | Mfloat32 => - do r <- freg_of dest; OK(Pcvtss2sd_fm r am :: k) + do r <- freg_of dest; OK(Pmovss_fm r am :: k) | Mfloat64 => do r <- freg_of dest; OK(Pmovsd_fm r am :: k) - | Mint64 => + | _ => Error (msg "Asmgen.transl_load") end. @@ -462,10 +485,10 @@ Definition transl_store (chunk: memory_chunk) | Mint32 => do r <- ireg_of src; OK(Pmov_mr am r :: k) | Mfloat32 => - do r <- freg_of src; OK(Pcvtsd2ss_mf am r :: k) + do r <- freg_of src; OK(Pmovss_mf am r :: k) | Mfloat64 => do r <- freg_of src; OK(Pmovsd_mf am r :: k) - | Mint64 => + | _ => Error (msg "Asmgen.transl_store") end. diff --git a/ia32/Asmgenproof.v b/ia32/Asmgenproof.v index 881375f0..eba710a1 100644 --- a/ia32/Asmgenproof.v +++ b/ia32/Asmgenproof.v @@ -168,11 +168,7 @@ Remark loadind_label: loadind base ofs ty dst k = OK c -> tail_nolabel k c. Proof. - unfold loadind; intros. destruct ty. - TailNoLabel. - destruct (preg_of dst); TailNoLabel. - discriminate. - TailNoLabel. + unfold loadind; intros. destruct ty; try discriminate; destruct (preg_of dst); TailNoLabel. Qed. Remark storeind_label: @@ -180,11 +176,7 @@ Remark storeind_label: storeind src base ofs ty k = OK c -> tail_nolabel k c. Proof. - unfold storeind; intros. destruct ty. - TailNoLabel. - destruct (preg_of src); TailNoLabel. - discriminate. - TailNoLabel. + unfold storeind; intros. destruct ty; try discriminate; destruct (preg_of src); TailNoLabel. Qed. Remark mk_setcc_base_label: @@ -220,6 +212,8 @@ Proof. destruct (Int.eq_dec i Int.zero); TailNoLabel. destruct c0; simpl; TailNoLabel. destruct c0; simpl; TailNoLabel. + destruct c0; simpl; TailNoLabel. + destruct c0; simpl; TailNoLabel. Qed. Remark transl_op_label: @@ -230,6 +224,7 @@ Proof. unfold transl_op; intros. destruct op; TailNoLabel. destruct (Int.eq_dec i Int.zero); TailNoLabel. destruct (Float.eq_dec f Float.zero); TailNoLabel. + destruct (Float32.eq_dec f Float32.zero); TailNoLabel. eapply tail_nolabel_trans. eapply transl_cond_label; eauto. eapply mk_setcc_label. Qed. diff --git a/ia32/Asmgenproof1.v b/ia32/Asmgenproof1.v index b3c815b6..7d71d1a2 100644 --- a/ia32/Asmgenproof1.v +++ b/ia32/Asmgenproof1.v @@ -107,7 +107,7 @@ Lemma mk_mov_correct: exists rs2, exec_straight ge fn c rs1 m k rs2 m /\ rs2#rd = rs1#rs - /\ forall r, data_preg r = true -> r <> ST0 -> r <> rd -> rs2#r = rs1#r. + /\ forall r, data_preg r = true -> r <> rd -> rs2#r = rs1#r. Proof. unfold mk_mov; intros. destruct rd; try (monadInv H); destruct rs; monadInv H. @@ -117,12 +117,6 @@ Proof. (* movd *) econstructor. split. apply exec_straight_one. simpl. eauto. auto. split. Simplifs. intros; Simplifs. -(* getfp0 *) - econstructor. split. apply exec_straight_one. simpl. eauto. auto. - split. Simplifs. intros; Simplifs. -(* setfp0 *) - econstructor. split. apply exec_straight_one. simpl. eauto. auto. - split. Simplifs. intros; Simplifs. Qed. (** Properties of division *) @@ -288,27 +282,10 @@ Proof. set (addr := Addrmode (Some base) None (inl (ident * int) ofs)) in *. assert (eval_addrmode ge addr rs = Val.add rs#base (Vint ofs)). unfold addr. simpl. rewrite Int.add_commut; rewrite Int.add_zero; auto. - destruct ty; simpl in H0. - (* int *) - monadInv H. - rewrite (ireg_of_eq _ _ EQ). econstructor. - split. apply exec_straight_one. simpl. unfold exec_load. rewrite H1. rewrite H0. - eauto. auto. - intuition Simplifs. - (* float *) - exists (nextinstr_nf (rs#(preg_of dst) <- v)). - split. destruct (preg_of dst); inv H; apply exec_straight_one; simpl; auto. - unfold exec_load. rewrite H1; rewrite H0; auto. - unfold exec_load. rewrite H1; rewrite H0; auto. - intuition Simplifs. - (* long *) - inv H. - (* single *) - monadInv H. - rewrite (freg_of_eq _ _ EQ). econstructor. - split. apply exec_straight_one. simpl. unfold exec_load. rewrite H1. rewrite H0. - eauto. auto. - intuition Simplifs. + exists (nextinstr_nf (rs#(preg_of dst) <- v)); split. +- destruct ty; try discriminate; destruct (preg_of dst); inv H; simpl in H0; + apply exec_straight_one; auto; simpl; unfold exec_load; rewrite H1, H0; auto. +- intuition Simplifs. Qed. Lemma storeind_correct: @@ -319,33 +296,15 @@ Lemma storeind_correct: exec_straight ge fn c rs m k rs' m' /\ forall r, data_preg r = true -> preg_notin r (destroyed_by_setstack ty) -> rs'#r = rs#r. Proof. +Local Transparent destroyed_by_setstack. unfold storeind; intros. set (addr := Addrmode (Some base) None (inl (ident * int) ofs)) in *. assert (eval_addrmode ge addr rs = Val.add rs#base (Vint ofs)). unfold addr. simpl. rewrite Int.add_commut; rewrite Int.add_zero; auto. - destruct ty; simpl in H0. - (* int *) - monadInv H. - rewrite (ireg_of_eq _ _ EQ) in H0. econstructor. - split. apply exec_straight_one. simpl. unfold exec_store. rewrite H1. rewrite H0. - eauto. auto. - intros; Simplifs. - (* float *) - destruct (preg_of src); inv H. - econstructor; split. apply exec_straight_one. - simpl. unfold exec_store. rewrite H1; rewrite H0. eauto. auto. - intros. apply nextinstr_nf_inv1; auto. - econstructor; split. apply exec_straight_one. - simpl. unfold exec_store. rewrite H1; rewrite H0. eauto. auto. - intros. simpl. Simplifs. - (* long *) - inv H. - (* single *) - monadInv H. - rewrite (freg_of_eq _ _ EQ) in H0. econstructor. - split. apply exec_straight_one. simpl. unfold exec_store. rewrite H1. rewrite H0. - simpl. eauto. auto. - intros. destruct H2. Simplifs. + destruct ty; try discriminate; destruct (preg_of src); inv H; simpl in H0; + (econstructor; split; + [apply exec_straight_one; [simpl; unfold exec_store; rewrite H1, H0; eauto|auto] + |simpl; intros; unfold undef_regs; repeat Simplifs]). Qed. (** Translation of addressing modes *) @@ -546,6 +505,21 @@ Proof. intros. Simplifs. Qed. +Lemma compare_floats32_spec: + forall rs n1 n2, + let rs' := nextinstr (compare_floats32 (Vsingle n1) (Vsingle n2) rs) in + rs'#ZF = Val.of_bool (negb (Float32.cmp Cne n1 n2)) + /\ rs'#CF = Val.of_bool (negb (Float32.cmp Cge n1 n2)) + /\ rs'#PF = Val.of_bool (negb (Float32.cmp Ceq n1 n2 || Float32.cmp Clt n1 n2 || Float32.cmp Cgt n1 n2)) + /\ (forall r, data_preg r = true -> rs'#r = rs#r). +Proof. + intros. unfold rs'; unfold compare_floats32. + split. auto. + split. auto. + split. auto. + intros. Simplifs. +Qed. + Definition eval_extcond (xc: extcond) (rs: regset) : option bool := match xc with | Cond_base c => @@ -664,8 +638,104 @@ Proof. destruct (Float.cmp Cge n1 n2); auto. Qed. +Lemma testcond_for_float32_comparison_correct: + forall c n1 n2 rs, + eval_extcond (testcond_for_condition (Ccompfs c)) + (nextinstr (compare_floats32 (Vsingle (swap_floats c n1 n2)) + (Vsingle (swap_floats c n2 n1)) rs)) = + Some(Float32.cmp c n1 n2). +Proof. + intros. + generalize (compare_floats32_spec rs (swap_floats c n1 n2) (swap_floats c n2 n1)). + set (rs' := nextinstr (compare_floats32 (Vsingle (swap_floats c n1 n2)) + (Vsingle (swap_floats c n2 n1)) rs)). + intros [A [B [C D]]]. + unfold eval_extcond, eval_testcond. rewrite A; rewrite B; rewrite C. + destruct c; simpl. +(* eq *) + rewrite Float32.cmp_ne_eq. + caseEq (Float32.cmp Ceq n1 n2); intros. + auto. + simpl. destruct (Float32.cmp Clt n1 n2 || Float32.cmp Cgt n1 n2); auto. +(* ne *) + rewrite Float32.cmp_ne_eq. + caseEq (Float32.cmp Ceq n1 n2); intros. + auto. + simpl. destruct (Float32.cmp Clt n1 n2 || Float32.cmp Cgt n1 n2); auto. +(* lt *) + rewrite <- (Float32.cmp_swap Cge n1 n2). + rewrite <- (Float32.cmp_swap Cne n1 n2). + simpl. + rewrite Float32.cmp_ne_eq. rewrite Float32.cmp_le_lt_eq. + caseEq (Float32.cmp Clt n1 n2); intros; simpl. + caseEq (Float32.cmp Ceq n1 n2); intros; simpl. + elimtype False. eapply Float32.cmp_lt_eq_false; eauto. + auto. + destruct (Float32.cmp Ceq n1 n2); auto. +(* le *) + rewrite <- (Float32.cmp_swap Cge n1 n2). simpl. + destruct (Float32.cmp Cle n1 n2); auto. +(* gt *) + rewrite Float32.cmp_ne_eq. rewrite Float32.cmp_ge_gt_eq. + caseEq (Float32.cmp Cgt n1 n2); intros; simpl. + caseEq (Float32.cmp Ceq n1 n2); intros; simpl. + elimtype False. eapply Float32.cmp_gt_eq_false; eauto. + auto. + destruct (Float32.cmp Ceq n1 n2); auto. +(* ge *) + destruct (Float32.cmp Cge n1 n2); auto. +Qed. + +Lemma testcond_for_neg_float32_comparison_correct: + forall c n1 n2 rs, + eval_extcond (testcond_for_condition (Cnotcompfs c)) + (nextinstr (compare_floats32 (Vsingle (swap_floats c n1 n2)) + (Vsingle (swap_floats c n2 n1)) rs)) = + Some(negb(Float32.cmp c n1 n2)). +Proof. + intros. + generalize (compare_floats32_spec rs (swap_floats c n1 n2) (swap_floats c n2 n1)). + set (rs' := nextinstr (compare_floats32 (Vsingle (swap_floats c n1 n2)) + (Vsingle (swap_floats c n2 n1)) rs)). + intros [A [B [C D]]]. + unfold eval_extcond, eval_testcond. rewrite A; rewrite B; rewrite C. + destruct c; simpl. +(* eq *) + rewrite Float32.cmp_ne_eq. + caseEq (Float32.cmp Ceq n1 n2); intros. + auto. + simpl. destruct (Float32.cmp Clt n1 n2 || Float32.cmp Cgt n1 n2); auto. +(* ne *) + rewrite Float32.cmp_ne_eq. + caseEq (Float32.cmp Ceq n1 n2); intros. + auto. + simpl. destruct (Float32.cmp Clt n1 n2 || Float32.cmp Cgt n1 n2); auto. +(* lt *) + rewrite <- (Float32.cmp_swap Cge n1 n2). + rewrite <- (Float32.cmp_swap Cne n1 n2). + simpl. + rewrite Float32.cmp_ne_eq. rewrite Float32.cmp_le_lt_eq. + caseEq (Float32.cmp Clt n1 n2); intros; simpl. + caseEq (Float32.cmp Ceq n1 n2); intros; simpl. + elimtype False. eapply Float32.cmp_lt_eq_false; eauto. + auto. + destruct (Float32.cmp Ceq n1 n2); auto. +(* le *) + rewrite <- (Float32.cmp_swap Cge n1 n2). simpl. + destruct (Float32.cmp Cle n1 n2); auto. +(* gt *) + rewrite Float32.cmp_ne_eq. rewrite Float32.cmp_ge_gt_eq. + caseEq (Float32.cmp Cgt n1 n2); intros; simpl. + caseEq (Float32.cmp Ceq n1 n2); intros; simpl. + elimtype False. eapply Float32.cmp_gt_eq_false; eauto. + auto. + destruct (Float32.cmp Ceq n1 n2); auto. +(* ge *) + destruct (Float32.cmp Cge n1 n2); auto. +Qed. + Remark swap_floats_commut: - forall c x y, swap_floats c (Vfloat x) (Vfloat y) = Vfloat (swap_floats c x y). + forall (A B: Type) c (f: A -> B) x y, swap_floats c (f x) (f y) = f (swap_floats c x y). Proof. intros. destruct c; auto. Qed. @@ -679,7 +749,18 @@ Proof. assert (DFL: undef_regs (CR ZF :: CR CF :: CR PF :: CR SF :: CR OF :: nil) rs r = rs r). simpl. Simplifs. unfold compare_floats; destruct vx; destruct vy; auto. Simplifs. -Qed. +Qed. + +Remark compare_floats32_inv: + forall vx vy rs r, + r <> CR ZF -> r <> CR CF -> r <> CR PF -> r <> CR SF -> r <> CR OF -> + compare_floats32 vx vy rs r = rs r. +Proof. + intros. + assert (DFL: undef_regs (CR ZF :: CR CF :: CR PF :: CR SF :: CR OF :: nil) rs r = rs r). + simpl. Simplifs. + unfold compare_floats32; destruct vx; destruct vy; auto. Simplifs. +Qed. Lemma transl_cond_correct: forall cond args k c rs m, @@ -740,6 +821,24 @@ Proof. split. destruct (rs x); destruct (rs x0); simpl; auto. repeat rewrite swap_floats_commut. apply testcond_for_neg_float_comparison_correct. intros. Simplifs. apply compare_floats_inv; auto with asmgen. +(* compfs *) + simpl. rewrite (freg_of_eq _ _ EQ). rewrite (freg_of_eq _ _ EQ1). + exists (nextinstr (compare_floats32 (swap_floats c0 (rs x) (rs x0)) (swap_floats c0 (rs x0) (rs x)) rs)). + split. apply exec_straight_one. + destruct c0; simpl; auto. + unfold nextinstr. rewrite Pregmap.gss. rewrite compare_floats32_inv; auto with asmgen. + split. destruct (rs x); destruct (rs x0); simpl; auto. + repeat rewrite swap_floats_commut. apply testcond_for_float32_comparison_correct. + intros. Simplifs. apply compare_floats32_inv; auto with asmgen. +(* notcompfs *) + simpl. rewrite (freg_of_eq _ _ EQ). rewrite (freg_of_eq _ _ EQ1). + exists (nextinstr (compare_floats32 (swap_floats c0 (rs x) (rs x0)) (swap_floats c0 (rs x0) (rs x)) rs)). + split. apply exec_straight_one. + destruct c0; simpl; auto. + unfold nextinstr. rewrite Pregmap.gss. rewrite compare_floats32_inv; auto with asmgen. + split. destruct (rs x); destruct (rs x0); simpl; auto. + repeat rewrite swap_floats_commut. apply testcond_for_neg_float32_comparison_correct. + intros. Simplifs. apply compare_floats32_inv; auto with asmgen. (* maskzero *) simpl. rewrite (ireg_of_eq _ _ EQ). econstructor. split. apply exec_straight_one. simpl; eauto. auto. @@ -909,11 +1008,13 @@ Transparent destroyed_by_op. destruct op; simpl in TR; ArgsInv; simpl in EV; try (inv EV); try (apply SAME; TranslOp; fail). (* move *) exploit mk_mov_correct; eauto. intros [rs2 [A [B C]]]. - apply SAME. exists rs2. split. eauto. split. simpl. auto. intros. destruct H; auto. + apply SAME. exists rs2. eauto. (* intconst *) apply SAME. destruct (Int.eq_dec i Int.zero). subst i. TranslOp. TranslOp. (* floatconst *) apply SAME. destruct (Float.eq_dec f Float.zero). subst f. TranslOp. TranslOp. +(* singleconst *) + apply SAME. destruct (Float32.eq_dec f Float32.zero). subst f. TranslOp. TranslOp. (* cast8signed *) apply SAME. eapply mk_intconv_correct; eauto. (* cast8unsigned *) @@ -963,6 +1064,10 @@ Transparent destroyed_by_op. apply SAME. TranslOp. rewrite H0; auto. (* floatofint *) apply SAME. TranslOp. rewrite H0; auto. +(* intofsingle *) + apply SAME. TranslOp. rewrite H0; auto. +(* singleofint *) + apply SAME. TranslOp. rewrite H0; auto. (* condition *) exploit transl_cond_correct; eauto. intros [rs2 [P [Q R]]]. exploit mk_setcc_correct; eauto. intros [rs3 [S [T U]]]. diff --git a/ia32/ConstpropOp.vp b/ia32/ConstpropOp.vp index b27f405a..8c7f01fa 100644 --- a/ia32/ConstpropOp.vp +++ b/ia32/ConstpropOp.vp @@ -174,10 +174,15 @@ Definition make_moduimm n (r1 r2: reg) := end. Definition make_mulfimm (n: float) (r r1 r2: reg) := - if Float.eq_dec n (Float.floatofint (Int.repr 2)) + if Float.eq_dec n (Float.of_int (Int.repr 2)) then (Oaddf, r :: r :: nil) else (Omulf, r1 :: r2 :: nil). +Definition make_mulfsimm (n: float32) (r r1 r2: reg) := + if Float32.eq_dec n (Float32.of_int (Int.repr 2)) + then (Oaddfs, r :: r :: nil) + else (Omulfs, r1 :: r2 :: nil). + Definition make_cast8signed (r: reg) (a: aval) := if vincl a (Sgn 8) then (Omove, r :: nil) else (Ocast8signed, r :: nil). Definition make_cast8unsigned (r: reg) (a: aval) := @@ -186,10 +191,6 @@ Definition make_cast16signed (r: reg) (a: aval) := if vincl a (Sgn 16) then (Omove, r :: nil) else (Ocast16signed, r :: nil). Definition make_cast16unsigned (r: reg) (a: aval) := if vincl a (Uns 16) then (Omove, r :: nil) else (Ocast16unsigned, r :: nil). -Definition make_singleoffloat (r: reg) (a: aval) := - if vincl a Fsingle && generate_float_constants tt - then (Omove, r :: nil) - else (Osingleoffloat, r :: nil). Nondetfunction op_strength_reduction (op: operation) (args: list reg) (vl: list aval) := @@ -217,9 +218,10 @@ Nondetfunction op_strength_reduction | Olea addr, args, vl => let (addr', args') := addr_strength_reduction addr args vl in (Olea addr', args') - | Osingleoffloat, r1 :: nil, v1 :: nil => make_singleoffloat r1 v1 | Ocmp c, args, vl => make_cmp c args vl | Omulf, r1 :: r2 :: nil, v1 :: F n2 :: nil => make_mulfimm n2 r1 r1 r2 | Omulf, r1 :: r2 :: nil, F n1 :: v2 :: nil => make_mulfimm n1 r2 r1 r2 + | Omulfs, r1 :: r2 :: nil, v1 :: FS n2 :: nil => make_mulfsimm n2 r1 r1 r2 + | Omulfs, r1 :: r2 :: nil, FS n1 :: v2 :: nil => make_mulfsimm n1 r2 r1 r2 | _, _, _ => (op, args) end. diff --git a/ia32/ConstpropOpproof.v b/ia32/ConstpropOpproof.v index d9eea2bc..6adb26fe 100644 --- a/ia32/ConstpropOpproof.v +++ b/ia32/ConstpropOpproof.v @@ -77,6 +77,10 @@ Ltac SimplVM := let E := fresh in assert (E: v = Vfloat n) by (inversion H; auto); rewrite E in *; clear H; SimplVM + | [ H: vmatch _ ?v (FS ?n) |- _ ] => + let E := fresh in + assert (E: v = Vsingle n) by (inversion H; auto); + rewrite E in *; clear H; SimplVM | [ H: vmatch _ ?v (Ptr(Gl ?id ?ofs)) |- _ ] => let E := fresh in assert (E: Val.lessdef v (Genv.symbol_address ge id ofs)) by (eapply vmatch_ptr_gl; eauto); @@ -368,7 +372,7 @@ Lemma make_mulfimm_correct: exists v, eval_operation ge (Vptr sp Int.zero) op e##args m = Some v /\ Val.lessdef (Val.mulf e#r1 e#r2) v. Proof. intros; unfold make_mulfimm. - destruct (Float.eq_dec n (Float.floatofint (Int.repr 2))); intros. + destruct (Float.eq_dec n (Float.of_int (Int.repr 2))); intros. simpl. econstructor; split. eauto. rewrite H; subst n. destruct (e#r1); simpl; auto. rewrite Float.mul2_add; auto. simpl. econstructor; split; eauto. @@ -381,13 +385,40 @@ Lemma make_mulfimm_correct_2: exists v, eval_operation ge (Vptr sp Int.zero) op e##args m = Some v /\ Val.lessdef (Val.mulf e#r1 e#r2) v. Proof. intros; unfold make_mulfimm. - destruct (Float.eq_dec n (Float.floatofint (Int.repr 2))); intros. + destruct (Float.eq_dec n (Float.of_int (Int.repr 2))); intros. simpl. econstructor; split. eauto. rewrite H; subst n. destruct (e#r2); simpl; auto. rewrite Float.mul2_add; auto. rewrite Float.mul_commut; auto. simpl. econstructor; split; eauto. Qed. +Lemma make_mulfsimm_correct: + forall n r1 r2, + e#r2 = Vsingle n -> + let (op, args) := make_mulfsimm n r1 r1 r2 in + exists v, eval_operation ge (Vptr sp Int.zero) op e##args m = Some v /\ Val.lessdef (Val.mulfs e#r1 e#r2) v. +Proof. + intros; unfold make_mulfsimm. + destruct (Float32.eq_dec n (Float32.of_int (Int.repr 2))); intros. + simpl. econstructor; split. eauto. rewrite H; subst n. + destruct (e#r1); simpl; auto. rewrite Float32.mul2_add; auto. + simpl. econstructor; split; eauto. +Qed. + +Lemma make_mulfsimm_correct_2: + forall n r1 r2, + e#r1 = Vsingle n -> + let (op, args) := make_mulfsimm n r2 r1 r2 in + exists v, eval_operation ge (Vptr sp Int.zero) op e##args m = Some v /\ Val.lessdef (Val.mulfs e#r1 e#r2) v. +Proof. + intros; unfold make_mulfsimm. + destruct (Float32.eq_dec n (Float32.of_int (Int.repr 2))); intros. + simpl. econstructor; split. eauto. rewrite H; subst n. + destruct (e#r2); simpl; auto. rewrite Float32.mul2_add; auto. + rewrite Float32.mul_commut; auto. + simpl. econstructor; split; eauto. +Qed. + Lemma make_cast8signed_correct: forall r x, vmatch bc e#r x -> @@ -444,21 +475,6 @@ Proof. econstructor; split; simpl; eauto. Qed. -Lemma make_singleoffloat_correct: - forall r x, - vmatch bc e#r x -> - let (op, args) := make_singleoffloat r x in - exists v, eval_operation ge (Vptr sp Int.zero) op e##args m = Some v /\ Val.lessdef (Val.singleoffloat e#r) v. -Proof. - intros; unfold make_singleoffloat. - destruct (vincl x Fsingle && generate_float_constants tt) eqn:INCL. - InvBooleans. exists e#r; split; auto. - assert (V: vmatch bc e#r Fsingle). - { eapply vmatch_ge; eauto. apply vincl_ge; auto. } - inv V; simpl; auto. rewrite Float.singleoffloat_of_single by auto. auto. - econstructor; split; simpl; eauto. -Qed. - Lemma op_strength_reduction_correct: forall op args vl v, vl = map (fun r => AE.get r ae) args -> @@ -510,14 +526,16 @@ Proof. exploit addr_strength_reduction_correct; eauto. destruct (addr_strength_reduction addr args0 vl0) as [addr' args']. auto. -(* singleoffloat *) - InvApproxRegs; SimplVM; inv H0. apply make_singleoffloat_correct; auto. (* cond *) inv H0. apply make_cmp_correct; auto. (* mulf *) InvApproxRegs; SimplVM; inv H0. rewrite <- H2. apply make_mulfimm_correct; auto. InvApproxRegs; SimplVM; inv H0. fold (Val.mulf (Vfloat n1) e#r2). rewrite <- H2. apply make_mulfimm_correct_2; auto. +(* mulfs *) + InvApproxRegs; SimplVM; inv H0. rewrite <- H2. apply make_mulfsimm_correct; auto. + InvApproxRegs; SimplVM; inv H0. fold (Val.mulfs (Vsingle n1) e#r2). + rewrite <- H2. apply make_mulfsimm_correct_2; auto. (* default *) exists v; auto. Qed. diff --git a/ia32/Machregs.v b/ia32/Machregs.v index da80a6ed..a9f2b6c4 100644 --- a/ia32/Machregs.v +++ b/ia32/Machregs.v @@ -43,11 +43,8 @@ Global Opaque mreg_eq. Definition mreg_type (r: mreg): typ := match r with - | AX => Tint | BX => Tint | CX => Tint | DX => Tint - | SI => Tint | DI => Tint | BP => Tint - | X0 => Tfloat | X1 => Tfloat | X2 => Tfloat | X3 => Tfloat - | X4 => Tfloat | X5 => Tfloat | X6 => Tfloat | X7 => Tfloat - | FP0 => Tfloat + | AX | BX | CX | DX | SI | DI | BP => Tany32 + | X0 | X1 | X2 | X3 | X4 | X5 | X6 | X7 | FP0 => Tany64 end. Local Open Scope positive_scope. @@ -76,7 +73,6 @@ Definition is_stack_reg (r: mreg) : bool := Definition destroyed_by_op (op: operation): list mreg := match op with - | Omove => FP0 :: nil | Ocast8signed | Ocast8unsigned | Ocast16signed | Ocast16unsigned => AX :: nil | Omulhs => AX :: DX :: nil | Omulhu => AX :: DX :: nil @@ -95,9 +91,7 @@ Definition destroyed_by_load (chunk: memory_chunk) (addr: addressing): list mreg Definition destroyed_by_store (chunk: memory_chunk) (addr: addressing): list mreg := match chunk with | Mint8signed | Mint8unsigned => AX :: CX :: nil - | Mint16signed | Mint16unsigned | Mint32 | Mint64 => nil - | Mfloat32 => X7 :: nil - | Mfloat64 => FP0 :: nil + | _ => nil end. Definition destroyed_by_cond (cond: condition): list mreg := @@ -116,9 +110,7 @@ Definition destroyed_by_builtin (ef: external_function): list mreg := | EF_memcpy sz al => if zle sz 32 then CX :: X7 :: nil else CX :: SI :: DI :: nil | EF_vstore (Mint8unsigned|Mint8signed) => AX :: CX :: nil - | EF_vstore Mfloat32 => X7 :: nil | EF_vstore_global (Mint8unsigned|Mint8signed) _ _ => AX :: nil - | EF_vstore_global Mfloat32 _ _ => X7 :: nil | EF_builtin id sg => if ident_eq id builtin_write16_reversed || ident_eq id builtin_write32_reversed @@ -127,12 +119,12 @@ Definition destroyed_by_builtin (ef: external_function): list mreg := end. Definition destroyed_at_function_entry: list mreg := - DX :: FP0 :: nil. (* must include destroyed_by_op Omove *) + (* must include [destroyed_by_setstack ty] *) + DX :: FP0 :: nil. Definition destroyed_by_setstack (ty: typ): list mreg := match ty with - | Tfloat => FP0 :: nil - | Tsingle => X7 :: FP0 :: nil + | Tfloat | Tsingle => FP0 :: nil | _ => nil end. @@ -190,6 +182,7 @@ Definition two_address_op (op: operation) : bool := | Omove => false | Ointconst _ => false | Ofloatconst _ => false + | Osingleconst _ => false | Oindirectsymbol _ => false | Ocast8signed => false | Ocast8unsigned => false @@ -228,9 +221,18 @@ Definition two_address_op (op: operation) : bool := | Osubf => true | Omulf => true | Odivf => true + | Onegfs => true + | Oabsfs => true + | Oaddfs => true + | Osubfs => true + | Omulfs => true + | Odivfs => true | Osingleoffloat => false + | Ofloatofsingle => false | Ointoffloat => false | Ofloatofint => false + | Ointofsingle => false + | Osingleofint => false | Omakelong => false | Olowlong => false | Ohighlong => false diff --git a/ia32/NeedOp.v b/ia32/NeedOp.v index 2c8698f9..52b9fcbe 100644 --- a/ia32/NeedOp.v +++ b/ia32/NeedOp.v @@ -37,6 +37,7 @@ Definition needs_of_operation (op: operation) (nv: nval): list nval := | Omove => op1 nv | Ointconst n => nil | Ofloatconst n => nil + | Osingleconst n => nil | Oindirectsymbol id => nil | Ocast8signed => op1 (sign_ext 8 nv) | Ocast8unsigned => op1 (zero_ext 8 nv) @@ -66,8 +67,10 @@ Definition needs_of_operation (op: operation) (nv: nval): list nval := | Olea addr => needs_of_addressing addr nv | Onegf | Oabsf => op1 (default nv) | Oaddf | Osubf | Omulf | Odivf => op2 (default nv) - | Osingleoffloat => op1 (singleoffloat nv) - | Ointoffloat | Ofloatofint => op1 (default nv) + | Onegfs | Oabsfs => op1 (default nv) + | Oaddfs | Osubfs | Omulfs | Odivfs => op2 (default nv) + | Osingleoffloat | Ofloatofsingle => op1 (default nv) + | Ointoffloat | Ofloatofint | Ointofsingle | Osingleofint => op1 (default nv) | Omakelong => op2 (default nv) | Olowlong | Ohighlong => op1 (default nv) | Ocmp c => needs_of_condition c @@ -81,7 +84,6 @@ Definition operation_is_redundant (op: operation) (nv: nval): bool := | Ocast16unsigned => zero_ext_redundant 16 nv | Oandimm n => andimm_redundant nv n | Oorimm n => orimm_redundant nv n - | Osingleoffloat => singleoffloat_redundant nv | _ => false end. @@ -165,7 +167,6 @@ Proof. - apply shruimm_sound; auto. - apply ror_sound; auto. - eapply needs_of_addressing_sound; eauto. -- apply singleoffloat_sound; auto. - destruct (eval_condition c args m) as [b|] eqn:EC; simpl in H2. erewrite needs_of_condition_sound by eauto. subst v; simpl. auto with na. @@ -186,7 +187,6 @@ Proof. - apply zero_ext_redundant_sound; auto. omega. - apply andimm_redundant_sound; auto. - apply orimm_redundant_sound; auto. -- apply singleoffloat_redundant_sound; auto. Qed. End SOUNDNESS. diff --git a/ia32/Op.v b/ia32/Op.v index e46c7400..14e4cbb1 100644 --- a/ia32/Op.v +++ b/ia32/Op.v @@ -42,8 +42,10 @@ Inductive condition : Type := | Ccompu: comparison -> condition (**r unsigned integer comparison *) | Ccompimm: comparison -> int -> condition (**r signed integer comparison with a constant *) | Ccompuimm: comparison -> int -> condition (**r unsigned integer comparison with a constant *) - | Ccompf: comparison -> condition (**r floating-point comparison *) + | Ccompf: comparison -> condition (**r 64-bit floating-point comparison *) | Cnotcompf: comparison -> condition (**r negation of a floating-point comparison *) + | Ccompfs: comparison -> condition (**r 32-bit floating-point comparison *) + | Cnotcompfs: comparison -> condition (**r negation of a floating-point comparison *) | Cmaskzero: int -> condition (**r test [(arg & constant) == 0] *) | Cmasknotzero: int -> condition. (**r test [(arg & constant) != 0] *) @@ -68,6 +70,7 @@ Inductive operation : Type := | Omove: operation (**r [rd = r1] *) | Ointconst: int -> operation (**r [rd] is set to the given integer constant *) | Ofloatconst: float -> operation (**r [rd] is set to the given float constant *) + | Osingleconst: float32 -> operation (**r [rd] is set to the given float constant *) | Oindirectsymbol: ident -> operation (**r [rd] is set to the address of the symbol *) (*c Integer arithmetic: *) | Ocast8signed: operation (**r [rd] is 8-bit sign extension of [r1] *) @@ -108,10 +111,19 @@ Inductive operation : Type := | Osubf: operation (**r [rd = r1 - r2] *) | Omulf: operation (**r [rd = r1 * r2] *) | Odivf: operation (**r [rd = r1 / r2] *) + | Onegfs: operation (**r [rd = - r1] *) + | Oabsfs: operation (**r [rd = abs(r1)] *) + | Oaddfs: operation (**r [rd = r1 + r2] *) + | Osubfs: operation (**r [rd = r1 - r2] *) + | Omulfs: operation (**r [rd = r1 * r2] *) + | Odivfs: operation (**r [rd = r1 / r2] *) | Osingleoffloat: operation (**r [rd] is [r1] truncated to single-precision float *) + | Ofloatofsingle: operation (**r [rd] is [r1] extended to double-precision float *) (*c Conversions between int and float: *) - | Ointoffloat: operation (**r [rd = signed_int_of_float(r1)] *) - | Ofloatofint: operation (**r [rd = float_of_signed_int(r1)] *) + | Ointoffloat: operation (**r [rd = signed_int_of_float64(r1)] *) + | Ofloatofint: operation (**r [rd = float64_of_signed_int(r1)] *) + | Ointofsingle: operation (**r [rd = signed_int_of_float32(r1)] *) + | Osingleofint: operation (**r [rd = float32_of_signed_int(r1)] *) (*c Manipulating 64-bit integers: *) | Omakelong: operation (**r [rd = r1 << 32 | r2] *) | Olowlong: operation (**r [rd = low-word(r1)] *) @@ -145,6 +157,7 @@ Definition eq_operation (x y: operation): {x=y} + {x<>y}. Proof. generalize Int.eq_dec; intro. generalize Float.eq_dec; intro. + generalize Float32.eq_dec; intro. generalize Int64.eq_dec; intro. decide equality. apply peq. @@ -169,6 +182,8 @@ Definition eval_condition (cond: condition) (vl: list val) (m: mem): option bool | Ccompuimm c n, v1 :: nil => Val.cmpu_bool (Mem.valid_pointer m) c v1 (Vint n) | Ccompf c, v1 :: v2 :: nil => Val.cmpf_bool c v1 v2 | Cnotcompf c, v1 :: v2 :: nil => option_map negb (Val.cmpf_bool c v1 v2) + | Ccompfs c, v1 :: v2 :: nil => Val.cmpfs_bool c v1 v2 + | Cnotcompfs c, v1 :: v2 :: nil => option_map negb (Val.cmpfs_bool c v1 v2) | Cmaskzero n, v1 :: nil => Val.maskzero_bool v1 n | Cmasknotzero n, v1 :: nil => option_map negb (Val.maskzero_bool v1 n) | _, _ => None @@ -204,6 +219,7 @@ Definition eval_operation | Omove, v1::nil => Some v1 | Ointconst n, nil => Some (Vint n) | Ofloatconst n, nil => Some (Vfloat n) + | Osingleconst n, nil => Some (Vsingle n) | Oindirectsymbol id, nil => Some (Genv.symbol_address genv id Int.zero) | Ocast8signed, v1 :: nil => Some (Val.sign_ext 8 v1) | Ocast8unsigned, v1 :: nil => Some (Val.zero_ext 8 v1) @@ -243,9 +259,18 @@ Definition eval_operation | Osubf, v1::v2::nil => Some(Val.subf v1 v2) | Omulf, v1::v2::nil => Some(Val.mulf v1 v2) | Odivf, v1::v2::nil => Some(Val.divf v1 v2) + | Onegfs, v1::nil => Some(Val.negfs v1) + | Oabsfs, v1::nil => Some(Val.absfs v1) + | Oaddfs, v1::v2::nil => Some(Val.addfs v1 v2) + | Osubfs, v1::v2::nil => Some(Val.subfs v1 v2) + | Omulfs, v1::v2::nil => Some(Val.mulfs v1 v2) + | Odivfs, v1::v2::nil => Some(Val.divfs v1 v2) | Osingleoffloat, v1::nil => Some(Val.singleoffloat v1) + | Ofloatofsingle, v1::nil => Some(Val.floatofsingle v1) | Ointoffloat, v1::nil => Val.intoffloat v1 | Ofloatofint, v1::nil => Val.floatofint v1 + | Ointofsingle, v1::nil => Val.intofsingle v1 + | Osingleofint, v1::nil => Val.singleofint v1 | Omakelong, v1::v2::nil => Some(Val.longofwords v1 v2) | Olowlong, v1::nil => Some(Val.loword v1) | Ohighlong, v1::nil => Some(Val.hiword v1) @@ -275,6 +300,8 @@ Definition type_of_condition (c: condition) : list typ := | Ccompuimm _ _ => Tint :: nil | Ccompf _ => Tfloat :: Tfloat :: nil | Cnotcompf _ => Tfloat :: Tfloat :: nil + | Ccompfs _ => Tsingle :: Tsingle :: nil + | Cnotcompfs _ => Tsingle :: Tsingle :: nil | Cmaskzero _ => Tint :: nil | Cmasknotzero _ => Tint :: nil end. @@ -295,7 +322,8 @@ Definition type_of_operation (op: operation) : list typ * typ := match op with | Omove => (nil, Tint) (* treated specially *) | Ointconst _ => (nil, Tint) - | Ofloatconst f => (nil, if Float.is_single_dec f then Tsingle else Tfloat) + | Ofloatconst f => (nil, Tfloat) + | Osingleconst f => (nil, Tsingle) | Oindirectsymbol _ => (nil, Tint) | Ocast8signed => (Tint :: nil, Tint) | Ocast8unsigned => (Tint :: nil, Tint) @@ -334,9 +362,18 @@ Definition type_of_operation (op: operation) : list typ * typ := | Osubf => (Tfloat :: Tfloat :: nil, Tfloat) | Omulf => (Tfloat :: Tfloat :: nil, Tfloat) | Odivf => (Tfloat :: Tfloat :: nil, Tfloat) + | Onegfs => (Tsingle :: nil, Tsingle) + | Oabsfs => (Tsingle :: nil, Tsingle) + | Oaddfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Osubfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Omulfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Odivfs => (Tsingle :: Tsingle :: nil, Tsingle) | Osingleoffloat => (Tfloat :: nil, Tsingle) + | Ofloatofsingle => (Tsingle :: nil, Tfloat) | Ointoffloat => (Tfloat :: nil, Tint) | Ofloatofint => (Tint :: nil, Tfloat) + | Ointofsingle => (Tsingle :: nil, Tint) + | Osingleofint => (Tint :: nil, Tsingle) | Omakelong => (Tint :: Tint :: nil, Tlong) | Olowlong => (Tlong :: nil, Tint) | Ohighlong => (Tlong :: nil, Tint) @@ -380,7 +417,8 @@ Proof with (try exact I). destruct op; simpl in H0; FuncInv; subst; simpl. congruence. exact I. - destruct (Float.is_single_dec f); auto. + exact I. + exact I. unfold Genv.symbol_address; destruct (Genv.find_symbol genv i)... destruct v0... destruct v0... @@ -422,8 +460,17 @@ Proof with (try exact I). destruct v0; destruct v1... destruct v0; destruct v1... destruct v0; destruct v1... - destruct v0... apply Float.singleoffloat_is_single. - destruct v0; simpl in H0; inv H0. destruct (Float.intoffloat f); inv H2... + destruct v0... + destruct v0... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0... + destruct v0... + destruct v0; simpl in H0; inv H0. destruct (Float.to_int f); inv H2... + destruct v0; simpl in H0; inv H0... + destruct v0; simpl in H0; inv H0. destruct (Float32.to_int f); inv H2... destruct v0; simpl in H0; inv H0... destruct v0; destruct v1... destruct v0... @@ -467,6 +514,8 @@ Definition negate_condition (cond: condition): condition := | Ccompuimm c n => Ccompuimm (negate_comparison c) n | Ccompf c => Cnotcompf c | Cnotcompf c => Ccompf c + | Ccompfs c => Cnotcompfs c + | Cnotcompfs c => Ccompfs c | Cmaskzero n => Cmasknotzero n | Cmasknotzero n => Cmaskzero n end. @@ -482,6 +531,8 @@ Proof. repeat (destruct vl; auto). apply Val.negate_cmpu_bool. repeat (destruct vl; auto). repeat (destruct vl; auto). destruct (Val.cmpf_bool c v v0) as [[]|]; auto. + repeat (destruct vl; auto). + repeat (destruct vl; auto). destruct (Val.cmpfs_bool c v v0) as [[]|]; auto. destruct vl; auto. destruct vl; auto. destruct vl; auto. destruct vl; auto. destruct (Val.maskzero_bool v i) as [[]|]; auto. Qed. @@ -608,61 +659,6 @@ Proof. destruct c; simpl; try congruence. reflexivity. Qed. -(** Checking whether two addressings, applied to the same arguments, produce - separated memory addresses. Used in [CSE]. *) - -Definition addressing_separated (chunk1: memory_chunk) (addr1: addressing) - (chunk2: memory_chunk) (addr2: addressing) : bool := - match addr1, addr2 with - | Aindexed ofs1, Aindexed ofs2 => - Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) - | Aglobal s1 ofs1, Aglobal s2 ofs2 => - if ident_eq s1 s2 then Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) else true - | Abased s1 ofs1, Abased s2 ofs2 => - if ident_eq s1 s2 then Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) else true - | Ainstack ofs1, Ainstack ofs2 => - Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) - | _, _ => false - end. - -Lemma addressing_separated_sound: - forall (F V: Type) (ge: Genv.t F V) sp chunk1 addr1 chunk2 addr2 vl b1 n1 b2 n2, - addressing_separated chunk1 addr1 chunk2 addr2 = true -> - eval_addressing ge sp addr1 vl = Some(Vptr b1 n1) -> - eval_addressing ge sp addr2 vl = Some(Vptr b2 n2) -> - b1 <> b2 \/ Int.unsigned n1 + size_chunk chunk1 <= Int.unsigned n2 \/ Int.unsigned n2 + size_chunk chunk2 <= Int.unsigned n1. -Proof. - unfold addressing_separated; intros. - generalize (size_chunk_pos chunk1) (size_chunk_pos chunk2); intros SZ1 SZ2. - destruct addr1; destruct addr2; try discriminate; simpl in *; FuncInv. -(* Aindexed *) - destruct v; simpl in *; inv H1; inv H2. - right. apply Int.no_overlap_sound; auto. -(* Aglobal *) - unfold Genv.symbol_address in *. - destruct (Genv.find_symbol ge i1) eqn:?; inv H2. - destruct (Genv.find_symbol ge i) eqn:?; inv H1. - destruct (ident_eq i i1). subst. - replace (Int.unsigned n1) with (Int.unsigned (Int.add Int.zero n1)). - replace (Int.unsigned n2) with (Int.unsigned (Int.add Int.zero n2)). - right. apply Int.no_overlap_sound; auto. - rewrite Int.add_commut; rewrite Int.add_zero; auto. - rewrite Int.add_commut; rewrite Int.add_zero; auto. - left. red; intros; elim n. subst. eapply Genv.genv_vars_inj; eauto. -(* Abased *) - unfold Genv.symbol_address in *. - destruct (Genv.find_symbol ge i1) eqn:?; simpl in *; try discriminate. - destruct v; inv H2. - destruct (Genv.find_symbol ge i) eqn:?; inv H1. - destruct (ident_eq i i1). subst. - rewrite (Int.add_commut i0 i3). rewrite (Int.add_commut i2 i3). - right. apply Int.no_overlap_sound; auto. - left. red; intros; elim n. subst. eapply Genv.genv_vars_inj; eauto. -(* Ainstack *) - destruct sp; simpl in *; inv H1; inv H2. - right. apply Int.no_overlap_sound; auto. -Qed. - (** * Invariance and compatibility properties. *) (** [eval_operation] and [eval_addressing] depend on a global environment @@ -770,6 +766,8 @@ Proof. eauto 3 using val_cmpu_bool_inject, Mem.valid_pointer_implies. inv H3; inv H2; simpl in H0; inv H0; auto. inv H3; inv H2; simpl in H0; inv H0; auto. + inv H3; inv H2; simpl in H0; inv H0; auto. + inv H3; inv H2; simpl in H0; inv H0; auto. inv H3; try discriminate; auto. inv H3; try discriminate; auto. Qed. @@ -853,7 +851,17 @@ Proof. inv H4; inv H2; simpl; auto. inv H4; inv H2; simpl; auto. inv H4; simpl; auto. - inv H4; simpl in H1; inv H1. simpl. destruct (Float.intoffloat f0); simpl in H2; inv H2. + inv H4; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; simpl; auto. + inv H4; simpl; auto. + inv H4; simpl in H1; inv H1. simpl. destruct (Float.to_int f0); simpl in H2; inv H2. + exists (Vint i); auto. + inv H4; simpl in H1; inv H1. simpl. TrivialExists. + inv H4; simpl in H1; inv H1. simpl. destruct (Float32.to_int f0); simpl in H2; inv H2. exists (Vint i); auto. inv H4; simpl in H1; inv H1. simpl. TrivialExists. inv H4; inv H2; simpl; auto. diff --git a/ia32/PrintAsm.ml b/ia32/PrintAsm.ml index 33e19f73..67fb80c1 100644 --- a/ia32/PrintAsm.ml +++ b/ia32/PrintAsm.ml @@ -353,7 +353,7 @@ let print_builtin_vload_common oc chunk addr res = fprintf oc " movl %a, %a\n" addressing addr ireg res2 end | Mfloat32, [FR res] -> - fprintf oc " cvtss2sd %a, %a\n" addressing addr freg res + fprintf oc " movss %a, %a\n" addressing addr freg res | Mfloat64, [FR res] -> fprintf oc " movsd %a, %a\n" addressing addr freg res | _ -> @@ -394,8 +394,7 @@ let print_builtin_vstore_common oc chunk addr src tmp = fprintf oc " movl %a, %a\n" ireg src2 addressing addr; fprintf oc " movl %a, %a\n" ireg src1 addressing addr' | Mfloat32, [FR src] -> - fprintf oc " cvtsd2ss %a, %%xmm7\n" freg src; - fprintf oc " movss %%xmm7, %a\n" addressing addr + fprintf oc " movss %a, %a\n" freg src addressing addr | Mfloat64, [FR src] -> fprintf oc " movsd %a, %a\n" freg src addressing addr | _ -> @@ -541,7 +540,8 @@ let print_builtin_inline oc name args res = (* Printing of instructions *) -let float_literals : (int * int64) list ref = ref [] +let float64_literals : (int * int64) list ref = ref [] +let float32_literals : (int * int32) list ref = ref [] let jumptables : (int * label list) list ref = ref [] let indirect_symbols : StringSet.t ref = ref StringSet.empty @@ -560,39 +560,38 @@ let print_instruction oc = function fprintf oc " movl L%a$non_lazy_ptr, %a\n" raw_symbol id ireg rd end else fprintf oc " movl $%a, %a\n" symbol id ireg rd - | Pmov_rm(rd, a) -> + | Pmov_rm(rd, a) | Pmov_rm_a(rd, a) -> fprintf oc " movl %a, %a\n" addressing a ireg rd - | Pmov_mr(a, r1) -> + | Pmov_mr(a, r1) | Pmov_mr_a(a, r1) -> fprintf oc " movl %a, %a\n" ireg r1 addressing a | Pmovsd_ff(rd, r1) -> fprintf oc " movapd %a, %a\n" freg r1 freg rd | Pmovsd_fi(rd, n) -> - let b = camlint64_of_coqint (Floats.Float.bits_of_double n) in + let b = camlint64_of_coqint (Floats.Float.to_bits n) in let lbl = new_label() in fprintf oc " movsd %a, %a %s %.18g\n" label lbl freg rd comment (camlfloat_of_coqfloat n); - float_literals := (lbl, b) :: !float_literals + float64_literals := (lbl, b) :: !float64_literals | Pmovsd_fm(rd, a) -> fprintf oc " movsd %a, %a\n" addressing a freg rd | Pmovsd_mf(a, r1) -> fprintf oc " movsd %a, %a\n" freg r1 addressing a - | Pfld_f(r1) -> - fprintf oc " subl $8, %%esp\n"; - cfi_adjust oc 8l; - fprintf oc " movsd %a, 0(%%esp)\n" freg r1; - fprintf oc " fldl 0(%%esp)\n"; - fprintf oc " addl $8, %%esp\n"; - cfi_adjust oc (-8l) - | Pfld_m(a) -> + | Pmovss_fi(rd, n) -> + let b = camlint_of_coqint (Floats.Float32.to_bits n) in + let lbl = new_label() in + fprintf oc " movss %a, %a %s %.18g\n" label lbl freg rd comment (camlfloat_of_coqfloat32 n); + float32_literals := (lbl, b) :: !float32_literals + | Pmovss_fm(rd, a) -> + fprintf oc " movss %a, %a\n" addressing a freg rd + | Pmovss_mf(a, r1) -> + fprintf oc " movss %a, %a\n" freg r1 addressing a + | Pfldl_m(a) -> fprintf oc " fldl %a\n" addressing a - | Pfstp_f(rd) -> - fprintf oc " subl $8, %%esp\n"; - cfi_adjust oc 8l; - fprintf oc " fstpl 0(%%esp)\n"; - fprintf oc " movsd 0(%%esp), %a\n" freg rd; - fprintf oc " addl $8, %%esp\n"; - cfi_adjust oc (-8l) - | Pfstp_m(a) -> + | Pfstpl_m(a) -> fprintf oc " fstpl %a\n" addressing a + | Pflds_m(a) -> + fprintf oc " flds %a\n" addressing a + | Pfstps_m(a) -> + fprintf oc " fstps %a\n" addressing a | Pxchg_rr(r1, r2) -> fprintf oc " xchgl %a, %a\n" ireg r1 ireg r2 (** Moves with conversion *) @@ -616,18 +615,18 @@ let print_instruction oc = function fprintf oc " movswl %a, %a\n" ireg16 r1 ireg rd | Pmovsw_rm(rd, a) -> fprintf oc " movswl %a, %a\n" addressing a ireg rd - | Pcvtss2sd_fm(rd, a) -> - fprintf oc " cvtss2sd %a, %a\n" addressing a freg rd | Pcvtsd2ss_ff(rd, r1) -> - fprintf oc " cvtsd2ss %a, %a\n" freg r1 freg rd; - fprintf oc " cvtss2sd %a, %a\n" freg rd freg rd - | Pcvtsd2ss_mf(a, r1) -> - fprintf oc " cvtsd2ss %a, %%xmm7\n" freg r1; - fprintf oc " movss %%xmm7, %a\n" addressing a + fprintf oc " cvtsd2ss %a, %a\n" freg r1 freg rd + | Pcvtss2sd_ff(rd, r1) -> + fprintf oc " cvtss2sd %a, %a\n" freg r1 freg rd | Pcvttsd2si_rf(rd, r1) -> fprintf oc " cvttsd2si %a, %a\n" freg r1 ireg rd | Pcvtsi2sd_fr(rd, r1) -> fprintf oc " cvtsi2sd %a, %a\n" ireg r1 freg rd + | Pcvttss2si_rf(rd, r1) -> + fprintf oc " cvttss2si %a, %a\n" freg r1 ireg rd + | Pcvtsi2ss_fr(rd, r1) -> + fprintf oc " cvtsi2ss %a, %a\n" ireg r1 freg rd (** Arithmetic and logical operations over integers *) | Plea(rd, a) -> fprintf oc " leal %a, %a\n" addressing a ireg rd @@ -713,6 +712,24 @@ let print_instruction oc = function fprintf oc " comisd %a, %a\n" freg r2 freg r1 | Pxorpd_f (rd) -> fprintf oc " xorpd %a, %a\n" freg rd freg rd + | Padds_ff(rd, r1) -> + fprintf oc " addss %a, %a\n" freg r1 freg rd + | Psubs_ff(rd, r1) -> + fprintf oc " subss %a, %a\n" freg r1 freg rd + | Pmuls_ff(rd, r1) -> + fprintf oc " mulss %a, %a\n" freg r1 freg rd + | Pdivs_ff(rd, r1) -> + fprintf oc " divss %a, %a\n" freg r1 freg rd + | Pnegs (rd) -> + need_masks := true; + fprintf oc " xorpd %a, %a\n" raw_symbol "__negs_mask" freg rd + | Pabss (rd) -> + need_masks := true; + fprintf oc " andpd %a, %a\n" raw_symbol "__abss_mask" freg rd + | Pcomiss_ff(r1, r2) -> + fprintf oc " comiss %a, %a\n" freg r2 freg r1 + | Pxorps_f (rd) -> + fprintf oc " xorpd %a, %a\n" freg rd freg rd (** Branches and calls *) | Pjmp_l(l) -> fprintf oc " jmp %a\n" label (transl_label l) @@ -785,8 +802,10 @@ let print_instruction oc = function assert false end -let print_literal oc (lbl, n) = +let print_literal64 oc (lbl, n) = fprintf oc "%a: .quad 0x%Lx\n" label lbl n +let print_literal32 oc (lbl, n) = + fprintf oc "%a: .long 0x%lx\n" label lbl n let print_jumptable oc (lbl, tbl) = fprintf oc "%a:" label lbl; @@ -796,7 +815,7 @@ let print_jumptable oc (lbl, tbl) = let print_function oc name fn = Hashtbl.clear current_function_labels; - float_literals := []; + float64_literals := []; float32_literals := []; jumptables := []; current_function_sig := fn.fn_sig; let (text, lit, jmptbl) = @@ -818,11 +837,13 @@ let print_function oc name fn = fprintf oc " .type %a, @function\n" symbol name; fprintf oc " .size %a, . - %a\n" symbol name symbol name end; - if !float_literals <> [] then begin + if !float64_literals <> [] || !float32_literals <> [] then begin section oc lit; print_align oc 8; - List.iter (print_literal oc) !float_literals; - float_literals := [] + List.iter (print_literal64 oc) !float64_literals; + float64_literals := []; + List.iter (print_literal32 oc) !float32_literals; + float32_literals := [] end; if !jumptables <> [] then begin section oc jmptbl; @@ -842,11 +863,11 @@ let print_init oc = function fprintf oc " .quad %Ld\n" (camlint64_of_coqint n) | Init_float32 n -> fprintf oc " .long %ld %s %.18g\n" - (camlint_of_coqint (Floats.Float.bits_of_single n)) + (camlint_of_coqint (Floats.Float32.to_bits n)) comment (camlfloat_of_coqfloat n) | Init_float64 n -> fprintf oc " .quad %Ld %s %.18g\n" - (camlint64_of_coqint (Floats.Float.bits_of_double n)) + (camlint64_of_coqint (Floats.Float.to_bits n)) comment (camlfloat_of_coqfloat n) | Init_space n -> if Z.gt n Z.zero then @@ -917,7 +938,11 @@ let print_program oc p = fprintf oc "%a: .quad 0x8000000000000000, 0\n" raw_symbol "__negd_mask"; fprintf oc "%a: .quad 0x7FFFFFFFFFFFFFFF, 0xFFFFFFFFFFFFFFFF\n" - raw_symbol "__absd_mask" + raw_symbol "__absd_mask"; + fprintf oc "%a: .long 0x80000000, 0, 0, 0\n" + raw_symbol "__negs_mask"; + fprintf oc "%a: .long 0x7FFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF\n" + raw_symbol "__abss_mask" end; if target = MacOS then begin fprintf oc " .section __IMPORT,__pointers,non_lazy_symbol_pointers\n"; diff --git a/ia32/PrintOp.ml b/ia32/PrintOp.ml index 193779e5..fb9a7cc5 100644 --- a/ia32/PrintOp.ml +++ b/ia32/PrintOp.ml @@ -64,6 +64,7 @@ let print_operation reg pp = function | Omove, [r1] -> reg pp r1 | Ointconst n, [] -> fprintf pp "%ld" (camlint_of_coqint n) | Ofloatconst n, [] -> fprintf pp "%F" (camlfloat_of_coqfloat n) + | Osingleconst n, [] -> fprintf pp "%Ff" (camlfloat_of_coqfloat32 n) | Oindirectsymbol id, [] -> fprintf pp "&%s" (extern_atom id) | Ocast8signed, [r1] -> fprintf pp "int8signed(%a)" reg r1 | Ocast8unsigned, [r1] -> fprintf pp "int8unsigned(%a)" reg r1 @@ -100,6 +101,7 @@ let print_operation reg pp = function | Omulf, [r1;r2] -> fprintf pp "%a *f %a" reg r1 reg r2 | Odivf, [r1;r2] -> fprintf pp "%a /f %a" reg r1 reg r2 | Osingleoffloat, [r1] -> fprintf pp "singleoffloat(%a)" reg r1 + | Ofloatofsingle, [r1] -> fprintf pp "floatofsingle(%a)" reg r1 | Ointoffloat, [r1] -> fprintf pp "intoffloat(%a)" reg r1 | Ofloatofint, [r1] -> fprintf pp "floatofint(%a)" reg r1 | Omakelong, [r1;r2] -> fprintf pp "makelong(%a,%a)" reg r1 reg r2 diff --git a/ia32/SelectOp.vp b/ia32/SelectOp.vp index 214608e4..b6aef725 100644 --- a/ia32/SelectOp.vp +++ b/ia32/SelectOp.vp @@ -346,6 +346,12 @@ Definition addf (e1 e2: expr) := Eop Oaddf (e1 ::: e2 ::: Enil). Definition subf (e1 e2: expr) := Eop Osubf (e1 ::: e2 ::: Enil). Definition mulf (e1 e2: expr) := Eop Omulf (e1 ::: e2 ::: Enil). +Definition negfs (e: expr) := Eop Onegfs (e ::: Enil). +Definition absfs (e: expr) := Eop Oabsfs (e ::: Enil). +Definition addfs (e1 e2: expr) := Eop Oaddfs (e1 ::: e2 ::: Enil). +Definition subfs (e1 e2: expr) := Eop Osubfs (e1 ::: e2 ::: Enil). +Definition mulfs (e1 e2: expr) := Eop Omulfs (e1 ::: e2 ::: Enil). + (** ** Comparisons *) Nondetfunction compimm (default: comparison -> int -> condition) @@ -405,6 +411,9 @@ Nondetfunction compu (c: comparison) (e1: expr) (e2: expr) := Definition compf (c: comparison) (e1: expr) (e2: expr) := Eop (Ocmp (Ccompf c)) (e1 ::: e2 ::: Enil). +Definition compfs (c: comparison) (e1: expr) (e2: expr) := + Eop (Ocmp (Ccompfs c)) (e1 ::: e2 ::: Enil). + (** ** Integer conversions *) Nondetfunction cast8unsigned (e: expr) := @@ -446,32 +455,50 @@ Nondetfunction cast16signed (e: expr) := (** Floating-point conversions *) Definition singleoffloat (e: expr) := Eop Osingleoffloat (e ::: Enil). +Definition floatofsingle (e: expr) := Eop Ofloatofsingle (e ::: Enil). + Definition intoffloat (e: expr) := Eop Ointoffloat (e ::: Enil). Nondetfunction floatofint (e: expr) := match e with - | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.floatofint n)) Enil + | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.of_int n)) Enil | _ => Eop Ofloatofint (e ::: Enil) end. Definition intuoffloat (e: expr) := Elet e - (Elet (Eop (Ofloatconst (Float.floatofintu Float.ox8000_0000)) Enil) + (Elet (Eop (Ofloatconst (Float.of_intu Float.ox8000_0000)) Enil) (Econdition (CEcond (Ccompf Clt) (Eletvar 1 ::: Eletvar 0 ::: Enil)) (intoffloat (Eletvar 1)) (addimm Float.ox8000_0000 (intoffloat (subf (Eletvar 1) (Eletvar 0))))))%nat. Nondetfunction floatofintu (e: expr) := match e with - | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.floatofintu n)) Enil + | Eop (Ointconst n) Enil => Eop (Ofloatconst (Float.of_intu n)) Enil | _ => - let f := Eop (Ofloatconst (Float.floatofintu Float.ox8000_0000)) Enil in + let f := Eop (Ofloatconst (Float.of_intu Float.ox8000_0000)) Enil in Elet e (Econdition (CEcond (Ccompuimm Clt Float.ox8000_0000) (Eletvar O ::: Enil)) (floatofint (Eletvar O)) (addf (floatofint (addimm (Int.neg Float.ox8000_0000) (Eletvar O))) f)) end. +Definition intofsingle (e: expr) := Eop Ointofsingle (e ::: Enil). + +Nondetfunction singleofint (e: expr) := + match e with + | Eop (Ointconst n) Enil => Eop (Osingleconst (Float32.of_int n)) Enil + | _ => Eop Osingleofint (e ::: Enil) + end. + +Definition intuofsingle (e: expr) := intuoffloat (floatofsingle e). + +Nondetfunction singleofintu (e: expr) := + match e with + | Eop (Ointconst n) Enil => Eop (Osingleconst (Float32.of_intu n)) Enil + | _ => singleoffloat (floatofintu e) + end. + (** ** Addressing modes *) Nondetfunction addressing (chunk: memory_chunk) (e: expr) := diff --git a/ia32/SelectOpproof.v b/ia32/SelectOpproof.v index 0fd6f7b8..50688621 100644 --- a/ia32/SelectOpproof.v +++ b/ia32/SelectOpproof.v @@ -567,6 +567,31 @@ Proof. red; intros; TrivialExists. Qed. +Theorem eval_negfs: unary_constructor_sound negfs Val.negfs. +Proof. + red; intros. TrivialExists. +Qed. + +Theorem eval_absfs: unary_constructor_sound absfs Val.absfs. +Proof. + red; intros. TrivialExists. +Qed. + +Theorem eval_addfs: binary_constructor_sound addfs Val.addfs. +Proof. + red; intros; TrivialExists. +Qed. + +Theorem eval_subfs: binary_constructor_sound subfs Val.subfs. +Proof. + red; intros; TrivialExists. +Qed. + +Theorem eval_mulfs: binary_constructor_sound mulfs Val.mulfs. +Proof. + red; intros; TrivialExists. +Qed. + Section COMP_IMM. Variable default: comparison -> int -> condition. @@ -674,6 +699,12 @@ Proof. intros; red; intros. unfold compf. TrivialExists. Qed. +Theorem eval_compfs: + forall c, binary_constructor_sound (compfs c) (Val.cmpfs c). +Proof. + intros; red; intros. unfold compfs. TrivialExists. +Qed. + Theorem eval_cast8signed: unary_constructor_sound cast8signed (Val.sign_ext 8). Proof. red; intros until x. unfold cast8signed. case (cast8signed_match a); intros; InvEval. @@ -711,6 +742,11 @@ Proof. red; intros. unfold singleoffloat. TrivialExists. Qed. +Theorem eval_floatofsingle: unary_constructor_sound floatofsingle Val.floatofsingle. +Proof. + red; intros. unfold floatofsingle. TrivialExists. +Qed. + Theorem eval_intoffloat: forall le a x y, eval_expr ge sp e m le a x -> @@ -738,10 +774,10 @@ Theorem eval_intuoffloat: exists v, eval_expr ge sp e m le (intuoffloat a) v /\ Val.lessdef y v. Proof. intros. destruct x; simpl in H0; try discriminate. - destruct (Float.intuoffloat f) as [n|] eqn:?; simpl in H0; inv H0. + destruct (Float.to_intu f) as [n|] eqn:?; simpl in H0; inv H0. exists (Vint n); split; auto. unfold intuoffloat. set (im := Int.repr Int.half_modulus). - set (fm := Float.floatofintu im). + set (fm := Float.of_intu im). assert (eval_expr ge sp e m (Vfloat fm :: Vfloat f :: le) (Eletvar (S O)) (Vfloat f)). constructor. auto. assert (eval_expr ge sp e m (Vfloat fm :: Vfloat f :: le) (Eletvar O) (Vfloat fm)). @@ -751,9 +787,9 @@ Proof. eapply eval_Econdition with (va := Float.cmp Clt f fm). eauto with evalexpr. destruct (Float.cmp Clt f fm) eqn:?. - exploit Float.intuoffloat_intoffloat_1; eauto. intro EQ. + exploit Float.to_intu_to_int_1; eauto. intro EQ. EvalOp. simpl. rewrite EQ; auto. - exploit Float.intuoffloat_intoffloat_2; eauto. + exploit Float.to_intu_to_int_2; eauto. change Float.ox8000_0000 with im. fold fm. intro EQ. set (t2 := subf (Eletvar (S O)) (Eletvar O)). set (t3 := intoffloat t2). @@ -778,25 +814,75 @@ Proof. intros until y; unfold floatofintu. case (floatofintu_match a); intros. InvEval. TrivialExists. destruct x; simpl in H0; try discriminate. inv H0. - exists (Vfloat (Float.floatofintu i)); split; auto. + exists (Vfloat (Float.of_intu i)); split; auto. econstructor. eauto. - set (fm := Float.floatofintu Float.ox8000_0000). + set (fm := Float.of_intu Float.ox8000_0000). assert (eval_expr ge sp e m (Vint i :: le) (Eletvar O) (Vint i)). constructor. auto. eapply eval_Econdition with (va := Int.ltu i Float.ox8000_0000). eauto with evalexpr. destruct (Int.ltu i Float.ox8000_0000) eqn:?. - rewrite Float.floatofintu_floatofint_1; auto. + rewrite Float.of_intu_of_int_1; auto. unfold floatofint. EvalOp. exploit (eval_addimm (Int.neg Float.ox8000_0000) (Vint i :: le) (Eletvar 0)); eauto. simpl. intros [v [A B]]. inv B. unfold addf. EvalOp. constructor. unfold floatofint. EvalOp. simpl; eauto. constructor. EvalOp. simpl; eauto. constructor. simpl; eauto. - fold fm. rewrite Float.floatofintu_floatofint_2; auto. + fold fm. rewrite Float.of_intu_of_int_2; auto. rewrite Int.sub_add_opp. auto. Qed. +Theorem eval_intofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.intofsingle x = Some y -> + exists v, eval_expr ge sp e m le (intofsingle a) v /\ Val.lessdef y v. +Proof. + intros; unfold intofsingle. TrivialExists. +Qed. + +Theorem eval_singleofint: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.singleofint x = Some y -> + exists v, eval_expr ge sp e m le (singleofint a) v /\ Val.lessdef y v. +Proof. + intros until y; unfold singleofint. case (singleofint_match a); intros; InvEval. + TrivialExists. + TrivialExists. +Qed. + +Theorem eval_intuofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.intuofsingle x = Some y -> + exists v, eval_expr ge sp e m le (intuofsingle a) v /\ Val.lessdef y v. +Proof. + intros. destruct x; simpl in H0; try discriminate. + destruct (Float32.to_intu f) as [n|] eqn:?; simpl in H0; inv H0. + unfold intuofsingle. apply eval_intuoffloat with (Vfloat (Float.of_single f)). + unfold floatofsingle. EvalOp. + simpl. change (Float.of_single f) with (Float32.to_double f). + erewrite Float32.to_intu_double; eauto. auto. +Qed. + +Theorem eval_singleofintu: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.singleofintu x = Some y -> + exists v, eval_expr ge sp e m le (singleofintu a) v /\ Val.lessdef y v. +Proof. + intros until y; unfold singleofintu. case (singleofintu_match a); intros. + InvEval. TrivialExists. + destruct x; simpl in H0; try discriminate. inv H0. + exploit eval_floatofintu. eauto. simpl. reflexivity. + intros (v & A & B). + exists (Val.singleoffloat v); split. + unfold singleoffloat; EvalOp. + inv B; simpl. rewrite Float32.of_intu_double. auto. +Qed. + Theorem eval_addressing: forall le chunk a v b ofs, eval_expr ge sp e m le a v -> diff --git a/ia32/Unusedglob1.ml b/ia32/Unusedglob1.ml index 8332a30c..eb0298bb 100644 --- a/ia32/Unusedglob1.ml +++ b/ia32/Unusedglob1.ml @@ -29,12 +29,14 @@ let referenced_builtin ef = let referenced_instr = function | Pmov_rm (_, a) | Pmov_mr (a, _) + | Pmov_rm_a (_, a) | Pmov_mr_a (a, _) | Pmovsd_fm (_, a) | Pmovsd_mf(a, _) - | Pfld_m a | Pfstp_m a + | Pmovss_fm (_, a) | Pmovss_mf(a, _) + | Pfldl_m a | Pflds_m a | Pfstpl_m a | Pfstps_m a | Pmovb_mr (a, _) | Pmovw_mr (a, _) | Pmovzb_rm (_, a) | Pmovsb_rm (_, a) | Pmovzw_rm (_, a) | Pmovsw_rm (_, a) - | Pcvtss2sd_fm (_, a) | Pcvtsd2ss_mf (a, _) | Plea (_, a) -> referenced_addr a + | Plea (_, a) -> referenced_addr a | Pjmp_s(s, _) -> [s] | Pcall_s(s, _) -> [s] | Pbuiltin(ef, args, res) -> referenced_builtin ef diff --git a/ia32/ValueAOp.v b/ia32/ValueAOp.v index 58b945f4..874c2be3 100644 --- a/ia32/ValueAOp.v +++ b/ia32/ValueAOp.v @@ -32,6 +32,8 @@ Definition eval_static_condition (cond: condition) (vl: list aval): abool := | Ccompuimm c n, v1 :: nil => cmpu_bool c v1 (I n) | Ccompf c, v1 :: v2 :: nil => cmpf_bool c v1 v2 | Cnotcompf c, v1 :: v2 :: nil => cnot (cmpf_bool c v1 v2) + | Ccompfs c, v1 :: v2 :: nil => cmpfs_bool c v1 v2 + | Cnotcompfs c, v1 :: v2 :: nil => cnot (cmpfs_bool c v1 v2) | Cmaskzero n, v1 :: nil => maskzero v1 n | Cmasknotzero n, v1 :: nil => cnot (maskzero v1 n) | _, _ => Bnone @@ -55,6 +57,7 @@ Definition eval_static_operation (op: operation) (vl: list aval): aval := | Omove, v1::nil => v1 | Ointconst n, nil => I n | Ofloatconst n, nil => if propagate_float_constants tt then F n else ftop + | Osingleconst n, nil => if propagate_float_constants tt then FS n else ftop | Oindirectsymbol id, nil => Ptr (Gl id Int.zero) | Ocast8signed, v1 :: nil => sign_ext 8 v1 | Ocast8unsigned, v1 :: nil => zero_ext 8 v1 @@ -93,9 +96,18 @@ Definition eval_static_operation (op: operation) (vl: list aval): aval := | Osubf, v1::v2::nil => subf v1 v2 | Omulf, v1::v2::nil => mulf v1 v2 | Odivf, v1::v2::nil => divf v1 v2 + | Onegfs, v1::nil => negfs v1 + | Oabsfs, v1::nil => absfs v1 + | Oaddfs, v1::v2::nil => addfs v1 v2 + | Osubfs, v1::v2::nil => subfs v1 v2 + | Omulfs, v1::v2::nil => mulfs v1 v2 + | Odivfs, v1::v2::nil => divfs v1 v2 | Osingleoffloat, v1::nil => singleoffloat v1 + | Ofloatofsingle, v1::nil => floatofsingle v1 | Ointoffloat, v1::nil => intoffloat v1 | Ofloatofint, v1::nil => floatofint v1 + | Ointofsingle, v1::nil => intofsingle v1 + | Osingleofint, v1::nil => singleofint v1 | Omakelong, v1::v2::nil => longofwords v1 v2 | Olowlong, v1::nil => loword v1 | Ohighlong, v1::nil => hiword v1 @@ -164,6 +176,7 @@ Proof. unfold eval_operation, eval_static_operation; intros; destruct op; InvHyps; eauto with va. destruct (propagate_float_constants tt); constructor. + destruct (propagate_float_constants tt); constructor. eapply eval_static_addressing_sound; eauto. apply of_optbool_sound. eapply eval_static_condition_sound; eauto. Qed. diff --git a/ia32/standard/Conventions1.v b/ia32/standard/Conventions1.v index e097e855..d1f7acd0 100644 --- a/ia32/standard/Conventions1.v +++ b/ia32/standard/Conventions1.v @@ -15,6 +15,7 @@ Require Import Coqlib. Require Import AST. +Require Import Events. Require Import Locations. (** * Classification of machine registers *) @@ -161,13 +162,13 @@ Proof. Qed. Lemma int_callee_save_type: - forall r, In r int_callee_save_regs -> mreg_type r = Tint. + forall r, In r int_callee_save_regs -> mreg_type r = Tany32. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. Lemma float_callee_save_type: - forall r, In r float_callee_save_regs -> mreg_type r = Tfloat. + forall r, In r float_callee_save_regs -> mreg_type r = Tany64. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. @@ -219,25 +220,20 @@ Qed. Definition loc_result (s: signature) : list mreg := match s.(sig_res) with | None => AX :: nil - | Some Tint => AX :: nil + | Some (Tint | Tany32) => AX :: nil | Some (Tfloat | Tsingle) => FP0 :: nil + | Some Tany64 => X0 :: nil | Some Tlong => DX :: AX :: nil end. -(* -(** The result location has the type stated in the signature. *) +(** The result registers have types compatible with that given in the signature. *) Lemma loc_result_type: forall sig, - mreg_type (loc_result sig) = - match sig.(sig_res) with None => Tint | Some ty => ty end. + subtype_list (proj_sig_res' sig) (map mreg_type (loc_result sig)) = true. Proof. - intros; unfold loc_result. - destruct (sig_res sig). - destruct t; reflexivity. - reflexivity. + intros. unfold proj_sig_res', loc_result. destruct (sig_res sig) as [[]|]; auto. Qed. -*) (** The result locations are caller-save registers *) @@ -246,9 +242,9 @@ Lemma loc_result_caller_save: In r (loc_result s) -> In r destroyed_at_call. Proof. intros. - assert (r = AX \/ r = DX \/ r = FP0). - unfold loc_result in H. destruct (sig_res s); [destruct t|idtac]; simpl in H; intuition. - destruct H0 as [A | [A | A]]; subst r; simpl; OrEq. + assert (r = AX \/ r = DX \/ r = FP0 \/ r = X0). + unfold loc_result in H. destruct (sig_res s) as [[]|]; simpl in H; intuition. + destruct H0 as [A | [A | [A | A]]]; subst r; simpl; OrEq. Qed. (** ** Location of function arguments *) @@ -263,6 +259,8 @@ Fixpoint loc_arguments_rec | Tfloat :: tys => S Outgoing ofs Tfloat :: loc_arguments_rec tys (ofs + 2) | Tsingle :: tys => S Outgoing ofs Tsingle :: loc_arguments_rec tys (ofs + 1) | Tlong :: tys => S Outgoing (ofs + 1) Tint :: S Outgoing ofs Tint :: loc_arguments_rec tys (ofs + 2) + | Tany32 :: tys => S Outgoing ofs Tany32 :: loc_arguments_rec tys (ofs + 1) + | Tany64 :: tys => S Outgoing ofs Tany64 :: loc_arguments_rec tys (ofs + 2) end. (** [loc_arguments s] returns the list of locations where to store arguments @@ -303,21 +301,18 @@ Remark loc_arguments_rec_charact: end. Proof. induction tyl; simpl loc_arguments_rec; intros. - destruct H. - destruct a. -- destruct H. subst l. split. omega. congruence. - exploit IHtyl; eauto. - destruct l; auto. destruct sl; auto. intuition omega. -- destruct H. subst l. split. omega. congruence. - exploit IHtyl; eauto. - destruct l; auto. destruct sl; auto. intuition omega. -- destruct H. subst l; split; [omega|congruence]. - destruct H. subst l; split; [omega|congruence]. - exploit IHtyl; eauto. - destruct l; auto. destruct sl; auto. intuition omega. -- destruct H. subst l. split. omega. congruence. - exploit IHtyl; eauto. - destruct l; auto. destruct sl; auto. intuition omega. +- destruct H. +- assert (REC: forall ofs1, In l (loc_arguments_rec tyl ofs1) -> ofs1 > ofs -> + match l with + | R _ => False + | S Local _ _ => False + | S Incoming _ _ => False + | S Outgoing ofs' ty => ofs' >= ofs /\ ty <> Tlong + end). + { intros. exploit IHtyl; eauto. destruct l; auto. destruct sl; intuition omega +. } + destruct a; simpl in H; repeat (destruct H); + ((eapply REC; eauto; omega) || (split; [omega|congruence])). Qed. Lemma loc_arguments_acceptable: @@ -357,16 +352,15 @@ Lemma loc_arguments_bounded: Proof. intros until ty. unfold loc_arguments, size_arguments. generalize (sig_args s) 0. induction l; simpl; intros. - destruct H. - destruct a. - destruct H. inv H. apply size_arguments_rec_above. auto. - destruct H. inv H. apply size_arguments_rec_above. auto. - destruct H. inv H. +- contradiction. +- Ltac decomp := + match goal with + | [ H: _ \/ _ |- _ ] => destruct H; decomp + | [ H: S _ _ _ = S _ _ _ |- _ ] => inv H + | _ => idtac + end. + destruct a; simpl in H; decomp; auto; try apply size_arguments_rec_above. simpl typesize. replace (z + 1 + 1) with (z + 2) by omega. apply size_arguments_rec_above. - destruct H. inv H. - simpl typesize. apply Zle_trans with (ofs + 2). omega. apply size_arguments_rec_above. - auto. - destruct H. inv H. apply size_arguments_rec_above. auto. + simpl typesize. apply Zle_trans with (ofs + 2). omega. apply size_arguments_rec_above. Qed. - diff --git a/lib/Camlcoq.ml b/lib/Camlcoq.ml index 8d6fd243..dd896367 100644 --- a/lib/Camlcoq.ml +++ b/lib/Camlcoq.ml @@ -320,9 +320,14 @@ let coqstring_of_camlstring s = (* Floats *) let coqfloat_of_camlfloat f = - Float.double_of_bits(coqint_of_camlint64(Int64.bits_of_float f)) + Float.of_bits(coqint_of_camlint64(Int64.bits_of_float f)) let camlfloat_of_coqfloat f = - Int64.float_of_bits(camlint64_of_coqint(Float.bits_of_double f)) + Int64.float_of_bits(camlint64_of_coqint(Float.to_bits f)) + +let coqfloat32_of_camlfloat f = + Float32.of_bits(coqint_of_camlint(Int32.bits_of_float f)) +let camlfloat_of_coqfloat32 f = + Int32.float_of_bits(camlint_of_coqint(Float32.to_bits f)) (* Int31 *) diff --git a/lib/Fappli_IEEE_extra.v b/lib/Fappli_IEEE_extra.v new file mode 100644 index 00000000..5194a644 --- /dev/null +++ b/lib/Fappli_IEEE_extra.v @@ -0,0 +1,1506 @@ +(* *********************************************************************) +(* *) +(* The Compcert verified compiler *) +(* *) +(* Xavier Leroy, INRIA Paris-Rocquencourt *) +(* Jacques-Henri Jourdan, INRIA Paris-Rocquencourt *) +(* *) +(* Copyright Institut National de Recherche en Informatique et en *) +(* Automatique. All rights reserved. This file is distributed *) +(* under the terms of the GNU General Public License as published by *) +(* the Free Software Foundation, either version 2 of the License, or *) +(* (at your option) any later version. This file is also distributed *) +(* under the terms of the INRIA Non-Commercial License Agreement. *) +(* *) +(* *********************************************************************) + +(** Additional operations and proofs about IEEE-754 binary + floating-point numbers, on top of the Flocq library. *) + +Require Import Psatz. +Require Import Bool. +Require Import Eqdep_dec. +Require Import Fcore. +Require Import Fcore_digits. +Require Import Fcalc_digits. +Require Import Fcalc_ops. +Require Import Fcalc_round. +Require Import Fcalc_bracket. +Require Import Fprop_Sterbenz. +Require Import Fappli_IEEE. +Require Import Fappli_rnd_odd. + +Local Open Scope Z_scope. + +Section Extra_ops. + +(** [prec] is the number of bits of the mantissa including the implicit one. + [emax] is the exponent of the infinities. + Typically p=24 and emax = 128 in single precision. *) + +Variable prec emax : Z. +Context (prec_gt_0_ : Prec_gt_0 prec). +Let emin := (3 - emax - prec)%Z. +Let fexp := FLT_exp emin prec. +Hypothesis Hmax : (prec < emax)%Z. +Let binary_float := binary_float prec emax. + +(** Remarks on [is_finite] *) + +Remark is_finite_not_is_nan: + forall (f: binary_float), is_finite _ _ f = true -> is_nan _ _ f = false. +Proof. + destruct f; reflexivity || discriminate. +Qed. + +Remark is_finite_strict_finite: + forall (f: binary_float), is_finite_strict _ _ f = true -> is_finite _ _ f = true. +Proof. + destruct f; reflexivity || discriminate. +Qed. + +(** Digression on FP numbers that cannot be [-0.0]. *) + +Definition is_finite_pos0 (f: binary_float) : bool := + match f with + | B754_zero s => negb s + | B754_infinity _ => false + | B754_nan _ _ => false + | B754_finite _ _ _ _ => true + end. + +Lemma Bsign_pos0: + forall x, is_finite_pos0 x = true -> Bsign _ _ x = Rlt_bool (B2R _ _ x) 0%R. +Proof. + intros. destruct x as [ [] | | | [] ex mx Bx ]; try discriminate; simpl. +- rewrite Rlt_bool_false; auto. lra. +- rewrite Rlt_bool_true; auto. apply F2R_lt_0_compat. compute; auto. +- rewrite Rlt_bool_false; auto. + assert ((F2R (Float radix2 (Z.pos ex) mx) > 0)%R) by + ( apply F2R_gt_0_compat; compute; auto ). + lra. +Qed. + +Theorem B2R_inj_pos0: + forall x y, + is_finite_pos0 x = true -> is_finite_pos0 y = true -> + B2R _ _ x = B2R _ _ y -> + x = y. +Proof. + intros. apply B2R_Bsign_inj. + destruct x; reflexivity||discriminate. + destruct y; reflexivity||discriminate. + auto. + rewrite ! Bsign_pos0 by auto. rewrite H1; auto. +Qed. + +(** ** Decidable equality *) + +Definition Beq_dec: forall (f1 f2: binary_float), {f1 = f2} + {f1 <> f2}. +Proof. + assert (UIP_bool: forall (b1 b2: bool) (e e': b1 = b2), e = e'). + { intros. apply UIP_dec. decide equality. } + Ltac try_not_eq := try solve [right; congruence]. + destruct f1 as [| |? []|], f2 as [| |? []|]; + try destruct b; try destruct b0; + try solve [left; auto]; try_not_eq. + destruct (positive_eq_dec x x0); try_not_eq; + subst; left; f_equal; f_equal; apply UIP_bool. + destruct (positive_eq_dec x x0); try_not_eq; + subst; left; f_equal; f_equal; apply UIP_bool. + destruct (positive_eq_dec m m0); try_not_eq; + destruct (Z_eq_dec e e1); try solve [right; intro H; inversion H; congruence]; + subst; left; f_equal; apply UIP_bool. + destruct (positive_eq_dec m m0); try_not_eq; + destruct (Z_eq_dec e e1); try solve [right; intro H; inversion H; congruence]; + subst; left; f_equal; apply UIP_bool. +Defined. + +(** ** Comparison *) + +(** [Some c] means ordered as per [c]; [None] means unordered. *) + +Definition Bcompare (f1 f2: binary_float): option comparison := + match f1, f2 with + | B754_nan _ _,_ | _,B754_nan _ _ => None + | B754_infinity true, B754_infinity true + | B754_infinity false, B754_infinity false => Some Eq + | B754_infinity true, _ => Some Lt + | B754_infinity false, _ => Some Gt + | _, B754_infinity true => Some Gt + | _, B754_infinity false => Some Lt + | B754_finite true _ _ _, B754_zero _ => Some Lt + | B754_finite false _ _ _, B754_zero _ => Some Gt + | B754_zero _, B754_finite true _ _ _ => Some Gt + | B754_zero _, B754_finite false _ _ _ => Some Lt + | B754_zero _, B754_zero _ => Some Eq + | B754_finite s1 m1 e1 _, B754_finite s2 m2 e2 _ => + match s1, s2 with + | true, false => Some Lt + | false, true => Some Gt + | false, false => + match Zcompare e1 e2 with + | Lt => Some Lt + | Gt => Some Gt + | Eq => Some (Pcompare m1 m2 Eq) + end + | true, true => + match Zcompare e1 e2 with + | Lt => Some Gt + | Gt => Some Lt + | Eq => Some (CompOpp (Pcompare m1 m2 Eq)) + end + end + end. + +Theorem Bcompare_finite_correct: + forall f1 f2, + is_finite _ _ f1 = true -> is_finite _ _ f2 = true -> + Bcompare f1 f2 = Some (Rcompare (B2R _ _ f1) (B2R _ _ f2)). +Proof. + Ltac apply_Rcompare := + match goal with + | [ |- Some Lt = Some (Rcompare _ _) ] => f_equal; symmetry; apply Rcompare_Lt + | [ |- Some Eq = Some (Rcompare _ _) ] => f_equal; symmetry; apply Rcompare_Eq + | [ |- Some Gt = Some (Rcompare _ _) ] => f_equal; symmetry; apply Rcompare_Gt + end. + unfold Bcompare; intros. + destruct f1, f2; try discriminate; unfold B2R, F2R, Fnum, Fexp, cond_Zopp; + try (replace 0%R with (Z2R 0 * bpow radix2 e)%R by (simpl Z2R; ring); + rewrite Rcompare_mult_r by (apply bpow_gt_0); rewrite Rcompare_Z2R). + apply_Rcompare; reflexivity. + destruct b0; reflexivity. + destruct b; reflexivity. + clear H H0. + apply andb_prop in e0; destruct e0; apply (canonic_canonic_mantissa _ _ false) in H. + apply andb_prop in e2; destruct e2; apply (canonic_canonic_mantissa _ _ false) in H1. + pose proof (Zcompare_spec e e1); unfold canonic, Fexp in H1, H. + assert (forall m1 m2 e1 e2, + let x := (Z2R (Zpos m1) * bpow radix2 e1)%R in + let y := (Z2R (Zpos m2) * bpow radix2 e2)%R in + (canonic_exp radix2 fexp x < canonic_exp radix2 fexp y)%Z -> (x < y)%R). + { + intros; apply Rnot_le_lt; intro; apply (ln_beta_le radix2) in H5. + unfold canonic_exp in H4. apply (fexp_monotone prec emax) in H5. + unfold fexp, emin in H4. omega. + apply Rmult_gt_0_compat; [apply (Z2R_lt 0); reflexivity|now apply bpow_gt_0]. + } + assert (forall m1 m2 e1 e2, (Z2R (- Zpos m1) * bpow radix2 e1 < Z2R (Zpos m2) * bpow radix2 e2)%R). + { + intros; apply (Rlt_trans _ 0%R). + replace 0%R with (0*bpow radix2 e0)%R by ring; apply Rmult_lt_compat_r; + [apply bpow_gt_0; reflexivity|now apply (Z2R_lt _ 0)]. + apply Rmult_gt_0_compat; [apply (Z2R_lt 0); reflexivity|now apply bpow_gt_0]. + } + destruct b, b0; try (now apply_Rcompare; apply H5); inversion H3; + try (apply_Rcompare; apply H4; rewrite H, H1 in H7; assumption); + try (apply_Rcompare; do 2 rewrite Z2R_opp, Ropp_mult_distr_l_reverse; + apply Ropp_lt_contravar; apply H4; rewrite H, H1 in H7; assumption); + rewrite H7, Rcompare_mult_r, Rcompare_Z2R by (apply bpow_gt_0); reflexivity. +Qed. + +Theorem Bcompare_swap: + forall x y, + Bcompare y x = match Bcompare x y with Some c => Some (CompOpp c) | None => None end. +Proof. + intros. + destruct x as [ ? | [] | ? ? | [] mx ex Bx ]; + destruct y as [ ? | [] | ? ? | [] my ey By ]; simpl; auto. +- rewrite <- (Zcompare_antisym ex ey). destruct (ex ?= ey)%Z; auto. + simpl. f_equal; f_equal. symmetry. apply Pcompare_antisym. +- rewrite <- (Zcompare_antisym ex ey). destruct (ex ?= ey)%Z; auto. + simpl. f_equal. symmetry. apply Pcompare_antisym. +Qed. + +(** ** Absolute value *) + +Definition Babs abs_nan (x: binary_float) : binary_float := + match x with + | B754_nan sx plx => + let '(sres, plres) := abs_nan sx plx in B754_nan _ _ sres plres + | B754_infinity sx => B754_infinity _ _ false + | B754_finite sx mx ex Hx => B754_finite _ _ false mx ex Hx + | B754_zero sx => B754_zero _ _ false + end. + +Theorem B2R_Babs : + forall abs_nan x, + B2R _ _ (Babs abs_nan x) = Rabs (B2R _ _ x). +Proof. + intros abs_nan [sx|sx|sx plx|sx mx ex Hx]; apply sym_eq ; try apply Rabs_R0. + simpl. destruct abs_nan. simpl. apply Rabs_R0. + simpl. rewrite <- F2R_abs. destruct sx; auto. +Qed. + +Theorem is_finite_Babs : + forall abs_nan x, + is_finite _ _ (Babs abs_nan x) = is_finite _ _ x. +Proof. + intros abs_nan [| | |] ; try easy. + intros s pl. + simpl. + now case abs_nan. +Qed. + +Theorem sign_Babs: + forall abs_nan x, + is_nan _ _ x = false -> + Bsign _ _ (Babs abs_nan x) = false. +Proof. + intros abs_nan [| | |]; reflexivity || discriminate. +Qed. + +Theorem Babs_idempotent : + forall abs_nan (x: binary_float), + is_nan _ _ x = false -> + Babs abs_nan (Babs abs_nan x) = Babs abs_nan x. +Proof. + now intros abs_nan [sx|sx|sx plx|sx mx ex Hx] ; auto. +Qed. + +Theorem Babs_opp: + forall abs_nan opp_nan x, + is_nan _ _ x = false -> + Babs abs_nan (Bopp _ _ opp_nan x) = Babs abs_nan x. +Proof. + intros abs_nan opp_nan [| | |]; reflexivity || discriminate. +Qed. + +(** ** Conversion from an integer to a FP number *) + +(** Integers that can be represented exactly as FP numbers. *) + +Definition integer_representable (n: Z): Prop := + Z.abs n <= 2^emax - 2^(emax - prec) /\ generic_format radix2 fexp (Z2R n). + +Let int_upper_bound_eq: 2^emax - 2^(emax - prec) = (2^prec - 1) * 2^(emax - prec). +Proof. + red in prec_gt_0_. + ring_simplify. rewrite <- (Zpower_plus radix2) by omega. f_equal. f_equal. omega. +Qed. + +Lemma integer_representable_n2p: + forall n p, + -2^prec < n < 2^prec -> 0 <= p -> p <= emax - prec -> + integer_representable (n * 2^p). +Proof. + intros; split. +- red in prec_gt_0_. replace (Z.abs (n * 2^p)) with (Z.abs n * 2^p). + rewrite int_upper_bound_eq. + apply Zmult_le_compat. zify; omega. apply (Zpower_le radix2); omega. + zify; omega. apply (Zpower_ge_0 radix2). + rewrite Z.abs_mul. f_equal. rewrite Z.abs_eq. auto. apply (Zpower_ge_0 radix2). +- apply generic_format_FLT. exists (Float radix2 n p). + unfold F2R; simpl. + split. rewrite <- Z2R_Zpower by auto. apply Z2R_mult. + split. zify; omega. + unfold emin; red in prec_gt_0_; omega. +Qed. + +Lemma integer_representable_2p: + forall p, + 0 <= p <= emax - 1 -> + integer_representable (2^p). +Proof. + intros; split. +- red in prec_gt_0_. + rewrite Z.abs_eq by (apply (Zpower_ge_0 radix2)). + apply Zle_trans with (2^(emax-1)). + apply (Zpower_le radix2); omega. + assert (2^emax = 2^(emax-1)*2). + { change 2 with (2^1) at 3. rewrite <- (Zpower_plus radix2) by omega. + f_equal. omega. } + assert (2^(emax - prec) <= 2^(emax - 1)). + { apply (Zpower_le radix2). omega. } + omega. +- red in prec_gt_0_. + apply generic_format_FLT. exists (Float radix2 1 p). + unfold F2R; simpl. + split. rewrite Rmult_1_l. rewrite <- Z2R_Zpower. auto. omega. + split. change 1 with (2^0). apply (Zpower_lt radix2). omega. auto. + unfold emin; omega. +Qed. + +Lemma integer_representable_opp: + forall n, integer_representable n -> integer_representable (-n). +Proof. + intros n (A & B); split. rewrite Z.abs_opp. auto. + rewrite Z2R_opp. apply generic_format_opp; auto. +Qed. + +Lemma integer_representable_n2p_wide: + forall n p, + -2^prec <= n <= 2^prec -> 0 <= p -> p < emax - prec -> + integer_representable (n * 2^p). +Proof. + intros. red in prec_gt_0_. + destruct (Z.eq_dec n (2^prec)); [idtac | destruct (Z.eq_dec n (-2^prec))]. +- rewrite e. rewrite <- (Zpower_plus radix2) by omega. + apply integer_representable_2p. omega. +- rewrite e. rewrite <- Zopp_mult_distr_l. apply integer_representable_opp. + rewrite <- (Zpower_plus radix2) by omega. + apply integer_representable_2p. omega. +- apply integer_representable_n2p; omega. +Qed. + +Lemma integer_representable_n: + forall n, -2^prec <= n <= 2^prec -> integer_representable n. +Proof. + red in prec_gt_0_. intros. + replace n with (n * 2^0) by (change (2^0) with 1; ring). + apply integer_representable_n2p_wide. auto. omega. omega. +Qed. + +Lemma round_int_no_overflow: + forall n, + Z.abs n <= 2^emax - 2^(emax-prec) -> + (Rabs (round radix2 fexp (round_mode mode_NE) (Z2R n)) < bpow radix2 emax)%R. +Proof. + intros. red in prec_gt_0_. + rewrite <- round_NE_abs. + apply Rle_lt_trans with (Z2R (2^emax - 2^(emax-prec))). + apply round_le_generic. apply fexp_correct; auto. apply valid_rnd_N. + apply generic_format_FLT. exists (Float radix2 (2^prec-1) (emax-prec)). + rewrite int_upper_bound_eq. unfold F2R; simpl. + split. rewrite <- Z2R_Zpower by omega. rewrite <- Z2R_mult. auto. + split. assert (0 < 2^prec) by (apply (Zpower_gt_0 radix2); omega). zify; omega. + unfold emin; omega. + rewrite <- Z2R_abs. apply Z2R_le. auto. + rewrite <- Z2R_Zpower by omega. apply Z2R_lt. simpl. + assert (0 < 2^(emax-prec)) by (apply (Zpower_gt_0 radix2); omega). + omega. + apply fexp_correct. auto. +Qed. + +(** Conversion from an integer. Round to nearest. *) + +Definition BofZ (n: Z) : binary_float := + binary_normalize prec emax prec_gt_0_ Hmax mode_NE n 0 false. + +Theorem BofZ_correct: + forall n, + if Rlt_bool (Rabs (round radix2 fexp (round_mode mode_NE) (Z2R n))) (bpow radix2 emax) + then + B2R prec emax (BofZ n) = round radix2 fexp (round_mode mode_NE) (Z2R n) /\ + is_finite _ _ (BofZ n) = true /\ + Bsign prec emax (BofZ n) = Zlt_bool n 0 + else + B2FF prec emax (BofZ n) = binary_overflow prec emax mode_NE (Zlt_bool n 0). +Proof. + intros. + generalize (binary_normalize_correct prec emax prec_gt_0_ Hmax mode_NE n 0 false). + fold emin; fold fexp; fold (BofZ n). + replace (F2R {| Fnum := n; Fexp := 0 |}) with (Z2R n). + destruct Rlt_bool. +- intros (A & B & C). split; [|split]. + + auto. + + auto. + + rewrite C. change 0%R with (Z2R 0). rewrite Rcompare_Z2R. + unfold Zlt_bool. auto. +- intros A; rewrite A. f_equal. change 0%R with (Z2R 0). + generalize (Zlt_bool_spec n 0); intros SPEC; inversion SPEC. + apply Rlt_bool_true; apply Z2R_lt; auto. + apply Rlt_bool_false; apply Z2R_le; auto. +- unfold F2R; simpl. ring. +Qed. + +Theorem BofZ_finite: + forall n, + Z.abs n <= 2^emax - 2^(emax-prec) -> + B2R _ _ (BofZ n) = round radix2 fexp (round_mode mode_NE) (Z2R n) + /\ is_finite _ _ (BofZ n) = true + /\ Bsign _ _ (BofZ n) = Zlt_bool n 0%Z. +Proof. + intros. + generalize (BofZ_correct n). rewrite Rlt_bool_true. auto. + apply round_int_no_overflow; auto. +Qed. + +Theorem BofZ_representable: + forall n, + integer_representable n -> + B2R _ _ (BofZ n) = Z2R n + /\ is_finite _ _ (BofZ n) = true + /\ Bsign _ _ (BofZ n) = (n + B2R _ _ (BofZ n) = Z2R n + /\ is_finite _ _ (BofZ n) = true + /\ Bsign _ _ (BofZ n) = Zlt_bool n 0%Z. +Proof. + intros. apply BofZ_representable. apply integer_representable_n; auto. +Qed. + +Lemma BofZ_finite_pos0: + forall n, + Z.abs n <= 2^emax - 2^(emax-prec) -> is_finite_pos0 (BofZ n) = true. +Proof. + intros. + generalize (binary_normalize_correct prec emax prec_gt_0_ Hmax mode_NE n 0 false). + fold emin; fold fexp; fold (BofZ n). + replace (F2R {| Fnum := n; Fexp := 0 |}) with (Z2R n) by + (unfold F2R; simpl; ring). + rewrite Rlt_bool_true by (apply round_int_no_overflow; auto). + intros (A & B & C). + destruct (BofZ n); auto; try discriminate. + simpl in *. rewrite C. change 0%R with (Z2R 0). rewrite Rcompare_Z2R. + generalize (Zcompare_spec n 0); intros SPEC; inversion SPEC; auto. + assert ((round radix2 fexp ZnearestE (Z2R n) <= -1)%R). + { change (-1)%R with (Z2R (-1)). + apply round_le_generic. apply fexp_correct. auto. apply valid_rnd_N. + apply (integer_representable_opp 1). + apply (integer_representable_2p 0). + red in prec_gt_0_; omega. + apply Z2R_le; omega. + } + lra. +Qed. + +Lemma BofZ_finite_equal: + forall x y, + Z.abs x <= 2^emax - 2^(emax-prec) -> + Z.abs y <= 2^emax - 2^(emax-prec) -> + B2R _ _ (BofZ x) = B2R _ _ (BofZ y) -> + BofZ x = BofZ y. +Proof. + intros. apply B2R_inj_pos0; auto; apply BofZ_finite_pos0; auto. +Qed. + +(** Commutation properties with addition, subtraction, multiplication. *) + +Theorem BofZ_plus: + forall nan p q, + integer_representable p -> integer_representable q -> + Bplus _ _ _ Hmax nan mode_NE (BofZ p) (BofZ q) = BofZ (p + q). +Proof. + intros. + destruct (BofZ_representable p) as (A & B & C); auto. + destruct (BofZ_representable q) as (D & E & F); auto. + generalize (Bplus_correct _ _ _ Hmax nan mode_NE (BofZ p) (BofZ q) B E). + fold emin; fold fexp. + rewrite A, D. rewrite <- Z2R_plus. + generalize (BofZ_correct (p + q)). destruct Rlt_bool. +- intros (P & Q & R) (U & V & W). + apply B2R_Bsign_inj; auto. + rewrite P, U; auto. + rewrite R, W, C, F. + change 0%R with (Z2R 0). rewrite Rcompare_Z2R. unfold Zlt_bool at 3. + generalize (Zcompare_spec (p + q) 0); intros SPEC; inversion SPEC; auto. + assert (EITHER: 0 <= p \/ 0 <= q) by omega. + destruct EITHER; [apply andb_false_intro1 | apply andb_false_intro2]; + apply Zlt_bool_false; auto. +- intros P (U & V). + apply B2FF_inj. + rewrite P, U, C. f_equal. rewrite C, F in V. + generalize (Zlt_bool_spec p 0) (Zlt_bool_spec q 0). rewrite <- V. + intros SPEC1 SPEC2; inversion SPEC1; inversion SPEC2; try congruence; symmetry. + apply Zlt_bool_true; omega. + apply Zlt_bool_false; omega. +Qed. + +Theorem BofZ_minus: + forall nan p q, + integer_representable p -> integer_representable q -> + Bminus _ _ _ Hmax nan mode_NE (BofZ p) (BofZ q) = BofZ (p - q). +Proof. + intros. + destruct (BofZ_representable p) as (A & B & C); auto. + destruct (BofZ_representable q) as (D & E & F); auto. + generalize (Bminus_correct _ _ _ Hmax nan mode_NE (BofZ p) (BofZ q) B E). + fold emin; fold fexp. + rewrite A, D. rewrite <- Z2R_minus. + generalize (BofZ_correct (p - q)). destruct Rlt_bool. +- intros (P & Q & R) (U & V & W). + apply B2R_Bsign_inj; auto. + rewrite P, U; auto. + rewrite R, W, C, F. + change 0%R with (Z2R 0). rewrite Rcompare_Z2R. unfold Zlt_bool at 3. + generalize (Zcompare_spec (p - q) 0); intros SPEC; inversion SPEC; auto. + assert (EITHER: 0 <= p \/ q < 0) by omega. + destruct EITHER; [apply andb_false_intro1 | apply andb_false_intro2]. + rewrite Zlt_bool_false; auto. + rewrite Zlt_bool_true; auto. +- intros P (U & V). + apply B2FF_inj. + rewrite P, U, C. f_equal. rewrite C, F in V. + generalize (Zlt_bool_spec p 0) (Zlt_bool_spec q 0). rewrite V. + intros SPEC1 SPEC2; inversion SPEC1; inversion SPEC2; symmetry. + rewrite <- H3 in H1; discriminate. + apply Zlt_bool_true; omega. + apply Zlt_bool_false; omega. + rewrite <- H3 in H1; discriminate. +Qed. + +Theorem BofZ_mult: + forall nan p q, + integer_representable p -> integer_representable q -> + 0 < q -> + Bmult _ _ _ Hmax nan mode_NE (BofZ p) (BofZ q) = BofZ (p * q). +Proof. + intros. + assert (SIGN: xorb (p + 2^prec <= Z.abs x -> + 0 <= p <= emax - 1 -> + Bmult _ _ _ Hmax nan mode_NE (BofZ x) (BofZ (2^p)) = BofZ (x * 2^p). +Proof. + intros. + destruct (Z.eq_dec x 0). +- subst x. apply BofZ_mult. + apply integer_representable_n. + generalize (Zpower_ge_0 radix2 prec). simpl; omega. + apply integer_representable_2p. auto. + apply (Zpower_gt_0 radix2). + omega. +- assert (Z2R x <> 0%R) by (apply (Z2R_neq _ _ n)). + destruct (BofZ_finite x H) as (A & B & C). + destruct (BofZ_representable (2^p)) as (D & E & F). + apply integer_representable_2p. auto. + assert (canonic_exp radix2 fexp (Z2R (x * 2^p)) = + canonic_exp radix2 fexp (Z2R x) + p). + { + unfold canonic_exp, fexp. rewrite Z2R_mult. + change (2^p) with (radix2^p). rewrite Z2R_Zpower by omega. + rewrite ln_beta_mult_bpow by auto. + assert (prec + 1 <= ln_beta radix2 (Z2R x)). + { rewrite <- (ln_beta_abs radix2 (Z2R x)). + rewrite <- (ln_beta_bpow radix2 prec). + apply ln_beta_le. + apply bpow_gt_0. rewrite <- Z2R_Zpower by (red in prec_gt_0_;omega). + rewrite <- Z2R_abs. apply Z2R_le; auto. } + unfold FLT_exp. + unfold emin; red in prec_gt_0_; zify; omega. + } + assert (forall m, round radix2 fexp m (Z2R x) * Z2R (2^p) = + round radix2 fexp m (Z2R (x * 2^p)))%R. + { + intros. unfold round, scaled_mantissa. rewrite H3. + rewrite Z2R_mult. rewrite Z.opp_add_distr. rewrite bpow_plus. + set (a := Z2R x); set (b := bpow radix2 (- canonic_exp radix2 fexp a)). + replace (a * Z2R (2^p) * (b * bpow radix2 (-p)))%R with (a * b)%R. + unfold F2R; simpl. rewrite Rmult_assoc. f_equal. + rewrite bpow_plus. f_equal. apply (Z2R_Zpower radix2). omega. + transitivity ((a * b) * (Z2R (2^p) * bpow radix2 (-p)))%R. + rewrite (Z2R_Zpower radix2). rewrite <- bpow_plus. + replace (p + -p) with 0 by omega. change (bpow radix2 0) with 1%R. ring. + omega. + ring. + } + assert (forall m x, + round radix2 fexp (round_mode m) (round radix2 fexp (round_mode m) x) = + round radix2 fexp (round_mode m) x). + { + intros. apply round_generic. apply valid_rnd_round_mode. + apply generic_format_round. apply fexp_correct; auto. + apply valid_rnd_round_mode. + } + assert (xorb (x 1 -> prec' > 1 -> prec' >= prec + 2 -> emin' <= emin - 2 -> + round radix2 fexp (Znearest choice) (round radix2 (FLT_exp emin' prec') Zrnd_odd x) = + round radix2 fexp (Znearest choice) x. +Proof. + intros. apply round_odd_prop. auto. apply fexp_correct; auto. + apply exists_NE_FLT. right; omega. + apply FLT_exp_valid. red; omega. + apply exists_NE_FLT. right; omega. + unfold fexp, FLT_exp; intros. zify; omega. +Qed. + +Corollary round_odd_fix: + forall x p choice, + prec > 1 -> + 0 <= p -> + (bpow radix2 (prec + p + 1) <= Rabs x)%R -> + round radix2 fexp (Znearest choice) (round radix2 (FIX_exp p) Zrnd_odd x) = + round radix2 fexp (Znearest choice) x. +Proof. + intros. destruct (Req_EM_T x 0%R). +- subst x. rewrite round_0. auto. apply valid_rnd_odd. +- set (prec' := ln_beta radix2 x - p). + set (emin' := emin - 2). + assert (PREC: ln_beta radix2 (bpow radix2 (prec + p + 1)) <= ln_beta radix2 x). + { rewrite <- (ln_beta_abs radix2 x). + apply ln_beta_le; auto. apply bpow_gt_0. } + rewrite ln_beta_bpow in PREC. + assert (CANON: canonic_exp radix2 (FLT_exp emin' prec') x = + canonic_exp radix2 (FIX_exp p) x). + { + unfold canonic_exp, FLT_exp, FIX_exp. + replace (ln_beta radix2 x - prec') with p by (unfold prec'; omega). + apply Z.max_l. unfold emin', emin. red in prec_gt_0_; omega. + } + assert (RND: round radix2 (FIX_exp p) Zrnd_odd x = + round radix2 (FLT_exp emin' prec') Zrnd_odd x). + { + unfold round, scaled_mantissa. rewrite CANON. auto. + } + rewrite RND. + apply round_odd_flt. auto. + unfold prec'. red in prec_gt_0_; omega. + unfold prec'. omega. + unfold emin'. omega. +Qed. + +Definition int_round_odd (x: Z) (p: Z) := + (if Z.eqb (x mod 2^p) 0 || Z.odd (x / 2^p) then x / 2^p else x / 2^p + 1) * 2^p. + +Lemma Zrnd_odd_int: + forall n p, 0 <= p -> + Zrnd_odd (Z2R n * bpow radix2 (-p)) * 2^p = + int_round_odd n p. +Proof. + intros. + assert (0 < 2^p) by (apply (Zpower_gt_0 radix2); omega). + assert (n = (n / 2^p) * 2^p + n mod 2^p) by (rewrite Zmult_comm; apply Z.div_mod; omega). + assert (0 <= n mod 2^p < 2^p) by (apply Z_mod_lt; omega). + unfold int_round_odd. set (q := n / 2^p) in *; set (r := n mod 2^p) in *. + f_equal. + pose proof (bpow_gt_0 radix2 (-p)). + assert (bpow radix2 p * bpow radix2 (-p) = 1)%R. + { rewrite <- bpow_plus. replace (p + -p) with 0 by omega. auto. } + assert (Z2R n * bpow radix2 (-p) = Z2R q + Z2R r * bpow radix2 (-p))%R. + { rewrite H1. rewrite Z2R_plus, Z2R_mult. + change (Z2R (2^p)) with (Z2R (radix2^p)). + rewrite Z2R_Zpower by omega. ring_simplify. + rewrite Rmult_assoc. rewrite H4. ring. } + assert (0 <= Z2R r < bpow radix2 p)%R. + { split. change 0%R with (Z2R 0). apply Z2R_le; omega. + rewrite <- Z2R_Zpower by omega. apply Z2R_lt; tauto. } + assert (0 <= Z2R r * bpow radix2 (-p) < 1)%R. + { generalize (bpow_gt_0 radix2 (-p)). intros. + split. apply Rmult_le_pos; lra. + rewrite <- H4. apply Rmult_lt_compat_r. auto. tauto. } + assert (Zfloor (Z2R n * bpow radix2 (-p)) = q). + { apply Zfloor_imp. rewrite H5. rewrite Z2R_plus. change (Z2R 1) with 1%R. lra. } + unfold Zrnd_odd. destruct Req_EM_T. +- assert (Z2R r * bpow radix2 (-p) = 0)%R. + { rewrite H8 in e. rewrite e in H5. lra. } + apply Rmult_integral in H9. destruct H9; [ | lra ]. + apply (eq_Z2R r 0) in H9. apply <- Z.eqb_eq in H9. rewrite H9. assumption. +- assert (Z2R r * bpow radix2 (-p) <> 0)%R. + { rewrite H8 in n0. lra. } + destruct (Z.eqb r 0) eqn:RZ. + apply Z.eqb_eq in RZ. rewrite RZ in H9. change (Z2R 0) with 0%R in H9. + rewrite Rmult_0_l in H9. congruence. + rewrite Zceil_floor_neq by lra. rewrite H8. + change Zeven with Z.even. rewrite Zodd_even_bool. destruct (Z.even q); auto. +Qed. + +Lemma int_round_odd_le: + forall p x y, 0 <= p -> + x <= y -> int_round_odd x p <= int_round_odd y p. +Proof. + intros. + assert (Zrnd_odd (Z2R x * bpow radix2 (-p)) <= Zrnd_odd (Z2R y * bpow radix2 (-p))). + { apply Zrnd_le. apply valid_rnd_odd. apply Rmult_le_compat_r. apply bpow_ge_0. + apply Z2R_le; auto. } + rewrite <- ! Zrnd_odd_int by auto. + apply Zmult_le_compat_r. auto. apply (Zpower_ge_0 radix2). +Qed. + +Lemma int_round_odd_exact: + forall p x, 0 <= p -> + (2^p | x) -> int_round_odd x p = x. +Proof. + intros. unfold int_round_odd. apply Znumtheory.Zdivide_mod in H0. + rewrite H0. simpl. rewrite Zmult_comm. symmetry. apply Z_div_exact_2. + apply Zlt_gt. apply (Zpower_gt_0 radix2). auto. auto. +Qed. + +Theorem BofZ_round_odd: + forall x p, + prec > 1 -> + Z.abs x <= 2^emax - 2^(emax-prec) -> + 0 <= p <= emax - prec -> + 2^(prec + p + 1) <= Z.abs x -> + BofZ x = BofZ (int_round_odd x p). +Proof. + intros x p PREC XRANGE PRANGE XGE. + assert (DIV: (2^p | 2^emax - 2^(emax - prec))). + { rewrite int_upper_bound_eq. apply Z.divide_mul_r. + exists (2^(emax - prec - p)). red in prec_gt_0_. + rewrite <- (Zpower_plus radix2) by omega. f_equal; omega. } + assert (YRANGE: Z.abs (int_round_odd x p) <= 2^emax - 2^(emax-prec)). + { apply Z.abs_le. split. + replace (-(2^emax - 2^(emax-prec))) with (int_round_odd (-(2^emax - 2^(emax-prec))) p). + apply int_round_odd_le; zify; omega. + apply int_round_odd_exact. omega. apply Z.divide_opp_r. auto. + replace (2^emax - 2^(emax-prec)) with (int_round_odd (2^emax - 2^(emax-prec)) p). + apply int_round_odd_le; zify; omega. + apply int_round_odd_exact. omega. auto. } + destruct (BofZ_finite x XRANGE) as (X1 & X2 & X3). + destruct (BofZ_finite (int_round_odd x p) YRANGE) as (Y1 & Y2 & Y3). + apply BofZ_finite_equal; auto. + rewrite X1, Y1. + assert (Z2R (int_round_odd x p) = round radix2 (FIX_exp p) Zrnd_odd (Z2R x)). + { + unfold round, scaled_mantissa, canonic_exp, FIX_exp. + rewrite <- Zrnd_odd_int by omega. + unfold F2R; simpl. rewrite Z2R_mult. f_equal. apply (Z2R_Zpower radix2). omega. + } + rewrite H. symmetry. apply round_odd_fix. auto. omega. + rewrite <- Z2R_Zpower. rewrite <- Z2R_abs. apply Z2R_le; auto. + red in prec_gt_0_; omega. +Qed. + +Lemma int_round_odd_shifts: + forall x p, 0 <= p -> + int_round_odd x p = + Z.shiftl (if Z.eqb (x mod 2^p) 0 then Z.shiftr x p else Z.lor (Z.shiftr x p) 1) p. +Proof. + intros. + unfold int_round_odd. rewrite Z.shiftl_mul_pow2 by auto. f_equal. + rewrite Z.shiftr_div_pow2 by auto. + destruct (x mod 2^p =? 0) eqn:E. auto. + assert (forall n, (if Z.odd n then n else n + 1) = Z.lor n 1). + { destruct n; simpl; auto. + destruct p0; auto. + destruct p0; auto. induction p0; auto. } + simpl. apply H0. +Qed. + +Lemma int_round_odd_bits: + forall x y p, 0 <= p -> + (forall i, 0 <= i < p -> Z.testbit y i = false) -> + Z.testbit y p = (if Z.eqb (x mod 2^p) 0 then Z.testbit x p else true) -> + (forall i, p < i -> Z.testbit y i = Z.testbit x i) -> + int_round_odd x p = y. +Proof. + intros until p; intros PPOS BELOW AT ABOVE. + rewrite int_round_odd_shifts by auto. + apply Z.bits_inj'. intros. + generalize (Zcompare_spec n p); intros SPEC; inversion SPEC. +- rewrite BELOW by auto. apply Z.shiftl_spec_low; auto. +- subst n. rewrite AT. rewrite Z.shiftl_spec_high by omega. + replace (p - p) with 0 by omega. + destruct (x mod 2^p =? 0). + + rewrite Z.shiftr_spec by omega. f_equal; omega. + + rewrite Z.lor_spec. apply orb_true_r. +- rewrite ABOVE by auto. rewrite Z.shiftl_spec_high by omega. + destruct (x mod 2^p =? 0). + rewrite Z.shiftr_spec by omega. f_equal; omega. + rewrite Z.lor_spec, Z.shiftr_spec by omega. + change 1 with (Z.ones 1). rewrite Z.ones_spec_high by omega. rewrite orb_false_r. + f_equal; omega. +Qed. + +(** ** Conversion from a FP number to an integer *) + +(** Always rounds toward zero. *) + +Definition ZofB (f: binary_float): option Z := + match f with + | B754_finite s m (Zpos e) _ => Some (cond_Zopp s (Zpos m) * Zpower_pos radix2 e)%Z + | B754_finite s m 0 _ => Some (cond_Zopp s (Zpos m)) + | B754_finite s m (Zneg e) _ => Some (cond_Zopp s (Zpos m / Zpower_pos radix2 e))%Z + | B754_zero _ => Some 0%Z + | _ => None + end. + +Theorem ZofB_correct: + forall f, + ZofB f = if is_finite _ _ f then Some (Ztrunc (B2R _ _ f)) else None. +Proof. + destruct f; simpl; auto. +- f_equal. symmetry. apply (Ztrunc_Z2R 0). +- destruct e; f_equal. + + unfold F2R; simpl. rewrite Rmult_1_r. rewrite Ztrunc_Z2R. auto. + + unfold F2R; simpl. rewrite <- Z2R_mult. rewrite Ztrunc_Z2R. auto. + + unfold F2R; simpl. rewrite Z2R_cond_Zopp. rewrite <- cond_Ropp_mult_l. + assert (EQ: forall x, Ztrunc (cond_Ropp b x) = cond_Zopp b (Ztrunc x)). + { + intros. destruct b; simpl; auto. apply Ztrunc_opp. + } + rewrite EQ. f_equal. + generalize (Zpower_pos_gt_0 2 p (refl_equal _)); intros. + rewrite Ztrunc_floor. symmetry. apply Zfloor_div. omega. + apply Rmult_le_pos. apply (Z2R_le 0). compute; congruence. + apply Rlt_le. apply Rinv_0_lt_compat. apply (Z2R_lt 0). auto. +Qed. + +(** Interval properties. *) + +Remark Ztrunc_range_pos: + forall x, 0 < Ztrunc x -> (Z2R (Ztrunc x) <= x < Z2R (Ztrunc x + 1)%Z)%R. +Proof. + intros. + rewrite Ztrunc_floor. split. apply Zfloor_lb. rewrite Z2R_plus. apply Zfloor_ub. + generalize (Rle_bool_spec 0%R x). intros RLE; inversion RLE; subst; clear RLE. + auto. + rewrite Ztrunc_ceil in H by lra. unfold Zceil in H. + assert (-x < 0)%R. + { apply Rlt_le_trans with (Z2R (Zfloor (-x)) + 1)%R. apply Zfloor_ub. + change 0%R with (Z2R 0). change 1%R with (Z2R 1). rewrite <- Z2R_plus. + apply Z2R_le. omega. } + lra. +Qed. + +Remark Ztrunc_range_zero: + forall x, Ztrunc x = 0 -> (-1 < x < 1)%R. +Proof. + intros; generalize (Rle_bool_spec 0%R x). intros RLE; inversion RLE; subst; clear RLE. +- rewrite Ztrunc_floor in H by auto. split. + + apply Rlt_le_trans with 0%R; auto. rewrite <- Ropp_0. apply Ropp_lt_contravar. apply Rlt_0_1. + + replace 1%R with (Z2R (Zfloor x) + 1)%R. apply Zfloor_ub. rewrite H. simpl. apply Rplus_0_l. +- rewrite Ztrunc_ceil in H by (apply Rlt_le; auto). split. + + apply Ropp_lt_cancel. rewrite Ropp_involutive. + replace 1%R with (Z2R (Zfloor (-x)) + 1)%R. apply Zfloor_ub. + unfold Zceil in H. replace (Zfloor (-x)) with 0 by omega. simpl. apply Rplus_0_l. + + apply Rlt_le_trans with 0%R; auto. apply Rle_0_1. +Qed. + +Theorem ZofB_range_pos: + forall f n, ZofB f = Some n -> 0 < n -> (Z2R n <= B2R _ _ f < Z2R (n + 1)%Z)%R. +Proof. + intros. rewrite ZofB_correct in H. destruct (is_finite prec emax f) eqn:FIN; inversion H. + apply Ztrunc_range_pos. congruence. +Qed. + +Theorem ZofB_range_neg: + forall f n, ZofB f = Some n -> n < 0 -> (Z2R (n - 1)%Z < B2R _ _ f <= Z2R n)%R. +Proof. + intros. rewrite ZofB_correct in H. destruct (is_finite prec emax f) eqn:FIN; inversion H. + set (x := B2R prec emax f) in *. set (y := (-x)%R). + assert (A: (Z2R (Ztrunc y) <= y < Z2R (Ztrunc y + 1)%Z)%R). + { apply Ztrunc_range_pos. unfold y. rewrite Ztrunc_opp. omega. } + destruct A as [B C]. + unfold y in B, C. rewrite Ztrunc_opp in B, C. + replace (- Ztrunc x + 1) with (- (Ztrunc x - 1)) in C by omega. + rewrite Z2R_opp in B, C. lra. +Qed. + +Theorem ZofB_range_zero: + forall f, ZofB f = Some 0 -> (-1 < B2R _ _ f < 1)%R. +Proof. + intros. rewrite ZofB_correct in H. destruct (is_finite prec emax f) eqn:FIN; inversion H. + apply Ztrunc_range_zero. auto. +Qed. + +Theorem ZofB_range_nonneg: + forall f n, ZofB f = Some n -> 0 <= n -> (-1 < B2R _ _ f < Z2R (n + 1)%Z)%R. +Proof. + intros. destruct (Z.eq_dec n 0). +- subst n. apply ZofB_range_zero. auto. +- destruct (ZofB_range_pos f n) as (A & B). auto. omega. + split; auto. apply Rlt_le_trans with (Z2R 0). simpl; lra. + apply Rle_trans with (Z2R n); auto. apply Z2R_le; auto. +Qed. + +(** For representable integers, [ZofB] is left inverse of [BofZ]. *) + +Theorem ZofBofZ_exact: + forall n, integer_representable n -> ZofB (BofZ n) = Some n. +Proof. + intros. destruct (BofZ_representable n H) as (A & B & C). + rewrite ZofB_correct. rewrite A, B. f_equal. apply Ztrunc_Z2R. +Qed. + +(** Compatibility with subtraction *) + +Remark Zfloor_minus: + forall x n, Zfloor (x - Z2R n) = Zfloor x - n. +Proof. + intros. apply Zfloor_imp. replace (Zfloor x - n + 1) with ((Zfloor x + 1) - n) by omega. + rewrite ! Z2R_minus. unfold Rminus. split. + apply Rplus_le_compat_r. apply Zfloor_lb. + apply Rplus_lt_compat_r. rewrite Z2R_plus. apply Zfloor_ub. +Qed. + +Theorem ZofB_minus: + forall minus_nan m f p q, + ZofB f = Some p -> 0 <= p < 2*q -> q <= 2^prec -> (Z2R q <= B2R _ _ f)%R -> + ZofB (Bminus _ _ _ Hmax minus_nan m f (BofZ q)) = Some (p - q). +Proof. + intros. + assert (Q: -2^prec <= q <= 2^prec). + { split; auto. generalize (Zpower_ge_0 radix2 prec); simpl; omega. } + assert (RANGE: (-1 < B2R _ _ f < Z2R (p + 1)%Z)%R) by (apply ZofB_range_nonneg; auto; omega). + rewrite ZofB_correct in H. destruct (is_finite prec emax f) eqn:FIN; try discriminate. + assert (PQ2: (Z2R (p + 1) <= Z2R q * 2)%R). + { change 2%R with (Z2R 2). rewrite <- Z2R_mult. apply Z2R_le. omega. } + assert (EXACT: round radix2 fexp (round_mode m) (B2R _ _ f - Z2R q)%R = (B2R _ _ f - Z2R q)%R). + { apply round_generic. apply valid_rnd_round_mode. + apply sterbenz_aux. apply FLT_exp_monotone. apply generic_format_B2R. + apply integer_representable_n. auto. lra. } + destruct (BofZ_exact q Q) as (A & B & C). + generalize (Bminus_correct _ _ _ Hmax minus_nan m f (BofZ q) FIN B). + rewrite Rlt_bool_true. +- fold emin; fold fexp. intros (D & E & F). + rewrite ZofB_correct. rewrite E. rewrite D. rewrite A. rewrite EXACT. + inversion H. f_equal. rewrite ! Ztrunc_floor. apply Zfloor_minus. + lra. lra. +- rewrite A. fold emin; fold fexp. rewrite EXACT. + apply Rle_lt_trans with (bpow radix2 prec). + apply Rle_trans with (Z2R q). apply Rabs_le. lra. + rewrite <- Z2R_Zpower. apply Z2R_le; auto. red in prec_gt_0_; omega. + apply bpow_lt. auto. +Qed. + +(** A variant of [ZofB] that bounds the range of representable integers. *) + +Definition ZofB_range (f: binary_float) (zmin zmax: Z): option Z := + match ZofB f with + | None => None + | Some z => if Zle_bool zmin z && Zle_bool z zmax then Some z else None + end. + +Theorem ZofB_range_correct: + forall f min max, + let n := Ztrunc (B2R _ _ f) in + ZofB_range f min max = + if is_finite _ _ f && Zle_bool min n && Zle_bool n max then Some n else None. +Proof. + intros. unfold ZofB_range. rewrite ZofB_correct. fold n. + destruct (is_finite prec emax f); auto. +Qed. + +Lemma ZofB_range_inversion: + forall f min max n, + ZofB_range f min max = Some n -> + min <= n /\ n <= max /\ ZofB f = Some n. +Proof. + intros. rewrite ZofB_range_correct in H. rewrite ZofB_correct. + destruct (is_finite prec emax f); try discriminate. + set (n1 := Ztrunc (B2R _ _ f)) in *. + destruct (min <=? n1) eqn:MIN; try discriminate. + destruct (n1 <=? max) eqn:MAX; try discriminate. + simpl in H. inversion H. subst n. + split. apply Zle_bool_imp_le; auto. + split. apply Zle_bool_imp_le; auto. + auto. +Qed. + +Theorem ZofB_range_minus: + forall minus_nan m f p q, + ZofB_range f 0 (2 * q - 1) = Some p -> q <= 2^prec -> (Z2R q <= B2R _ _ f)%R -> + ZofB_range (Bminus _ _ _ Hmax minus_nan m f (BofZ q)) (-q) (q - 1) = Some (p - q). +Proof. + intros. destruct (ZofB_range_inversion _ _ _ _ H) as (A & B & C). + set (f' := Bminus prec emax prec_gt_0_ Hmax minus_nan m f (BofZ q)). + assert (D: ZofB f' = Some (p - q)). + { apply ZofB_minus. auto. omega. auto. auto. } + unfold ZofB_range. rewrite D. rewrite Zle_bool_true by omega. rewrite Zle_bool_true by omega. auto. +Qed. + +(** ** Algebraic identities *) + +(** Commutativity of addition and multiplication *) + +Theorem Bplus_commut: + forall plus_nan mode (x y: binary_float), + plus_nan x y = plus_nan y x -> + Bplus _ _ _ Hmax plus_nan mode x y = Bplus _ _ _ Hmax plus_nan mode y x. +Proof. + intros until y; intros NAN. + pose proof (Bplus_correct _ _ _ Hmax plus_nan mode x y). + pose proof (Bplus_correct _ _ _ Hmax plus_nan mode y x). + unfold Bplus in *; destruct x; destruct y; auto. +- rewrite (eqb_sym b0 b). destruct (eqb b b0) eqn:EQB; auto. + f_equal; apply eqb_prop; auto. +- rewrite NAN; auto. +- rewrite (eqb_sym b0 b). destruct (eqb b b0) eqn:EQB. + f_equal; apply eqb_prop; auto. + rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- generalize (H (refl_equal _) (refl_equal _)); clear H. + generalize (H0 (refl_equal _) (refl_equal _)); clear H0. + fold emin. fold fexp. + set (x := B754_finite prec emax b0 m0 e1 e2). set (rx := B2R _ _ x). + set (y := B754_finite prec emax b m e e0). set (ry := B2R _ _ y). + rewrite (Rplus_comm ry rx). destruct Rlt_bool. + + intros (A1 & A2 & A3) (B1 & B2 & B3). + apply B2R_Bsign_inj; auto. rewrite <- B1 in A1. auto. + rewrite Z.add_comm. rewrite Z.min_comm. auto. + + intros (A1 & A2) (B1 & B2). apply B2FF_inj. rewrite B2 in B1. rewrite <- B1 in A1. auto. +Qed. + +Theorem Bmult_commut: + forall mult_nan mode (x y: binary_float), + mult_nan x y = mult_nan y x -> + Bmult _ _ _ Hmax mult_nan mode x y = Bmult _ _ _ Hmax mult_nan mode y x. +Proof. + intros until y; intros NAN. + pose proof (Bmult_correct _ _ _ Hmax mult_nan mode x y). + pose proof (Bmult_correct _ _ _ Hmax mult_nan mode y x). + unfold Bmult in *; destruct x; destruct y; auto. +- rewrite (xorb_comm b0 b); auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite (xorb_comm b0 b); auto. +- rewrite NAN; auto. +- rewrite (xorb_comm b0 b); auto. +- rewrite NAN; auto. +- rewrite (xorb_comm b0 b); auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite NAN; auto. +- rewrite (xorb_comm b0 b); auto. +- rewrite (xorb_comm b0 b); auto. +- rewrite NAN; auto. +- revert H H0. fold emin. fold fexp. + set (x := B754_finite prec emax b0 m0 e1 e2). set (rx := B2R _ _ x). + set (y := B754_finite prec emax b m e e0). set (ry := B2R _ _ y). + rewrite (Rmult_comm ry rx). destruct Rlt_bool. + + intros (A1 & A2 & A3) (B1 & B2 & B3). + apply B2R_Bsign_inj; auto. rewrite <- B1 in A1. auto. + rewrite ! Bsign_FF2B. f_equal. f_equal. apply xorb_comm. apply Pos.mul_comm. apply Z.add_comm. + + intros A B. apply B2FF_inj. etransitivity. eapply A. rewrite xorb_comm. auto. +Qed. + +(** Multiplication by 2 is diagonal addition. *) + +Theorem Bmult2_Bplus: + forall plus_nan mult_nan mode (f: binary_float), + (forall (x y: binary_float), + is_nan _ _ x = true -> is_finite _ _ y = true -> plus_nan x x = mult_nan x y) -> + Bplus _ _ _ Hmax plus_nan mode f f = Bmult _ _ _ Hmax mult_nan mode f (BofZ 2%Z). +Proof. + intros until f; intros NAN. + destruct (BofZ_representable 2) as (A & B & C). + apply (integer_representable_2p 1). red in prec_gt_0_; omega. + pose proof (Bmult_correct _ _ _ Hmax mult_nan mode f (BofZ 2%Z)). fold emin in H. + rewrite A, B, C in H. rewrite xorb_false_r in H. + destruct (is_finite _ _ f) eqn:FIN. +- pose proof (Bplus_correct _ _ _ Hmax plus_nan mode f f FIN FIN). fold emin in H0. + assert (EQ: (B2R prec emax f * Z2R 2%Z = B2R prec emax f + B2R prec emax f)%R). + { change (Z2R 2%Z) with 2%R. ring. } + rewrite <- EQ in H0. destruct Rlt_bool. + + destruct H0 as (P & Q & R). destruct H as (S & T & U). + apply B2R_Bsign_inj; auto. + rewrite P, S. auto. + rewrite R, U. + replace 0%R with (0 * Z2R 2%Z)%R by ring. rewrite Rcompare_mult_r. + rewrite andb_diag, orb_diag. destruct f; try discriminate; simpl. + rewrite Rcompare_Eq by auto. destruct mode; auto. + replace 0%R with (@F2R radix2 {| Fnum := 0%Z; Fexp := e |}). + rewrite Rcompare_F2R. destruct b; auto. + unfold F2R. simpl. ring. + change 0%R with (Z2R 0%Z). apply Z2R_lt. omega. + destruct (Bmult prec emax prec_gt_0_ Hmax mult_nan mode f (BofZ 2)); reflexivity || discriminate. + + destruct H0 as (P & Q). apply B2FF_inj. rewrite P, H. auto. +- destruct f; try discriminate. + + simpl Bplus. rewrite eqb_true. destruct (BofZ 2) eqn:B2; try discriminate; simpl in *. + assert ((0 = 2)%Z) by (apply eq_Z2R; auto). discriminate. + subst b0. rewrite xorb_false_r. auto. + auto. + + unfold Bplus, Bmult. rewrite <- NAN by auto. auto. +Qed. + +(** Divisions that can be turned into multiplications by an inverse *) + +Definition Bexact_inverse_mantissa := Z.iter (prec - 1) xO xH. + +Remark Bexact_inverse_mantissa_value: + Zpos Bexact_inverse_mantissa = 2 ^ (prec - 1). +Proof. + assert (REC: forall n, Z.pos (nat_iter n xO xH) = 2 ^ (Z.of_nat n)). + { induction n. reflexivity. + simpl nat_iter. transitivity (2 * Z.pos (nat_iter n xO xH)). reflexivity. + rewrite inj_S. rewrite IHn. unfold Z.succ. rewrite Zpower_plus by omega. + change (2 ^ 1) with 2. ring. } + red in prec_gt_0_. + unfold Bexact_inverse_mantissa. rewrite iter_nat_of_Z by omega. rewrite REC. + rewrite Zabs2Nat.id_abs. rewrite Z.abs_eq by omega. auto. +Qed. + +Remark Bexact_inverse_mantissa_digits2_Pnat: + digits2_Pnat Bexact_inverse_mantissa = Z.to_nat (prec - 1). +Proof. + assert (DIGITS: forall n, digits2_Pnat (nat_iter n xO xH) = n). + { induction n; simpl. auto. congruence. } + red in prec_gt_0_. + unfold Bexact_inverse_mantissa. rewrite iter_nat_of_Z by omega. rewrite DIGITS. + apply Zabs2Nat.abs_nat_nonneg. omega. +Qed. + +Remark bounded_Bexact_inverse: + forall e, + emin <= e <= emax - prec <-> bounded prec emax Bexact_inverse_mantissa e = true. +Proof. + intros. unfold bounded, canonic_mantissa. rewrite andb_true_iff. + rewrite <- Zeq_is_eq_bool. rewrite <- Zle_is_le_bool. + rewrite Bexact_inverse_mantissa_digits2_Pnat. + rewrite inj_S. red in prec_gt_0_. rewrite Z2Nat.id by omega. + split. +- intros; split. unfold FLT_exp. unfold emin in H. zify; omega. omega. +- intros [A B]. unfold FLT_exp in A. unfold emin. zify; omega. +Qed. + +Program Definition Bexact_inverse (f: binary_float) : option binary_float := + match f with + | B754_finite s m e B => + if positive_eq_dec m Bexact_inverse_mantissa then + let e' := -e - (prec - 1) * 2 in + if Z_le_dec emin e' then + if Z_le_dec e' emax then + Some(B754_finite _ _ s m e' _) + else None else None else None + | _ => None + end. +Next Obligation. + rewrite <- bounded_Bexact_inverse in B. rewrite <- bounded_Bexact_inverse. + unfold emin in *. omega. +Qed. + +Lemma Bexact_inverse_correct: + forall f f', Bexact_inverse f = Some f' -> + is_finite_strict _ _ f = true + /\ is_finite_strict _ _ f' = true + /\ B2R _ _ f' = (/ B2R _ _ f)%R + /\ B2R _ _ f <> 0%R + /\ Bsign _ _ f' = Bsign _ _ f. +Proof with (try discriminate). + intros f f' EI. unfold Bexact_inverse in EI. destruct f... + destruct (Pos.eq_dec m Bexact_inverse_mantissa)... + set (e' := -e - (prec - 1) * 2) in *. + destruct (Z_le_dec emin e')... + destruct (Z_le_dec e' emax)... + inversion EI; clear EI; subst f' m. + split. auto. split. auto. split. unfold B2R. rewrite Bexact_inverse_mantissa_value. + unfold F2R; simpl. rewrite Z2R_cond_Zopp. + rewrite <- ! cond_Ropp_mult_l. + red in prec_gt_0_. + replace (Z2R (2 ^ (prec - 1))) with (bpow radix2 (prec - 1)) + by (symmetry; apply (Z2R_Zpower radix2); omega). + rewrite <- ! bpow_plus. + replace (prec - 1 + e') with (- (prec - 1 + e)) by (unfold e'; omega). + rewrite bpow_opp. unfold cond_Ropp; destruct b; auto. + rewrite Ropp_inv_permute. auto. apply Rgt_not_eq. apply bpow_gt_0. + split. simpl. red; intros. apply F2R_eq_0_reg in H. destruct b; simpl in H; discriminate. + auto. +Qed. + +Theorem Bdiv_mult_inverse: + forall div_nan mult_nan mode x y z, + (forall (x y z: binary_float), + is_nan _ _ x = true -> is_finite _ _ y = true -> is_finite _ _ z = true -> + div_nan x y = mult_nan x z) -> + Bexact_inverse y = Some z -> + Bdiv _ _ _ Hmax div_nan mode x y = Bmult _ _ _ Hmax mult_nan mode x z. +Proof. + intros until z; intros NAN; intros. destruct (Bexact_inverse_correct _ _ H) as (A & B & C & D & E). + pose proof (Bmult_correct _ _ _ Hmax mult_nan mode x z). + fold emin in H0. fold fexp in H0. + pose proof (Bdiv_correct _ _ _ Hmax div_nan mode x y D). + fold emin in H1. fold fexp in H1. + unfold Rdiv in H1. rewrite <- C in H1. + destruct (is_finite _ _ x) eqn:FINX. +- destruct Rlt_bool. + + destruct H0 as (P & Q & R). destruct H1 as (S & T & U). + apply B2R_Bsign_inj; auto. + rewrite Q. simpl. apply is_finite_strict_finite; auto. + rewrite P, S. auto. + rewrite R, U, E. auto. + apply is_finite_not_is_nan; auto. + apply is_finite_not_is_nan. rewrite Q. simpl. apply is_finite_strict_finite; auto. + apply B2FF_inj. rewrite H0, H1. rewrite E. auto. +- destruct y; try discriminate. destruct z; try discriminate. + destruct x; try discriminate; simpl. + + simpl in E; congruence. + + erewrite NAN; eauto. +Qed. + +End Extra_ops. + +(** ** Conversions between two FP formats *) + +Section Conversions. + +Variable prec1 emax1 prec2 emax2 : Z. +Context (prec1_gt_0_ : Prec_gt_0 prec1) (prec2_gt_0_ : Prec_gt_0 prec2). +Let emin1 := (3 - emax1 - prec1)%Z. +Let fexp1 := FLT_exp emin1 prec1. +Let emin2 := (3 - emax2 - prec2)%Z. +Let fexp2 := FLT_exp emin2 prec2. +Hypothesis Hmax1 : (prec1 < emax1)%Z. +Hypothesis Hmax2 : (prec2 < emax2)%Z. +Let binary_float1 := binary_float prec1 emax1. +Let binary_float2 := binary_float prec2 emax2. + +Definition Bconv (conv_nan: bool -> nan_pl prec1 -> bool * nan_pl prec2) (md: mode) (f: binary_float1) : binary_float2 := + match f with + | B754_nan s pl => let '(s, pl) := conv_nan s pl in B754_nan _ _ s pl + | B754_infinity s => B754_infinity _ _ s + | B754_zero s => B754_zero _ _ s + | B754_finite s m e _ => binary_normalize _ _ _ Hmax2 md (cond_Zopp s (Zpos m)) e s + end. + +Theorem Bconv_correct: + forall conv_nan m f, + is_finite _ _ f = true -> + if Rlt_bool (Rabs (round radix2 fexp2 (round_mode m) (B2R _ _ f))) (bpow radix2 emax2) + then + B2R _ _ (Bconv conv_nan m f) = round radix2 fexp2 (round_mode m) (B2R _ _ f) + /\ is_finite _ _ (Bconv conv_nan m f) = true + /\ Bsign _ _ (Bconv conv_nan m f) = Bsign _ _ f + else + B2FF _ _ (Bconv conv_nan m f) = binary_overflow prec2 emax2 m (Bsign _ _ f). +Proof. + intros. destruct f; try discriminate. +- simpl. rewrite round_0. rewrite Rabs_R0. rewrite Rlt_bool_true. auto. + apply bpow_gt_0. apply valid_rnd_round_mode. +- generalize (binary_normalize_correct _ _ _ Hmax2 m (cond_Zopp b (Zpos m0)) e b). + fold emin2; fold fexp2. simpl. destruct Rlt_bool. + + intros (A & B & C). split. auto. split. auto. rewrite C. + destruct b; simpl. + rewrite Rcompare_Lt. auto. apply F2R_lt_0_compat. simpl. compute; auto. + rewrite Rcompare_Gt. auto. apply F2R_gt_0_compat. simpl. compute; auto. + + intros A. rewrite A. f_equal. destruct b. + apply Rlt_bool_true. apply F2R_lt_0_compat. simpl. compute; auto. + apply Rlt_bool_false. apply Rlt_le. apply Rgt_lt. apply F2R_gt_0_compat. simpl. compute; auto. +Qed. + +(** Converting a finite FP number to higher or equal precision preserves its value. *) + +Theorem Bconv_widen_exact: + (prec2 >= prec1)%Z -> (emax2 >= emax1)%Z -> + forall conv_nan m f, + is_finite _ _ f = true -> + B2R _ _ (Bconv conv_nan m f) = B2R _ _ f + /\ is_finite _ _ (Bconv conv_nan m f) = true + /\ Bsign _ _ (Bconv conv_nan m f) = Bsign _ _ f. +Proof. + intros PREC EMAX; intros. generalize (Bconv_correct conv_nan m f H). + assert (LT: (Rabs (B2R _ _ f) < bpow radix2 emax2)%R). + { + destruct f; try discriminate; simpl. + rewrite Rabs_R0. apply bpow_gt_0. + apply Rlt_le_trans with (bpow radix2 emax1). + rewrite F2R_cond_Zopp. rewrite abs_cond_Ropp. rewrite <- F2R_Zabs. simpl Z.abs. + eapply bounded_lt_emax; eauto. + apply bpow_le. omega. + } + assert (EQ: round radix2 fexp2 (round_mode m) (B2R prec1 emax1 f) = B2R prec1 emax1 f). + { + apply round_generic. apply valid_rnd_round_mode. eapply generic_inclusion_le. + 5: apply generic_format_B2R. apply fexp_correct; auto. apply fexp_correct; auto. + instantiate (1 := emax2). intros. unfold fexp2, FLT_exp. unfold emin2. zify; omega. + apply Rlt_le; auto. + } + rewrite EQ. rewrite Rlt_bool_true by auto. auto. +Qed. + +(** Conversion from integers and change of format *) + +Theorem Bconv_BofZ: + forall conv_nan n, + integer_representable prec1 emax1 n -> + Bconv conv_nan mode_NE (BofZ prec1 emax1 _ Hmax1 n) = BofZ prec2 emax2 _ Hmax2 n. +Proof. + intros. + destruct (BofZ_representable _ _ _ Hmax1 n H) as (A & B & C). + set (f := BofZ prec1 emax1 prec1_gt_0_ Hmax1 n) in *. + generalize (Bconv_correct conv_nan mode_NE f B). + unfold BofZ. + generalize (binary_normalize_correct _ _ _ Hmax2 mode_NE n 0 false). + fold emin2; fold fexp2. rewrite A. + replace (F2R {| Fnum := n; Fexp := 0 |}) with (Z2R n). + destruct Rlt_bool. +- intros (P & Q & R) (D & E & F). apply B2R_Bsign_inj; auto. + congruence. rewrite F, C, R. change 0%R with (Z2R 0). rewrite Rcompare_Z2R. + unfold Zlt_bool. auto. +- intros P Q. apply B2FF_inj. rewrite P, Q. rewrite C. f_equal. change 0%R with (Z2R 0). + generalize (Zlt_bool_spec n 0); intros LT; inversion LT. + rewrite Rlt_bool_true; auto. apply Z2R_lt; auto. + rewrite Rlt_bool_false; auto. apply Z2R_le; auto. +- unfold F2R; simpl. rewrite Rmult_1_r. auto. +Qed. + +(** Change of format (to higher precision) and conversion to integer. *) + +Theorem ZofB_Bconv: + prec2 >= prec1 -> emax2 >= emax1 -> + forall conv_nan m f n, + ZofB _ _ f = Some n -> ZofB _ _ (Bconv conv_nan m f) = Some n. +Proof. + intros. rewrite ZofB_correct in H1. destruct (is_finite _ _ f) eqn:FIN; inversion H1. + destruct (Bconv_widen_exact H H0 conv_nan m f) as (A & B & C). auto. + rewrite ZofB_correct. rewrite B. rewrite A. auto. +Qed. + +Theorem ZofB_range_Bconv: + forall min1 max1 min2 max2, + prec2 >= prec1 -> emax2 >= emax1 -> min2 <= min1 -> max1 <= max2 -> + forall conv_nan m f n, + ZofB_range _ _ f min1 max1 = Some n -> + ZofB_range _ _ (Bconv conv_nan m f) min2 max2 = Some n. +Proof. + intros. + destruct (ZofB_range_inversion _ _ _ _ _ _ H3) as (A & B & C). + unfold ZofB_range. erewrite ZofB_Bconv by eauto. + rewrite ! Zle_bool_true by omega. auto. +Qed. + +(** Change of format (to higher precision) and comparison. *) + +Theorem Bcompare_Bconv_widen: + prec2 >= prec1 -> emax2 >= emax1 -> + forall conv_nan m x y, + Bcompare _ _ (Bconv conv_nan m x) (Bconv conv_nan m y) = Bcompare _ _ x y. +Proof. + intros. destruct (is_finite _ _ x && is_finite _ _ y) eqn:FIN. +- apply andb_true_iff in FIN. destruct FIN. + destruct (Bconv_widen_exact H H0 conv_nan m x H1) as (A & B & C). + destruct (Bconv_widen_exact H H0 conv_nan m y H2) as (D & E & F). + rewrite ! Bcompare_finite_correct by auto. rewrite A, D. auto. +- generalize (Bconv_widen_exact H H0 conv_nan m x) + (Bconv_widen_exact H H0 conv_nan m y); intros P Q. + destruct x, y; try discriminate; simpl in P, Q; simpl; + repeat (match goal with |- context [conv_nan ?b ?pl] => destruct (conv_nan b pl) end); + auto. + destruct Q as (D & E & F); auto. + destruct (binary_normalize prec2 emax2 prec2_gt_0_ Hmax2 m (cond_Zopp b0 (Z.pos m0)) e b0); + discriminate || reflexivity. + destruct P as (A & B & C); auto. + destruct (binary_normalize prec2 emax2 prec2_gt_0_ Hmax2 m (cond_Zopp b (Z.pos m0)) e b); + try discriminate; simpl. destruct b; auto. destruct b, b1; auto. + destruct P as (A & B & C); auto. + destruct (binary_normalize prec2 emax2 prec2_gt_0_ Hmax2 m (cond_Zopp b (Z.pos m0)) e b); + try discriminate; simpl. destruct b; auto. + destruct b, b2; auto. +Qed. + +End Conversions. + +Section Compose_Conversions. + +Variable prec1 emax1 prec2 emax2 : Z. +Context (prec1_gt_0_ : Prec_gt_0 prec1) (prec2_gt_0_ : Prec_gt_0 prec2). +Let emin1 := (3 - emax1 - prec1)%Z. +Let fexp1 := FLT_exp emin1 prec1. +Let emin2 := (3 - emax2 - prec2)%Z. +Let fexp2 := FLT_exp emin2 prec2. +Hypothesis Hmax1 : (prec1 < emax1)%Z. +Hypothesis Hmax2 : (prec2 < emax2)%Z. +Let binary_float1 := binary_float prec1 emax1. +Let binary_float2 := binary_float prec2 emax2. + +(** Converting to a higher precision then down to the original format + is the identity. *) +Theorem Bconv_narrow_widen: + prec2 >= prec1 -> emax2 >= emax1 -> + forall narrow_nan widen_nan m f, + is_nan _ _ f = false -> + Bconv prec2 emax2 prec1 emax1 _ Hmax1 narrow_nan m (Bconv prec1 emax1 prec2 emax2 _ Hmax2 widen_nan m f) = f. +Proof. + intros. destruct (is_finite _ _ f) eqn:FIN. +- assert (EQ: round radix2 fexp1 (round_mode m) (B2R prec1 emax1 f) = B2R prec1 emax1 f). + { apply round_generic. apply valid_rnd_round_mode. apply generic_format_B2R. } + generalize (Bconv_widen_exact _ _ _ _ _ _ Hmax2 H H0 widen_nan m f FIN). + set (f' := Bconv prec1 emax1 prec2 emax2 _ Hmax2 widen_nan m f). + intros (A & B & C). + generalize (Bconv_correct _ _ _ _ _ Hmax1 narrow_nan m f' B). + fold emin1. fold fexp1. rewrite A, C, EQ. rewrite Rlt_bool_true. + intros (D & E & F). + apply B2R_Bsign_inj; auto. + destruct f; try discriminate; simpl. + rewrite Rabs_R0. apply bpow_gt_0. + rewrite F2R_cond_Zopp. rewrite abs_cond_Ropp. rewrite <- F2R_Zabs. simpl Z.abs. + eapply bounded_lt_emax; eauto. +- destruct f; try discriminate. simpl. auto. +Qed. + +End Compose_Conversions. + +(** Specialization to binary32 and binary64 formats. *) + +Require Import Fappli_IEEE_bits. + +Section B3264. + +Let prec32 : (0 < 24)%Z. +apply refl_equal. +Qed. + +Let emax32 : (24 < 128)%Z. +apply refl_equal. +Qed. + +Let prec64 : (0 < 53)%Z. +apply refl_equal. +Qed. + +Let emax64 : (53 < 1024)%Z. +apply refl_equal. +Qed. + +Definition b32_abs : (bool -> nan_pl 24 -> bool * nan_pl 24) -> binary32 -> binary32 := Babs 24 128. +Definition b32_eq_dec : forall (f1 f2: binary32), {f1=f2} + {f1<>f2} := Beq_dec 24 128. +Definition b32_compare : binary32 -> binary32 -> option comparison := Bcompare 24 128. +Definition b32_of_Z : Z -> binary32 := BofZ 24 128 prec32 emax32. +Definition b32_to_Z : binary32 -> option Z := ZofB 24 128. +Definition b32_to_Z_range : binary32 -> Z -> Z -> option Z := ZofB_range 24 128. +Definition b32_exact_inverse : binary32 -> option binary32 := Bexact_inverse 24 128 prec32. + +Definition b64_abs : (bool -> nan_pl 53 -> bool * nan_pl 53) -> binary64 -> binary64 := Babs 53 1024. +Definition b64_eq_dec : forall (f1 f2: binary64), {f1=f2} + {f1<>f2} := Beq_dec 53 1024. +Definition b64_compare : binary64 -> binary64 -> option comparison := Bcompare 53 1024. +Definition b64_of_Z : Z -> binary64 := BofZ 53 1024 prec64 emax64. +Definition b64_to_Z : binary64 -> option Z := ZofB 53 1024. +Definition b64_to_Z_range : binary64 -> Z -> Z -> option Z := ZofB_range 53 1024. +Definition b64_exact_inverse : binary64 -> option binary64 := Bexact_inverse 53 1024 prec64. + +Definition b64_of_b32 : (bool -> nan_pl 24 -> bool * nan_pl 53) -> mode -> binary32 -> binary64 := + Bconv 24 128 53 1024 prec32 prec64. +Definition b32_of_b64 : (bool -> nan_pl 53 -> bool * nan_pl 24) -> mode -> binary64 -> binary32 := + Bconv 53 1024 24 128 prec64 prec32. + +End B3264. diff --git a/lib/Floats.v b/lib/Floats.v index 35009d82..bbc2a927 100644 --- a/lib/Floats.v +++ b/lib/Floats.v @@ -16,47 +16,121 @@ (** Formalization of floating-point numbers, using the Flocq library. *) -Require Import Axioms. Require Import Coqlib. Require Import Integers. Require Import Fappli_IEEE. Require Import Fappli_IEEE_bits. +Require Import Fappli_IEEE_extra. Require Import Fcore. -Require Import Fcalc_round. -Require Import Fcalc_bracket. -Require Import Fprop_Sterbenz. Require Import Program. Require Archi. Close Scope R_scope. -Definition float := binary64. (**r the type of IEE754 doubles *) +Definition float := binary64. (**r the type of IEE754 double-precision FP numbers *) +Definition float32 := binary32. (**r the type of IEE754 single-precision FP numbers *) + +(** Boolean-valued comparisons *) + +Definition cmp_of_comparison (c: comparison) (x: option Datatypes.comparison) : bool := + match c with + | Ceq => + match x with Some Eq => true | _ => false end + | Cne => + match x with Some Eq => false | _ => true end + | Clt => + match x with Some Lt => true | _ => false end + | Cle => + match x with Some(Lt|Eq) => true | _ => false end + | Cgt => + match x with Some Gt => true | _ => false end + | Cge => + match x with Some(Gt|Eq) => true | _ => false end + end. + +Lemma cmp_of_comparison_swap: + forall c x, + cmp_of_comparison (swap_comparison c) x = + cmp_of_comparison c (match x with None => None | Some x => Some (CompOpp x) end). +Proof. + intros. destruct c; destruct x as [[]|]; reflexivity. +Qed. + +Lemma cmp_of_comparison_ne_eq: + forall x, cmp_of_comparison Cne x = negb (cmp_of_comparison Ceq x). +Proof. + intros. destruct x as [[]|]; reflexivity. +Qed. + +Lemma cmp_of_comparison_lt_eq_false: + forall x, cmp_of_comparison Clt x = true -> cmp_of_comparison Ceq x = true -> False. +Proof. + destruct x as [[]|]; simpl; intros; discriminate. +Qed. + +Lemma cmp_of_comparison_le_lt_eq: + forall x, cmp_of_comparison Cle x = cmp_of_comparison Clt x || cmp_of_comparison Ceq x. +Proof. + destruct x as [[]|]; reflexivity. +Qed. + +Lemma cmp_of_comparison_gt_eq_false: + forall x, cmp_of_comparison Cgt x = true -> cmp_of_comparison Ceq x = true -> False. +Proof. + destruct x as [[]|]; simpl; intros; discriminate. +Qed. + +Lemma cmp_of_comparison_ge_gt_eq: + forall x, cmp_of_comparison Cge x = cmp_of_comparison Cgt x || cmp_of_comparison Ceq x. +Proof. + destruct x as [[]|]; reflexivity. +Qed. + +Lemma cmp_of_comparison_lt_gt_false: + forall x, cmp_of_comparison Clt x = true -> cmp_of_comparison Cgt x = true -> False. +Proof. + destruct x as [[]|]; simpl; intros; discriminate. +Qed. + +(** Function used to parse floats *) + +Program Definition build_from_parsed + (prec:Z) (emax:Z) (prec_gt_0 : Prec_gt_0 prec) (Hmax:prec < emax) + (base:positive) (intPart:positive) (expPart:Z) := + match expPart return _ with + | Z0 => + binary_normalize prec emax prec_gt_0 Hmax mode_NE (Zpos intPart) Z0 false + | Zpos p => + binary_normalize prec emax prec_gt_0 Hmax mode_NE ((Zpos intPart) * Zpower_pos (Zpos base) p) Z0 false + | Zneg p => + let exp := Zpower_pos (Zpos base) p in + match exp return 0 < exp -> _ with + | Zneg _ | Z0 => _ + | Zpos p => + fun _ => + FF2B prec emax _ (proj1 (Bdiv_correct_aux prec emax prec_gt_0 Hmax mode_NE false intPart Z0 false p Z0)) + end _ + end. +Next Obligation. + apply Zpower_pos_gt_0. reflexivity. +Qed. + +Local Notation __ := (refl_equal Datatypes.Lt). + +Local Hint Extern 1 (Prec_gt_0 _) => exact (refl_equal Datatypes.Lt). +Local Hint Extern 1 (_ < _) => exact (refl_equal Datatypes.Lt). + +(** * Double-precision FP numbers *) Module Float. -Definition zero: float := B754_zero _ _ false. (**r the float [+0.0] *) +(** ** NaN payload manipulations *) -Definition eq_dec: forall (f1 f2: float), {f1 = f2} + {f1 <> f2}. -Proof. - Ltac try_not_eq := try solve [right; congruence]. - destruct f1 as [| |? []|], f2 as [| |? []|]; - try destruct b; try destruct b0; - try solve [left; auto]; try_not_eq. - destruct (positive_eq_dec x x0); try_not_eq; - subst; left; f_equal; f_equal; apply proof_irr. - destruct (positive_eq_dec x x0); try_not_eq; - subst; left; f_equal; f_equal; apply proof_irr. - destruct (positive_eq_dec m m0); try_not_eq; - destruct (Z_eq_dec e e1); try solve [right; intro H; inv H; congruence]; - subst; left; rewrite (proof_irr e0 e2); auto. - destruct (positive_eq_dec m m0); try_not_eq; - destruct (Z_eq_dec e e1); try solve [right; intro H; inv H; congruence]; - subst; left; rewrite (proof_irr e0 e2); auto. -Defined. +(** The following definitions are not part of the IEEE754 standard but + apply to all architectures supported by CompCert. *) + +(** Transform a Nan payload to a quiet Nan payload. *) -(* Transform a Nan payload to a quiet Nan payload. - This is not part of the IEEE754 standard, but shared between all - architectures of Compcert. *) Program Definition transform_quiet_pl (pl:nan_pl 53) : nan_pl 53 := Pos.lor pl (nat_iter 51 xO xH). Next Obligation. @@ -91,39 +165,10 @@ Proof. simpl. apply lor_idempotent. Qed. -(** Arithmetic operations *) - -(* The Nan payload operations for neg and abs is not part of the IEEE754 - standard, but shared between all architectures of Compcert. *) -Definition neg_pl (s:bool) (pl:nan_pl 53) := (negb s, pl). -Definition abs_pl (s:bool) (pl:nan_pl 53) := (false, pl). - -Definition neg: float -> float := b64_opp neg_pl. (**r opposite (change sign) *) -Definition abs (x: float): float := (**r absolute value (set sign to [+]) *) - match x with - | B754_nan s pl => let '(s, pl) := abs_pl s pl in B754_nan _ _ s pl - | B754_infinity _ => B754_infinity _ _ false - | B754_finite _ m e H => B754_finite _ _ false m e H - | B754_zero _ => B754_zero _ _ false - end. - -Definition binary_normalize64 (m e:Z) (s:bool): float := - binary_normalize 53 1024 eq_refl eq_refl mode_NE m e s. +(** Nan payload operations for single <-> double conversions. *) -Definition binary_normalize64_correct (m e:Z) (s:bool) := - binary_normalize_correct 53 1024 eq_refl eq_refl mode_NE m e s. -Global Opaque binary_normalize64_correct. - -Definition binary_normalize32 (m e:Z) (s:bool) : binary32 := - binary_normalize 24 128 eq_refl eq_refl mode_NE m e s. - -Definition binary_normalize32_correct (m e:Z) (s:bool) := - binary_normalize_correct 24 128 eq_refl eq_refl mode_NE m e s. -Global Opaque binary_normalize32_correct. - -(* The Nan payload operations for single <-> double conversions are not part of - the IEEE754 standard, but shared between all architectures of Compcert. *) -Definition floatofbinary32_pl (s:bool) (pl:nan_pl 24) : (bool * nan_pl 53). +Definition of_single_pl (s:bool) (pl:nan_pl 24) : (bool * nan_pl 53). +Proof. Proof. refine (s, transform_quiet_pl (exist _ (Pos.shiftl_nat (proj1_sig pl) 29) _)). abstract ( @@ -133,7 +178,8 @@ Proof. zify; omega). Defined. -Definition binary32offloat_pl (s:bool) (pl:nan_pl 53) : (bool * nan_pl 24). +Definition to_single_pl (s:bool) (pl:nan_pl 53) : (bool * nan_pl 24). +Proof. Proof. refine (s, exist _ (Pos.shiftr_nat (proj1_sig (transform_quiet_pl pl)) 29) _). abstract ( @@ -144,128 +190,14 @@ Proof. rewrite !H, <- !NPeano.Nat.sub_add_distr; zify; omega). Defined. -Definition floatofbinary32 (f: binary32) : float := (**r single precision embedding in double precision *) - match f with - | B754_nan s pl => let '(s, pl) := floatofbinary32_pl s pl in B754_nan _ _ s pl - | B754_infinity s => B754_infinity _ _ s - | B754_zero s => B754_zero _ _ s - | B754_finite s m e _ => - binary_normalize64 (cond_Zopp s (Zpos m)) e s - end. - -Definition binary32offloat (f: float) : binary32 := (**r conversion to single precision *) - match f with - | B754_nan s pl => let '(s, pl) := binary32offloat_pl s pl in B754_nan _ _ s pl - | B754_infinity s => B754_infinity _ _ s - | B754_zero s => B754_zero _ _ s - | B754_finite s m e _ => - binary_normalize32 (cond_Zopp s (Zpos m)) e s - end. - -Definition singleoffloat (f: float): float := (**r conversion to single precision, embedded in double *) - floatofbinary32 (binary32offloat f). - -Definition Zoffloat (f:float): option Z := (**r conversion to Z *) - match f with - | B754_finite s m (Zpos e) _ => Some (cond_Zopp s (Zpos m) * Zpower_pos radix2 e) - | B754_finite s m 0 _ => Some (cond_Zopp s (Zpos m)) - | B754_finite s m (Zneg e) _ => Some (cond_Zopp s (Zpos m / Zpower_pos radix2 e)) - | B754_zero _ => Some 0 - | _ => None - end. - -Definition intoffloat (f:float): option int := (**r conversion to signed 32-bit int *) - match Zoffloat f with - | Some n => - if Zle_bool Int.min_signed n && Zle_bool n Int.max_signed then - Some (Int.repr n) - else - None - | None => None - end. - -Definition intuoffloat (f:float): option int := (**r conversion to unsigned 32-bit int *) - match Zoffloat f with - | Some n => - if Zle_bool 0 n && Zle_bool n Int.max_unsigned then - Some (Int.repr n) - else - None - | None => None - end. - -Definition longoffloat (f:float): option int64 := (**r conversion to signed 64-bit int *) - match Zoffloat f with - | Some n => - if Zle_bool Int64.min_signed n && Zle_bool n Int64.max_signed then - Some (Int64.repr n) - else - None - | None => None - end. - -Definition longuoffloat (f:float): option int64 := (**r conversion to unsigned 64-bit int *) - match Zoffloat f with - | Some n => - if Zle_bool 0 n && Zle_bool n Int64.max_unsigned then - Some (Int64.repr n) - else - None - | None => None - end. - -(* Functions used to parse floats *) -Program Definition build_from_parsed - (prec:Z) (emax:Z) (prec_gt_0 :Prec_gt_0 prec) (Hmax:prec < emax) - (base:positive) (intPart:positive) (expPart:Z) := - match expPart return _ with - | Z0 => - binary_normalize prec emax prec_gt_0 Hmax mode_NE (Zpos intPart) Z0 false - | Zpos p => - binary_normalize prec emax prec_gt_0 Hmax mode_NE ((Zpos intPart) * Zpower_pos (Zpos base) p) Z0 false - | Zneg p => - let exp := Zpower_pos (Zpos base) p in - match exp return 0 < exp -> _ with - | Zneg _ | Z0 => _ - | Zpos p => - fun _ => - FF2B prec emax _ (proj1 (Bdiv_correct_aux prec emax prec_gt_0 Hmax mode_NE false intPart Z0 false p Z0)) - end _ - end. -Next Obligation. -apply Zpower_pos_gt_0. -reflexivity. -Qed. - -Definition build_from_parsed64 (base:positive) (intPart:positive) (expPart:Z) : float := - build_from_parsed 53 1024 eq_refl eq_refl base intPart expPart. +(** NaN payload operations for opposite and absolute value. *) -Definition build_from_parsed32 (base:positive) (intPart:positive) (expPart:Z) : float := - floatofbinary32 (build_from_parsed 24 128 eq_refl eq_refl base intPart expPart). - -Definition floatofint (n:int): float := (**r conversion from signed 32-bit int *) - binary_normalize64 (Int.signed n) 0 false. -Definition floatofintu (n:int): float:= (**r conversion from unsigned 32-bit int *) - binary_normalize64 (Int.unsigned n) 0 false. - -Definition floatoflong (n:int64): float := (**r conversion from signed 64-bit int *) - binary_normalize64 (Int64.signed n) 0 false. -Definition floatoflongu (n:int64): float:= (**r conversion from unsigned 64-bit int *) - binary_normalize64 (Int64.unsigned n) 0 false. - -Definition singleofint (n:int): float := (**r conversion from signed 32-bit int to single-precision float *) - floatofbinary32 (binary_normalize32 (Int.signed n) 0 false). -Definition singleofintu (n:int): float:= (**r conversion from unsigned 32-bit int to single-precision float *) - floatofbinary32 (binary_normalize32 (Int.unsigned n) 0 false). - -Definition singleoflong (n:int64): float := (**r conversion from signed 64-bit int to single-precision float *) - floatofbinary32 (binary_normalize32 (Int64.signed n) 0 false). -Definition singleoflongu (n:int64): float:= (**r conversion from unsigned 64-bit int to single-precision float *) - floatofbinary32 (binary_normalize32 (Int64.unsigned n) 0 false). +Definition neg_pl (s:bool) (pl:nan_pl 53) := (negb s, pl). +Definition abs_pl (s:bool) (pl:nan_pl 53) := (false, pl). -(* The Nan payload operations for two-argument arithmetic operations are not part of - the IEEE754 standard, but all architectures of Compcert share a similar - NaN behavior, parameterized by: +(** The NaN payload operations for two-argument arithmetic operations + are not part of the IEEE754 standard, but all architectures of + Compcert share a similar NaN behavior, parameterized by: - a "default" payload which occurs when an operation generates a NaN from non-NaN arguments; - a choice function determining which of the payload arguments to choose, @@ -274,78 +206,67 @@ Definition singleoflongu (n:int64): float:= (**r conversion from unsigned 64-bit Definition binop_pl (x y: binary64) : bool*nan_pl 53 := match x, y with | B754_nan s1 pl1, B754_nan s2 pl2 => - if Archi.choose_binop_pl s1 pl1 s2 pl2 + if Archi.choose_binop_pl_64 s1 pl1 s2 pl2 then (s2, transform_quiet_pl pl2) else (s1, transform_quiet_pl pl1) | B754_nan s1 pl1, _ => (s1, transform_quiet_pl pl1) | _, B754_nan s2 pl2 => (s2, transform_quiet_pl pl2) - | _, _ => Archi.default_pl + | _, _ => Archi.default_pl_64 end. +(** ** Operations over double-precision floats *) + +Definition zero: float := B754_zero _ _ false. (**r the float [+0.0] *) + +Definition eq_dec: forall (f1 f2: float), {f1 = f2} + {f1 <> f2} := b64_eq_dec. + +(** Arithmetic operations *) + +Definition neg: float -> float := b64_opp neg_pl. (**r opposite (change sign) *) +Definition abs: float -> float := b64_abs abs_pl. (**r absolute value (set sign to [+]) *) Definition add: float -> float -> float := b64_plus binop_pl mode_NE. (**r addition *) Definition sub: float -> float -> float := b64_minus binop_pl mode_NE. (**r subtraction *) Definition mul: float -> float -> float := b64_mult binop_pl mode_NE. (**r multiplication *) Definition div: float -> float -> float := b64_div binop_pl mode_NE. (**r division *) +Definition cmp (c:comparison) (f1 f2: float) : bool := (**r comparison *) + cmp_of_comparison c (b64_compare f1 f2). -Definition order_float (f1 f2:float): option Datatypes.comparison := - match f1, f2 with - | B754_nan _ _,_ | _,B754_nan _ _ => None - | B754_infinity true, B754_infinity true - | B754_infinity false, B754_infinity false => Some Eq - | B754_infinity true, _ => Some Lt - | B754_infinity false, _ => Some Gt - | _, B754_infinity true => Some Gt - | _, B754_infinity false => Some Lt - | B754_finite true _ _ _, B754_zero _ => Some Lt - | B754_finite false _ _ _, B754_zero _ => Some Gt - | B754_zero _, B754_finite true _ _ _ => Some Gt - | B754_zero _, B754_finite false _ _ _ => Some Lt - | B754_zero _, B754_zero _ => Some Eq - | B754_finite s1 m1 e1 _, B754_finite s2 m2 e2 _ => - match s1, s2 with - | true, false => Some Lt - | false, true => Some Gt - | false, false => - match Zcompare e1 e2 with - | Lt => Some Lt - | Gt => Some Gt - | Eq => Some (Pcompare m1 m2 Eq) - end - | true, true => - match Zcompare e1 e2 with - | Lt => Some Gt - | Gt => Some Lt - | Eq => Some (CompOpp (Pcompare m1 m2 Eq)) - end - end - end. +(** Conversions *) -Definition cmp (c:comparison) (f1 f2:float) : bool := (**r comparison *) - match c with - | Ceq => - match order_float f1 f2 with Some Eq => true | _ => false end - | Cne => - match order_float f1 f2 with Some Eq => false | _ => true end - | Clt => - match order_float f1 f2 with Some Lt => true | _ => false end - | Cle => - match order_float f1 f2 with Some(Lt|Eq) => true | _ => false end - | Cgt => - match order_float f1 f2 with Some Gt => true | _ => false end - | Cge => - match order_float f1 f2 with Some(Gt|Eq) => true | _ => false end - end. +Definition of_single: float32 -> float := b64_of_b32 of_single_pl mode_NE. +Definition to_single: float -> float32 := b32_of_b64 to_single_pl mode_NE. + +Definition to_int (f:float): option int := (**r conversion to signed 32-bit int *) + option_map Int.repr (b64_to_Z_range f Int.min_signed Int.max_signed). +Definition to_intu (f:float): option int := (**r conversion to unsigned 32-bit int *) + option_map Int.repr (b64_to_Z_range f 0 Int.max_unsigned). +Definition to_long (f:float): option int64 := (**r conversion to signed 64-bit int *) + option_map Int64.repr (b64_to_Z_range f Int64.min_signed Int64.max_signed). +Definition to_longu (f:float): option int64 := (**r conversion to unsigned 64-bit int *) + option_map Int64.repr (b64_to_Z_range f 0 Int64.max_unsigned). + +Definition of_int (n:int): float := (**r conversion from signed 32-bit int *) + b64_of_Z (Int.signed n). +Definition of_intu (n:int): float:= (**r conversion from unsigned 32-bit int *) + b64_of_Z (Int.unsigned n). + +Definition of_long (n:int64): float := (**r conversion from signed 64-bit int *) + b64_of_Z (Int64.signed n). +Definition of_longu (n:int64): float:= (**r conversion from unsigned 64-bit int *) + b64_of_Z (Int64.unsigned n). + +Definition from_parsed (base:positive) (intPart:positive) (expPart:Z) : float := + build_from_parsed 53 1024 __ __ base intPart expPart. (** Conversions between floats and their concrete in-memory representation - as a sequence of 64 bits (double precision) or 32 bits (single precision). *) + as a sequence of 64 bits. *) -Definition bits_of_double (f: float): int64 := Int64.repr (bits_of_b64 f). -Definition double_of_bits (b: int64): float := b64_of_bits (Int64.unsigned b). +Definition to_bits (f: float): int64 := Int64.repr (bits_of_b64 f). +Definition of_bits (b: int64): float := b64_of_bits (Int64.unsigned b). -Definition bits_of_single (f: float) : int := Int.repr (bits_of_b32 (binary32offloat f)). -Definition single_of_bits (b: int): float := floatofbinary32 (b32_of_bits (Int.unsigned b)). +Definition from_words (hi lo: int) : float := of_bits (Int64.ofwords hi lo). -Definition from_words (hi lo: int) : float := double_of_bits (Int64.ofwords hi lo). +(** ** Properties *) (** Below are the only properties of floating-point arithmetic that we rely on in the compiler proof. *) @@ -362,75 +283,37 @@ Ltac smart_omega := compute_this Int.min_signed; compute_this Int.max_signed; compute_this Int64.modulus; compute_this Int64.half_modulus; compute_this Int64.max_unsigned; - compute_this (Zpower_pos 2 1024); compute_this (Zpower_pos 2 53); compute_this (Zpower_pos 2 52); + compute_this (Zpower_pos 2 1024); compute_this (Zpower_pos 2 53); compute_this (Zpower_pos 2 52); compute_this (Zpower_pos 2 32); zify; omega. -Lemma floatofbinary32_exact : - forall f, is_finite_strict _ _ f = true -> - is_finite_strict _ _ (floatofbinary32 f) = true /\ B2R _ _ f = B2R _ _ (floatofbinary32 f). -Proof. - destruct f as [ | | |s m e]; try discriminate; intro. - pose proof (binary_normalize64_correct (cond_Zopp s (Zpos m)) e s). - match goal with [H0:if Rlt_bool (Rabs ?x) _ then _ else _ |- _ /\ ?y = _] => assert (x=y)%R end. - apply round_generic; [now apply valid_rnd_round_mode|]. - apply (generic_inclusion_ln_beta _ (FLT_exp (3 - 128 - 24) 24)). - intro; eapply Zle_trans; [apply Zle_max_compat_l | apply Zle_max_compat_r]; omega. - apply generic_format_canonic; apply canonic_canonic_mantissa; apply (proj1 (andb_prop _ _ e0)). - rewrite H1, Rlt_bool_true in H0; intuition; unfold floatofbinary32, binary_normalize64. - match goal with [ |- _ _ _ ?x = true ] => destruct x end; try discriminate. - symmetry in H2; apply F2R_eq_0_reg in H2; destruct s; discriminate. - reflexivity. - eapply Rlt_trans. - unfold B2R; rewrite <- F2R_Zabs, abs_cond_Zopp; eapply bounded_lt_emax; now apply e0. - now apply bpow_lt. -Qed. - -Lemma binary32offloatofbinary32_num : - forall f, is_nan _ _ f = false -> - binary32offloat (floatofbinary32 f) = f. -Proof. - intros f Hnan; pose proof (floatofbinary32_exact f); destruct f as [ | | |s m e]; try reflexivity. - discriminate. - specialize (H eq_refl); destruct H. - destruct (floatofbinary32 (B754_finite 24 128 s m e e0)) as [ | | |s1 m1 e1]; try discriminate. - unfold binary32offloat. - pose proof (binary_normalize32_correct (cond_Zopp s1 (Zpos m1)) e1 s1). - unfold B2R at 2 in H0; cbv iota zeta beta in H0; rewrite <- H0, round_generic in H1. - rewrite Rlt_bool_true in H1. - unfold binary_normalize32. - apply B2R_inj; intuition; match goal with [|- _ _ _ ?f = true] => destruct f end; try discriminate. - symmetry in H2; apply F2R_eq_0_reg in H2; destruct s; discriminate. - reflexivity. - unfold B2R; rewrite <- F2R_Zabs, abs_cond_Zopp; eapply bounded_lt_emax; apply e0. - now apply valid_rnd_round_mode. - now apply generic_format_B2R. -Qed. - -Lemma floatofbinary32offloatofbinary32_pl: +Lemma of_single_to_single_pl: forall s pl, - prod_rect (fun _ => _) floatofbinary32_pl (prod_rect (fun _ => _) binary32offloat_pl (floatofbinary32_pl s pl)) = floatofbinary32_pl s pl. + prod_rect (fun _ => _) of_single_pl (prod_rect (fun _ => _) to_single_pl (of_single_pl s pl)) = of_single_pl s pl. Proof. - destruct pl. unfold binary32offloat_pl, floatofbinary32_pl. + destruct pl. unfold of_single_pl, to_single_pl. unfold transform_quiet_pl, proj1_sig. simpl. f_equal. apply nan_payload_fequal. unfold Pos.shiftr_nat. simpl. rewrite !lor_idempotent. reflexivity. Qed. -Lemma floatofbinary32offloatofbinary32 : - forall f, floatofbinary32 (binary32offloat (floatofbinary32 f)) = floatofbinary32 f. +Lemma of_single_to_single_of_single: + forall f, of_single (to_single (of_single f)) = of_single f. Proof. - destruct f; try (rewrite binary32offloatofbinary32_num; tauto). - unfold floatofbinary32, binary32offloat. - rewrite <- floatofbinary32offloatofbinary32_pl at 2. + intros. unfold of_single, to_single, b64_of_b32, b32_of_b64. + destruct (is_nan _ _ f) eqn:ISNAN. +- destruct f; try discriminate. + unfold Bconv. + rewrite <- of_single_to_single_pl at 2. reflexivity. +- rewrite (Bconv_narrow_widen 24); auto; omega. Qed. -Lemma binary32offloatofbinary32offloat_pl: +Lemma to_single_of_single_pl: forall s pl, - prod_rect (fun _ => _) binary32offloat_pl (prod_rect (fun _ => _) floatofbinary32_pl (binary32offloat_pl s pl)) = binary32offloat_pl s pl. + prod_rect (fun _ => _) to_single_pl (prod_rect (fun _ => _) of_single_pl (to_single_pl s pl)) = to_single_pl s pl. Proof. - destruct pl. unfold binary32offloat_pl, floatofbinary32_pl. unfold prod_rect. + destruct pl. unfold of_single_pl, to_single_pl. unfold prod_rect. f_equal. apply nan_payload_fequal. rewrite transform_quiet_pl_idempotent. unfold transform_quiet_pl, proj1_sig. @@ -444,293 +327,132 @@ Proof. rewrite !nat_iter_succ_r with (f:=Pos.div2). auto. Qed. -Lemma binary32offloatofbinary32offloat : - forall f, binary32offloat (floatofbinary32 (binary32offloat f)) = binary32offloat f. +Lemma to_single_of_single_to_single: + forall f, to_single (of_single (to_single f)) = to_single f. Proof. - destruct f; try (rewrite binary32offloatofbinary32_num; simpl; tauto). - unfold floatofbinary32, binary32offloat. - rewrite <- binary32offloatofbinary32offloat_pl at 2. + intros. unfold to_single, of_single, b32_of_b64, b64_of_b32. + destruct (is_nan _ _ f) eqn:ISNAN. +- destruct f; try discriminate. + unfold Bconv. + rewrite <- to_single_of_single_pl at 2. reflexivity. - rewrite binary32offloatofbinary32_num; simpl. auto. - unfold binary_normalize32. - pose proof (binary_normalize32_correct (cond_Zopp b (Z.pos m)) e b). - destruct binary_normalize; auto. simpl in H. - destruct Rlt_bool in H. intuition. - unfold binary_overflow in H. destruct n. - destruct overflow_to_inf in H; discriminate. +- rewrite (Bconv_narrow_widen 24); auto. omega. omega. + set (f' := Bconv 53 1024 24 128 __ __ to_single_pl mode_NE f). + destruct f; try discriminate; try reflexivity. + exploit (Bconv_correct 53 1024 24 128 __ __ to_single_pl mode_NE + (B754_finite 53 1024 b m e e0)). auto. + destruct Rlt_bool. + intros (A & B & C). apply is_finite_not_is_nan; auto. + fold f'. intros A. destruct f'; auto. + simpl in A. unfold binary_overflow in A. + destruct overflow_to_inf in A; destruct n; discriminate. Qed. -Theorem singleoffloat_idem: - forall f, singleoffloat (singleoffloat f) = singleoffloat f. -Proof. - intros; unfold singleoffloat; rewrite binary32offloatofbinary32offloat; reflexivity. -Qed. +(** Commutativity properties of addition and multiplication. *) -Theorem singleoflong_idem: - forall n, singleoffloat (singleoflong n) = singleoflong n. +Theorem add_commut: + forall x y, is_nan _ _ x = false \/ is_nan _ _ y = false -> add x y = add y x. Proof. - intros; unfold singleoffloat, singleoflong. rewrite floatofbinary32offloatofbinary32; reflexivity. + intros. apply Bplus_commut. + destruct x, y; try reflexivity. simpl in H. intuition congruence. Qed. -Theorem singleoflongu_idem: - forall n, singleoffloat (singleoflongu n) = singleoflongu n. +Theorem mul_commut: + forall x y, is_nan _ _ x = false \/ is_nan _ _ y = false -> mul x y = mul y x. Proof. - intros; unfold singleoffloat, singleoflongu. rewrite floatofbinary32offloatofbinary32; reflexivity. + intros. apply Bmult_commut. + destruct x, y; try reflexivity. simpl in H. intuition congruence. Qed. -Definition is_single (f: float) : Prop := exists s, f = floatofbinary32 s. - -Theorem singleoffloat_is_single: - forall f, is_single (singleoffloat f). -Proof. - intros. exists (binary32offloat f); auto. -Qed. +(** Multiplication by 2 is diagonal addition. *) -Theorem singleoffloat_of_single: - forall f, is_single f -> singleoffloat f = f. +Theorem mul2_add: + forall f, add f f = mul f (of_int (Int.repr 2%Z)). Proof. - intros. destruct H as [s EQ]. subst f. unfold singleoffloat. - apply floatofbinary32offloatofbinary32. + intros. apply Bmult2_Bplus. + intros. destruct x; try discriminate. simpl. + transitivity (b, transform_quiet_pl n). + destruct Archi.choose_binop_pl_64; auto. + destruct y; auto || discriminate. Qed. -Theorem is_single_dec: forall f, {is_single f} + {~is_single f}. -Proof. - intros. case (eq_dec (singleoffloat f) f); intros. - unfold singleoffloat in e. left. exists (binary32offloat f). auto. - right; red; intros; elim n. apply singleoffloat_of_single; auto. -Defined. +(** Divisions that can be turned into multiplication by an inverse. *) -(** Commutativity properties of addition and multiplication. *) +Definition exact_inverse : float -> option float := b64_exact_inverse. -Theorem add_commut: - forall x y, is_nan _ _ x = false \/ is_nan _ _ y = false -> add x y = add y x. -Proof. - intros x y NAN. unfold add, b64_plus. - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE x y). - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE y x). - unfold Bplus in *; destruct x; destruct y; auto. -- rewrite (eqb_sym b0 b). destruct (eqb b b0) eqn:EQB; auto. f_equal; apply eqb_prop; auto. -- rewrite (eqb_sym b0 b). destruct (eqb b b0) eqn:EQB. - f_equal; apply eqb_prop; auto. - auto. -- simpl in NAN; intuition congruence. -- exploit H; auto. clear H. exploit H0; auto. clear H0. - set (x := B754_finite 53 1024 b0 m0 e1 e2). - set (rx := B2R 53 1024 x). - set (y := B754_finite 53 1024 b m e e0). - set (ry := B2R 53 1024 y). - rewrite (Rplus_comm ry rx). destruct Rlt_bool. - intros (A1 & A2 & A3) (B1 & B2 & B3). - apply B2R_Bsign_inj; auto. rewrite <- B1 in A1. auto. - rewrite Z.add_comm. rewrite Z.min_comm. auto. - intros (A1 & A2) (B1 & B2). apply B2FF_inj. rewrite B2 in B1. rewrite <- B1 in A1. auto. -Qed. - -Theorem mul_commut: - forall x y, is_nan _ _ x = false \/ is_nan _ _ y = false -> mul x y = mul y x. +Theorem div_mul_inverse: + forall x y z, exact_inverse y = Some z -> div x y = mul x z. Proof. - intros x y NAN. unfold mul, b64_mult. - pose proof (Bmult_correct 53 1024 eq_refl eq_refl binop_pl mode_NE x y). - pose proof (Bmult_correct 53 1024 eq_refl eq_refl binop_pl mode_NE y x). - unfold Bmult in *; destruct x; destruct y; auto. -- f_equal. apply xorb_comm. -- f_equal. apply xorb_comm. -- f_equal. apply xorb_comm. -- f_equal. apply xorb_comm. -- simpl in NAN. intuition congruence. -- f_equal. apply xorb_comm. -- f_equal. apply xorb_comm. -- set (x := B754_finite 53 1024 b0 m0 e1 e2) in *. - set (rx := B2R 53 1024 x) in *. - set (y := B754_finite 53 1024 b m e e0) in *. - set (ry := B2R 53 1024 y) in *. - rewrite (Rmult_comm ry rx) in *. destruct Rlt_bool. - destruct H as (A1 & A2 & A3); destruct H0 as (B1 & B2 & B3). - apply B2R_Bsign_inj; auto. rewrite <- B1 in A1. auto. - rewrite ! Bsign_FF2B. f_equal. f_equal. apply xorb_comm. apply Pos.mul_comm. apply Z.add_comm. - apply B2FF_inj. etransitivity. eapply H. rewrite xorb_comm. auto. + intros. apply Bdiv_mult_inverse; auto. + intros. destruct x0; try discriminate. simpl. + transitivity (b, transform_quiet_pl n). + destruct y0; reflexivity || discriminate. + destruct z0; reflexivity || discriminate. Qed. (** Properties of comparisons. *) -Theorem order_float_finite_correct: - forall f1 f2, is_finite _ _ f1 = true -> is_finite _ _ f2 = true -> - match order_float f1 f2 with - | Some c => Rcompare (B2R _ _ f1) (B2R _ _ f2) = c - | None => False - end. -Proof. - Ltac apply_Rcompare := - match goal with - | [ |- Rcompare _ _ = Lt ] => apply Rcompare_Lt - | [ |- Rcompare _ _ = Eq ] => apply Rcompare_Eq - | [ |- Rcompare _ _ = Gt ] => apply Rcompare_Gt - end. - unfold order_float; intros. - destruct f1, f2; try discriminate; unfold B2R, F2R, Fnum, Fexp, cond_Zopp; - try (replace 0%R with (Z2R 0 * bpow radix2 e)%R by (simpl Z2R; ring); - rewrite Rcompare_mult_r by (apply bpow_gt_0); rewrite Rcompare_Z2R). - apply_Rcompare; reflexivity. - destruct b0; reflexivity. - destruct b; reflexivity. - clear H H0. - apply andb_prop in e0; destruct e0; apply (canonic_canonic_mantissa _ _ false) in H. - apply andb_prop in e2; destruct e2; apply (canonic_canonic_mantissa _ _ false) in H1. - pose proof (Zcompare_spec e e1); unfold canonic, Fexp in H1, H. - assert (forall m1 m2 e1 e2, - let x := (Z2R (Zpos m1) * bpow radix2 e1)%R in - let y := (Z2R (Zpos m2) * bpow radix2 e2)%R in - canonic_exp radix2 (FLT_exp (3-1024-53) 53) x < canonic_exp radix2 (FLT_exp (3-1024-53) 53) y -> (x < y)%R). - intros; apply Rnot_le_lt; intro; apply (ln_beta_le radix2) in H5. - apply (fexp_monotone 53 1024) in H5; unfold canonic_exp in H4; omega. - apply Rmult_gt_0_compat; [apply (Z2R_lt 0); reflexivity|now apply bpow_gt_0]. - assert (forall m1 m2 e1 e2, (Z2R (- Zpos m1) * bpow radix2 e1 < Z2R (Zpos m2) * bpow radix2 e2)%R). - intros; apply (Rlt_trans _ 0%R). - replace 0%R with (0*bpow radix2 e0)%R by ring; apply Rmult_lt_compat_r; - [apply bpow_gt_0; reflexivity|now apply (Z2R_lt _ 0)]. - apply Rmult_gt_0_compat; [apply (Z2R_lt 0); reflexivity|now apply bpow_gt_0]. - destruct b, b0; try (now apply_Rcompare; apply H5); inversion H3; - try (apply_Rcompare; apply H4; rewrite H, H1 in H7; assumption); - try (apply_Rcompare; do 2 rewrite Z2R_opp, Ropp_mult_distr_l_reverse; - apply Ropp_lt_contravar; apply H4; rewrite H, H1 in H7; assumption); - rewrite H7, Rcompare_mult_r, Rcompare_Z2R by (apply bpow_gt_0); reflexivity. -Qed. - Theorem cmp_swap: - forall c x y, Float.cmp (swap_comparison c) x y = Float.cmp c y x. + forall c x y, cmp (swap_comparison c) x y = cmp c y x. Proof. - destruct c, x, y; simpl; try destruct b; try destruct b0; try reflexivity; - rewrite <- (Zcompare_antisym e e1); destruct (e ?= e1); try reflexivity; - change Eq with (CompOpp Eq); rewrite <- (Pcompare_antisym m m0 Eq); - simpl; destruct (Pcompare m m0 Eq); reflexivity. + unfold cmp, b64_compare; intros. rewrite (Bcompare_swap _ _ x y). + apply cmp_of_comparison_swap. Qed. Theorem cmp_ne_eq: forall f1 f2, cmp Cne f1 f2 = negb (cmp Ceq f1 f2). Proof. - unfold cmp; intros; destruct (order_float f1 f2) as [ [] | ]; reflexivity. + intros; apply cmp_of_comparison_ne_eq. Qed. Theorem cmp_lt_eq_false: forall f1 f2, cmp Clt f1 f2 = true -> cmp Ceq f1 f2 = true -> False. Proof. - unfold cmp; intros; destruct (order_float f1 f2) as [ [] | ]; discriminate. + intros f1 f2; apply cmp_of_comparison_lt_eq_false. Qed. Theorem cmp_le_lt_eq: forall f1 f2, cmp Cle f1 f2 = cmp Clt f1 f2 || cmp Ceq f1 f2. Proof. - unfold cmp; intros; destruct (order_float f1 f2) as [ [] | ]; reflexivity. + intros f1 f2; apply cmp_of_comparison_le_lt_eq. Qed. -Corollary cmp_gt_eq_false: +Theorem cmp_gt_eq_false: forall x y, cmp Cgt x y = true -> cmp Ceq x y = true -> False. Proof. - intros; rewrite <- cmp_swap in H; rewrite <- cmp_swap in H0; - eapply cmp_lt_eq_false; now eauto. + intros f1 f2; apply cmp_of_comparison_gt_eq_false. Qed. -Corollary cmp_ge_gt_eq: +Theorem cmp_ge_gt_eq: forall f1 f2, cmp Cge f1 f2 = cmp Cgt f1 f2 || cmp Ceq f1 f2. Proof. - intros. - change Cge with (swap_comparison Cle); change Cgt with (swap_comparison Clt); - change Ceq with (swap_comparison Ceq). - repeat rewrite cmp_swap. - now apply cmp_le_lt_eq. + intros f1 f2; apply cmp_of_comparison_ge_gt_eq. Qed. Theorem cmp_lt_gt_false: forall f1 f2, cmp Clt f1 f2 = true -> cmp Cgt f1 f2 = true -> False. Proof. - unfold cmp; intros; destruct (order_float f1 f2) as [ [] | ]; discriminate. + intros f1 f2; apply cmp_of_comparison_lt_gt_false. Qed. (** Properties of conversions to/from in-memory representation. - The double-precision conversions are bijective (one-to-one). - The single-precision conversions lose precision exactly - as described by [singleoffloat] rounding. *) + The conversions are bijective (one-to-one). *) -Theorem double_of_bits_of_double: - forall f, double_of_bits (bits_of_double f) = f. +Theorem of_to_bits: + forall f, of_bits (to_bits f) = f. Proof. - intros; unfold double_of_bits, bits_of_double, bits_of_b64, b64_of_bits. + intros; unfold of_bits, to_bits, bits_of_b64, b64_of_bits. rewrite Int64.unsigned_repr, binary_float_of_bits_of_binary_float; [reflexivity|]. - destruct f. - simpl; try destruct b; vm_compute; split; congruence. - simpl; try destruct b; vm_compute; split; congruence. - destruct n as [p Hp]. - simpl. rewrite Z.ltb_lt in Hp. - apply Zlt_succ_le with (m:=52) in Hp. - apply Zpower_le with (r:=radix2) in Hp. - edestruct Fcore_digits.digits2_Pnat_correct. - rewrite Zpower_nat_Z in H0. - eapply Z.lt_le_trans in Hp; eauto. - unfold join_bits; destruct b. - compute_this ((2 ^ 11 + 2047) * 2 ^ 52). smart_omega. - compute_this ((0 + 2047) * 2 ^ 52). smart_omega. - unfold bits_of_binary_float, join_bits. - destruct (andb_prop _ _ e0); apply Zle_bool_imp_le in H0; apply Zeq_bool_eq in H; unfold FLT_exp in H. - match goal with [H:Zmax ?x ?y = e|-_] => pose proof (Zle_max_l x y); pose proof (Zle_max_r x y) end. - rewrite H, Fcalc_digits.Z_of_nat_S_digits2_Pnat in *. - lapply (Fcalc_digits.Zpower_gt_Zdigits radix2 53 (Zpos m)). intro. - unfold radix2, radix_val, Zabs in H3. - pose proof (Zle_bool_spec (2 ^ 52) (Zpos m)). - assert (Zpos m > 0); [vm_compute; exact eq_refl|]. - compute_this (2^11); compute_this (2^(11-1)). - inversion H4; fold (2^52) in *; destruct H6; destruct b; now smart_omega. - change Fcalc_digits.radix2 with radix2 in H1; omega. -Qed. - -Theorem single_of_bits_of_single: - forall f, single_of_bits (bits_of_single f) = singleoffloat f. -Proof. - intros; unfold single_of_bits, bits_of_single, bits_of_b32, b32_of_bits. - rewrite Int.unsigned_repr, binary_float_of_bits_of_binary_float; [reflexivity|]. - destruct (binary32offloat f). - simpl; try destruct b; vm_compute; split; congruence. - simpl; try destruct b; vm_compute; split; congruence. - destruct n as [p Hp]. - simpl. rewrite Z.ltb_lt in Hp. - apply Zlt_succ_le with (m:=23) in Hp. - apply Zpower_le with (r:=radix2) in Hp. - edestruct Fcore_digits.digits2_Pnat_correct. - rewrite Zpower_nat_Z in H0. - eapply Z.lt_le_trans in Hp; eauto. - compute_this (radix2^23). - unfold join_bits; destruct b. - compute_this ((2 ^ 8 + 255) * 2 ^ 23). smart_omega. - compute_this ((0 + 255) * 2 ^ 23). smart_omega. - unfold bits_of_binary_float, join_bits. - destruct (andb_prop _ _ e0); apply Zle_bool_imp_le in H0; apply Zeq_bool_eq in H. - unfold FLT_exp in H. - match goal with [H:Zmax ?x ?y = e|-_] => pose proof (Zle_max_l x y); pose proof (Zle_max_r x y) end. - rewrite H, Fcalc_digits.Z_of_nat_S_digits2_Pnat in *. - lapply (Fcalc_digits.Zpower_gt_Zdigits radix2 24 (Zpos m)). intro. - unfold radix2, radix_val, Zabs in H3. - pose proof (Zle_bool_spec (2 ^ 23) (Zpos m)). - compute_this (2^23); compute_this (2^24); compute_this (2^8); compute_this (2^(8-1)). - assert (Zpos m > 0); [exact eq_refl|]. - inversion H4; destruct b; now smart_omega. - change Fcalc_digits.radix2 with radix2 in H1; omega. + generalize (bits_of_binary_float_range 52 11 __ __ f). + change (2^(52+11+1)) with (Int64.max_unsigned + 1). omega. Qed. -Theorem bits_of_singleoffloat: - forall f, bits_of_single (singleoffloat f) = bits_of_single f. +Theorem to_of_bits: + forall b, to_bits (of_bits b) = b. Proof. - intro; unfold singleoffloat, bits_of_single; rewrite binary32offloatofbinary32offloat; reflexivity. -Qed. - -Theorem singleoffloat_of_bits: - forall b, singleoffloat (single_of_bits b) = single_of_bits b. -Proof. - intro; unfold singleoffloat, single_of_bits; rewrite floatofbinary32offloatofbinary32; reflexivity. -Qed. - -Theorem single_of_bits_is_single: - forall b, is_single (single_of_bits b). -Proof. - intros. exists (b32_of_bits (Int.unsigned b)); auto. + intros; unfold of_bits, to_bits, bits_of_b64, b64_of_bits. + rewrite bits_of_binary_float_of_bits. apply Int64.repr_unsigned. + apply Int64.unsigned_range. Qed. (** Conversions between floats and unsigned ints can be defined @@ -740,279 +462,103 @@ Qed. Definition ox8000_0000 := Int.repr Int.half_modulus. (**r [0x8000_0000] *) -Lemma round_exact: - forall n, -2^53 < n < 2^53 -> - round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (Z2R n) = Z2R n. -Proof. - intros; rewrite round_generic; [reflexivity|now apply valid_rnd_round_mode|]. - apply generic_format_FLT; exists (Float radix2 n 0). - unfold F2R, Fnum, Fexp, bpow; rewrite Rmult_1_r; intuition. - pose proof (Zabs_spec n); now smart_omega. -Qed. - -Lemma binary_normalize64_exact: - forall n, -2^53 < n < 2^53 -> - B2R _ _ (binary_normalize64 n 0 false) = Z2R n /\ - is_finite _ _ (binary_normalize64 n 0 false) = true. -Proof. - intros; pose proof (binary_normalize64_correct n 0 false). - unfold F2R, Fnum, Fexp, bpow in H0; rewrite Rmult_1_r, round_exact, Rlt_bool_true in H0; try now intuition. - rewrite <- Z2R_abs; apply Z2R_lt; pose proof (Zabs_spec n); now smart_omega. -Qed. - -Theorem floatofintu_floatofint_1: +Theorem of_intu_of_int_1: forall x, Int.ltu x ox8000_0000 = true -> - floatofintu x = floatofint x. + of_intu x = of_int x. Proof. - unfold floatofintu, floatofint, Int.signed, Int.ltu; intro. + unfold of_intu, of_int, Int.signed, Int.ltu; intro. change (Int.unsigned ox8000_0000) with Int.half_modulus. destruct (zlt (Int.unsigned x) Int.half_modulus); now intuition. Qed. -Theorem floatofintu_floatofint_2: +Theorem of_intu_of_int_2: forall x, Int.ltu x ox8000_0000 = false -> - floatofintu x = add (floatofint (Int.sub x ox8000_0000)) - (floatofintu ox8000_0000). + of_intu x = add (of_int (Int.sub x ox8000_0000)) (of_intu ox8000_0000). Proof. - unfold floatofintu, floatofint, Int.signed, Int.ltu, Int.sub; intros. - pose proof (Int.unsigned_range x). - compute_this (Int.unsigned ox8000_0000). - destruct (zlt (Int.unsigned x) 2147483648); try discriminate. - rewrite Int.unsigned_repr by smart_omega. - destruct (zlt ((Int.unsigned x) - 2147483648) Int.half_modulus). - unfold add, b64_plus. - match goal with [|- _ = Bplus _ _ _ _ _ _ ?x ?y] => - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE x y) end. - do 2 rewrite (fun x H => proj1 (binary_normalize64_exact x H)) in H1 by smart_omega. - do 2 rewrite (fun x H => proj2 (binary_normalize64_exact x H)) in H1 by smart_omega. - rewrite <- Z2R_plus, round_exact in H1 by smart_omega. - rewrite Rlt_bool_true in H1; - replace (Int.unsigned x - 2147483648 + 2147483648) with (Int.unsigned x) in * by ring. - apply B2R_inj. - destruct (binary_normalize64_exact (Int.unsigned x)); [now smart_omega|]. - match goal with [|- _ _ _ ?f = _] => destruct f end; intuition. - exfalso; simpl in H2; change 0%R with (Z2R 0) in H2; apply eq_Z2R in H2; omega. - try (change (53 ?= 1024) with Lt in H1). (* for Coq 8.4 *) - simpl Zcompare in *. - match goal with [|- _ _ _ ?f = _] => destruct f end; intuition. - exfalso; simpl in H0; change 0%R with (Z2R 0) in H0; apply eq_Z2R in H0; omega. - rewrite (fun x H => proj1 (binary_normalize64_exact x H)) by smart_omega; now intuition. - rewrite <- Z2R_Zpower, <- Z2R_abs by omega; apply Z2R_lt; - pose proof (Zabs_spec (Int.unsigned x)); now smart_omega. - exfalso; now smart_omega. + unfold add, b64_plus, of_intu, of_int, b64_of_Z; intros. + set (y := Int.sub x ox8000_0000). + pose proof (Int.unsigned_range x); pose proof (Int.signed_range y). + assert (Ry: integer_representable 53 1024 (Int.signed y)). + { apply integer_representable_n; auto; smart_omega. } + assert (R8: integer_representable 53 1024 (Int.unsigned ox8000_0000)). + { apply integer_representable_2p with (p := 31);auto; smart_omega. } + rewrite BofZ_plus by auto. + f_equal. + unfold Int.ltu in H. destruct zlt in H; try discriminate. + unfold y, Int.sub. rewrite Int.signed_repr. omega. + compute_this (Int.unsigned ox8000_0000); smart_omega. Qed. -Lemma Zoffloat_correct: - forall f, - match Zoffloat f with - | Some n => - is_finite _ _ f = true /\ - Z2R n = round radix2 (FIX_exp 0) (round_mode mode_ZR) (B2R _ _ f) - | None => - is_finite _ _ f = false - end. -Proof. - destruct f; try now intuition. - simpl B2R. rewrite round_0. now intuition. now apply valid_rnd_round_mode. - destruct e. split. reflexivity. - rewrite round_generic. symmetry. now apply Rmult_1_r. - now apply valid_rnd_round_mode. - apply generic_format_FIX. exists (Float radix2 (cond_Zopp b (Zpos m)) 0). split; reflexivity. - split; [reflexivity|]. - rewrite round_generic, Z2R_mult, Z2R_Zpower_pos, <- bpow_powerRZ; - [reflexivity|now apply valid_rnd_round_mode|apply generic_format_F2R; discriminate]. - rewrite (inbetween_float_ZR_sign _ _ _ ((Zpos m) / Zpower_pos radix2 p) - (new_location (Zpower_pos radix2 p) (Zpos m mod Zpower_pos radix2 p) loc_Exact)). - unfold B2R, F2R, Fnum, Fexp, canonic_exp, bpow, FIX_exp, Zoffloat, radix2, radix_val. - pose proof (Rlt_bool_spec (Z2R (cond_Zopp b (Zpos m)) * / Z2R (Zpower_pos 2 p)) 0). - inversion H; rewrite <- (Rmult_0_l (bpow radix2 (Zneg p))) in H1. - apply Rmult_lt_reg_r in H1. apply (lt_Z2R _ 0) in H1. - destruct b; [split; [|ring_simplify];reflexivity|discriminate]. - now apply bpow_gt_0. - apply Rmult_le_reg_r in H1. apply (le_Z2R 0) in H1. - destruct b; [destruct H1|split; [|ring_simplify]]; reflexivity. - now apply (bpow_gt_0 radix2 (Zneg p)). - unfold canonic_exp, FIX_exp; replace 0 with (Zneg p + Zpos p) by apply Zplus_opp_r. - apply (inbetween_float_new_location radix2 _ _ _ _ (Zpos p)); [reflexivity|]. - apply inbetween_Exact; unfold B2R, F2R, Fnum, Fexp; destruct b. - rewrite Rabs_left; [simpl; ring_simplify; reflexivity|]. - replace 0%R with (0*(bpow radix2 (Zneg p)))%R by ring; apply Rmult_gt_compat_r. - now apply bpow_gt_0. - apply (Z2R_lt _ 0); reflexivity. - apply Rabs_right; replace 0%R with (0*(bpow radix2 (Zneg p)))%R by ring; apply Rgt_ge. - apply Rmult_gt_compat_r; [now apply bpow_gt_0|apply (Z2R_lt 0); reflexivity]. -Qed. - -Theorem intoffloat_correct: - forall f, - match intoffloat f with - | Some n => - is_finite _ _ f = true /\ - Z2R (Int.signed n) = round radix2 (FIX_exp 0) (round_mode mode_ZR) (B2R _ _ f) - | None => - is_finite _ _ f = false \/ - (B2R _ _ f <= Z2R (Zpred Int.min_signed)\/ - Z2R (Zsucc Int.max_signed) <= B2R _ _ f)%R - end. -Proof. - intro; pose proof (Zoffloat_correct f); unfold intoffloat; destruct (Zoffloat f). - pose proof (Zle_bool_spec Int.min_signed z); pose proof (Zle_bool_spec z Int.max_signed). - compute_this Int.min_signed; compute_this Int.max_signed; destruct H. - inversion H0; [inversion H1|]. - rewrite <- (Int.signed_repr z) in H2 by smart_omega; split; assumption. - right; right; eapply Rle_trans; [apply Z2R_le; apply Zlt_le_succ; now apply H6|]. - rewrite H2, round_ZR_pos. - unfold round, scaled_mantissa, canonic_exp, FIX_exp, F2R, Fnum, Fexp; simpl bpow. - do 2 rewrite Rmult_1_r; now apply Zfloor_lb. - apply Rnot_lt_le; intro; apply Rlt_le in H7; apply (round_le radix2 (FIX_exp 0) (round_mode mode_ZR)) in H7; - rewrite <- H2, round_0 in H7; [apply (le_Z2R _ 0) in H7; now smart_omega|now apply valid_rnd_round_mode]. - right; left; eapply Rle_trans; [|apply (Z2R_le z); simpl; omega]. - rewrite H2, round_ZR_neg. - unfold round, scaled_mantissa, canonic_exp, FIX_exp, F2R, Fnum, Fexp; simpl bpow. - do 2 rewrite Rmult_1_r; now apply Zceil_ub. - apply Rnot_lt_le; intro; apply Rlt_le in H5; apply (round_le radix2 (FIX_exp 0) (round_mode mode_ZR)) in H5. - rewrite <- H2, round_0 in H5; [apply (le_Z2R 0) in H5; omega|now apply valid_rnd_round_mode]. - left; assumption. -Qed. - -Theorem intuoffloat_correct: - forall f, - match intuoffloat f with - | Some n => - is_finite _ _ f = true /\ - Z2R (Int.unsigned n) = round radix2 (FIX_exp 0) (round_mode mode_ZR) (B2R _ _ f) - | None => - is_finite _ _ f = false \/ - (B2R _ _ f <= -1 \/ - Z2R (Zsucc Int.max_unsigned) <= B2R _ _ f)%R - end. -Proof. - intro; pose proof (Zoffloat_correct f); unfold intuoffloat; destruct (Zoffloat f). - pose proof (Zle_bool_spec 0 z); pose proof (Zle_bool_spec z Int.max_unsigned). - compute_this Int.max_unsigned; destruct H. - inversion H0. inversion H1. - rewrite <- (Int.unsigned_repr z) in H2 by smart_omega; split; assumption. - right; right; eapply Rle_trans; [apply Z2R_le; apply Zlt_le_succ; now apply H6|]. - rewrite H2, round_ZR_pos. - unfold round, scaled_mantissa, canonic_exp, FIX_exp, F2R, Fnum, Fexp; simpl bpow; - do 2 rewrite Rmult_1_r; now apply Zfloor_lb. - apply Rnot_lt_le; intro; apply Rlt_le in H7; eapply (round_le radix2 (FIX_exp 0) (round_mode mode_ZR)) in H7; - rewrite <- H2, round_0 in H7; [apply (le_Z2R _ 0) in H7; now smart_omega|now apply valid_rnd_round_mode]. - right; left; eapply Rle_trans; [|change (-1)%R with (Z2R (-1)); apply (Z2R_le z); omega]. - rewrite H2, round_ZR_neg; unfold round, scaled_mantissa, canonic_exp, FIX_exp, F2R, Fnum, Fexp; simpl bpow. - do 2 rewrite Rmult_1_r; now apply Zceil_ub. - apply Rnot_lt_le; intro; apply Rlt_le in H5; apply (round_le radix2 (FIX_exp 0) (round_mode mode_ZR)) in H5; - rewrite <- H2, round_0 in H5; [apply (le_Z2R 0) in H5; omega|now apply valid_rnd_round_mode]. - left; assumption. -Qed. - -Lemma intuoffloat_interval: - forall f n, - intuoffloat f = Some n -> - (-1 < B2R _ _ f < Z2R (Zsucc Int.max_unsigned))%R. -Proof. - intro; pose proof (intuoffloat_correct f); destruct (intuoffloat f); try discriminate; destruct H. - destruct f; try discriminate; intros. - simpl B2R; change 0%R with (Z2R 0); change (-1)%R with (Z2R (-1)); split; apply Z2R_lt; reflexivity. - pose proof (Int.unsigned_range i). - unfold round, scaled_mantissa, B2R, F2R, Fnum, Fexp in H0 |- *; simpl bpow in H0; do 2 rewrite Rmult_1_r in H0; - apply eq_Z2R in H0. - split; apply Rnot_le_lt; intro. - rewrite Ztrunc_ceil in H0; - [apply Zceil_le in H3; change (-1)%R with (Z2R (-1)) in H3; rewrite Zceil_Z2R in H3; omega|]. - eapply Rle_trans; [now apply H3|apply (Z2R_le (-1) 0); discriminate]. - rewrite Ztrunc_floor in H0; [apply Zfloor_le in H3; rewrite Zfloor_Z2R in H3; now smart_omega|]. - eapply Rle_trans; [|now apply H3]; apply (Z2R_le 0); discriminate. -Qed. - -Theorem intuoffloat_intoffloat_1: +Theorem to_intu_to_int_1: forall x n, - cmp Clt x (floatofintu ox8000_0000) = true -> - intuoffloat x = Some n -> - intoffloat x = Some n. -Proof. - intros; unfold cmp in H; pose proof (order_float_finite_correct x (floatofintu ox8000_0000)). - destruct (order_float x (floatofintu ox8000_0000)); try destruct c; try discriminate. - pose proof (intuoffloat_correct x); rewrite H0 in H2; destruct H2. - specialize (H1 H2 eq_refl); pose proof (intoffloat_correct x); destruct (intoffloat x). - f_equal; rewrite <- (proj2 H4) in H3; apply eq_Z2R in H3. - pose proof (eq_refl (Int.repr (Int.unsigned n))); rewrite H3 in H5 at 1. - rewrite Int.repr_signed, Int.repr_unsigned in H5; assumption. - destruct H4; [rewrite H2 in H4; discriminate|]. - apply intuoffloat_interval in H0; exfalso; destruct H0, H4. - eapply Rlt_le_trans in H0; [|now apply H4]; apply (lt_Z2R (-1)) in H0; discriminate. - apply Rcompare_Lt_inv in H1; eapply Rle_lt_trans in H1; [|now apply H4]. - unfold floatofintu in H1; rewrite (fun x H => proj1 (binary_normalize64_exact x H)) in H1; - [apply lt_Z2R in H1; discriminate|split; reflexivity]. -Qed. - -Lemma Zfloor_minus : - forall x n, Zfloor(x-Z2R n) = Zfloor(x)-n. -Proof. - intros; apply Zfloor_imp; replace (Zfloor x - n + 1) with (Zfloor x + 1 - n) by ring; do 2 rewrite Z2R_minus. - split; - [apply Rplus_le_compat_r; now apply Zfloor_lb| - apply Rplus_lt_compat_r; rewrite Z2R_plus; now apply Zfloor_ub]. -Qed. - -Theorem intuoffloat_intoffloat_2: + cmp Clt x (of_intu ox8000_0000) = true -> + to_intu x = Some n -> + to_int x = Some n. +Proof. + intros. unfold to_intu in H0. + destruct (b64_to_Z_range x 0 Int.max_unsigned) as [p|] eqn:E; simpl in H0; inv H0. + unfold b64_to_Z_range in E. exploit ZofB_range_inversion; eauto. intros (A & B & C). + unfold to_int, b64_to_Z_range. unfold ZofB_range. rewrite C. + rewrite Zle_bool_true by smart_omega. rewrite Zle_bool_true; auto. + exploit (BofZ_exact 53 1024 __ __ (Int.unsigned ox8000_0000)). + vm_compute; intuition congruence. + set (y := of_intu ox8000_0000) in *. + change (BofZ 53 1024 eq_refl eq_refl (Int.unsigned ox8000_0000)) with y. + intros (EQy & FINy & SIGNy). + assert (FINx: is_finite _ _ x = true). + { rewrite ZofB_correct in C. destruct (is_finite _ _ x) eqn:FINx; congruence. } + destruct (zeq p 0). + subst p; smart_omega. + destruct (ZofB_range_pos 53 1024 __ __ x p C) as [P Q]. omega. + assert (CMP: b64_compare x y = Some Lt). + { unfold cmp, cmp_of_comparison in H. destruct (b64_compare x y) as [[]|]; auto; discriminate. } + unfold b64_compare in CMP. rewrite Bcompare_finite_correct in CMP by auto. + inv CMP. apply Rcompare_Lt_inv in H1. rewrite EQy in H1. + assert (p < Int.unsigned ox8000_0000). + { apply lt_Z2R. eapply Rle_lt_trans; eauto. } + change Int.max_signed with (Int.unsigned ox8000_0000 - 1). omega. +Qed. + +Theorem to_intu_to_int_2: forall x n, - cmp Clt x (floatofintu ox8000_0000) = false -> - intuoffloat x = Some n -> - intoffloat (sub x (floatofintu ox8000_0000)) = Some (Int.sub n ox8000_0000). -Proof. - assert (B2R _ _ (floatofintu ox8000_0000) = Z2R (Int.unsigned ox8000_0000)). - apply (fun x H => proj1 (binary_normalize64_exact x H)); split; reflexivity. - intros; unfold cmp in H0; pose proof (order_float_finite_correct x (floatofintu ox8000_0000)). - destruct (order_float x (floatofintu ox8000_0000)); try destruct c; try discriminate; - pose proof (intuoffloat_correct x); rewrite H1 in H3; destruct H3; specialize (H2 H3 eq_refl). - apply Rcompare_Eq_inv in H2; apply B2R_inj in H2. - subst x; vm_compute in H1; injection H1; intro; subst n; vm_compute; reflexivity. - destruct x; try discriminate H3; - [rewrite H in H2; simpl B2R in H2; apply (eq_Z2R 0) in H2; discriminate|reflexivity]. - reflexivity. - rewrite H in H2; apply Rcompare_Gt_inv in H2; pose proof (intuoffloat_interval _ _ H1). - unfold sub, b64_minus. - exploit (Bminus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE x (floatofintu ox8000_0000)); [assumption|reflexivity|]; intro. - rewrite H, round_generic in H6. - match goal with [H6:if Rlt_bool ?x ?y then _ else _|-_] => - pose proof (Rlt_bool_spec x y); destruct (Rlt_bool x y) end. - destruct H6 as [? []]. - match goal with [|- _ ?y = _] => pose proof (intoffloat_correct y); destruct (intoffloat y) end. - destruct H10. - f_equal; rewrite <- (Int.repr_signed i); unfold Int.sub; f_equal; apply eq_Z2R. - rewrite Z2R_minus, H11, H4. - unfold round, scaled_mantissa, F2R, Fexp, Fnum, round_mode; simpl bpow; repeat rewrite Rmult_1_r; - rewrite <- Z2R_minus; f_equal. - rewrite (Ztrunc_floor (B2R _ _ x)), <- Zfloor_minus, <- Ztrunc_floor; - [f_equal; assumption|apply Rle_0_minus; left; assumption|]. - left; eapply Rlt_trans; [|now apply H2]; apply (Z2R_lt 0); reflexivity. - try (change (0 ?= 53) with Lt in H6,H8). (* for Coq 8.4 *) - try (change (53 ?= 1024) with Lt in H6,H8). (* for Coq 8.4 *) - exfalso; simpl Zcompare in H6, H8; rewrite H6, H8 in H10. - destruct H10 as [|[]]; [discriminate|..]. - eapply Rle_trans in H10; [|apply Rle_0_minus; left; assumption]; apply (le_Z2R 0) in H10; apply H10; reflexivity. - eapply Rle_lt_trans in H10; [|apply Rplus_lt_compat_r; now apply (proj2 H5)]. - rewrite <- Z2R_opp, <- Z2R_plus in H10; apply lt_Z2R in H10; discriminate. - exfalso; inversion H7; rewrite Rabs_right in H8. - eapply Rle_lt_trans in H8. apply Rle_not_lt in H8; [assumption|apply (bpow_le _ 31); discriminate]. - change (bpow radix2 31) with (Z2R(Zsucc Int.max_unsigned - Int.unsigned ox8000_0000)); rewrite Z2R_minus. - apply Rplus_lt_compat_r; exact (proj2 H5). - apply Rle_ge; apply Rle_0_minus; left; assumption. - now apply valid_rnd_round_mode. - apply Fprop_Sterbenz.sterbenz_aux; [now apply fexp_monotone|now apply generic_format_B2R| |]. - rewrite <- H; now apply generic_format_B2R. - destruct H5; split; left; assumption. - now destruct H2. + cmp Clt x (of_intu ox8000_0000) = false -> + to_intu x = Some n -> + to_int (sub x (of_intu ox8000_0000)) = Some (Int.sub n ox8000_0000). +Proof. + intros. unfold to_intu in H0. + destruct (b64_to_Z_range x 0 Int.max_unsigned) as [p|] eqn:E; simpl in H0; inv H0. + unfold b64_to_Z_range in E. exploit ZofB_range_inversion; eauto. intros (A & B & C). + exploit (BofZ_exact 53 1024 __ __ (Int.unsigned ox8000_0000)). + vm_compute; intuition congruence. + set (y := of_intu ox8000_0000) in *. + change (BofZ 53 1024 __ __ (Int.unsigned ox8000_0000)) with y. + intros (EQy & FINy & SIGNy). + assert (FINx: is_finite _ _ x = true). + { rewrite ZofB_correct in C. destruct (is_finite _ _ x) eqn:FINx; congruence. } + assert (GE: (B2R _ _ x >= Z2R (Int.unsigned ox8000_0000))%R). + { rewrite <- EQy. unfold cmp, cmp_of_comparison, b64_compare in H. + rewrite Bcompare_finite_correct in H by auto. + destruct (Rcompare (B2R 53 1024 x) (B2R 53 1024 y)) eqn:CMP. + apply Req_ge; apply Rcompare_Eq_inv; auto. + discriminate. + apply Rgt_ge; apply Rcompare_Gt_inv; auto. + } + assert (EQ: b64_to_Z_range (sub x y) Int.min_signed Int.max_signed = Some (p - Int.unsigned ox8000_0000)). + { + apply ZofB_range_minus. exact E. + compute_this (Int.unsigned ox8000_0000). smart_omega. + apply Rge_le; auto. + } + unfold to_int; rewrite EQ. simpl. f_equal. unfold Int.sub. f_equal. f_equal. + symmetry; apply Int.unsigned_repr. omega. Qed. (** Conversions from ints to floats can be defined as bitwise manipulations over the in-memory representation. This is what the PowerPC port does. The trick is that [from_words 0x4330_0000 x] is the float - [2^52 + floatofintu x]. *) + [2^52 + of_intu x]. *) Definition ox4330_0000 := Int.repr 1127219200. (**r [0x4330_0000] *) @@ -1032,55 +578,42 @@ Qed. Lemma from_words_value: forall x, - B2R _ _ (from_words ox4330_0000 x) = - (bpow radix2 52 + Z2R (Int.unsigned x))%R /\ - is_finite _ _ (from_words ox4330_0000 x) = true. + B2R _ _ (from_words ox4330_0000 x) = (bpow radix2 52 + Z2R (Int.unsigned x))%R + /\ is_finite _ _ (from_words ox4330_0000 x) = true + /\ Bsign _ _ (from_words ox4330_0000 x) = false. Proof. - intros; unfold from_words, double_of_bits, b64_of_bits, binary_float_of_bits. - rewrite B2R_FF2B. rewrite is_finite_FF2B. + intros; unfold from_words, of_bits, b64_of_bits, binary_float_of_bits. + rewrite B2R_FF2B, is_finite_FF2B, Bsign_FF2B. unfold binary_float_of_bits_aux; rewrite split_bits_or; simpl; pose proof (Int.unsigned_range x). destruct (Int.unsigned x + Zpower_pos 2 52) eqn:?. exfalso; now smart_omega. - simpl; rewrite <- Heqz; unfold F2R; simpl. - rewrite <- (Z2R_plus 4503599627370496), Rmult_1_r. - split; [f_equal; compute_this (Zpower_pos 2 52); ring | reflexivity]. - assert (Zneg p < 0) by reflexivity. + simpl; rewrite <- Heqz; unfold F2R; simpl. split; auto. + rewrite <- (Z2R_plus 4503599627370496), Rmult_1_r. f_equal. rewrite Zplus_comm. auto. exfalso; now smart_omega. Qed. -Theorem floatofintu_from_words: - forall x, - floatofintu x = - sub (from_words ox4330_0000 x) (from_words ox4330_0000 Int.zero). +Lemma from_words_eq: + forall x, from_words ox4330_0000 x = BofZ 53 1024 __ __ (2^52 + Int.unsigned x). Proof. - intros; destruct (Int.eq_dec x Int.zero); [subst; vm_compute; reflexivity|]. - assert (Int.unsigned x <> 0). - intro; destruct n; rewrite <- (Int.repr_unsigned x), H; reflexivity. + intros. pose proof (Int.unsigned_range x). - pose proof (binary_normalize64_exact (Int.unsigned x)). destruct H1; [smart_omega|]. - unfold floatofintu, sub, b64_minus. - match goal with [|- _ = Bminus _ _ _ _ _ _ ?x ?y] => - pose proof (Bminus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE x y) end. - apply (fun f x y => f x y) in H3; try apply (fun x => proj2 (from_words_value x)). - do 2 rewrite (fun x => proj1 (from_words_value x)) in H3. - rewrite Int.unsigned_zero in H3. - replace (bpow radix2 52 + Z2R (Int.unsigned x) - - (bpow radix2 52 + Z2R 0))%R with (Z2R (Int.unsigned x)) in H3 by (simpl; ring). - rewrite round_exact in H3 by smart_omega. - match goal with [H3:if Rlt_bool ?x ?y then _ else _ |- _] => - pose proof (Rlt_bool_spec x y); destruct (Rlt_bool x y) end; destruct H3 as [? []]. - try (change (53 ?= 1024) with Lt in H3,H5). (* for Coq 8.4 *) - simpl Zcompare in *; apply B2R_inj; - try match goal with [H':B2R _ _ ?f = _ , H'':is_finite _ _ ?f = true |- is_finite_strict _ _ ?f = true] => - destruct f; [ - simpl in H'; change 0%R with (Z2R 0) in H'; apply eq_Z2R in H'; now destruct (H (eq_sym H')) | - discriminate H'' | discriminate H'' | reflexivity - ] - end. - rewrite H3; assumption. - inversion H4; change (bpow radix2 1024) with (Z2R (radix2 ^ 1024)) in H5; rewrite <- Z2R_abs in H5. - apply le_Z2R in H5; pose proof (Zabs_spec (Int.unsigned x)); - exfalso; now smart_omega. + destruct (from_words_value x) as (A & B & C). + destruct (BofZ_exact 53 1024 __ __ (2^52 + Int.unsigned x)) as (D & E & F). + smart_omega. + apply B2R_Bsign_inj; auto. + rewrite A, D. rewrite Z2R_plus. auto. + rewrite C, F. symmetry. apply Zlt_bool_false. smart_omega. +Qed. + +Theorem of_intu_from_words: + forall x, + of_intu x = sub (from_words ox4330_0000 x) (from_words ox4330_0000 Int.zero). +Proof. + intros. pose proof (Int.unsigned_range x). + rewrite ! from_words_eq. unfold sub, b64_minus. rewrite BofZ_minus. + unfold of_intu, b64_of_Z. f_equal. rewrite Int.unsigned_zero. omega. + apply integer_representable_n; auto; smart_omega. + apply integer_representable_n; auto; rewrite Int.unsigned_zero; smart_omega. Qed. Lemma ox8000_0000_signed_unsigned: @@ -1095,337 +628,19 @@ Proof. apply Int.eqm_add; [now apply Int.eqm_refl|exists 1;reflexivity]. Qed. -Theorem floatofint_from_words: +Theorem of_int_from_words: forall x, - floatofint x = - sub (from_words ox4330_0000 (Int.add x ox8000_0000)) - (from_words ox4330_0000 ox8000_0000). -Proof. -Local Transparent Int.repr Int64.repr. - intros; destruct (Int.eq_dec x Int.zero); [subst; vm_compute; reflexivity|]. - assert (Int.signed x <> 0). - intro; destruct n; rewrite <- (Int.repr_signed x), H; reflexivity. - pose proof (Int.signed_range x). - pose proof (binary_normalize64_exact (Int.signed x)); destruct H1; [now smart_omega|]. - unfold floatofint, sub, b64_minus. - match goal with [|- _ = Bminus _ _ _ _ _ _ ?x ?y] => - pose proof (Bminus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE x y) end. - apply (fun f x y => f x y) in H3; try apply (fun x => proj2 (from_words_value x)). - do 2 rewrite (fun x => proj1 (from_words_value x)) in H3. - replace (bpow radix2 52 + Z2R (Int.unsigned (Int.add x ox8000_0000)) - - (bpow radix2 52 + Z2R (Int.unsigned ox8000_0000)))%R with (Z2R (Int.signed x)) in H3 - by (rewrite ox8000_0000_signed_unsigned; rewrite Z2R_plus; simpl; ring). - rewrite round_exact in H3 by smart_omega. - match goal with [H3:if Rlt_bool ?x ?y then _ else _ |- _] => - pose proof (Rlt_bool_spec x y); destruct (Rlt_bool x y) end; destruct H3 as [? []]. - try (change (0 ?= 53) with Lt in H3,H5). (* for Coq 8.4 *) - try (change (53 ?= 1024) with Lt in H3,H5). (* for Coq 8.4 *) - simpl Zcompare in *; apply B2R_inj; - try match goal with [H':B2R _ _ ?f = _ , H'':is_finite _ _ ?f = true |- is_finite_strict _ _ ?f = true] => - destruct f; [ - simpl in H'; change 0%R with (Z2R 0) in H'; apply eq_Z2R in H'; now destruct (H (eq_sym H')) | - discriminate H'' | discriminate H'' | reflexivity - ] - end. - rewrite H3; assumption. - inversion H4; unfold bpow in H5; rewrite <- Z2R_abs in H5; - apply le_Z2R in H5; pose proof (Zabs_spec (Int.signed x)); exfalso; now smart_omega. -Qed. - -(** Conversions from 32-bit integers to single-precision floats can - be decomposed into a conversion to a double-precision float, - followed by a [singleoffloat] normalization. No double rounding occurs. *) - -Lemma is_finite_strict_ge_1: - forall (f: binary32), - is_finite _ _ f = true -> - (1 <= Rabs (B2R _ _ f))%R -> - is_finite_strict _ _ f = true. -Proof. - intros. destruct f; auto. simpl in H0. - change 0%R with (Z2R 0) in H0. - change 1%R with (Z2R 1) in H0. - rewrite <- Z2R_abs in H0. - exploit le_Z2R; eauto. -Qed. - -Lemma single_float_of_int: - forall n, - -2^53 < n < 2^53 -> - singleoffloat (binary_normalize64 n 0 false) = floatofbinary32 (binary_normalize32 n 0 false). -Proof. - intros. unfold singleoffloat. f_equal. - assert (EITHER: n = 0 \/ Z.abs n > 0) by (destruct n; compute; auto). - destruct EITHER as [EQ|GT]. - subst n; reflexivity. - exploit binary_normalize64_exact; eauto. intros [A B]. - destruct (binary_normalize64 n 0 false) as [ | | | s m e] eqn:B64; simpl in *. -- assert (0 = n) by (apply eq_Z2R; auto). subst n. simpl in GT. omegaContradiction. -- discriminate. -- discriminate. -- set (n1 := cond_Zopp s (Z.pos m)) in *. - generalize (binary_normalize32_correct n1 e s). - fold (binary_normalize32 n1 e s). intros C. - generalize (binary_normalize32_correct n 0 false). - fold (binary_normalize32 n 0 false). intros D. - assert (A': @F2R radix2 {| Fnum := n; Fexp := 0 |} = Z2R n). - { unfold F2R. apply Rmult_1_r. } - rewrite A in C. rewrite A' in D. - destruct (Rlt_bool - (Rabs - (round radix2 (FLT_exp (3 - 128 - 24) 24) (round_mode mode_NE) - (Z2R n))) (bpow radix2 128)). -+ destruct C as [C1 [C2 _]]; destruct D as [D1 [D2 _]]. - assert (1 <= Rabs (round radix2 (FLT_exp (3 - 128 - 24) 24) (round_mode mode_NE) (Z2R n)))%R. - { apply abs_round_ge_generic. - apply fexp_correct. red. omega. - apply valid_rnd_round_mode. - apply generic_format_bpow with (e := 0). compute. congruence. - rewrite <- Z2R_abs. change 1%R with (Z2R 1). apply Z2R_le. omega. } - apply B2R_inj. - apply is_finite_strict_ge_1; auto. rewrite C1; auto. - apply is_finite_strict_ge_1; auto. rewrite D1; auto. - congruence. -+ apply B2FF_inj. congruence. -Qed. - -Theorem singleofint_floatofint: - forall n, singleofint n = singleoffloat (floatofint n). -Proof. - intros. symmetry. apply single_float_of_int. - generalize (Int.signed_range n). smart_omega. -Qed. - -Theorem singleofintu_floatofintu: - forall n, singleofintu n = singleoffloat (floatofintu n). -Proof. - intros. symmetry. apply single_float_of_int. - generalize (Int.unsigned_range n). smart_omega. -Qed. - -Theorem mul2_add: - forall f, add f f = mul f (floatofint (Int.repr 2%Z)). -Proof. - intros. unfold add, b64_plus, mul, b64_mult. - destruct (is_finite_strict _ _ f) eqn:EQFINST. - - assert (EQFIN:is_finite _ _ f = true) by (destruct f; simpl in *; congruence). - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE f f EQFIN EQFIN). - pose proof (Bmult_correct 53 1024 eq_refl eq_refl binop_pl mode_NE f - (floatofint (Int.repr 2%Z))). - rewrite <- double, Rmult_comm in H. - replace (B2R 53 1024 (floatofint (Int.repr 2))) with 2%R in H0 by (compute; field). - destruct Rlt_bool. - + destruct H0 as [? []], H as [? []]. - rewrite EQFIN in H1. - apply B2R_Bsign_inj; auto. - etransitivity. apply H. symmetry. apply H0. - etransitivity. apply H4. symmetry. etransitivity. apply H2. - destruct Bmult; try reflexivity; discriminate. - simpl. rewrite xorb_false_r. - erewrite <- Rmult_0_l, Rcompare_mult_r. - destruct f; try discriminate EQFINST. - simpl. unfold F2R. - erewrite <- Rmult_0_l, Rcompare_mult_r. - rewrite Rcompare_Z2R with (y:=0). - destruct b; reflexivity. - apply bpow_gt_0. - apply (Z2R_lt 0 2). omega. - + destruct H. - apply B2FF_inj. - etransitivity. apply H. - symmetry. etransitivity. apply H0. - f_equal. destruct Bsign; reflexivity. - - destruct f as [[]|[]| |]; try discriminate; simpl. - auto. auto. auto. auto. - destruct (Archi.choose_binop_pl b n b n); auto. -Qed. - -Program Definition pow2_float (b:bool) (e:Z) (H:-1023 < e < 1023) : float := - B754_finite _ _ b (nat_iter 52 xO xH) (e-52) _. -Next Obligation. - unfold Fappli_IEEE.bounded, canonic_mantissa. - rewrite andb_true_iff, Zle_bool_true by omega. split; auto. - apply Zeq_bool_true. unfold FLT_exp. simpl Z.of_nat. - apply Z.max_case_strong; omega. -Qed. - -Theorem mul_div_pow2: - forall b e f H H', - mul f (pow2_float b e H) = div f (pow2_float b (-e) H'). -Proof. - intros. unfold mul, b64_mult, div, b64_div. - pose proof (Bmult_correct 53 1024 eq_refl eq_refl binop_pl mode_NE f (pow2_float b e H)). - pose proof (Bdiv_correct 53 1024 eq_refl eq_refl binop_pl mode_NE f (pow2_float b (-e) H')). - lapply H1. clear H1. intro. - change (is_finite 53 1024 (pow2_float b e H)) with true in H0. - unfold Rdiv in H1. - replace (/ B2R 53 1024 (pow2_float b (-e) H'))%R - with (B2R 53 1024 (pow2_float b e H)) in H1. - destruct (is_finite _ _ f) eqn:EQFIN. - - destruct Rlt_bool. - + destruct H0 as [? []], H1 as [? []]. - apply B2R_Bsign_inj; auto. - etransitivity. apply H0. symmetry. apply H1. - etransitivity. apply H3. destruct Bmult; try discriminate H2; reflexivity. - symmetry. etransitivity. apply H5. destruct Bdiv; try discriminate H4; reflexivity. - reflexivity. - + apply B2FF_inj. - etransitivity. apply H0. symmetry. etransitivity. apply H1. - reflexivity. - - destruct f; try discriminate EQFIN; auto. - - simpl. - assert ((4503599627370496 * bpow radix2 (e - 52))%R = - (/ (4503599627370496 * bpow radix2 (- e - 52)))%R). - { etransitivity. symmetry. apply (bpow_plus radix2 52). - symmetry. etransitivity. apply f_equal. symmetry. apply (bpow_plus radix2 52). - rewrite <- bpow_opp. f_equal. ring. } - destruct b. unfold cond_Zopp. - rewrite !F2R_Zopp, <- Ropp_inv_permute. f_equal. auto. - intro. apply F2R_eq_0_reg in H3. omega. - apply H2. - - simpl. intro. apply F2R_eq_0_reg in H2. - destruct b; simpl in H2; omega. -Qed. - -Definition exact_inverse_mantissa := nat_iter 52 xO xH. - -Program Definition exact_inverse (f: float) : option float := - match f with - | B754_finite s m e B => - if peq m exact_inverse_mantissa then - if zlt (-1023) (e + 52) then - if zlt (e + 52) 1023 then - Some(B754_finite _ _ s m (-e - 104) _) - else None else None else None - | _ => None - end. -Next Obligation. - unfold Fappli_IEEE.bounded, canonic_mantissa. apply andb_true_iff; split. - simpl Z.of_nat. apply Zeq_bool_true. unfold FLT_exp. apply Z.max_case_strong; omega. - apply Zle_bool_true. omega. -Qed. - -Remark B754_finite_eq: - forall s1 m1 e1 B1 s2 m2 e2 B2, - s1 = s2 -> m1 = m2 -> e1 = e2 -> - B754_finite _ _ s1 m1 e1 B1 = (B754_finite _ _ s2 m2 e2 B2 : float). + of_int x = sub (from_words ox4330_0000 (Int.add x ox8000_0000)) + (from_words ox4330_0000 ox8000_0000). Proof. - intros. subst. f_equal. apply proof_irrelevance. -Qed. - -Theorem div_mul_inverse: - forall x y z, exact_inverse y = Some z -> div x y = mul x z. -Proof with (try discriminate). - unfold exact_inverse; intros. destruct y... - destruct (peq m exact_inverse_mantissa)... - destruct (zlt (-1023) (e + 52))... - destruct (zlt (e + 52) 1023)... - inv H. - set (n := - e - 52). - assert (RNG1: -1023 < n < 1023) by (unfold n; omega). - assert (RNG2: -1023 < -n < 1023) by (unfold n; omega). - symmetry. - transitivity (mul x (pow2_float b n RNG1)). - f_equal. apply B754_finite_eq; auto. unfold n; omega. - transitivity (div x (pow2_float b (-n) RNG2)). - apply mul_div_pow2. - f_equal. apply B754_finite_eq; auto. unfold n; omega. -Qed. - -Theorem floatoflongu_decomp: - forall l, floatoflongu l = - add (mul (floatofintu (Int64.hiword l)) (pow2_float false 32 (conj eq_refl eq_refl))) - (floatofintu (Int64.loword l)). -Proof. - intros. - unfold floatofintu. - pose proof (Int.unsigned_range (Int64.loword l)). - pose proof (Int.unsigned_range (Int64.hiword l)). - pose proof (Int64.unsigned_range l). - compute_this Int.modulus. - destruct (binary_normalize64_exact (Int.unsigned (Int64.loword l))); - [compute_this (2 ^ 53); omega|]. - destruct (binary_normalize64_exact (Int.unsigned (Int64.hiword l))); - [compute_this (2 ^ 53); omega|]. - unfold mul, b64_mult. - pose proof (Bmult_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (binary_normalize64 (Int.unsigned (Int64.hiword l)) 0 false) - (pow2_float false 32 (conj eq_refl eq_refl))). - rewrite H4 in H6. - replace (B2R 53 1024 (pow2_float false 32 (conj eq_refl eq_refl))) - with (Z2R 4294967296)%R in H6 by (compute; field). - rewrite <- Z2R_mult in H6. - rewrite round_generic in H6. - - rewrite Rlt_bool_true in H6. - + rewrite H5 in H6. - destruct H6 as [? [? ?]]. - { unfold add, b64_plus. - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (Bmult 53 1024 eq_refl eq_refl binop_pl mode_NE - (binary_normalize64 (Int.unsigned (Int64.hiword l)) 0 false) - (pow2_float false 32 (conj eq_refl eq_refl))) - (binary_normalize64 (Int.unsigned (Int64.loword l)) 0 false) H7 H3). - rewrite H6, H2, <- Z2R_plus in H9. - change 4294967296 with (two_p 32) in H9. - rewrite <- Int64.ofwords_add', Int64.ofwords_recompose in H9. - assert (Rabs (round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (Z2R (Int64.unsigned l))) < - bpow radix2 1024)%R. - { rewrite <- round_NE_abs by (apply fexp_correct; reflexivity). - eapply Rle_lt_trans with (Z2R (two_p 64)). 2:apply Z2R_lt; reflexivity. - erewrite <- round_generic. - - apply round_le. apply fexp_correct; reflexivity. - apply (valid_rnd_round_mode mode_NE). - rewrite <- Z2R_abs. apply Z2R_le. change (two_p 64) with Int64.modulus. zify; omega. - - apply (valid_rnd_round_mode mode_NE). - - apply (generic_format_bpow radix2 _ 64). compute. discriminate. } - rewrite Rlt_bool_true in H9 by auto. - unfold floatoflongu, binary_normalize64. - pose proof (binary_normalize64_correct (Int64.unsigned l) 0 false). - replace (F2R (beta:=radix2) {| Fnum := Int64.unsigned l; Fexp := 0 |}) - with (Z2R (Int64.unsigned l)) in H11 - by (unfold F2R, Fexp, Fnum, bpow; field). - rewrite Rlt_bool_true in H11 by auto. - destruct (Int64.eq_dec l Int64.zero). subst. reflexivity. - destruct H11, H9. - assert (1 <= round radix2 (FLT_exp (3 - 1024 - 53) 53) (round_mode mode_NE) - (Z2R (Int64.unsigned l)))%R. - { erewrite <- round_generic with (x:=1%R). - apply round_le. apply fexp_correct. reflexivity. apply valid_rnd_round_mode. - assert (Int64.unsigned l <> 0). - { contradict n. rewrite <- (Int64.repr_unsigned l), n. auto. } - apply (Z2R_le 1). omega. - apply valid_rnd_round_mode. - apply (generic_format_bpow _ _ 0). compute. discriminate. } - unfold binary_normalize64 in *. - apply B2R_inj. - + destruct H12, (binary_normalize 53 1024 eq_refl eq_refl mode_NE (Int64.unsigned l) 0 false); try discriminate. - unfold B2R in H11. rewrite <- H11 in H14. apply (le_Z2R 1 0) in H14. omega. - auto. - + destruct H13; match goal with Hf0:is_finite _ _ ?f0 = true, - Hf1:B2R _ _ ?f1 = _ |- - is_finite_strict _ _ ?f = true => - change f0 with f in Hf0; change f1 with f in Hf1; - destruct f - end; try discriminate. - unfold B2R in H9. rewrite <- H9 in H14. apply (le_Z2R 1 0) in H14. omega. - auto. - + rewrite H11. symmetry. apply H9. } - + rewrite <- Z2R_abs. - apply (Z2R_lt _ (radix2 ^ 1024)). - compute_this (radix2 ^ 1024); zify; omega. - - apply valid_rnd_round_mode. - - destruct (Z.eq_dec (Int.unsigned (Int64.hiword l)) 0). - rewrite e. apply generic_format_0. - apply generic_format_FLT_FLX. - + apply Rle_trans with (bpow radix2 0). apply bpow_le. omega. - rewrite <- Z2R_abs. apply (Z2R_le 1). - clear - n. zify; omega. - + apply generic_format_FLX. - eexists {| Fnum := Int.unsigned (Int64.hiword l); Fexp := 32 |}. - unfold F2R, Fnum, Fexp. split. - rewrite Z2R_mult. auto. - compute_this (radix2 ^ 53). zify; omega. + intros. + pose proof (Int.signed_range x). + rewrite ! from_words_eq. rewrite ox8000_0000_signed_unsigned. + change (Int.unsigned ox8000_0000) with Int.half_modulus. + unfold sub, b64_minus. rewrite BofZ_minus. + unfold of_int, b64_of_Z. f_equal. omega. + apply integer_representable_n; auto; smart_omega. + apply integer_representable_n; auto; smart_omega. Qed. Definition ox4530_0000 := Int.repr 1160773632. (**r [0x4530_0000] *) @@ -1446,359 +661,694 @@ Qed. Lemma from_words_value': forall x, - B2R _ _ (from_words ox4530_0000 x) = - (bpow radix2 84 + Z2R (Int.unsigned x * two_p 32))%R /\ - is_finite _ _ (from_words ox4530_0000 x) = true. + B2R _ _ (from_words ox4530_0000 x) = (bpow radix2 84 + Z2R (Int.unsigned x * two_p 32))%R + /\ is_finite _ _ (from_words ox4530_0000 x) = true + /\ Bsign _ _ (from_words ox4530_0000 x) = false. Proof. - intros; unfold from_words, double_of_bits, b64_of_bits, binary_float_of_bits. - rewrite B2R_FF2B. rewrite is_finite_FF2B. + intros; unfold from_words, of_bits, b64_of_bits, binary_float_of_bits. + rewrite B2R_FF2B, is_finite_FF2B, Bsign_FF2B. unfold binary_float_of_bits_aux; rewrite split_bits_or'; simpl; pose proof (Int.unsigned_range x). destruct (Int.unsigned x + Zpower_pos 2 52) eqn:?. exfalso; now smart_omega. - simpl; rewrite <- Heqz; unfold F2R; simpl. + simpl; rewrite <- Heqz; unfold F2R; simpl. split; auto. rewrite <- (Z2R_plus 19342813113834066795298816), <- (Z2R_mult _ 4294967296). - split; [f_equal; compute_this (Zpower_pos 2 52); - compute_this (two_power_pos 32); ring | reflexivity]. + f_equal; compute_this (Zpower_pos 2 52); compute_this (two_power_pos 32); ring. assert (Zneg p < 0) by reflexivity. exfalso; now smart_omega. Qed. -Theorem floatoflongu_from_words: +Lemma from_words_eq': + forall x, from_words ox4530_0000 x = BofZ 53 1024 __ __ (2^84 + Int.unsigned x * 2^32). +Proof. + intros. + pose proof (Int.unsigned_range x). + destruct (from_words_value' x) as (A & B & C). + destruct (BofZ_representable 53 1024 __ __ (2^84 + Int.unsigned x * 2^32)) as (D & E & F). + replace (2^84 + Int.unsigned x * 2^32) + with ((2^52 + Int.unsigned x) * 2^32) by ring. + apply integer_representable_n2p; auto. smart_omega. omega. omega. + apply B2R_Bsign_inj; auto. + rewrite A, D. rewrite <- Z2R_Zpower by omega. rewrite <- Z2R_plus. auto. + rewrite C, F. symmetry. apply Zlt_bool_false. + compute_this (2^84); compute_this (2^32); omega. +Qed. + +Theorem of_longu_from_words: forall l, - floatoflongu l = + of_longu l = add (sub (from_words ox4530_0000 (Int64.hiword l)) (from_words ox4530_0000 (Int.repr (two_p 20)))) (from_words ox4330_0000 (Int64.loword l)). Proof. intros. - pose proof (Int64.unsigned_range l). + pose proof (Int64.unsigned_range l). pose proof (Int.unsigned_range (Int64.hiword l)). - destruct (from_words_value (Int64.loword l)). - destruct (from_words_value' (Int64.hiword l)). - destruct (from_words_value' (Int.repr (two_p 20))). - unfold sub, b64_minus. - pose proof (Bminus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (from_words ox4530_0000 (Int64.hiword l)) - (from_words ox4530_0000 (Int.repr (two_p 20))) H4 H6). - rewrite round_generic in H7. - - rewrite H3, H5 in H7. - replace (bpow radix2 84 + Z2R (Int.unsigned (Int64.hiword l) * two_p 32) - - (bpow radix2 84 + Z2R (Int.unsigned (Int.repr (two_p 20)) * two_p 32)))%R - with (Z2R (Int.unsigned (Int64.hiword l) * two_p 32 - two_p 52)) in H7. - + rewrite Rlt_bool_true in H7. - * { destruct H7 as [? []]. - unfold floatoflongu, binary_normalize64. - pose proof (binary_normalize64_correct (Int64.unsigned l) 0 false). - replace (F2R (beta:=radix2) {| Fnum := Int64.unsigned l; Fexp := 0 |}) - with (Z2R (Int64.unsigned l)) in H10 - by (unfold F2R, Fexp, Fnum, bpow; field). - assert (Rabs (round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (Z2R (Int64.unsigned l))) < - bpow radix2 1024)%R. - { rewrite <- round_NE_abs by (apply fexp_correct; reflexivity). - eapply Rle_lt_trans with (Z2R (two_p 64)). 2:apply Z2R_lt; reflexivity. - erewrite <- round_generic. - - apply round_le. apply fexp_correct; reflexivity. - apply (valid_rnd_round_mode mode_NE). - rewrite <- Z2R_abs. apply Z2R_le. change (two_p 64) with Int64.modulus. zify; omega. - - apply (valid_rnd_round_mode mode_NE). - - apply (generic_format_bpow radix2 _ 64). compute. discriminate. } - rewrite Rlt_bool_true in H10 by auto. - unfold add, b64_plus. - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (Bminus 53 1024 eq_refl eq_refl binop_pl mode_NE - (from_words ox4530_0000 (Int64.hiword l)) - (from_words ox4530_0000 (Int.repr (two_p 20)))) - (from_words ox4330_0000 (Int64.loword l)) H8 H2). - change (bpow radix2 52) with (Z2R (two_p 52)) in H1. - rewrite H7, H1, <- !Z2R_plus in H12. - replace (Int.unsigned (Int64.hiword l) * two_p 32 - two_p 52 + - (two_p 52 + Int.unsigned (Int64.loword l))) - with (Int.unsigned (Int64.hiword l) * two_p 32 + Int.unsigned (Int64.loword l)) - in H12 by ring. - rewrite <- Int64.ofwords_add', Int64.ofwords_recompose, Rlt_bool_true in H12 by auto. - destruct (Z.eq_dec (Int64.unsigned l) 0). - - apply (f_equal Int64.repr) in e. rewrite Int64.repr_unsigned in e. - subst. reflexivity. - - destruct H12 as [? []], H10 as [? []]. - assert (1 <= Rabs (round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (Z2R (Int64.unsigned l)))) %R. - { rewrite <- round_NE_abs, <- Z2R_abs by (apply fexp_correct; reflexivity). - erewrite <- round_generic with (x := 1%R). - 3:eapply (generic_format_bpow _ _ 0). - apply round_le, (Z2R_le 1). - apply fexp_correct; reflexivity. apply (valid_rnd_round_mode mode_NE). - zify; omega. - apply (valid_rnd_round_mode mode_NE). - compute; discriminate. } - eapply B2R_inj. - + destruct binary_normalize; try discriminate H15. - unfold B2R in H10. rewrite <- H10, Rabs_R0 in H17. apply (le_Z2R 1 0) in H17. omega. - auto. - + match goal with Hf0:is_finite _ _ ?f0 = true, - Hf1:B2R _ _ ?f1 = _ |- - is_finite_strict _ _ ?f = true => - change f0 with f in Hf0; change f1 with f in Hf1; - destruct f - end; try discriminate H13. - unfold B2R in H12. rewrite <- H12, Rabs_R0 in H17. apply (le_Z2R 1 0) in H17. omega. - auto. - + etransitivity; eauto. } - * rewrite <- Z2R_abs. apply (Z2R_lt _ (2^1024)). - compute_this Int.modulus; compute_this (two_p 32); - compute_this (two_p 52); compute_this (2^1024). - clear - H0. zify; omega. - + rewrite Z2R_minus, Int.unsigned_repr, <- two_p_is_exp, !Z2R_mult. - ring_simplify. reflexivity. - omega. omega. compute; split; discriminate. - - apply valid_rnd_round_mode. - - apply sterbenz. - + apply FLT_exp_monotone. - + apply generic_format_B2R. - + apply generic_format_B2R. - + rewrite H3, H5, Int.unsigned_repr by (compute; split; discriminate). - unfold bpow. rewrite <- !Z2R_plus, <- (Z2R_mult 2). - compute_this (Z.pow_pos radix2 84); - compute_this (two_p 20 * two_p 32); compute_this (two_p 32); - compute_this (Int.modulus). - change (19342813113834066795298816 + 4503599627370496) - with (9671406559168833211334656 * 2). - unfold Rdiv. rewrite Z2R_mult, Rmult_assoc, Rinv_r, Rmult_1_r by (apply (Z2R_neq 2 0); omega). - split; apply Z2R_le; omega. -Qed. - -Theorem floatoflong_decomp: - forall l, floatoflong l = - add (mul (floatofint (Int64.hiword l)) (pow2_float false 32 (conj eq_refl eq_refl))) - (floatofintu (Int64.loword l)). -Proof. - intros. - unfold floatofintu, floatofint. - destruct (binary_normalize64_exact (Int.signed (Int64.hiword l))). - { pose proof (Int.signed_range (Int64.hiword l)). - revert H. generalize (Int.signed (Int64.hiword l)). - change (forall z : Z, -2147483648 <= z <= 2147483647 -> - 9007199254740992 < z < 9007199254740992). - intros. omega. } - destruct (binary_normalize64_exact (Int.unsigned (Int64.loword l))). - { pose proof (Int.unsigned_range (Int64.loword l)). - revert H1. generalize (Int.unsigned (Int64.loword l)). - change (forall z : Z, 0 <= z < 4294967296 -> - 9007199254740992 < z < 9007199254740992). - intros. omega. } - unfold mul, b64_mult. - pose proof (Bmult_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (binary_normalize64 (Int.signed (Int64.hiword l)) 0 false) - (pow2_float false 32 (conj eq_refl eq_refl))). - rewrite H in H3. - remember (B2R 53 1024 (pow2_float false 32 (conj eq_refl eq_refl))). - compute in Heqr. - change 4503599627370496%R with (Z2R (1048576*4294967296)) in Heqr. - change 1048576%R with (Z2R 1048576) in Heqr. - rewrite Z2R_mult in Heqr. - assert (r = Z2R 4294967296). - { rewrite Heqr. field. change 0%R with (Z2R 0). intro. apply eq_Z2R in H4. discriminate. } - clear Heqr. subst. - pose proof (Int.signed_range (Int64.hiword l)). - change Int.min_signed with (-2147483648) in H4. - change Int.max_signed with 2147483647 in H4. - rewrite <- Z2R_mult in H3. - rewrite round_generic in H3. - - destruct (Rlt_bool_spec (Rabs (Z2R (Int.signed (Int64.hiword l) * 4294967296))) - (bpow radix2 1024)). - + rewrite H0 in H3. - destruct H3 as [? [? ?]]. - { unfold add, b64_plus. - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (Bmult 53 1024 eq_refl eq_refl binop_pl mode_NE - (binary_normalize64 (Int.signed (Int64.hiword l)) 0 false) - (pow2_float false 32 (conj eq_refl eq_refl))) - (binary_normalize64 (Int.unsigned (Int64.loword l)) 0 false) H6 H2). - rewrite H3, H1, <- Z2R_plus in H8. - change 4294967296 with (two_p 32) in H8. - rewrite <- Int64.ofwords_add'', Int64.ofwords_recompose in H8. - destruct (Rlt_bool_spec - (Rabs - (round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (Z2R (Int64.signed l)))) - (bpow radix2 1024)). - - unfold floatoflong. unfold binary_normalize64. - pose proof (binary_normalize64_correct (Int64.signed l) 0 false). - replace (F2R (beta:=radix2) {| Fnum := Int64.signed l; Fexp := 0 |}) - with (Z2R (Int64.signed l)) in H10 - by (unfold F2R, Fexp, Fnum, bpow; field). - rewrite Rlt_bool_true in H10 by auto. - destruct (Int64.eq_dec l Int64.zero). subst. reflexivity. - destruct H10, H8. - assert (1 <= round radix2 (FLT_exp (3 - 1024 - 53) 53) (round_mode mode_NE) - (Z2R (Zabs (Int64.signed l))))%R. - { erewrite <- round_generic with (x:=1%R). - apply round_le. apply fexp_correct. reflexivity. apply valid_rnd_round_mode. - assert (Int64.signed l <> 0). - { contradict n. rewrite <- (Int64.repr_signed l), n. auto. } - change 1%R with (Z2R 1). apply Z2R_le. - zify. omega. apply valid_rnd_round_mode. - apply (generic_format_bpow _ _ 0). compute. discriminate. } - rewrite Z2R_abs in H13. - rewrite round_NE_abs in H13 by (apply fexp_correct; reflexivity). - change ZnearestE with (round_mode mode_NE) in H13. - unfold binary_normalize64 in *. - apply B2R_inj. - + destruct H11, (binary_normalize 53 1024 eq_refl eq_refl mode_NE (Int64.signed l) 0 false); try discriminate. - unfold B2R in H10. rewrite <- H10, Rabs_R0 in H13. apply (le_Z2R 1 0) in H13. omega. - auto. - + destruct H12; match goal with Hf0:is_finite _ _ ?f0 = true, - Hf1:B2R _ _ ?f1 = _ |- - is_finite_strict _ _ ?f = true => - change f0 with f in Hf0; change f1 with f in Hf1; - destruct f - end; try discriminate. - unfold B2R in H8. rewrite <- H8, Rabs_R0 in H13. apply (le_Z2R 1 0) in H13. omega. - auto. - + rewrite H10. symmetry. apply H8. - - exfalso. - eapply Rle_trans with (r3:=round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (bpow radix2 64)) in H9. - rewrite round_generic in H9. - + eapply le_bpow in H9. omega. - + apply valid_rnd_round_mode. - + apply generic_format_bpow. compute. discriminate. - + rewrite <- round_NE_abs. 2:apply fexp_correct; reflexivity. - apply round_le. apply fexp_correct; reflexivity. apply valid_rnd_round_mode. - rewrite <- Z2R_abs. change (bpow radix2 64)%R with (Z2R Int64.modulus). - apply Z2R_le. - destruct (Int64.signed_range l). - assert (-Int64.modulus < Int64.min_signed) by reflexivity. - assert (Int64.max_signed < Int64.modulus) by reflexivity. - zify. omega. } - + exfalso. - rewrite <- Z2R_abs in H5. - change (bpow radix2 1024) with (Z2R (radix2 ^ 1024)) in H5. - apply le_Z2R in H5. assert (radix2 ^ 1024 < 18446744073709551616) by (zify; omega). - discriminate. - - apply valid_rnd_round_mode. - - destruct (Z.eq_dec (Int.signed (Int64.hiword l)) 0). - rewrite e. apply generic_format_0. - apply generic_format_FLT_FLX. - + apply Rle_trans with (bpow radix2 0). apply bpow_le. omega. - change (bpow radix2 0) with (Z2R 1). rewrite <- Z2R_abs. apply Z2R_le. - clear - n H4. zify; omega. - + apply generic_format_FLX. - eexists {| Fnum := Int.signed (Int64.hiword l); Fexp := 32 |}. - unfold F2R, Fnum, Fexp. split. - rewrite Z2R_mult. auto. - change (radix2 ^ 53) with 9007199254740992. - clear -n H4. zify; omega. -Qed. - -Theorem floatoflong_from_words: + pose proof (Int.unsigned_range (Int64.loword l)). + rewrite ! from_words_eq, ! from_words_eq'. + set (p20 := Int.unsigned (Int.repr (two_p 20))). + set (x := Int64.unsigned l) in *; + set (xl := Int.unsigned (Int64.loword l)) in *; + set (xh := Int.unsigned (Int64.hiword l)) in *. + unfold sub, b64_minus. rewrite BofZ_minus. + replace (2^84 + xh * 2^32 - (2^84 + p20 * 2^32)) + with ((xh - p20) * 2^32) by ring. + unfold add, b64_plus. rewrite BofZ_plus. + unfold of_longu, b64_of_Z. f_equal. + rewrite <- (Int64.ofwords_recompose l) at 1. rewrite Int64.ofwords_add'. + fold xh; fold xl. compute_this (two_p 32); compute_this p20; ring. + apply integer_representable_n2p; auto. + compute_this p20; smart_omega. omega. omega. + apply integer_representable_n; auto; smart_omega. + replace (2^84 + xh * 2^32) with ((2^52 + xh) * 2^32) by ring. + apply integer_representable_n2p; auto. smart_omega. omega. omega. + change (2^84 + p20 * 2^32) with ((2^52 + 1048576) * 2^32). + apply integer_representable_n2p; auto. omega. omega. +Qed. + +Theorem of_long_from_words: forall l, - floatoflong l = + of_long l = add (sub (from_words ox4530_0000 (Int.add (Int64.hiword l) ox8000_0000)) (from_words ox4530_0000 (Int.repr (two_p 20+two_p 31)))) (from_words ox4330_0000 (Int64.loword l)). Proof. intros. - pose proof (Int64.signed_range l); - compute_this (Int64.min_signed); compute_this (Int64.max_signed). - pose proof (Int.unsigned_range (Int.add (Int64.hiword l) ox8000_0000)). - destruct (from_words_value (Int64.loword l)). - destruct (from_words_value' (Int.add (Int64.hiword l) ox8000_0000)). - destruct (from_words_value' (Int.repr (two_p 20+two_p 31))). - unfold sub, b64_minus. - pose proof (Bminus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (from_words ox4530_0000 (Int.add (Int64.hiword l) ox8000_0000)) - (from_words ox4530_0000 (Int.repr (two_p 20+two_p 31))) H4 H6). - rewrite round_generic in H7. - - rewrite H3, H5, ox8000_0000_signed_unsigned in H7. - replace (bpow radix2 84 + Z2R ((Int.signed (Int64.hiword l) + Int.half_modulus) * two_p 32) - - (bpow radix2 84 + Z2R (Int.unsigned (Int.repr (two_p 20+two_p 31)) * two_p 32)))%R - with (Z2R (Int.unsigned (Int.add (Int64.hiword l) ox8000_0000) * two_p 32 -two_p 52-two_p 63)) in H7. - + rewrite Rlt_bool_true in H7. - * { destruct H7 as [? []]. - unfold floatoflong, binary_normalize64. - pose proof (binary_normalize64_correct (Int64.signed l) 0 false). - replace (F2R (beta:=radix2) {| Fnum := Int64.signed l; Fexp := 0 |}) - with (Z2R (Int64.signed l)) in H10 - by (unfold F2R, Fexp, Fnum, bpow; field). - assert (Rabs (round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (Z2R (Int64.signed l))) < - bpow radix2 1024)%R. - { rewrite <- round_NE_abs by (apply fexp_correct; reflexivity). - eapply Rle_lt_trans with (Z2R (two_p 64)). 2:apply Z2R_lt; reflexivity. - erewrite <- round_generic. - - apply round_le. apply fexp_correct; reflexivity. - apply (valid_rnd_round_mode mode_NE). - rewrite <- Z2R_abs. apply Z2R_le. - compute_this (two_p 64). zify; omega. - - apply (valid_rnd_round_mode mode_NE). - - apply (generic_format_bpow radix2 _ 64). compute. discriminate. } - rewrite Rlt_bool_true in H10 by auto. - unfold add, b64_plus. - pose proof (Bplus_correct 53 1024 eq_refl eq_refl binop_pl mode_NE - (Bminus 53 1024 eq_refl eq_refl binop_pl mode_NE - (from_words ox4530_0000 (Int.add (Int64.hiword l) ox8000_0000)) - (from_words ox4530_0000 (Int.repr (two_p 20 + two_p 31)))) - (from_words ox4330_0000 (Int64.loword l)) H8 H2). - change (bpow radix2 52) with (Z2R (two_p 52)) in H1. - rewrite H7, H1, <- !Z2R_plus, ox8000_0000_signed_unsigned in H12. - change (two_p 63) with (Int.half_modulus * two_p 32) in H12. - replace ((Int.signed (Int64.hiword l) + Int.half_modulus) * - two_p 32 - two_p 52 - (Int.half_modulus * two_p 32) + - (two_p 52 + Int.unsigned (Int64.loword l))) - with (Int.signed (Int64.hiword l) * two_p 32 + Int.unsigned (Int64.loword l)) - in H12 by ring. - rewrite <- Int64.ofwords_add'', Int64.ofwords_recompose, Rlt_bool_true in H12 by auto. - destruct (Z.eq_dec (Int64.signed l) 0). - - apply (f_equal Int64.repr) in e. rewrite Int64.repr_signed in e. - subst. reflexivity. - - destruct H12 as [? []], H10 as [? []]. - assert (1 <= Rabs (round radix2 (FLT_exp (3 - 1024 - 53) 53) - (round_mode mode_NE) (Z2R (Int64.signed l)))) %R. - { rewrite <- round_NE_abs, <- Z2R_abs by (apply fexp_correct; reflexivity). - erewrite <- round_generic with (x := 1%R). - 3:eapply (generic_format_bpow _ _ 0). - apply round_le, (Z2R_le 1). - apply fexp_correct; reflexivity. apply (valid_rnd_round_mode mode_NE). - zify; omega. - apply (valid_rnd_round_mode mode_NE). - compute; discriminate. } - eapply B2R_inj. - + destruct binary_normalize; try discriminate H15. - unfold B2R in H10. rewrite <- H10, Rabs_R0 in H17. apply (le_Z2R 1 0) in H17. omega. - auto. - + match goal with Hf0:is_finite _ _ ?f0 = true, - Hf1:B2R _ _ ?f1 = _ |- - is_finite_strict _ _ ?f = true => - change f0 with f in Hf0; change f1 with f in Hf1; - destruct f - end; try discriminate H13. - unfold B2R in H12. rewrite <- H12, Rabs_R0 in H17. apply (le_Z2R 1 0) in H17. omega. - auto. - + etransitivity; eauto. } - * rewrite <- Z2R_abs. apply (Z2R_lt _ (2^1024)). - compute_this Int.modulus; compute_this (two_p 32); - compute_this (two_p 52); compute_this (two_p 63); compute_this (2^1024). - clear - H0. zify; omega. - + rewrite ox8000_0000_signed_unsigned, !Z2R_minus. - compute_this (Z2R (Int.unsigned (Int.repr (two_p 20 + two_p 31)) * two_p 32)). - compute_this (Z2R (two_p 52)). compute_this (Z2R (two_p 63)). ring. - - apply valid_rnd_round_mode. - - apply sterbenz. - + apply FLT_exp_monotone. - + apply generic_format_B2R. - + apply generic_format_B2R. - + rewrite H3, H5, Int.unsigned_repr by (compute; split; discriminate). - unfold bpow. rewrite <- !Z2R_plus, <- (Z2R_mult 2). - compute_this (Z.pow_pos radix2 84); compute_this (Z.pow_pos radix2 84); - compute_this ((two_p 20 + two_p 31) * two_p 32); compute_this (two_p 32); - compute_this (Int.modulus). - change (19342813113834066795298816 + 9227875636482146304) - with (9671411170854851638722560 * 2). - unfold Rdiv. rewrite Z2R_mult, Rmult_assoc, Rinv_r, Rmult_1_r by (apply (Z2R_neq 2 0); omega). - split; apply Z2R_le; omega. + pose proof (Int64.signed_range l). + pose proof (Int.signed_range (Int64.hiword l)). + pose proof (Int.unsigned_range (Int64.loword l)). + rewrite ! from_words_eq, ! from_words_eq'. + set (p := Int.unsigned (Int.repr (two_p 20 + two_p 31))). + set (x := Int64.signed l) in *; + set (xl := Int.unsigned (Int64.loword l)) in *; + set (xh := Int.signed (Int64.hiword l)) in *. + rewrite ox8000_0000_signed_unsigned. fold xh. + unfold sub, b64_minus. rewrite BofZ_minus. + replace (2^84 + (xh + Int.half_modulus) * 2^32 - (2^84 + p * 2^32)) + with ((xh - 2^20) * 2^32) + by (compute_this p; compute_this Int.half_modulus; ring). + unfold add, b64_plus. rewrite BofZ_plus. + unfold of_long, b64_of_Z. f_equal. + rewrite <- (Int64.ofwords_recompose l) at 1. rewrite Int64.ofwords_add''. + fold xh; fold xl. compute_this (two_p 32); ring. + apply integer_representable_n2p; auto. + compute_this (2^20); smart_omega. omega. omega. + apply integer_representable_n; auto; smart_omega. + replace (2^84 + (xh + Int.half_modulus) * 2^32) + with ((2^52 + xh + Int.half_modulus) * 2^32) + by (compute_this Int.half_modulus; ring). + apply integer_representable_n2p; auto. smart_omega. omega. omega. + change (2^84 + p * 2^32) with ((2^52 + p) * 2^32). + apply integer_representable_n2p; auto. + compute_this p; smart_omega. omega. Qed. -Global Opaque - zero eq_dec neg abs singleoffloat intoffloat intuoffloat floatofint floatofintu - add sub mul div cmp bits_of_double double_of_bits bits_of_single single_of_bits from_words. +(** Conversions from 64-bit integers can be expressed in terms of + conversions from their 32-bit halves. *) + +Theorem of_longu_decomp: + forall l, + of_longu l = add (mul (of_intu (Int64.hiword l)) (b64_of_Z (2^32))) + (of_intu (Int64.loword l)). +Proof. + intros. + unfold of_longu, of_intu, b64_of_Z, add, mul, b64_plus, b64_mult. + pose proof (Int.unsigned_range (Int64.loword l)). + pose proof (Int.unsigned_range (Int64.hiword l)). + pose proof (Int64.unsigned_range l). + set (x := Int64.unsigned l) in *. + set (yl := Int.unsigned (Int64.loword l)) in *. + set (yh := Int.unsigned (Int64.hiword l)) in *. + assert (DECOMP: x = yh * 2^32 + yl). + { unfold x. rewrite <- (Int64.ofwords_recompose l). apply Int64.ofwords_add'. } + rewrite BofZ_mult. rewrite BofZ_plus. rewrite DECOMP; auto. + apply integer_representable_n2p; auto. smart_omega. omega. omega. + apply integer_representable_n; auto; smart_omega. + apply integer_representable_n; auto; smart_omega. + apply integer_representable_n; auto; smart_omega. + compute; auto. +Qed. + +Theorem of_long_decomp: + forall l, + of_long l = add (mul (of_int (Int64.hiword l)) (b64_of_Z (2^32))) + (of_intu (Int64.loword l)). +Proof. + intros. + unfold of_long, of_int, of_intu, b64_of_Z, add, mul, b64_plus, b64_mult. + pose proof (Int.unsigned_range (Int64.loword l)). + pose proof (Int.signed_range (Int64.hiword l)). + pose proof (Int64.signed_range l). + set (x := Int64.signed l) in *. + set (yl := Int.unsigned (Int64.loword l)) in *. + set (yh := Int.signed (Int64.hiword l)) in *. + assert (DECOMP: x = yh * 2^32 + yl). + { unfold x. rewrite <- (Int64.ofwords_recompose l), Int64.ofwords_add''. auto. } + rewrite BofZ_mult. rewrite BofZ_plus. rewrite DECOMP; auto. + apply integer_representable_n2p; auto. smart_omega. omega. omega. + apply integer_representable_n; auto; smart_omega. + apply integer_representable_n; auto; smart_omega. + apply integer_representable_n; auto. compute; intuition congruence. + compute; auto. +Qed. + +(** Conversions from unsigned longs can be expressed in terms of conversions from signed longs. + If the unsigned long is too big, a round-to-odd must be performed on it + to avoid double rounding. *) + +Theorem of_longu_of_long_1: + forall x, + Int64.ltu x (Int64.repr Int64.half_modulus) = true -> + of_longu x = of_long x. +Proof. + unfold of_longu, of_long, Int64.signed, Int64.ltu; intro. + change (Int64.unsigned (Int64.repr Int64.half_modulus)) with Int64.half_modulus. + destruct (zlt (Int64.unsigned x) Int64.half_modulus); now intuition. +Qed. + +Theorem of_longu_of_long_2: + forall x, + Int64.ltu x (Int64.repr Int64.half_modulus) = false -> + of_longu x = mul (of_long (Int64.or (Int64.shru x Int64.one) + (Int64.and x Int64.one))) + (of_int (Int.repr 2)). +Proof. + intros. change (of_int (Int.repr 2)) with (BofZ 53 1024 __ __ (2^1)). + pose proof (Int64.unsigned_range x). + unfold Int64.ltu in H. + change (Int64.unsigned (Int64.repr Int64.half_modulus)) with (2^63) in H. + destruct (zlt (Int64.unsigned x) (2^63)); inv H. + assert (Int64.modulus <= 2^1024 - 2^(1024-53)) by (vm_compute; intuition congruence). + set (n := Int64.or (Int64.shru x Int64.one) (Int64.and x Int64.one)). + assert (NB: forall i, 0 <= i < 64 -> + Int64.testbit n i = + if zeq i 0 then Int64.testbit x 1 || Int64.testbit x 0 + else if zeq i 63 then false else Int64.testbit x (i + 1)). + { intros; unfold n; autorewrite with ints; auto. rewrite Int64.unsigned_one. + rewrite Int64.bits_one. compute_this Int64.zwordsize. + destruct (zeq i 0); simpl proj_sumbool. + rewrite zlt_true by omega. rewrite andb_true_r. subst i; auto. + rewrite andb_false_r, orb_false_r. + destruct (zeq i 63). subst i. apply zlt_false; omega. + apply zlt_true; omega. } + assert (NB2: forall i, 0 <= i -> + Z.testbit (Int64.signed n * 2^1) i = + if zeq i 0 then false else + if zeq i 1 then Int64.testbit x 1 || Int64.testbit x 0 else + Int64.testbit x i). + { intros. rewrite Z.mul_pow2_bits by omega. destruct (zeq i 0). + apply Z.testbit_neg_r; omega. + rewrite Int64.bits_signed by omega. compute_this Int64.zwordsize. + destruct (zlt (i-1) 64). + rewrite NB by omega. destruct (zeq i 1). + subst. rewrite dec_eq_true by auto. auto. + rewrite dec_eq_false by omega. destruct (zeq (i - 1) 63). + symmetry. apply Int64.bits_above. compute_this Int64.zwordsize; omega. + f_equal; omega. + rewrite NB by omega. rewrite dec_eq_false by omega. rewrite dec_eq_true by auto. + rewrite dec_eq_false by omega. symmetry. apply Int64.bits_above. compute_this Int64.zwordsize; omega. + } + assert (EQ: Int64.signed n * 2 = int_round_odd (Int64.unsigned x) 1). + { + symmetry. apply (int_round_odd_bits 53 1024). omega. + intros. rewrite NB2 by omega. replace i with 0 by omega. auto. + rewrite NB2 by omega. rewrite dec_eq_false by omega. rewrite dec_eq_true. + rewrite orb_comm. unfold Int64.testbit. change (2^1) with 2. + destruct (Z.testbit (Int64.unsigned x) 0) eqn:B0; + [rewrite Z.testbit_true in B0 by omega|rewrite Z.testbit_false in B0 by omega]; + change (2^0) with 1 in B0; rewrite Zdiv_1_r in B0; rewrite B0; auto. + intros. rewrite NB2 by omega. rewrite ! dec_eq_false by omega. auto. + } + unfold mul, of_long, of_longu, b64_mult, b64_of_Z. + rewrite BofZ_mult_2p. +- change (2^1) with 2. rewrite EQ. apply BofZ_round_odd with (p := 1). ++ omega. ++ apply Zle_trans with Int64.modulus; trivial. smart_omega. ++ omega. ++ apply Zle_trans with (2^63). compute; intuition congruence. xomega. +- apply Zle_trans with Int64.modulus; trivial. + pose proof (Int64.signed_range n). + compute_this Int64.min_signed; compute_this Int64.max_signed; + compute_this Int64.modulus; xomega. +- assert (2^63 <= int_round_odd (Int64.unsigned x) 1). + { change (2^63) with (int_round_odd (2^63) 1). apply (int_round_odd_le 0 0); omega. } + rewrite <- EQ in H1. compute_this (2^63). compute_this (2^53). xomega. +- omega. +Qed. End Float. + +(** * Single-precision FP numbers *) + +Module Float32. + +(** ** NaN payload manipulations *) + +Program Definition transform_quiet_pl (pl:nan_pl 24) : nan_pl 24 := + Pos.lor pl (nat_iter 22 xO xH). +Next Obligation. + destruct pl. + simpl. rewrite Z.ltb_lt in *. + assert (forall x, S (Fcore_digits.digits2_Pnat x) = Pos.to_nat (Pos.size x)). + { induction x0; simpl; auto; rewrite IHx0; zify; omega. } + fold (Z.of_nat (S (Fcore_digits.digits2_Pnat (Pos.lor x 4194304)))). + rewrite H, positive_nat_Z, Psize_log_inf, <- Zlog2_log_inf in *. clear H. + change (Z.pos (Pos.lor x 4194304)) with (Z.lor (Z.pos x) 4194304). + rewrite Z.log2_lor by (zify; omega). + apply Z.max_case. auto. simpl. omega. +Qed. + +Lemma transform_quiet_pl_idempotent: + forall pl, transform_quiet_pl (transform_quiet_pl pl) = transform_quiet_pl pl. +Proof. + intros []; simpl; intros. apply Float.nan_payload_fequal. + simpl. apply Float.lor_idempotent. +Qed. + +Definition neg_pl (s:bool) (pl:nan_pl 24) := (negb s, pl). +Definition abs_pl (s:bool) (pl:nan_pl 24) := (false, pl). + +Definition binop_pl (x y: binary32) : bool*nan_pl 24 := + match x, y with + | B754_nan s1 pl1, B754_nan s2 pl2 => + if Archi.choose_binop_pl_32 s1 pl1 s2 pl2 + then (s2, transform_quiet_pl pl2) + else (s1, transform_quiet_pl pl1) + | B754_nan s1 pl1, _ => (s1, transform_quiet_pl pl1) + | _, B754_nan s2 pl2 => (s2, transform_quiet_pl pl2) + | _, _ => Archi.default_pl_32 + end. + +(** ** Operations over single-precision floats *) + +Definition zero: float32 := B754_zero _ _ false. (**r the float [+0.0] *) + +Definition eq_dec: forall (f1 f2: float32), {f1 = f2} + {f1 <> f2} := b32_eq_dec. + +(** Arithmetic operations *) + +Definition neg: float32 -> float32 := b32_opp neg_pl. (**r opposite (change sign) *) +Definition abs: float32 -> float32 := b32_abs abs_pl. (**r absolute value (set sign to [+]) *) +Definition add: float32 -> float32 -> float32 := b32_plus binop_pl mode_NE. (**r addition *) +Definition sub: float32 -> float32 -> float32 := b32_minus binop_pl mode_NE. (**r subtraction *) +Definition mul: float32 -> float32 -> float32 := b32_mult binop_pl mode_NE. (**r multiplication *) +Definition div: float32 -> float32 -> float32 := b32_div binop_pl mode_NE. (**r division *) +Definition cmp (c:comparison) (f1 f2: float32) : bool := (**r comparison *) + cmp_of_comparison c (b32_compare f1 f2). + +(** Conversions *) + +Definition of_double : float -> float32 := Float.to_single. +Definition to_double : float32 -> float := Float.of_single. + +Definition to_int (f:float32): option int := (**r conversion to signed 32-bit int *) + option_map Int.repr (b32_to_Z_range f Int.min_signed Int.max_signed). +Definition to_intu (f:float32): option int := (**r conversion to unsigned 32-bit int *) + option_map Int.repr (b32_to_Z_range f 0 Int.max_unsigned). +Definition to_long (f:float32): option int64 := (**r conversion to signed 64-bit int *) + option_map Int64.repr (b32_to_Z_range f Int64.min_signed Int64.max_signed). +Definition to_longu (f:float32): option int64 := (**r conversion to unsigned 64-bit int *) + option_map Int64.repr (b32_to_Z_range f 0 Int64.max_unsigned). + +Definition of_int (n:int): float32 := (**r conversion from signed 32-bit int to single-precision float *) + b32_of_Z (Int.signed n). +Definition of_intu (n:int): float32 := (**r conversion from unsigned 32-bit int to single-precision float *) + b32_of_Z (Int.unsigned n). + +Definition of_long (n:int64): float32 := (**r conversion from signed 64-bit int to single-precision float *) + b32_of_Z (Int64.signed n). +Definition of_longu (n:int64): float32 := (**r conversion from unsigned 64-bit int to single-precision float *) + b32_of_Z (Int64.unsigned n). + +Definition from_parsed (base:positive) (intPart:positive) (expPart:Z) : float32 := + build_from_parsed 24 128 __ __ base intPart expPart. + +(** Conversions between floats and their concrete in-memory representation + as a sequence of 32 bits. *) + +Definition to_bits (f: float32) : int := Int.repr (bits_of_b32 f). +Definition of_bits (b: int): float32 := b32_of_bits (Int.unsigned b). + +(** ** Properties *) + +(** Commutativity properties of addition and multiplication. *) + +Theorem add_commut: + forall x y, is_nan _ _ x = false \/ is_nan _ _ y = false -> add x y = add y x. +Proof. + intros. apply Bplus_commut. + destruct x, y; try reflexivity. simpl in H. intuition congruence. +Qed. + +Theorem mul_commut: + forall x y, is_nan _ _ x = false \/ is_nan _ _ y = false -> mul x y = mul y x. +Proof. + intros. apply Bmult_commut. + destruct x, y; try reflexivity. simpl in H. intuition congruence. +Qed. + +(** Multiplication by 2 is diagonal addition. *) + +Theorem mul2_add: + forall f, add f f = mul f (of_int (Int.repr 2%Z)). +Proof. + intros. apply Bmult2_Bplus. + intros. destruct x; try discriminate. simpl. + transitivity (b, transform_quiet_pl n). + destruct Archi.choose_binop_pl_32; auto. + destruct y; auto || discriminate. +Qed. + +(** Divisions that can be turned into multiplication by an inverse. *) + +Definition exact_inverse : float32 -> option float32 := b32_exact_inverse. + +Theorem div_mul_inverse: + forall x y z, exact_inverse y = Some z -> div x y = mul x z. +Proof. + intros. apply Bdiv_mult_inverse; auto. + intros. destruct x0; try discriminate. simpl. + transitivity (b, transform_quiet_pl n). + destruct y0; reflexivity || discriminate. + destruct z0; reflexivity || discriminate. +Qed. + +(** Properties of comparisons. *) + +Theorem cmp_swap: + forall c x y, cmp (swap_comparison c) x y = cmp c y x. +Proof. + unfold cmp, b32_compare; intros. rewrite (Bcompare_swap _ _ x y). + apply cmp_of_comparison_swap. +Qed. + +Theorem cmp_ne_eq: + forall f1 f2, cmp Cne f1 f2 = negb (cmp Ceq f1 f2). +Proof. + intros; apply cmp_of_comparison_ne_eq. +Qed. + +Theorem cmp_lt_eq_false: + forall f1 f2, cmp Clt f1 f2 = true -> cmp Ceq f1 f2 = true -> False. +Proof. + intros f1 f2; apply cmp_of_comparison_lt_eq_false. +Qed. + +Theorem cmp_le_lt_eq: + forall f1 f2, cmp Cle f1 f2 = cmp Clt f1 f2 || cmp Ceq f1 f2. +Proof. + intros f1 f2; apply cmp_of_comparison_le_lt_eq. +Qed. + +Theorem cmp_gt_eq_false: + forall x y, cmp Cgt x y = true -> cmp Ceq x y = true -> False. +Proof. + intros f1 f2; apply cmp_of_comparison_gt_eq_false. +Qed. + +Theorem cmp_ge_gt_eq: + forall f1 f2, cmp Cge f1 f2 = cmp Cgt f1 f2 || cmp Ceq f1 f2. +Proof. + intros f1 f2; apply cmp_of_comparison_ge_gt_eq. +Qed. + +Theorem cmp_lt_gt_false: + forall f1 f2, cmp Clt f1 f2 = true -> cmp Cgt f1 f2 = true -> False. +Proof. + intros f1 f2; apply cmp_of_comparison_lt_gt_false. +Qed. + +Theorem cmp_double: + forall f1 f2 c, cmp c f1 f2 = Float.cmp c (to_double f1) (to_double f2). +Proof. + unfold cmp, Float.cmp; intros. f_equal. symmetry. apply Bcompare_Bconv_widen. + red; omega. omega. omega. +Qed. + +(** Properties of conversions to/from in-memory representation. + The conversions are bijective (one-to-one). *) + +Theorem of_to_bits: + forall f, of_bits (to_bits f) = f. +Proof. + intros; unfold of_bits, to_bits, bits_of_b32, b32_of_bits. + rewrite Int.unsigned_repr, binary_float_of_bits_of_binary_float; [reflexivity|]. + generalize (bits_of_binary_float_range 23 8 __ __ f). + change (2^(23+8+1)) with (Int.max_unsigned + 1). omega. +Qed. + +Theorem to_of_bits: + forall b, to_bits (of_bits b) = b. +Proof. + intros; unfold of_bits, to_bits, bits_of_b32, b32_of_bits. + rewrite bits_of_binary_float_of_bits. apply Int.repr_unsigned. + apply Int.unsigned_range. +Qed. + +(** Conversions from 32-bit integers to single-precision floats can + be decomposed into a conversion to a double-precision float, + followed by a [Float32.of_double] conversion. No double rounding occurs. *) + +Theorem of_int_double: + forall n, of_int n = of_double (Float.of_int n). +Proof. + intros. symmetry. apply Bconv_BofZ. + apply integer_representable_n; auto. generalize (Int.signed_range n); Float.smart_omega. +Qed. + +Theorem of_intu_double: + forall n, of_intu n = of_double (Float.of_intu n). +Proof. + intros. symmetry. apply Bconv_BofZ. + apply integer_representable_n; auto. generalize (Int.unsigned_range n); Float.smart_omega. +Qed. + +(** Conversion of single-precision floats to integers can be decomposed + into a [Float32.to_double] extension, followed by a double-precision-to-int + conversion. *) + +Theorem to_int_double: + forall f n, to_int f = Some n -> Float.to_int (to_double f) = Some n. +Proof. + intros. + unfold to_int in H. + destruct (b32_to_Z_range f Int.min_signed Int.max_signed) as [n'|] eqn:E; inv H. + unfold Float.to_int, to_double, Float.of_single, b64_to_Z_range, b64_of_b32. + erewrite ZofB_range_Bconv; eauto. auto. omega. omega. omega. omega. +Qed. + +Theorem to_intu_double: + forall f n, to_intu f = Some n -> Float.to_intu (to_double f) = Some n. +Proof. + intros. + unfold to_intu in H. + destruct (b32_to_Z_range f 0 Int.max_unsigned) as [n'|] eqn:E; inv H. + unfold Float.to_intu, to_double, Float.of_single, b64_to_Z_range, b64_of_b32. + erewrite ZofB_range_Bconv; eauto. auto. omega. omega. omega. omega. +Qed. + +Theorem to_long_double: + forall f n, to_long f = Some n -> Float.to_long (to_double f) = Some n. +Proof. + intros. + unfold to_long in H. + destruct (b32_to_Z_range f Int64.min_signed Int64.max_signed) as [n'|] eqn:E; inv H. + unfold Float.to_long, to_double, Float.of_single, b64_to_Z_range, b64_of_b32. + erewrite ZofB_range_Bconv; eauto. auto. omega. omega. omega. omega. +Qed. + +Theorem to_longu_double: + forall f n, to_longu f = Some n -> Float.to_longu (to_double f) = Some n. +Proof. + intros. + unfold to_longu in H. + destruct (b32_to_Z_range f 0 Int64.max_unsigned) as [n'|] eqn:E; inv H. + unfold Float.to_longu, to_double, Float.of_single, b64_to_Z_range, b64_of_b32. + erewrite ZofB_range_Bconv; eauto. auto. omega. omega. omega. omega. +Qed. + +(** Conversions from 64-bit integers to single-precision floats can be expressed + as conversion to a double-precision float followed by a [Float32.of_double] conversion. + To avoid double rounding when the integer is large (above [2^53]), a round + to odd must be performed on the integer before conversion to double-precision float. *) + +Lemma int_round_odd_plus: + forall p n, 0 <= p -> + int_round_odd n p = Z.land (Z.lor n (Z.land n (2^p-1) + (2^p-1))) (-(2^p)). +Proof. + intros. + assert (POS: 0 < 2^p) by (apply (Zpower_gt_0 radix2); auto). + assert (A: Z.land n (2^p-1) = n mod 2^p). + { rewrite <- Z.land_ones by auto. f_equal. rewrite Z.ones_equiv. omega. } + rewrite A. + assert (B: 0 <= n mod 2^p < 2^p). + { apply Z_mod_lt. omega. } + set (m := n mod 2^p + (2^p-1)) in *. + assert (C: m / 2^p = if zeq (n mod 2^p) 0 then 0 else 1). + { unfold m. destruct (zeq (n mod 2^p) 0). + rewrite e. apply Zdiv_small. omega. + eapply Zdiv_unique with (n mod 2^p - 1). ring. omega. } + assert (D: Z.testbit m p = if zeq (n mod 2^p) 0 then false else true). + { destruct (zeq (n mod 2^p) 0). + apply Z.testbit_false; auto. rewrite C; auto. + apply Z.testbit_true; auto. rewrite C; auto. } + assert (E: forall i, p < i -> Z.testbit m i = false). + { intros. apply Z.testbit_false. omega. + replace (m / 2^i) with 0. auto. symmetry. apply Zdiv_small. + unfold m. split. omega. apply Zlt_le_trans with (2 * 2^p). omega. + change 2 with (2^1) at 1. rewrite <- (Zpower_plus radix2) by omega. + apply Zpower_le. omega. } + assert (F: forall i, 0 <= i -> Z.testbit (-2^p) i = if zlt i p then false else true). + { intros. rewrite Z.bits_opp by auto. rewrite <- Z.ones_equiv. + destruct (zlt i p). + rewrite Z.ones_spec_low by omega. auto. + rewrite Z.ones_spec_high by omega. auto. } + apply int_round_odd_bits; auto. + - intros. rewrite Z.land_spec, F, zlt_true by omega. apply andb_false_r. + - rewrite Z.land_spec, Z.lor_spec, D, F, zlt_false, andb_true_r by omega. + destruct (Z.eqb (n mod 2^p) 0) eqn:Z. + rewrite Z.eqb_eq in Z. rewrite Z, zeq_true. apply orb_false_r. + rewrite Z.eqb_neq in Z. rewrite zeq_false by auto. apply orb_true_r. + - intros. rewrite Z.land_spec, Z.lor_spec, E, F, zlt_false, andb_true_r by omega. + apply orb_false_r. +Qed. + +Lemma of_long_round_odd: + forall n conv_nan, + 2^36 <= Z.abs n < 2^64 -> + b32_of_Z n = b32_of_b64 conv_nan mode_NE (b64_of_Z (Z.land (Z.lor n ((Z.land n 2047) + 2047)) (-2048))). +Proof. + intros. rewrite <- (int_round_odd_plus 11) by omega. + assert (-2^64 <= int_round_odd n 11). + { change (-2^64) with (int_round_odd (-2^64) 11). apply (int_round_odd_le 0 0); xomega. } + assert (int_round_odd n 11 <= 2^64). + { change (2^64) with (int_round_odd (2^64) 11). apply (int_round_odd_le 0 0); xomega. } + unfold b32_of_Z, b32_of_b64, b64_of_Z. + rewrite Bconv_BofZ. + apply BofZ_round_odd with (p := 11). + omega. + apply Zle_trans with (2^64). omega. compute; intuition congruence. + omega. + exact (proj1 H). + unfold int_round_odd. apply integer_representable_n2p_wide. auto. omega. + unfold int_round_odd in H0, H1. + split; (apply Zmult_le_reg_r with (2^11); [compute; auto | assumption]). + omega. + omega. +Qed. + +Theorem of_longu_double_1: + forall n, + Int64.unsigned n <= 2^53 -> + of_longu n = of_double (Float.of_longu n). +Proof. + intros. symmetry; apply Bconv_BofZ. apply integer_representable_n; auto. + pose proof (Int64.unsigned_range n); omega. +Qed. + +Theorem of_longu_double_2: + forall n, + 2^36 <= Int64.unsigned n -> + of_longu n = of_double (Float.of_longu + (Int64.and (Int64.or n + (Int64.add (Int64.and n (Int64.repr 2047)) + (Int64.repr 2047))) + (Int64.repr (-2048)))). +Proof. + intros. + pose proof (Int64.unsigned_range n). + unfold of_longu. erewrite of_long_round_odd. + unfold of_double, Float.to_single. instantiate (1 := Float.to_single_pl). + f_equal. unfold Float.of_longu. f_equal. + set (n' := Z.land (Z.lor (Int64.unsigned n) (Z.land (Int64.unsigned n) 2047 + 2047)) (-2048)). + assert (int_round_odd (Int64.unsigned n) 11 = n') by (apply int_round_odd_plus; omega). + assert (0 <= n'). + { rewrite <- H1. change 0 with (int_round_odd 0 11). apply (int_round_odd_le 0 0); omega. } + assert (n' < Int64.modulus). + { apply Zle_lt_trans with (int_round_odd (Int64.modulus - 1) 11). + rewrite <- H1. apply (int_round_odd_le 0 0); omega. + compute; auto. } + rewrite <- (Int64.unsigned_repr n') by (unfold Int64.max_unsigned; omega). + f_equal. Int64.bit_solve. rewrite Int64.testbit_repr by auto. unfold n'. + rewrite Z.land_spec, Z.lor_spec. f_equal. f_equal. + unfold Int64.testbit. rewrite Int64.add_unsigned. + fold (Int64.testbit (Int64.repr + (Int64.unsigned (Int64.and n (Int64.repr 2047)) + + Int64.unsigned (Int64.repr 2047))) i). + rewrite Int64.testbit_repr by auto. f_equal. f_equal. unfold Int64.and. + symmetry. apply Int64.unsigned_repr. change 2047 with (Z.ones 11). + rewrite Z.land_ones by omega. + exploit (Z_mod_lt (Int64.unsigned n) (2^11)). compute; auto. + assert (2^11 < Int64.max_unsigned) by (compute; auto). omega. + apply Int64.same_bits_eqm; auto. exists (-1); auto. + split. xomega. change (2^64) with Int64.modulus. xomega. +Qed. + +Theorem of_long_double_1: + forall n, + Z.abs (Int64.signed n) <= 2^53 -> + of_long n = of_double (Float.of_long n). +Proof. + intros. symmetry; apply Bconv_BofZ. apply integer_representable_n; auto. xomega. +Qed. + +Theorem of_long_double_2: + forall n, + 2^36 <= Z.abs (Int64.signed n) -> + of_long n = of_double (Float.of_long + (Int64.and (Int64.or n + (Int64.add (Int64.and n (Int64.repr 2047)) + (Int64.repr 2047))) + (Int64.repr (-2048)))). +Proof. + intros. + pose proof (Int64.signed_range n). + unfold of_long. erewrite of_long_round_odd. + unfold of_double, Float.to_single. instantiate (1 := Float.to_single_pl). + f_equal. unfold Float.of_long. f_equal. + set (n' := Z.land (Z.lor (Int64.signed n) (Z.land (Int64.signed n) 2047 + 2047)) (-2048)). + assert (int_round_odd (Int64.signed n) 11 = n') by (apply int_round_odd_plus; omega). + assert (Int64.min_signed <= n'). + { rewrite <- H1. change Int64.min_signed with (int_round_odd Int64.min_signed 11). apply (int_round_odd_le 0 0); omega. } + assert (n' <= Int64.max_signed). + { apply Zle_trans with (int_round_odd Int64.max_signed 11). + rewrite <- H1. apply (int_round_odd_le 0 0); omega. + compute; intuition congruence. } + rewrite <- (Int64.signed_repr n') by omega. + f_equal. Int64.bit_solve. rewrite Int64.testbit_repr by auto. unfold n'. + rewrite Z.land_spec, Z.lor_spec. f_equal. f_equal. + rewrite Int64.bits_signed by omega. rewrite zlt_true by omega. auto. + unfold Int64.testbit. rewrite Int64.add_unsigned. + fold (Int64.testbit (Int64.repr + (Int64.unsigned (Int64.and n (Int64.repr 2047)) + + Int64.unsigned (Int64.repr 2047))) i). + rewrite Int64.testbit_repr by auto. f_equal. f_equal. unfold Int64.and. + change (Int64.unsigned (Int64.repr 2047)) with 2047. + change 2047 with (Z.ones 11). rewrite ! Z.land_ones by omega. + rewrite Int64.unsigned_repr. apply Int64.eqmod_mod_eq. + apply Zlt_gt. apply (Zpower_gt_0 radix2); omega. + apply Int64.eqmod_divides with (2^64). apply Int64.eqm_signed_unsigned. + exists (2^(64-11)); auto. + exploit (Z_mod_lt (Int64.unsigned n) (2^11)). compute; auto. + assert (2^11 < Int64.max_unsigned) by (compute; auto). omega. + apply Int64.same_bits_eqm; auto. exists (-1); auto. + split. auto. assert (-2^64 < Int64.min_signed) by (compute; auto). + assert (Int64.max_signed < 2^64) by (compute; auto). + xomega. +Qed. + +End Float32. + +Global Opaque + Float.zero Float.eq_dec Float.neg Float.abs Float.of_single Float.to_single + Float.of_int Float.of_intu Float.of_long Float.of_longu + Float.to_int Float.to_intu Float.to_long Float.to_longu + Float.add Float.sub Float.mul Float.div Float.cmp + Float.to_bits Float.of_bits Float.from_words. + +Global Opaque + Float32.zero Float32.eq_dec Float32.neg Float32.abs + Float32.of_int Float32.of_intu Float32.of_long Float32.of_longu + Float32.to_int Float32.to_intu Float32.to_long Float32.to_longu + Float32.add Float32.sub Float32.mul Float32.div Float32.cmp + Float32.to_bits Float32.of_bits. + diff --git a/powerpc/Archi.v b/powerpc/Archi.v index 0b871d93..070f7cc9 100644 --- a/powerpc/Archi.v +++ b/powerpc/Archi.v @@ -25,10 +25,18 @@ Definition big_endian := true. Notation align_int64 := 8%Z (only parsing). Notation align_float64 := 8%Z (only parsing). -Program Definition default_pl : bool * nan_pl 53 := +Program Definition default_pl_64 : bool * nan_pl 53 := (false, nat_iter 51 xO xH). -Definition choose_binop_pl (s1: bool) (pl1: nan_pl 53) (s2: bool) (pl2: nan_pl 53) := +Definition choose_binop_pl_64 (s1: bool) (pl1: nan_pl 53) (s2: bool) (pl2: nan_pl 53) := false. (**r always choose first NaN *) -Global Opaque big_endian default_pl choose_binop_pl. +Program Definition default_pl_32 : bool * nan_pl 24 := + (false, nat_iter 22 xO xH). + +Definition choose_binop_pl_32 (s1: bool) (pl1: nan_pl 24) (s2: bool) (pl2: nan_pl 24) := + false. (**r always choose first NaN *) + +Global Opaque big_endian + default_pl_64 choose_binop_pl_64 + default_pl_32 choose_binop_pl_32. diff --git a/powerpc/Asm.v b/powerpc/Asm.v index aba78d45..a7e5eafe 100644 --- a/powerpc/Asm.v +++ b/powerpc/Asm.v @@ -165,20 +165,29 @@ Inductive instruction : Type := | Pextsh: ireg -> ireg -> instruction (**r 16-bit sign extension *) | Pfreeframe: Z -> int -> instruction (**r deallocate stack frame and restore previous frame *) | Pfabs: freg -> freg -> instruction (**r float absolute value *) + | Pfabss: freg -> freg -> instruction (**r float absolute value *) | Pfadd: freg -> freg -> freg -> instruction (**r float addition *) + | Pfadds: freg -> freg -> freg -> instruction (**r float addition *) | Pfcmpu: freg -> freg -> instruction (**r float comparison *) | Pfcti: ireg -> freg -> instruction (**r float-to-signed-int conversion *) | Pfdiv: freg -> freg -> freg -> instruction (**r float division *) + | Pfdivs: freg -> freg -> freg -> instruction (**r float division *) | Pfmake: freg -> ireg -> ireg -> instruction (**r build a float from 2 ints *) | Pfmr: freg -> freg -> instruction (**r float move *) | Pfmul: freg -> freg -> freg -> instruction (**r float multiply *) + | Pfmuls: freg -> freg -> freg -> instruction (**r float multiply *) | Pfneg: freg -> freg -> instruction (**r float negation *) + | Pfnegs: freg -> freg -> instruction (**r float negation *) | Pfrsp: freg -> freg -> instruction (**r float round to single precision *) + | Pfxdp: freg -> freg -> instruction (**r float extend to double precision (pseudo) *) | Pfsub: freg -> freg -> freg -> instruction (**r float subtraction *) + | Pfsubs: freg -> freg -> freg -> instruction (**r float subtraction *) | Plbz: ireg -> constant -> ireg -> instruction (**r load 8-bit unsigned int *) | Plbzx: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Plfd: freg -> constant -> ireg -> instruction (**r load 64-bit float *) | Plfdx: freg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) + | Plfd_a: freg -> constant -> ireg -> instruction (**r load 64-bit quantity to float reg *) + | Plfdx_a: freg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Plfs: freg -> constant -> ireg -> instruction (**r load 32-bit float *) | Plfsx: freg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Plha: ireg -> constant -> ireg -> instruction (**r load 16-bit signed int *) @@ -186,8 +195,11 @@ Inductive instruction : Type := | Plhz: ireg -> constant -> ireg -> instruction (**r load 16-bit unsigned int *) | Plhzx: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Plfi: freg -> float -> instruction (**r load float constant *) + | Plfis: freg -> float32 -> instruction (**r load float constant *) | Plwz: ireg -> constant -> ireg -> instruction (**r load 32-bit int *) | Plwzx: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) + | Plwz_a: ireg -> constant -> ireg -> instruction (**r load 32-bit quantity to int reg *) + | Plwzx_a: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Pmfcrbit: ireg -> crbit -> instruction (**r move condition bit to reg *) | Pmflr: ireg -> instruction (**r move LR to reg *) | Pmr: ireg -> ireg -> instruction (**r integer move *) @@ -196,7 +208,7 @@ Inductive instruction : Type := | Pmulli: ireg -> ireg -> constant -> instruction (**r integer multiply immediate *) | Pmullw: ireg -> ireg -> ireg -> instruction (**r integer multiply *) | Pmulhw: ireg -> ireg -> ireg -> instruction (**r multiply high signed *) - | Pmulhwu: ireg -> ireg -> ireg -> instruction (**r multiply high signed *) + | Pmulhwu: ireg -> ireg -> ireg -> instruction (**r multiply high signed *) | Pnand: ireg -> ireg -> ireg -> instruction (**r bitwise not-and *) | Pnor: ireg -> ireg -> ireg -> instruction (**r bitwise not-or *) | Por: ireg -> ireg -> ireg -> instruction (**r bitwise or *) @@ -213,12 +225,16 @@ Inductive instruction : Type := | Pstbx: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Pstfd: freg -> constant -> ireg -> instruction (**r store 64-bit float *) | Pstfdx: freg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) + | Pstfd_a: freg -> constant -> ireg -> instruction (**r store 64-bit quantity from float reg *) + | Pstfdx_a: freg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Pstfs: freg -> constant -> ireg -> instruction (**r store 32-bit float *) | Pstfsx: freg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Psth: ireg -> constant -> ireg -> instruction (**r store 16-bit int *) | Psthx: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Pstw: ireg -> constant -> ireg -> instruction (**r store 32-bit int *) | Pstwx: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) + | Pstw_a: ireg -> constant -> ireg -> instruction (**r store 32-bit quantity from int reg *) + | Pstwx_a: ireg -> ireg -> ireg -> instruction (**r same, with 2 index regs *) | Psubfc: ireg -> ireg -> ireg -> instruction (**r reversed integer subtraction *) | Psubfe: ireg -> ireg -> ireg -> instruction (**r reversed integer subtraction with carry *) | Psubfic: ireg -> ireg -> constant -> instruction (**r integer subtraction from immediate *) @@ -645,26 +661,40 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out end | Pfabs rd r1 => Next (nextinstr (rs#rd <- (Val.absf rs#r1))) m + | Pfabss rd r1 => + Next (nextinstr (rs#rd <- (Val.absfs rs#r1))) m | Pfadd rd r1 r2 => Next (nextinstr (rs#rd <- (Val.addf rs#r1 rs#r2))) m + | Pfadds rd r1 r2 => + Next (nextinstr (rs#rd <- (Val.addfs rs#r1 rs#r2))) m | Pfcmpu r1 r2 => Next (nextinstr (compare_float rs rs#r1 rs#r2)) m | Pfcti rd r1 => Next (nextinstr (rs#FPR13 <- Vundef #rd <- (Val.maketotal (Val.intoffloat rs#r1)))) m | Pfdiv rd r1 r2 => Next (nextinstr (rs#rd <- (Val.divf rs#r1 rs#r2))) m + | Pfdivs rd r1 r2 => + Next (nextinstr (rs#rd <- (Val.divfs rs#r1 rs#r2))) m | Pfmake rd r1 r2 => Next (nextinstr (rs#rd <- (Val.floatofwords rs#r1 rs#r2))) m | Pfmr rd r1 => Next (nextinstr (rs#rd <- (rs#r1))) m | Pfmul rd r1 r2 => Next (nextinstr (rs#rd <- (Val.mulf rs#r1 rs#r2))) m + | Pfmuls rd r1 r2 => + Next (nextinstr (rs#rd <- (Val.mulfs rs#r1 rs#r2))) m | Pfneg rd r1 => Next (nextinstr (rs#rd <- (Val.negf rs#r1))) m + | Pfnegs rd r1 => + Next (nextinstr (rs#rd <- (Val.negfs rs#r1))) m | Pfrsp rd r1 => Next (nextinstr (rs#rd <- (Val.singleoffloat rs#r1))) m + | Pfxdp rd r1 => + Next (nextinstr (rs#rd <- (Val.floatofsingle rs#r1))) m | Pfsub rd r1 r2 => Next (nextinstr (rs#rd <- (Val.subf rs#r1 rs#r2))) m + | Pfsubs rd r1 r2 => + Next (nextinstr (rs#rd <- (Val.subfs rs#r1 rs#r2))) m | Plbz rd cst r1 => load1 Mint8unsigned rd cst r1 rs m | Plbzx rd r1 r2 => @@ -673,6 +703,10 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out load1 Mfloat64 rd cst r1 rs m | Plfdx rd r1 r2 => load2 Mfloat64 rd r1 r2 rs m + | Plfd_a rd cst r1 => + load1 Many64 rd cst r1 rs m + | Plfdx_a rd r1 r2 => + load2 Many64 rd r1 r2 rs m | Plfs rd cst r1 => load1 Mfloat32 rd cst r1 rs m | Plfsx rd r1 r2 => @@ -687,10 +721,16 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out load2 Mint16unsigned rd r1 r2 rs m | Plfi rd f => Next (nextinstr (rs #GPR12 <- Vundef #rd <- (Vfloat f))) m + | Plfis rd f => + Next (nextinstr (rs #GPR12 <- Vundef #rd <- (Vsingle f))) m | Plwz rd cst r1 => load1 Mint32 rd cst r1 rs m | Plwzx rd r1 r2 => load2 Mint32 rd r1 r2 rs m + | Plwz_a rd cst r1 => + load1 Many32 rd cst r1 rs m + | Plwzx_a rd r1 r2 => + load2 Many32 rd r1 r2 rs m | Pmfcrbit rd bit => Next (nextinstr (rs#rd <- (rs#(reg_of_crbit bit)))) m | Pmflr rd => @@ -742,16 +782,14 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out store1 Mfloat64 rd cst r1 rs m | Pstfdx rd r1 r2 => store2 Mfloat64 rd r1 r2 rs m + | Pstfd_a rd cst r1 => + store1 Many64 rd cst r1 rs m + | Pstfdx_a rd r1 r2 => + store2 Many64 rd r1 r2 rs m | Pstfs rd cst r1 => - match store1 Mfloat32 rd cst r1 rs m with - | Next rs' m' => Next (rs'#FPR13 <- Vundef) m' - | Stuck => Stuck - end + store1 Mfloat32 rd cst r1 rs m | Pstfsx rd r1 r2 => - match store2 Mfloat32 rd r1 r2 rs m with - | Next rs' m' => Next (rs'#FPR13 <- Vundef) m' - | Stuck => Stuck - end + store2 Mfloat32 rd r1 r2 rs m | Psth rd cst r1 => store1 Mint16unsigned rd cst r1 rs m | Psthx rd r1 r2 => @@ -760,6 +798,10 @@ Definition exec_instr (f: function) (i: instruction) (rs: regset) (m: mem) : out store1 Mint32 rd cst r1 rs m | Pstwx rd r1 r2 => store2 Mint32 rd r1 r2 rs m + | Pstw_a rd cst r1 => + store1 Many32 rd cst r1 rs m + | Pstwx_a rd r1 r2 => + store2 Many32 rd r1 r2 rs m | Psubfc rd r1 r2 => Next (nextinstr (rs#rd <- (Val.sub rs#r2 rs#r1) #CARRY <- (Val.add_carry rs#r2 (Val.notint rs#r1) Vone))) m diff --git a/powerpc/Asmgen.v b/powerpc/Asmgen.v index 6b66686e..5ca770d1 100644 --- a/powerpc/Asmgen.v +++ b/powerpc/Asmgen.v @@ -122,40 +122,32 @@ Definition rolm (r1 r2: ireg) (amount mask: int) (k: code) := (** Accessing slots in the stack frame. *) +Definition accessind {A: Type} + (instr1: A -> constant -> ireg -> instruction) + (instr2: A -> ireg -> ireg -> instruction) + (base: ireg) (ofs: int) (r: A) (k: code) := + if Int.eq (high_s ofs) Int.zero + then instr1 r (Cint ofs) base :: k + else loadimm GPR0 ofs (instr2 r base GPR0 :: k). + Definition loadind (base: ireg) (ofs: int) (ty: typ) (dst: mreg) (k: code) := - match ty with - | Tint => - do r <- ireg_of dst; - OK (if Int.eq (high_s ofs) Int.zero then - Plwz r (Cint ofs) base :: k - else - loadimm GPR0 ofs (Plwzx r base GPR0 :: k)) - | Tfloat => - do r <- freg_of dst; - OK (if Int.eq (high_s ofs) Int.zero then - Plfd r (Cint ofs) base :: k - else - loadimm GPR0 ofs (Plfdx r base GPR0 :: k)) - | Tlong | Tsingle => - Error (msg "Asmgen.loadind") + match ty, preg_of dst with + | Tint, IR r => OK(accessind Plwz Plwzx base ofs r k) + | Tany32, IR r => OK(accessind Plwz_a Plwzx_a base ofs r k) + | Tsingle, FR r => OK(accessind Plfs Plfsx base ofs r k) + | Tfloat, FR r => OK(accessind Plfd Plfdx base ofs r k) + | Tany64, FR r => OK(accessind Plfd_a Plfdx_a base ofs r k) + | _, _ => Error (msg "Asmgen.loadind") end. Definition storeind (src: mreg) (base: ireg) (ofs: int) (ty: typ) (k: code) := - match ty with - | Tint => - do r <- ireg_of src; - OK (if Int.eq (high_s ofs) Int.zero then - Pstw r (Cint ofs) base :: k - else - loadimm GPR0 ofs (Pstwx r base GPR0 :: k)) - | Tfloat => - do r <- freg_of src; - OK (if Int.eq (high_s ofs) Int.zero then - Pstfd r (Cint ofs) base :: k - else - loadimm GPR0 ofs (Pstfdx r base GPR0 :: k)) - | Tlong | Tsingle => - Error (msg "Asmgen.storeind") + match ty, preg_of src with + | Tint, IR r => OK(accessind Pstw Pstwx base ofs r k) + | Tany32, IR r => OK(accessind Pstw_a Pstwx_a base ofs r k) + | Tsingle, FR r => OK(accessind Pstfs Pstfsx base ofs r k) + | Tfloat, FR r => OK(accessind Pstfd Pstfdx base ofs r k) + | Tany64, FR r => OK(accessind Pstfd_a Pstfdx_a base ofs r k) + | _, _ => Error (msg "Asmgen.storeind") end. (** Constructor for a floating-point comparison. The PowerPC has @@ -340,6 +332,8 @@ Definition transl_op do r <- ireg_of res; OK (loadimm r n k) | Ofloatconst f, nil => do r <- freg_of res; OK (Plfi r f :: k) + | Osingleconst f, nil => + do r <- freg_of res; OK (Plfis r f :: k) | Oaddrsymbol s ofs, nil => do r <- ireg_of res; OK (if symbol_is_small_data s ofs then @@ -477,9 +471,30 @@ Definition transl_op | Odivf, a1 :: a2 :: nil => do r1 <- freg_of a1; do r2 <- freg_of a2; do r <- freg_of res; OK (Pfdiv r r1 r2 :: k) + | Onegfs, a1 :: nil => + do r1 <- freg_of a1; do r <- freg_of res; + OK (Pfnegs r r1 :: k) + | Oabsfs, a1 :: nil => + do r1 <- freg_of a1; do r <- freg_of res; + OK (Pfabss r r1 :: k) + | Oaddfs, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; do r <- freg_of res; + OK (Pfadds r r1 r2 :: k) + | Osubfs, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; do r <- freg_of res; + OK (Pfsubs r r1 r2 :: k) + | Omulfs, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; do r <- freg_of res; + OK (Pfmuls r r1 r2 :: k) + | Odivfs, a1 :: a2 :: nil => + do r1 <- freg_of a1; do r2 <- freg_of a2; do r <- freg_of res; + OK (Pfdivs r r1 r2 :: k) | Osingleoffloat, a1 :: nil => do r1 <- freg_of a1; do r <- freg_of res; OK (Pfrsp r r1 :: k) + | Ofloatofsingle, a1 :: nil => + do r1 <- freg_of a1; do r <- freg_of res; + OK (Pfxdp r r1 :: k) | Ointoffloat, a1 :: nil => do r1 <- freg_of a1; do r <- ireg_of res; OK (Pfcti r r1 :: k) @@ -566,7 +581,7 @@ Definition transl_load (chunk: memory_chunk) (addr: addressing) | Mfloat64 => do r <- freg_of dst; transl_memory_access (Plfd r) (Plfdx r) addr args GPR12 k - | Mint64 => + | _ => Error (msg "Asmgen.transl_load") end. @@ -589,7 +604,7 @@ Definition transl_store (chunk: memory_chunk) (addr: addressing) | Mfloat64 => do r <- freg_of src; transl_memory_access (Pstfd r) (Pstfdx r) addr args temp k - | Mint64 => + | _ => Error (msg "Asmgen.transl_store") end. diff --git a/powerpc/Asmgenproof.v b/powerpc/Asmgenproof.v index 879d7524..913fb50a 100644 --- a/powerpc/Asmgenproof.v +++ b/powerpc/Asmgenproof.v @@ -196,17 +196,21 @@ Remark loadind_label: forall base ofs ty dst k c, loadind base ofs ty dst k = OK c -> tail_nolabel k c. Proof. - unfold loadind; intros. - destruct ty; destruct (Int.eq (high_s ofs) Int.zero); + unfold loadind, accessind; intros. + destruct ty; try discriminate; + destruct (preg_of dst); try discriminate; + destruct (Int.eq (high_s ofs) Int.zero); TailNoLabel; eapply tail_nolabel_trans; TailNoLabel. Qed. Remark storeind_label: forall base ofs ty src k c, - storeind base src ofs ty k = OK c -> tail_nolabel k c. + storeind src base ofs ty k = OK c -> tail_nolabel k c. Proof. - unfold storeind; intros. - destruct ty; destruct (Int.eq (high_s ofs) Int.zero); + unfold storeind, accessind; intros. + destruct ty; try discriminate; + destruct (preg_of src); try discriminate; + destruct (Int.eq (high_s ofs) Int.zero); TailNoLabel; eapply tail_nolabel_trans; TailNoLabel. Qed. diff --git a/powerpc/Asmgenproof1.v b/powerpc/Asmgenproof1.v index cfeb8232..e1ab9a15 100644 --- a/powerpc/Asmgenproof1.v +++ b/powerpc/Asmgenproof1.v @@ -435,6 +435,36 @@ Qed. (** Indexed memory loads. *) +Lemma accessind_load_correct: + forall (A: Type) (inj: A -> preg) + (instr1: A -> constant -> ireg -> instruction) + (instr2: A -> ireg -> ireg -> instruction) + (base: ireg) ofs rx chunk v (rs: regset) m k, + (forall rs m r1 cst r2, + exec_instr ge fn (instr1 r1 cst r2) rs m = load1 ge chunk (inj r1) cst r2 rs m) -> + (forall rs m r1 r2 r3, + exec_instr ge fn (instr2 r1 r2 r3) rs m = load2 chunk (inj r1) r2 r3 rs m) -> + Mem.loadv chunk m (Val.add rs#base (Vint ofs)) = Some v -> + base <> GPR0 -> inj rx <> PC -> + exists rs', + exec_straight ge fn (accessind instr1 instr2 base ofs rx k) rs m k rs' m + /\ rs'#(inj rx) = v + /\ forall r, r <> PC -> r <> inj rx -> r <> GPR0 -> rs'#r = rs#r. +Proof. + intros. unfold accessind. destruct (Int.eq (high_s ofs) Int.zero). +- econstructor; split. apply exec_straight_one. + rewrite H. unfold load1. rewrite gpr_or_zero_not_zero by auto. simpl. + rewrite H1. eauto. unfold nextinstr. repeat Simplif. + split. unfold nextinstr. repeat Simplif. + intros. repeat Simplif. +- exploit (loadimm_correct GPR0 ofs); eauto. intros [rs' [P [Q R]]]. + econstructor; split. eapply exec_straight_trans. eexact P. + apply exec_straight_one. rewrite H0. unfold load2. rewrite Q, R by auto with asmgen. + rewrite H1. reflexivity. unfold nextinstr. repeat Simplif. + split. repeat Simplif. + intros. repeat Simplif. +Qed. + Lemma loadind_correct: forall (base: ireg) ofs ty dst k (rs: regset) m v c, loadind base ofs ty dst k = OK c -> @@ -445,38 +475,44 @@ Lemma loadind_correct: /\ rs'#(preg_of dst) = v /\ forall r, r <> PC -> r <> preg_of dst -> r <> GPR0 -> rs'#r = rs#r. Proof. -Opaque Int.eq. - unfold loadind; intros. destruct ty; monadInv H; simpl in H0. -(* integer *) - rewrite (ireg_of_eq _ _ EQ). - destruct (Int.eq (high_s ofs) Int.zero). - (* one load *) - econstructor; split. apply exec_straight_one. simpl. - unfold load1. rewrite gpr_or_zero_not_zero; auto. simpl. rewrite H0. eauto. auto. - intuition Simpl. - (* loadimm + load *) - exploit (loadimm_correct GPR0 ofs); eauto. intros [rs' [A [B C]]]. - exists (nextinstr (rs'#x <- v)); split. - eapply exec_straight_trans. eexact A. apply exec_straight_one; auto. - simpl. unfold load2. rewrite C; auto with asmgen. rewrite B. rewrite H0. auto. - intuition Simpl. -(* float *) - rewrite (freg_of_eq _ _ EQ). - destruct (Int.eq (high_s ofs) Int.zero). - (* one load *) - econstructor; split. apply exec_straight_one. simpl. - unfold load1. rewrite gpr_or_zero_not_zero; auto. simpl. rewrite H0. eauto. auto. - intuition Simpl. - (* loadimm + load *) - exploit (loadimm_correct GPR0 ofs); eauto. intros [rs' [A [B C]]]. - exists (nextinstr (rs'#x <- v)); split. - eapply exec_straight_trans. eexact A. apply exec_straight_one; auto. - simpl. unfold load2. rewrite C; auto with asmgen. rewrite B. rewrite H0. auto. - intuition Simpl. + unfold loadind; intros. destruct ty; try discriminate; destruct (preg_of dst); inv H; simpl in H0. + apply accessind_load_correct with (inj := IR) (chunk := Mint32); auto with asmgen. + apply accessind_load_correct with (inj := FR) (chunk := Mfloat64); auto with asmgen. + apply accessind_load_correct with (inj := FR) (chunk := Mfloat32); auto with asmgen. + apply accessind_load_correct with (inj := IR) (chunk := Many32); auto with asmgen. + apply accessind_load_correct with (inj := FR) (chunk := Many64); auto with asmgen. Qed. (** Indexed memory stores. *) +Lemma accessind_store_correct: + forall (A: Type) (inj: A -> preg) + (instr1: A -> constant -> ireg -> instruction) + (instr2: A -> ireg -> ireg -> instruction) + (base: ireg) ofs rx chunk (rs: regset) m m' k, + (forall rs m r1 cst r2, + exec_instr ge fn (instr1 r1 cst r2) rs m = store1 ge chunk (inj r1) cst r2 rs m) -> + (forall rs m r1 r2 r3, + exec_instr ge fn (instr2 r1 r2 r3) rs m = store2 chunk (inj r1) r2 r3 rs m) -> + Mem.storev chunk m (Val.add rs#base (Vint ofs)) (rs (inj rx)) = Some m' -> + base <> GPR0 -> inj rx <> PC -> inj rx <> GPR0 -> + exists rs', + exec_straight ge fn (accessind instr1 instr2 base ofs rx k) rs m k rs' m' + /\ forall r, r <> PC -> r <> GPR0 -> rs'#r = rs#r. +Proof. + intros. unfold accessind. destruct (Int.eq (high_s ofs) Int.zero). +- econstructor; split. apply exec_straight_one. + rewrite H. unfold store1. rewrite gpr_or_zero_not_zero by auto. simpl. + rewrite H1. eauto. unfold nextinstr. repeat Simplif. + intros. repeat Simplif. +- exploit (loadimm_correct GPR0 ofs); eauto. intros [rs' [P [Q R]]]. + econstructor; split. eapply exec_straight_trans. eexact P. + apply exec_straight_one. rewrite H0. unfold store2. + rewrite Q. rewrite R by auto with asmgen. rewrite R by auto. + rewrite H1. reflexivity. unfold nextinstr. repeat Simplif. + intros. repeat Simplif. +Qed. + Lemma storeind_correct: forall (base: ireg) ofs ty src k (rs: regset) m m' c, storeind src base ofs ty k = OK c -> @@ -488,33 +524,12 @@ Lemma storeind_correct: Proof. unfold storeind; intros. assert (preg_of src <> GPR0) by auto with asmgen. - destruct ty; monadInv H; simpl in H0. -(* integer *) - rewrite (ireg_of_eq _ _ EQ) in *. - destruct (Int.eq (high_s ofs) Int.zero). - (* one store *) - econstructor; split. apply exec_straight_one. simpl. - unfold store1. rewrite gpr_or_zero_not_zero; auto. simpl. rewrite H0. eauto. auto. - intros; Simpl. - (* loadimm + store *) - exploit (loadimm_correct GPR0 ofs); eauto. intros [rs' [A [B C]]]. - exists (nextinstr rs'); split. - eapply exec_straight_trans. eexact A. apply exec_straight_one; auto. - simpl. unfold store2. rewrite B. rewrite ! C; auto with asmgen. rewrite H0. auto. - intuition Simpl. -(* float *) - rewrite (freg_of_eq _ _ EQ) in *. - destruct (Int.eq (high_s ofs) Int.zero). - (* one store *) - econstructor; split. apply exec_straight_one. simpl. - unfold store1. rewrite gpr_or_zero_not_zero; auto. simpl. rewrite H0. eauto. auto. - intuition Simpl. - (* loadimm + store *) - exploit (loadimm_correct GPR0 ofs); eauto. intros [rs' [A [B C]]]. - exists (nextinstr rs'); split. - eapply exec_straight_trans. eexact A. apply exec_straight_one; auto. - simpl. unfold store2. rewrite B. rewrite ! C; auto with asmgen. rewrite H0. auto. - intuition Simpl. + destruct ty; try discriminate; destruct (preg_of src) ; inv H; simpl in H0. + apply accessind_store_correct with (inj := IR) (chunk := Mint32); auto with asmgen. + apply accessind_store_correct with (inj := FR) (chunk := Mfloat64); auto with asmgen. + apply accessind_store_correct with (inj := FR) (chunk := Mfloat32); auto with asmgen. + apply accessind_store_correct with (inj := IR) (chunk := Many32); auto with asmgen. + apply accessind_store_correct with (inj := FR) (chunk := Many64); auto with asmgen. Qed. (** Float comparisons. *) @@ -1210,15 +1225,7 @@ Local Transparent destroyed_by_store. - (* Mint32 *) eapply BASE; eauto; erewrite ireg_of_eq by eauto; auto. - (* Mfloat32 *) - rewrite (freg_of_eq _ _ EQ) in H1. - eapply transl_memory_access_correct. eauto. eauto. eauto. - intros. econstructor; split. apply exec_straight_one. - simpl. unfold store1. rewrite H. rewrite H2; auto with asmgen. rewrite H1. eauto. auto. -Local Transparent destroyed_by_store. - simpl; intros. destruct H5 as [A [B C]]. Simpl. apply H2; auto with asmgen. destruct TEMP0; congruence. - intros. econstructor; split. apply exec_straight_one. - simpl. unfold store2. rewrite H. rewrite H2; auto with asmgen. rewrite H1. eauto. auto. - simpl; intros. destruct H5 as [A [B C]]. Simpl. apply H2; auto with asmgen. destruct TEMP0; congruence. + eapply BASE; eauto; erewrite freg_of_eq by eauto; auto. - (* Mfloat64 *) eapply BASE; eauto; erewrite freg_of_eq by eauto; auto. Qed. diff --git a/powerpc/ConstpropOp.vp b/powerpc/ConstpropOp.vp index e1e19601..bba0fad4 100644 --- a/powerpc/ConstpropOp.vp +++ b/powerpc/ConstpropOp.vp @@ -136,18 +136,19 @@ Definition make_xorimm (n: int) (r: reg) := else (Oxorimm n, r :: nil). Definition make_mulfimm (n: float) (r r1 r2: reg) := - if Float.eq_dec n (Float.floatofint (Int.repr 2)) + if Float.eq_dec n (Float.of_int (Int.repr 2)) then (Oaddf, r :: r :: nil) else (Omulf, r1 :: r2 :: nil). +Definition make_mulfsimm (n: float32) (r r1 r2: reg) := + if Float32.eq_dec n (Float32.of_int (Int.repr 2)) + then (Oaddfs, r :: r :: nil) + else (Omulfs, r1 :: r2 :: nil). + Definition make_cast8signed (r: reg) (a: aval) := if vincl a (Sgn 8) then (Omove, r :: nil) else (Ocast8signed, r :: nil). Definition make_cast16signed (r: reg) (a: aval) := if vincl a (Sgn 16) then (Omove, r :: nil) else (Ocast16signed, r :: nil). -Definition make_singleoffloat (r: reg) (a: aval) := - if vincl a Fsingle && generate_float_constants tt - then (Omove, r :: nil) - else (Osingleoffloat, r :: nil). Nondetfunction op_strength_reduction (op: operation) (args: list reg) (vl: list aval) := @@ -174,10 +175,11 @@ Nondetfunction op_strength_reduction | Oshl, r1 :: r2 :: nil, v1 :: I n2 :: nil => make_shlimm n2 r1 r2 | Oshr, r1 :: r2 :: nil, v1 :: I n2 :: nil => make_shrimm n2 r1 r2 | Oshru, r1 :: r2 :: nil, v1 :: I n2 :: nil => make_shruimm n2 r1 r2 - | Osingleoffloat, r1 :: nil, v1 :: nil => make_singleoffloat r1 v1 | Ocmp c, args, vl => make_cmp c args vl | Omulf, r1 :: r2 :: nil, v1 :: F n2 :: nil => make_mulfimm n2 r1 r1 r2 | Omulf, r1 :: r2 :: nil, F n1 :: v2 :: nil => make_mulfimm n1 r2 r1 r2 + | Omulfs, r1 :: r2 :: nil, v1 :: FS n2 :: nil => make_mulfsimm n2 r1 r1 r2 + | Omulfs, r1 :: r2 :: nil, FS n1 :: v2 :: nil => make_mulfsimm n1 r2 r1 r2 | _, _, _ => (op, args) end. diff --git a/powerpc/ConstpropOpproof.v b/powerpc/ConstpropOpproof.v index 584865ad..8498868a 100644 --- a/powerpc/ConstpropOpproof.v +++ b/powerpc/ConstpropOpproof.v @@ -80,6 +80,10 @@ Ltac SimplVM := let E := fresh in assert (E: v = Vfloat n) by (inversion H; auto); rewrite E in *; clear H; SimplVM + | [ H: vmatch _ ?v (FS ?n) |- _ ] => + let E := fresh in + assert (E: v = Vsingle n) by (inversion H; auto); + rewrite E in *; clear H; SimplVM | [ H: vmatch _ ?v (Ptr(Gl ?id ?ofs)) |- _ ] => let E := fresh in assert (E: Val.lessdef v (Genv.symbol_address ge id ofs)) by (eapply vmatch_ptr_gl; eauto); @@ -315,7 +319,7 @@ Lemma make_mulfimm_correct: exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulf rs#r1 rs#r2) v. Proof. intros; unfold make_mulfimm. - destruct (Float.eq_dec n (Float.floatofint (Int.repr 2))); intros. + destruct (Float.eq_dec n (Float.of_int (Int.repr 2))); intros. simpl. econstructor; split. eauto. rewrite H; subst n. destruct (rs#r1); simpl; auto. rewrite Float.mul2_add; auto. simpl. econstructor; split; eauto. @@ -328,13 +332,40 @@ Lemma make_mulfimm_correct_2: exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulf rs#r1 rs#r2) v. Proof. intros; unfold make_mulfimm. - destruct (Float.eq_dec n (Float.floatofint (Int.repr 2))); intros. + destruct (Float.eq_dec n (Float.of_int (Int.repr 2))); intros. simpl. econstructor; split. eauto. rewrite H; subst n. destruct (rs#r2); simpl; auto. rewrite Float.mul2_add; auto. rewrite Float.mul_commut; auto. simpl. econstructor; split; eauto. Qed. +Lemma make_mulfsimm_correct: + forall n r1 r2, + rs#r2 = Vsingle n -> + let (op, args) := make_mulfsimm n r1 r1 r2 in + exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulfs rs#r1 rs#r2) v. +Proof. + intros; unfold make_mulfsimm. + destruct (Float32.eq_dec n (Float32.of_int (Int.repr 2))); intros. + simpl. econstructor; split. eauto. rewrite H; subst n. + destruct (rs#r1); simpl; auto. rewrite Float32.mul2_add; auto. + simpl. econstructor; split; eauto. +Qed. + +Lemma make_mulfsimm_correct_2: + forall n r1 r2, + rs#r1 = Vsingle n -> + let (op, args) := make_mulfsimm n r2 r1 r2 in + exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.mulfs rs#r1 rs#r2) v. +Proof. + intros; unfold make_mulfsimm. + destruct (Float32.eq_dec n (Float32.of_int (Int.repr 2))); intros. + simpl. econstructor; split. eauto. rewrite H; subst n. + destruct (rs#r2); simpl; auto. rewrite Float32.mul2_add; auto. + rewrite Float32.mul_commut; auto. + simpl. econstructor; split; eauto. +Qed. + Lemma make_cast8signed_correct: forall r x, vmatch bc rs#r x -> @@ -363,21 +394,6 @@ Proof. econstructor; split; simpl; eauto. Qed. -Lemma make_singleoffloat_correct: - forall r x, - vmatch bc rs#r x -> - let (op, args) := make_singleoffloat r x in - exists v, eval_operation ge (Vptr sp Int.zero) op rs##args m = Some v /\ Val.lessdef (Val.singleoffloat rs#r) v. -Proof. - intros; unfold make_singleoffloat. - destruct (vincl x Fsingle && generate_float_constants tt) eqn:INCL. - InvBooleans. exists rs#r; split; auto. - assert (V: vmatch bc rs#r Fsingle). - { eapply vmatch_ge; eauto. apply vincl_ge; auto. } - inv V; simpl; auto. rewrite Float.singleoffloat_of_single by auto. auto. - econstructor; split; simpl; eauto. -Qed. - Lemma op_strength_reduction_correct: forall op args vl v, vl = map (fun r => AE.get r ae) args -> @@ -424,14 +440,16 @@ Proof. InvApproxRegs; SimplVM; inv H0. apply make_shrimm_correct; auto. (* shru *) InvApproxRegs; SimplVM; inv H0. apply make_shruimm_correct; auto. -(* singleoffloat *) - InvApproxRegs; SimplVM; inv H0. apply make_singleoffloat_correct; auto. (* cmp *) inv H0. apply make_cmp_correct; auto. (* mulf *) InvApproxRegs; SimplVM; inv H0. rewrite <- H2. apply make_mulfimm_correct; auto. InvApproxRegs; SimplVM; inv H0. fold (Val.mulf (Vfloat n1) rs#r2). rewrite <- H2. apply make_mulfimm_correct_2; auto. +(* mulfs *) + InvApproxRegs; SimplVM; inv H0. rewrite <- H2. apply make_mulfsimm_correct; auto. + InvApproxRegs; SimplVM; inv H0. fold (Val.mulfs (Vsingle n1) rs#r2). + rewrite <- H2. apply make_mulfsimm_correct_2; auto. (* default *) exists v; auto. Qed. diff --git a/powerpc/Machregs.v b/powerpc/Machregs.v index baad496d..f7ed7793 100644 --- a/powerpc/Machregs.v +++ b/powerpc/Machregs.v @@ -54,23 +54,13 @@ Global Opaque mreg_eq. Definition mreg_type (r: mreg): typ := match r with - | R3 => Tint | R4 => Tint | R5 => Tint | R6 => Tint - | R7 => Tint | R8 => Tint | R9 => Tint | R10 => Tint - | R11 => Tint | R12 => Tint - | R14 => Tint | R15 => Tint | R16 => Tint - | R17 => Tint | R18 => Tint | R19 => Tint | R20 => Tint - | R21 => Tint | R22 => Tint | R23 => Tint | R24 => Tint - | R25 => Tint | R26 => Tint | R27 => Tint | R28 => Tint - | R29 => Tint | R30 => Tint | R31 => Tint - | F0 => Tfloat - | F1 => Tfloat | F2 => Tfloat | F3 => Tfloat | F4 => Tfloat - | F5 => Tfloat | F6 => Tfloat | F7 => Tfloat | F8 => Tfloat - | F9 => Tfloat | F10 => Tfloat | F11 => Tfloat | F12 => Tfloat - | F13 => Tfloat | F14 => Tfloat | F15 => Tfloat - | F16 => Tfloat | F17 => Tfloat | F18 => Tfloat | F19 => Tfloat - | F20 => Tfloat | F21 => Tfloat | F22 => Tfloat | F23 => Tfloat - | F24 => Tfloat | F25 => Tfloat | F26 => Tfloat | F27 => Tfloat - | F28 => Tfloat | F29 => Tfloat | F30 => Tfloat | F31 => Tfloat + | R3 | R4 | R5 | R6 | R7 | R8 | R9 | R10 | R11 | R12 + | R14 | R15 | R16 | R17 | R18 | R19 | R20 | R21 | R22 | R23 | R24 + | R25 | R26 | R27 | R28 | R29 | R30 | R31 => Tany32 + | F0 | F1 | F2 | F3 | F4 | F5 | F6 | F7 + | F8 | F9 | F10 | F11 | F12 | F13 | F14 | F15 + | F16 | F17 | F18 | F19 | F20 | F21 | F22 | F23 + | F24 | F25 | F26 | F27 | F28 | F29 | F30 | F31 => Tany64 end. Open Scope positive_scope. @@ -112,6 +102,7 @@ Definition is_stack_reg (r: mreg) : bool := false. Definition destroyed_by_op (op: operation): list mreg := match op with | Ofloatconst _ => R12 :: nil + | Osingleconst _ => R12 :: nil | Ointoffloat => F13 :: nil | _ => nil end. @@ -120,10 +111,7 @@ Definition destroyed_by_load (chunk: memory_chunk) (addr: addressing): list mreg R12 :: nil. Definition destroyed_by_store (chunk: memory_chunk) (addr: addressing): list mreg := - match chunk with - | Mfloat32 => R11 :: R12 :: F13 :: nil - | _ => R11 :: R12 :: nil - end. + R11 :: R12 :: nil. Definition destroyed_by_cond (cond: condition): list mreg := nil. @@ -135,21 +123,16 @@ Definition destroyed_by_builtin (ef: external_function): list mreg := match ef with | EF_builtin _ _ => F13 :: nil | EF_vload _ => nil - | EF_vstore Mfloat32 => F13 :: nil | EF_vstore _ => nil | EF_vload_global _ _ _ => R11 :: nil | EF_vstore_global Mint64 _ _ => R10 :: R11 :: R12 :: nil - | EF_vstore_global Mfloat32 _ _ => R11 :: R12 :: F13 :: nil | EF_vstore_global _ _ _ => R11 :: R12 :: nil | EF_memcpy _ _ => R11 :: R12 :: F13 :: nil | _ => nil end. Definition destroyed_by_setstack (ty: typ): list mreg := - match ty with - | Tsingle => F13 :: nil - | _ => nil - end. + nil. Definition destroyed_at_function_entry: list mreg := nil. diff --git a/powerpc/NeedOp.v b/powerpc/NeedOp.v index d43af3c8..e1307492 100644 --- a/powerpc/NeedOp.v +++ b/powerpc/NeedOp.v @@ -25,6 +25,7 @@ Definition needs_of_operation (op: operation) (nv: nval): list nval := | Omove => op1 nv | Ointconst n => nil | Ofloatconst n => nil + | Osingleconst n => nil | Oaddrsymbol id ofs => nil | Oaddrstack ofs => nil | Ocast8signed => op1 (sign_ext 8 nv) @@ -52,7 +53,9 @@ Definition needs_of_operation (op: operation) (nv: nval): list nval := | Oroli amount mask => op1 (default nv) | Onegf | Oabsf => op1 (default nv) | Oaddf | Osubf | Omulf | Odivf => op2 (default nv) - | Osingleoffloat => op1 (singleoffloat nv) + | Onegfs | Oabsfs => op1 (default nv) + | Oaddfs | Osubfs | Omulfs | Odivfs => op2 (default nv) + | Osingleoffloat | Ofloatofsingle => op1 (default nv) | Ointoffloat => op1 (default nv) | Ofloatofwords | Omakelong => op2 (default nv) | Olowlong | Ohighlong => op1 (default nv) @@ -66,7 +69,6 @@ Definition operation_is_redundant (op: operation) (nv: nval): bool := | Oandimm n => andimm_redundant nv n | Oorimm n => orimm_redundant nv n | Orolm amount mask => rolm_redundant nv amount mask - | Osingleoffloat => singleoffloat_redundant nv | _ => false end. @@ -136,7 +138,6 @@ Proof. - apply or_sound; auto. apply notint_sound; rewrite bitwise_idem; auto. - apply shrimm_sound; auto. - apply rolm_sound; auto. -- apply singleoffloat_sound; auto. - destruct (eval_condition c args m) as [b|] eqn:EC; simpl in H2. erewrite needs_of_condition_sound by eauto. subst v; simpl. auto with na. @@ -156,7 +157,6 @@ Proof. - apply andimm_redundant_sound; auto. - apply orimm_redundant_sound; auto. - apply rolm_redundant_sound; auto. -- apply singleoffloat_redundant_sound; auto. Qed. End SOUNDNESS. diff --git a/powerpc/Op.v b/powerpc/Op.v index 17cf0725..dbec2755 100644 --- a/powerpc/Op.v +++ b/powerpc/Op.v @@ -54,6 +54,7 @@ Inductive operation : Type := | Omove: operation (**r [rd = r1] *) | Ointconst: int -> operation (**r [rd] is set to the given integer constant *) | Ofloatconst: float -> operation (**r [rd] is set to the given float constant *) + | Osingleconst: float32 -> operation (**r [rd] is set to the given float constant *) | Oaddrsymbol: ident -> int -> operation (**r [rd] is set to the the address of the symbol plus the offset *) | Oaddrstack: int -> operation (**r [rd] is set to the stack pointer plus the given offset *) (*c Integer arithmetic: *) @@ -96,7 +97,14 @@ Inductive operation : Type := | Osubf: operation (**r [rd = r1 - r2] *) | Omulf: operation (**r [rd = r1 * r2] *) | Odivf: operation (**r [rd = r1 / r2] *) + | Onegfs: operation (**r [rd = - r1] *) + | Oabsfs: operation (**r [rd = abs(r1)] *) + | Oaddfs: operation (**r [rd = r1 + r2] *) + | Osubfs: operation (**r [rd = r1 - r2] *) + | Omulfs: operation (**r [rd = r1 * r2] *) + | Odivfs: operation (**r [rd = r1 / r2] *) | Osingleoffloat: operation (**r [rd] is [r1] truncated to single-precision float *) + | Ofloatofsingle: operation (**r [rd] is [r1] extended to double-precision float *) (*c Conversions between int and float: *) | Ointoffloat: operation (**r [rd = signed_int_of_float(r1)] *) | Ofloatofwords: operation (**r [rd = float_of_words(r1,r2)] *) @@ -130,7 +138,7 @@ Defined. Definition eq_operation (x y: operation): {x=y} + {x<>y}. Proof. generalize Int.eq_dec; intro. - generalize Float.eq_dec; intro. + generalize Float.eq_dec Float32.eq_dec; intros. assert (forall (x y: ident), {x=y}+{x<>y}). exact peq. generalize eq_condition; intro. decide equality. @@ -172,6 +180,7 @@ Definition eval_operation | Omove, v1::nil => Some v1 | Ointconst n, nil => Some (Vint n) | Ofloatconst n, nil => Some (Vfloat n) + | Osingleconst n, nil => Some (Vsingle n) | Oaddrsymbol s ofs, nil => Some (Genv.symbol_address genv s ofs) | Oaddrstack ofs, nil => Some (Val.add sp (Vint ofs)) | Ocast8signed, v1::nil => Some (Val.sign_ext 8 v1) @@ -213,7 +222,14 @@ Definition eval_operation | Osubf, v1::v2::nil => Some(Val.subf v1 v2) | Omulf, v1::v2::nil => Some(Val.mulf v1 v2) | Odivf, v1::v2::nil => Some(Val.divf v1 v2) + | Onegfs, v1::nil => Some(Val.negfs v1) + | Oabsfs, v1::nil => Some(Val.absfs v1) + | Oaddfs, v1::v2::nil => Some(Val.addfs v1 v2) + | Osubfs, v1::v2::nil => Some(Val.subfs v1 v2) + | Omulfs, v1::v2::nil => Some(Val.mulfs v1 v2) + | Odivfs, v1::v2::nil => Some(Val.divfs v1 v2) | Osingleoffloat, v1::nil => Some(Val.singleoffloat v1) + | Ofloatofsingle, v1::nil => Some(Val.floatofsingle v1) | Ointoffloat, v1::nil => Val.intoffloat v1 | Ofloatofwords, v1::v2::nil => Some(Val.floatofwords v1 v2) | Omakelong, v1::v2::nil => Some(Val.longofwords v1 v2) @@ -265,7 +281,8 @@ Definition type_of_operation (op: operation) : list typ * typ := match op with | Omove => (nil, Tint) (* treated specially *) | Ointconst _ => (nil, Tint) - | Ofloatconst f => (nil, if Float.is_single_dec f then Tsingle else Tfloat) + | Ofloatconst f => (nil, Tfloat) + | Osingleconst f => (nil, Tsingle) | Oaddrsymbol _ _ => (nil, Tint) | Oaddrstack _ => (nil, Tint) | Ocast8signed => (Tint :: nil, Tint) @@ -306,7 +323,14 @@ Definition type_of_operation (op: operation) : list typ * typ := | Osubf => (Tfloat :: Tfloat :: nil, Tfloat) | Omulf => (Tfloat :: Tfloat :: nil, Tfloat) | Odivf => (Tfloat :: Tfloat :: nil, Tfloat) + | Onegfs => (Tsingle :: nil, Tsingle) + | Oabsfs => (Tsingle :: nil, Tsingle) + | Oaddfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Osubfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Omulfs => (Tsingle :: Tsingle :: nil, Tsingle) + | Odivfs => (Tsingle :: Tsingle :: nil, Tsingle) | Osingleoffloat => (Tfloat :: nil, Tsingle) + | Ofloatofsingle => (Tsingle :: nil, Tfloat) | Ointoffloat => (Tfloat :: nil, Tint) | Ofloatofwords => (Tint :: Tint :: nil, Tfloat) | Omakelong => (Tint :: Tint :: nil, Tlong) @@ -343,7 +367,8 @@ Proof with (try exact I). destruct op; simpl in H0; FuncInv; subst; simpl. congruence. exact I. - destruct (Float.is_single_dec f); auto. + auto. + auto. unfold Genv.symbol_address. destruct (Genv.find_symbol genv i)... destruct sp... destruct v0... @@ -386,8 +411,15 @@ Proof with (try exact I). destruct v0; destruct v1... destruct v0; destruct v1... destruct v0; destruct v1... - destruct v0... simpl. apply Float.singleoffloat_is_single. - destruct v0; simpl in H0; inv H0. destruct (Float.intoffloat f); inv H2... + destruct v0... + destruct v0... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0; destruct v1... + destruct v0... + destruct v0... + destruct v0; simpl in H0; inv H0. destruct (Float.to_int f); inv H2... destruct v0; destruct v1... destruct v0; destruct v1... destruct v0... @@ -521,35 +553,6 @@ Proof. rewrite Val.add_assoc. auto. Qed. -(** Transformation of addressing modes with two operands or more - into an equivalent arithmetic operation. This is used in the [Reload] - pass when a store instruction cannot be reloaded directly because - it runs out of temporary registers. *) - -(** For the PowerPC, there is only one binary addressing mode: [Aindexed2]. - The corresponding operation is [Oadd]. *) - -Definition op_for_binary_addressing (addr: addressing) : operation := Oadd. - -Lemma eval_op_for_binary_addressing: - forall (F V: Type) (ge: Genv.t F V) sp addr args v m, - (length args >= 2)%nat -> - eval_addressing ge sp addr args = Some v -> - eval_operation ge sp (op_for_binary_addressing addr) args m = Some v. -Proof. - intros. - destruct addr; simpl in H0; FuncInv; simpl in H; try omegaContradiction. - simpl; congruence. -Qed. - -Lemma type_op_for_binary_addressing: - forall addr, - (length (type_of_addressing addr) >= 2)%nat -> - type_of_operation (op_for_binary_addressing addr) = (type_of_addressing addr, Tint). -Proof. - intros. destruct addr; simpl in H; reflexivity || omegaContradiction. -Qed. - (** Operations that are so cheap to recompute that CSE should not factor them out. *) Definition is_trivial_op (op: operation) : bool := @@ -578,61 +581,6 @@ Proof. destruct c; simpl; auto; discriminate. Qed. -(** Checking whether two addressings, applied to the same arguments, produce - separated memory addresses. Used in [CSE]. *) - -Definition addressing_separated (chunk1: memory_chunk) (addr1: addressing) - (chunk2: memory_chunk) (addr2: addressing) : bool := - match addr1, addr2 with - | Aindexed ofs1, Aindexed ofs2 => - Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) - | Aglobal s1 ofs1, Aglobal s2 ofs2 => - if ident_eq s1 s2 then Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) else true - | Abased s1 ofs1, Abased s2 ofs2 => - if ident_eq s1 s2 then Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) else true - | Ainstack ofs1, Ainstack ofs2 => - Int.no_overlap ofs1 (size_chunk chunk1) ofs2 (size_chunk chunk2) - | _, _ => false - end. - -Lemma addressing_separated_sound: - forall (F V: Type) (ge: Genv.t F V) sp chunk1 addr1 chunk2 addr2 vl b1 n1 b2 n2, - addressing_separated chunk1 addr1 chunk2 addr2 = true -> - eval_addressing ge sp addr1 vl = Some(Vptr b1 n1) -> - eval_addressing ge sp addr2 vl = Some(Vptr b2 n2) -> - b1 <> b2 \/ Int.unsigned n1 + size_chunk chunk1 <= Int.unsigned n2 \/ Int.unsigned n2 + size_chunk chunk2 <= Int.unsigned n1. -Proof. - unfold addressing_separated; intros. - generalize (size_chunk_pos chunk1) (size_chunk_pos chunk2); intros SZ1 SZ2. - destruct addr1; destruct addr2; try discriminate; simpl in *; FuncInv. -(* Aindexed *) - destruct v; simpl in *; inv H1; inv H2. - right. apply Int.no_overlap_sound; auto. -(* Aglobal *) - unfold Genv.symbol_address in *. - destruct (Genv.find_symbol ge i1) eqn:?; inv H2. - destruct (Genv.find_symbol ge i) eqn:?; inv H1. - destruct (ident_eq i i1). subst. - replace (Int.unsigned n1) with (Int.unsigned (Int.add Int.zero n1)). - replace (Int.unsigned n2) with (Int.unsigned (Int.add Int.zero n2)). - right. apply Int.no_overlap_sound; auto. - rewrite Int.add_commut; rewrite Int.add_zero; auto. - rewrite Int.add_commut; rewrite Int.add_zero; auto. - left. red; intros; elim n. subst. eapply Genv.genv_vars_inj; eauto. -(* Abased *) - unfold Genv.symbol_address in *. - destruct (Genv.find_symbol ge i1) eqn:?; simpl in *; try discriminate. - destruct v; inv H2. - destruct (Genv.find_symbol ge i) eqn:?; inv H1. - destruct (ident_eq i i1). subst. - rewrite (Int.add_commut i0 i3). rewrite (Int.add_commut i2 i3). - right. apply Int.no_overlap_sound; auto. - left. red; intros; elim n. subst. eapply Genv.genv_vars_inj; eauto. -(* Ainstack *) - destruct sp; simpl in *; inv H1; inv H2. - right. apply Int.no_overlap_sound; auto. -Qed. - (** * Invariance and compatibility properties. *) (** [eval_operation] and [eval_addressing] depend on a global environment @@ -719,6 +667,8 @@ Ltac InvInject := inv H; InvInject | [ H: val_inject _ (Vfloat _) _ |- _ ] => inv H; InvInject + | [ H: val_inject _ (Vsingle _) _ |- _ ] => + inv H; InvInject | [ H: val_inject _ (Vptr _ _) _ |- _ ] => inv H; InvInject | [ H: val_list_inject _ nil _ |- _ ] => @@ -806,7 +756,14 @@ Proof. inv H4; inv H2; simpl; auto. inv H4; inv H2; simpl; auto. inv H4; simpl; auto. - inv H4; simpl in H1; inv H1. simpl. destruct (Float.intoffloat f0); simpl in H2; inv H2. + inv H4; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; inv H2; simpl; auto. + inv H4; simpl; auto. + inv H4; simpl; auto. + inv H4; simpl in H1; inv H1. simpl. destruct (Float.to_int f0); simpl in H2; inv H2. exists (Vint i); auto. inv H4; inv H2; simpl; auto. inv H4; inv H2; simpl; auto. diff --git a/powerpc/PrintAsm.ml b/powerpc/PrintAsm.ml index b90b9f21..e3f07244 100644 --- a/powerpc/PrintAsm.ml +++ b/powerpc/PrintAsm.ml @@ -395,11 +395,11 @@ let print_builtin_vload_common oc chunk base offset res = fprintf oc " lhz %a, %a(%a)\n" ireg res constant offset ireg base | Mint16signed, IR res -> fprintf oc " lha %a, %a(%a)\n" ireg res constant offset ireg base - | Mint32, IR res -> + | (Mint32 | Many32), IR res -> fprintf oc " lwz %a, %a(%a)\n" ireg res constant offset ireg base | Mfloat32, FR res -> fprintf oc " lfs %a, %a(%a)\n" freg res constant offset ireg base - | Mfloat64, FR res -> + | (Mfloat64 | Many64), FR res -> fprintf oc " lfd %a, %a(%a)\n" freg res constant offset ireg base (* Mint64 is special-cased below *) | _ -> @@ -451,12 +451,11 @@ let print_builtin_vstore_common oc chunk base offset src = fprintf oc " stb %a, %a(%a)\n" ireg src constant offset ireg base | (Mint16signed | Mint16unsigned), IR src -> fprintf oc " sth %a, %a(%a)\n" ireg src constant offset ireg base - | Mint32, IR src -> + | (Mint32 | Many32), IR src -> fprintf oc " stw %a, %a(%a)\n" ireg src constant offset ireg base | Mfloat32, FR src -> - fprintf oc " frsp %a, %a\n" freg FPR13 freg src; - fprintf oc " stfs %a, %a(%a)\n" freg FPR13 constant offset ireg base - | Mfloat64, FR src -> + fprintf oc " stfs %a, %a(%a)\n" freg src constant offset ireg base + | (Mfloat64 | Many64), FR src -> fprintf oc " stfd %a, %a(%a)\n" freg src constant offset ireg base (* Mint64 is special-cased below *) | _ -> @@ -512,11 +511,11 @@ let align n a = (n + a - 1) land (-a) let rec next_arg_locations ir fr ofs = function | [] -> (ir, fr, ofs) - | Tint :: l -> + | (Tint | Tany32) :: l -> if ir < 8 then next_arg_locations (ir + 1) fr ofs l else next_arg_locations ir fr (ofs + 4) l - | (Tfloat | Tsingle) :: l -> + | (Tfloat | Tsingle | Tany64) :: l -> if fr < 8 then next_arg_locations ir (fr + 1) ofs l else next_arg_locations ir fr (align ofs 8 + 8) l @@ -676,6 +675,7 @@ let short_cond_branch tbl pc lbl_dest = (* Printing of instructions *) let float_literals : (int * int64) list ref = ref [] +let float32_literals : (int * int32) list ref = ref [] let jumptables : (int * label list) list ref = ref [] let print_instruction oc tbl pc fallthrough = function @@ -804,10 +804,12 @@ let print_instruction oc tbl pc fallthrough = function fprintf oc " addi %a, %a, %ld\n" ireg GPR1 ireg GPR1 sz else fprintf oc " lwz %a, %ld(%a)\n" ireg GPR1 ofs ireg GPR1 - | Pfabs(r1, r2) -> + | Pfabs(r1, r2) | Pfabss(r1, r2) -> fprintf oc " fabs %a, %a\n" freg r1 freg r2 | Pfadd(r1, r2, r3) -> fprintf oc " fadd %a, %a, %a\n" freg r1 freg r2 freg r3 + | Pfadds(r1, r2, r3) -> + fprintf oc " fadds %a, %a, %a\n" freg r1 freg r2 freg r3 | Pfcmpu(r1, r2) -> fprintf oc " fcmpu %a, %a, %a\n" creg 0 freg r1 freg r2 | Pfcti(r1, r2) -> @@ -821,6 +823,8 @@ let print_instruction oc tbl pc fallthrough = function fprintf oc "%s end pseudoinstr fcti\n" comment | Pfdiv(r1, r2, r3) -> fprintf oc " fdiv %a, %a, %a\n" freg r1 freg r2 freg r3 + | Pfdivs(r1, r2, r3) -> + fprintf oc " fdivs %a, %a, %a\n" freg r1 freg r2 freg r3 | Pfmake(rd, r1, r2) -> fprintf oc "%s begin pseudoinstr %a = fmake(%a, %a)\n" comment freg rd ireg r1 ireg r2; @@ -835,25 +839,37 @@ let print_instruction oc tbl pc fallthrough = function fprintf oc " fmr %a, %a\n" freg r1 freg r2 | Pfmul(r1, r2, r3) -> fprintf oc " fmul %a, %a, %a\n" freg r1 freg r2 freg r3 - | Pfneg(r1, r2) -> + | Pfmuls(r1, r2, r3) -> + fprintf oc " fmuls %a, %a, %a\n" freg r1 freg r2 freg r3 + | Pfneg(r1, r2) | Pfnegs(r1, r2) -> fprintf oc " fneg %a, %a\n" freg r1 freg r2 | Pfrsp(r1, r2) -> fprintf oc " frsp %a, %a\n" freg r1 freg r2 + | Pfxdp(r1, r2) -> + if r1 <> r2 then + fprintf oc " fmr %a, %a\n" freg r1 freg r2 | Pfsub(r1, r2, r3) -> fprintf oc " fsub %a, %a, %a\n" freg r1 freg r2 freg r3 + | Pfsubs(r1, r2, r3) -> + fprintf oc " fsubs %a, %a, %a\n" freg r1 freg r2 freg r3 | Plbz(r1, c, r2) -> fprintf oc " lbz %a, %a(%a)\n" ireg r1 constant c ireg r2 | Plbzx(r1, r2, r3) -> fprintf oc " lbzx %a, %a, %a\n" ireg r1 ireg r2 ireg r3 - | Plfd(r1, c, r2) -> + | Plfd(r1, c, r2) | Plfd_a(r1, c, r2) -> fprintf oc " lfd %a, %a(%a)\n" freg r1 constant c ireg r2 - | Plfdx(r1, r2, r3) -> + | Plfdx(r1, r2, r3) | Plfdx_a(r1, r2, r3) -> fprintf oc " lfdx %a, %a, %a\n" freg r1 ireg r2 ireg r3 | Plfi(r1, c) -> let lbl = new_label() in fprintf oc " addis %a, 0, %a\n" ireg GPR12 label_high lbl; fprintf oc " lfd %a, %a(%a) %s %.18g\n" freg r1 label_low lbl ireg GPR12 comment (camlfloat_of_coqfloat c); - float_literals := (lbl, camlint64_of_coqint (Floats.Float.bits_of_double c)) :: !float_literals; + float_literals := (lbl, camlint64_of_coqint (Floats.Float.to_bits c)) :: !float_literals; + | Plfis(r1, c) -> + let lbl = new_label() in + fprintf oc " addis %a, 0, %a\n" ireg GPR12 label_high lbl; + fprintf oc " lfs %a, %a(%a) %s %.18g\n" freg r1 label_low lbl ireg GPR12 comment (camlfloat_of_coqfloat32 c); + float32_literals := (lbl, camlint_of_coqint (Floats.Float32.to_bits c)) :: !float32_literals; | Plfs(r1, c, r2) -> fprintf oc " lfs %a, %a(%a)\n" freg r1 constant c ireg r2 | Plfsx(r1, r2, r3) -> @@ -866,9 +882,9 @@ let print_instruction oc tbl pc fallthrough = function fprintf oc " lhz %a, %a(%a)\n" ireg r1 constant c ireg r2 | Plhzx(r1, r2, r3) -> fprintf oc " lhzx %a, %a, %a\n" ireg r1 ireg r2 ireg r3 - | Plwz(r1, c, r2) -> + | Plwz(r1, c, r2) | Plwz_a(r1, c, r2) -> fprintf oc " lwz %a, %a(%a)\n" ireg r1 constant c ireg r2 - | Plwzx(r1, r2, r3) -> + | Plwzx(r1, r2, r3) | Plwzx_a(r1, r2, r3) -> fprintf oc " lwzx %a, %a, %a\n" ireg r1 ireg r2 ireg r3 | Pmfcrbit(r1, bit) -> fprintf oc " mfcr %a\n" ireg r1; @@ -924,23 +940,21 @@ let print_instruction oc tbl pc fallthrough = function fprintf oc " stb %a, %a(%a)\n" ireg r1 constant c ireg r2 | Pstbx(r1, r2, r3) -> fprintf oc " stbx %a, %a, %a\n" ireg r1 ireg r2 ireg r3 - | Pstfd(r1, c, r2) -> + | Pstfd(r1, c, r2) | Pstfd_a(r1, c, r2) -> fprintf oc " stfd %a, %a(%a)\n" freg r1 constant c ireg r2 - | Pstfdx(r1, r2, r3) -> + | Pstfdx(r1, r2, r3) | Pstfdx_a(r1, r2, r3) -> fprintf oc " stfdx %a, %a, %a\n" freg r1 ireg r2 ireg r3 | Pstfs(r1, c, r2) -> - fprintf oc " frsp %a, %a\n" freg FPR13 freg r1; - fprintf oc " stfs %a, %a(%a)\n" freg FPR13 constant c ireg r2 + fprintf oc " stfs %a, %a(%a)\n" freg r1 constant c ireg r2 | Pstfsx(r1, r2, r3) -> - fprintf oc " frsp %a, %a\n" freg FPR13 freg r1; - fprintf oc " stfsx %a, %a, %a\n" freg FPR13 ireg r2 ireg r3 + fprintf oc " stfsx %a, %a, %a\n" freg r1 ireg r2 ireg r3 | Psth(r1, c, r2) -> fprintf oc " sth %a, %a(%a)\n" ireg r1 constant c ireg r2 | Psthx(r1, r2, r3) -> fprintf oc " sthx %a, %a, %a\n" ireg r1 ireg r2 ireg r3 - | Pstw(r1, c, r2) -> + | Pstw(r1, c, r2) | Pstw_a(r1, c, r2) -> fprintf oc " stw %a, %a(%a)\n" ireg r1 constant c ireg r2 - | Pstwx(r1, r2, r3) -> + | Pstwx(r1, r2, r3) | Pstwx_a(r1, r2, r3) -> fprintf oc " stwx %a, %a, %a\n" ireg r1 ireg r2 ireg r3 | Psubfc(r1, r2, r3) -> fprintf oc " subfc %a, %a, %a\n" ireg r1 ireg r2 ireg r3 @@ -1061,11 +1075,14 @@ let rec print_instructions oc tbl pc fallthrough = function (* Print the code for a function *) -let print_literal oc (lbl, n) = +let print_literal64 oc (lbl, n) = let nlo = Int64.to_int32 n and nhi = Int64.to_int32(Int64.shift_right_logical n 32) in fprintf oc "%a: .long 0x%lx, 0x%lx\n" label lbl nhi nlo +let print_literal32 oc (lbl, n) = + fprintf oc "%a: .long 0x%lx\n" label lbl n + let print_jumptable oc (lbl, tbl) = fprintf oc "%a:" label lbl; List.iter @@ -1075,6 +1092,7 @@ let print_jumptable oc (lbl, tbl) = let print_function oc name fn = Hashtbl.clear current_function_labels; float_literals := []; + float32_literals := []; jumptables := []; current_function_sig := fn.fn_sig; let (text, lit, jmptbl) = @@ -1095,11 +1113,12 @@ let print_function oc name fn = cfi_endproc oc; fprintf oc " .type %a, @function\n" symbol name; fprintf oc " .size %a, . - %a\n" symbol name symbol name; - if !float_literals <> [] then begin + if !float_literals <> [] || !float32_literals <> [] then begin section oc lit; fprintf oc " .balign 8\n"; - List.iter (print_literal oc) !float_literals; - float_literals := [] + List.iter (print_literal64 oc) !float_literals; + List.iter (print_literal32 oc) !float32_literals; + float_literals := []; float32_literals := [] end; if !jumptables <> [] then begin section oc jmptbl; @@ -1131,10 +1150,10 @@ let print_init oc = function (Int64.logand b 0xFFFFFFFFL) | Init_float32 n -> fprintf oc " .long 0x%lx %s %.18g\n" - (camlint_of_coqint (Floats.Float.bits_of_single n)) + (camlint_of_coqint (Floats.Float32.to_bits n)) comment (camlfloat_of_coqfloat n) | Init_float64 n -> - let b = camlint64_of_coqint (Floats.Float.bits_of_double n) in + let b = camlint64_of_coqint (Floats.Float.to_bits n) in fprintf oc " .long 0x%Lx, 0x%Lx %s %.18g\n" (Int64.shift_right_logical b 32) (Int64.logand b 0xFFFFFFFFL) diff --git a/powerpc/SelectOp.vp b/powerpc/SelectOp.vp index 371a08a4..70b1feb6 100644 --- a/powerpc/SelectOp.vp +++ b/powerpc/SelectOp.vp @@ -374,6 +374,12 @@ Definition addf (e1 e2: expr) := Eop Oaddf (e1 ::: e2 ::: Enil). Definition subf (e1 e2: expr) := Eop Osubf (e1 ::: e2 ::: Enil). Definition mulf (e1 e2: expr) := Eop Omulf (e1 ::: e2 ::: Enil). +Definition negfs (e: expr) := Eop Onegfs (e ::: Enil). +Definition absfs (e: expr) := Eop Oabsfs (e ::: Enil). +Definition addfs (e1 e2: expr) := Eop Oaddfs (e1 ::: e2 ::: Enil). +Definition subfs (e1 e2: expr) := Eop Osubfs (e1 ::: e2 ::: Enil). +Definition mulfs (e1 e2: expr) := Eop Omulfs (e1 ::: e2 ::: Enil). + (** ** Comparisons *) Nondetfunction compimm (default: comparison -> int -> condition) @@ -433,6 +439,10 @@ Nondetfunction compu (c: comparison) (e1: expr) (e2: expr) := Definition compf (c: comparison) (e1: expr) (e2: expr) := Eop (Ocmp (Ccompf c)) (e1 ::: e2 ::: Enil). +Definition compfs (c: comparison) (e1: expr) (e2: expr) := + Eop (Ocmp (Ccompf c)) (Eop Ofloatofsingle (e1 ::: Enil) ::: + Eop Ofloatofsingle (e2 ::: Enil) ::: Enil). + (** ** Integer conversions *) Definition cast8unsigned (e: expr) := andimm (Int.repr 255) e. @@ -457,7 +467,7 @@ Definition intoffloat (e: expr) := Eop Ointoffloat (e ::: Enil). Definition intuoffloat (e: expr) := Elet e - (Elet (Eop (Ofloatconst (Float.floatofintu Float.ox8000_0000)) Enil) + (Elet (Eop (Ofloatconst (Float.of_intu Float.ox8000_0000)) Enil) (Econdition (CEcond (Ccompf Clt) (Eletvar 1 ::: Eletvar 0 ::: Enil)) (intoffloat (Eletvar 1)) (addimm Float.ox8000_0000 (intoffloat (subf (Eletvar 1) (Eletvar 0))))))%nat. @@ -465,7 +475,7 @@ Definition intuoffloat (e: expr) := Nondetfunction floatofintu (e: expr) := match e with | Eop (Ointconst n) Enil => - Eop (Ofloatconst (Float.floatofintu n)) Enil + Eop (Ofloatconst (Float.of_intu n)) Enil | _ => subf (Eop Ofloatofwords (Eop (Ointconst Float.ox4330_0000) Enil ::: e ::: Enil)) (Eop (Ofloatconst (Float.from_words Float.ox4330_0000 Int.zero)) Enil) @@ -474,14 +484,27 @@ Nondetfunction floatofintu (e: expr) := Nondetfunction floatofint (e: expr) := match e with | Eop (Ointconst n) Enil => - Eop (Ofloatconst (Float.floatofint n)) Enil + Eop (Ofloatconst (Float.of_int n)) Enil | _ => subf (Eop Ofloatofwords (Eop (Ointconst Float.ox4330_0000) Enil ::: addimm Float.ox8000_0000 e ::: Enil)) (Eop (Ofloatconst (Float.from_words Float.ox4330_0000 Float.ox8000_0000)) Enil) end. +Definition intofsingle (e: expr) := + intoffloat (Eop Ofloatofsingle (e ::: Enil)). + +Definition singleofint (e: expr) := + Eop Osingleoffloat (floatofint e ::: Enil). + +Definition intuofsingle (e: expr) := + intuoffloat (Eop Ofloatofsingle (e ::: Enil)). + +Definition singleofintu (e: expr) := + Eop Osingleoffloat (floatofintu e ::: Enil). + Definition singleoffloat (e: expr) := Eop Osingleoffloat (e ::: Enil). +Definition floatofsingle (e: expr) := Eop Ofloatofsingle (e ::: Enil). (** ** Recognition of addressing modes for load and store operations *) diff --git a/powerpc/SelectOpproof.v b/powerpc/SelectOpproof.v index cb48d51a..8311b82c 100644 --- a/powerpc/SelectOpproof.v +++ b/powerpc/SelectOpproof.v @@ -639,6 +639,31 @@ Proof. red; intros; TrivialExists. Qed. +Theorem eval_negfs: unary_constructor_sound negfs Val.negfs. +Proof. + red; intros. TrivialExists. +Qed. + +Theorem eval_absfs: unary_constructor_sound absfs Val.absfs. +Proof. + red; intros. TrivialExists. +Qed. + +Theorem eval_addfs: binary_constructor_sound addfs Val.addfs. +Proof. + red; intros; TrivialExists. +Qed. + +Theorem eval_subfs: binary_constructor_sound subfs Val.subfs. +Proof. + red; intros; TrivialExists. +Qed. + +Theorem eval_mulfs: binary_constructor_sound mulfs Val.mulfs. +Proof. + red; intros; TrivialExists. +Qed. + Section COMP_IMM. Variable default: comparison -> int -> condition. @@ -746,6 +771,18 @@ Proof. intros; red; intros. unfold compf. TrivialExists. Qed. +Theorem eval_compfs: + forall c, binary_constructor_sound (compfs c) (Val.cmpfs c). +Proof. + intros; red; intros. unfold compfs. + replace (Val.cmpfs c x y) with + (Val.cmpf c (Val.floatofsingle x) (Val.floatofsingle y)). + TrivialExists. constructor. EvalOp. simpl; reflexivity. + constructor. EvalOp. simpl; reflexivity. constructor. + auto. + destruct x; auto. destruct y; auto. unfold Val.cmpf, Val.cmpfs; simpl. + rewrite Float32.cmp_double. auto. +Qed. Theorem eval_cast8signed: unary_constructor_sound cast8signed (Val.sign_ext 8). Proof. @@ -778,6 +815,11 @@ Proof. red; intros. unfold singleoffloat. TrivialExists. Qed. +Theorem eval_floatofsingle: unary_constructor_sound floatofsingle Val.floatofsingle. +Proof. + red; intros. unfold floatofsingle. TrivialExists. +Qed. + Theorem eval_intoffloat: forall le a x y, eval_expr ge sp e m le a x -> @@ -794,10 +836,10 @@ Theorem eval_intuoffloat: exists v, eval_expr ge sp e m le (intuoffloat a) v /\ Val.lessdef y v. Proof. intros. destruct x; simpl in H0; try discriminate. - destruct (Float.intuoffloat f) as [n|] eqn:?; simpl in H0; inv H0. + destruct (Float.to_intu f) as [n|] eqn:?; simpl in H0; inv H0. exists (Vint n); split; auto. unfold intuoffloat. set (im := Int.repr Int.half_modulus). - set (fm := Float.floatofintu im). + set (fm := Float.of_intu im). assert (eval_expr ge sp e m (Vfloat fm :: Vfloat f :: le) (Eletvar (S O)) (Vfloat f)). constructor. auto. assert (eval_expr ge sp e m (Vfloat fm :: Vfloat f :: le) (Eletvar O) (Vfloat fm)). @@ -807,9 +849,9 @@ Proof. eapply eval_Econdition with (va := Float.cmp Clt f fm). eauto with evalexpr. destruct (Float.cmp Clt f fm) eqn:?. - exploit Float.intuoffloat_intoffloat_1; eauto. intro EQ. + exploit Float.to_intu_to_int_1; eauto. intro EQ. EvalOp. simpl. rewrite EQ; auto. - exploit Float.intuoffloat_intoffloat_2; eauto. + exploit Float.to_intu_to_int_2; eauto. change Float.ox8000_0000 with im. fold fm. intro EQ. set (t2 := subf (Eletvar (S O)) (Eletvar O)). set (t3 := intoffloat t2). @@ -834,7 +876,7 @@ Proof. intros until y. unfold floatofint. destruct (floatofint_match a); intros. InvEval. TrivialExists. rename e0 into a. destruct x; simpl in H0; inv H0. - exists (Vfloat (Float.floatofint i)); split; auto. + exists (Vfloat (Float.of_int i)); split; auto. set (t1 := addimm Float.ox8000_0000 a). set (t2 := Eop Ofloatofwords (Eop (Ointconst Float.ox4330_0000) Enil ::: t1 ::: Enil)). set (t3 := Eop (Ofloatconst (Float.from_words Float.ox4330_0000 Float.ox8000_0000)) Enil). @@ -844,7 +886,7 @@ Proof. unfold t2. EvalOp. constructor. EvalOp. simpl; eauto. constructor. eauto. constructor. unfold eval_operation. eauto. instantiate (2 := t3). unfold t3. EvalOp. simpl; eauto. - intros [v2 [A2 B2]]. simpl in B2. inv B2. rewrite Float.floatofint_from_words. auto. + intros [v2 [A2 B2]]. simpl in B2. inv B2. rewrite Float.of_int_from_words. auto. Qed. Theorem eval_floatofintu: @@ -856,7 +898,7 @@ Proof. intros until y. unfold floatofintu. destruct (floatofintu_match a); intros. InvEval. TrivialExists. rename e0 into a. destruct x; simpl in H0; inv H0. - exists (Vfloat (Float.floatofintu i)); split; auto. + exists (Vfloat (Float.of_intu i)); split; auto. unfold floatofintu. set (t2 := Eop Ofloatofwords (Eop (Ointconst Float.ox4330_0000) Enil ::: a ::: Enil)). set (t3 := Eop (Ofloatconst (Float.from_words Float.ox4330_0000 Int.zero)) Enil). @@ -864,7 +906,73 @@ Proof. unfold t2. EvalOp. constructor. EvalOp. simpl; eauto. constructor. eauto. constructor. unfold eval_operation. eauto. instantiate (2 := t3). unfold t3. EvalOp. simpl; eauto. - intros [v2 [A2 B2]]. simpl in B2. inv B2. rewrite Float.floatofintu_from_words. auto. + intros [v2 [A2 B2]]. simpl in B2. inv B2. rewrite Float.of_intu_from_words. auto. +Qed. + +Theorem eval_intofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.intofsingle x = Some y -> + exists v, eval_expr ge sp e m le (intofsingle a) v /\ Val.lessdef y v. +Proof. + intros; unfold intofsingle. + assert (Val.intoffloat (Val.floatofsingle x) = Some y). + { destruct x; simpl in H0; try discriminate. + destruct (Float32.to_int f) eqn:F; inv H0. + apply Float32.to_int_double in F. + simpl. unfold Float32.to_double in F; rewrite F; auto. + } + apply eval_intoffloat with (Val.floatofsingle x); auto. EvalOp. +Qed. + +Theorem eval_singleofint: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.singleofint x = Some y -> + exists v, eval_expr ge sp e m le (singleofint a) v /\ Val.lessdef y v. +Proof. + intros. unfold singleofint. + assert (exists z, Val.floatofint x = Some z /\ y = Val.singleoffloat z). + { + destruct x; inv H0. simpl. exists (Vfloat (Float.of_int i)); simpl; split; auto. + f_equal. apply Float32.of_int_double. + } + destruct H1 as (z & A & B). subst y. + exploit eval_floatofint; eauto. intros (v & C & D). + exists (Val.singleoffloat v); split. EvalOp. inv D; auto. +Qed. + +Theorem eval_intuofsingle: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.intuofsingle x = Some y -> + exists v, eval_expr ge sp e m le (intuofsingle a) v /\ Val.lessdef y v. +Proof. + intros; unfold intuofsingle. + assert (Val.intuoffloat (Val.floatofsingle x) = Some y). + { destruct x; simpl in H0; try discriminate. + destruct (Float32.to_intu f) eqn:F; inv H0. + apply Float32.to_intu_double in F. + simpl. unfold Float32.to_double in F; rewrite F; auto. + } + apply eval_intuoffloat with (Val.floatofsingle x); auto. EvalOp. +Qed. + +Theorem eval_singleofintu: + forall le a x y, + eval_expr ge sp e m le a x -> + Val.singleofintu x = Some y -> + exists v, eval_expr ge sp e m le (singleofintu a) v /\ Val.lessdef y v. +Proof. + intros. unfold singleofintu. + assert (exists z, Val.floatofintu x = Some z /\ y = Val.singleoffloat z). + { + destruct x; inv H0. simpl. exists (Vfloat (Float.of_intu i)); simpl; split; auto. + f_equal. apply Float32.of_intu_double. + } + destruct H1 as (z & A & B). subst y. + exploit eval_floatofintu; eauto. intros (v & C & D). + exists (Val.singleoffloat v); split. EvalOp. inv D; auto. Qed. Theorem eval_addressing: diff --git a/powerpc/Unusedglob1.ml b/powerpc/Unusedglob1.ml index 49c0774e..2d3efe39 100644 --- a/powerpc/Unusedglob1.ml +++ b/powerpc/Unusedglob1.ml @@ -42,18 +42,22 @@ let referenced_instr = function | Pcmpwi(_, c) | Plbz(_, c, _) | Plfd(_, c, _) + | Plfd_a(_, c, _) | Plfs(_, c, _) | Plha(_, c, _) | Plhz(_, c, _) | Plwz(_, c, _) + | Plwz_a(_, c, _) | Pmulli(_, _, c) | Pori(_, _, c) | Poris(_, _, c) | Pstb(_, c, _) | Pstfd(_, c, _) + | Pstfd_a(_, c, _) | Pstfs(_, c, _) | Psth(_, c, _) | Pstw(_, c, _) + | Pstw_a(_, c, _) | Psubfic(_, _, c) | Pxori(_, _, c) | Pxoris(_, _, c) -> referenced_constant c diff --git a/powerpc/ValueAOp.v b/powerpc/ValueAOp.v index 77463f4b..a5a1db80 100644 --- a/powerpc/ValueAOp.v +++ b/powerpc/ValueAOp.v @@ -52,6 +52,7 @@ Definition eval_static_operation (op: operation) (vl: list aval): aval := | Omove, v1::nil => v1 | Ointconst n, nil => I n | Ofloatconst n, nil => if propagate_float_constants tt then F n else ftop + | Osingleconst n, nil => if propagate_float_constants tt then FS n else ftop | Oaddrsymbol id ofs, nil => Ptr (Gl id ofs) | Oaddrstack ofs, nil => Ptr (Stk ofs) | Ocast8signed, v1 :: nil => sign_ext 8 v1 @@ -92,7 +93,14 @@ Definition eval_static_operation (op: operation) (vl: list aval): aval := | Osubf, v1::v2::nil => subf v1 v2 | Omulf, v1::v2::nil => mulf v1 v2 | Odivf, v1::v2::nil => divf v1 v2 + | Onegfs, v1::nil => negfs v1 + | Oabsfs, v1::nil => absfs v1 + | Oaddfs, v1::v2::nil => addfs v1 v2 + | Osubfs, v1::v2::nil => subfs v1 v2 + | Omulfs, v1::v2::nil => mulfs v1 v2 + | Odivfs, v1::v2::nil => divfs v1 v2 | Osingleoffloat, v1::nil => singleoffloat v1 + | Ofloatofsingle, v1::nil => floatofsingle v1 | Ointoffloat, v1::nil => intoffloat v1 | Ofloatofwords, v1::v2::nil => floatofwords v1 v2 | Omakelong, v1::v2::nil => longofwords v1 v2 @@ -163,6 +171,7 @@ Proof. unfold eval_operation, eval_static_operation; intros; destruct op; InvHyps; eauto with va. destruct (propagate_float_constants tt); constructor. + destruct (propagate_float_constants tt); constructor. rewrite Int.add_zero_l; eauto with va. fold (Val.sub (Vint i) a1). auto with va. apply floatofwords_sound; auto. diff --git a/powerpc/eabi/Conventions1.v b/powerpc/eabi/Conventions1.v index 2db1f737..866e73d6 100644 --- a/powerpc/eabi/Conventions1.v +++ b/powerpc/eabi/Conventions1.v @@ -15,6 +15,7 @@ Require Import Coqlib. Require Import AST. +Require Import Events. Require Import Locations. (** * Classification of machine registers *) @@ -180,13 +181,13 @@ Proof. Qed. Lemma int_callee_save_type: - forall r, In r int_callee_save_regs -> mreg_type r = Tint. + forall r, In r int_callee_save_regs -> mreg_type r = Tany32. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. Lemma float_callee_save_type: - forall r, In r float_callee_save_regs -> mreg_type r = Tfloat. + forall r, In r float_callee_save_regs -> mreg_type r = Tany64. Proof. intro. simpl; ElimOrEq; reflexivity. Qed. @@ -239,12 +240,21 @@ Qed. Definition loc_result (s: signature) : list mreg := match s.(sig_res) with | None => R3 :: nil - | Some Tint => R3 :: nil - | Some (Tfloat | Tsingle) => F1 :: nil + | Some (Tint | Tany32) => R3 :: nil + | Some (Tfloat | Tsingle | Tany64) => F1 :: nil | Some Tlong => R3 :: R4 :: nil end. -(** The result location is a caller-save register *) +(** The result registers have types compatible with that given in the signature. *) + +Lemma loc_result_type: + forall sig, + subtype_list (proj_sig_res' sig) (map mreg_type (loc_result sig)) = true. +Proof. + intros. unfold proj_sig_res', loc_result. destruct (sig_res sig) as [[]|]; auto. +Qed. + +(** The result locations are caller-save registers *) Lemma loc_result_caller_save: forall (s: signature) (r: mreg), @@ -278,18 +288,18 @@ Fixpoint loc_arguments_rec (tyl: list typ) (ir fr ofs: Z) {struct tyl} : list loc := match tyl with | nil => nil - | Tint :: tys => + | (Tint | Tany32) as ty :: tys => match list_nth_z int_param_regs ir with | None => - S Outgoing ofs Tint :: loc_arguments_rec tys ir fr (ofs + 1) + S Outgoing ofs ty :: loc_arguments_rec tys ir fr (ofs + 1) | Some ireg => R ireg :: loc_arguments_rec tys (ir + 1) fr ofs end - | (Tfloat | Tsingle) :: tys => + | (Tfloat | Tsingle | Tany64) as ty :: tys => match list_nth_z float_param_regs fr with | None => let ofs := align ofs 2 in - S Outgoing ofs Tfloat :: loc_arguments_rec tys ir fr (ofs + 2) + S Outgoing ofs ty :: loc_arguments_rec tys ir fr (ofs + 2) | Some freg => R freg :: loc_arguments_rec tys ir (fr + 1) ofs end @@ -316,12 +326,12 @@ Definition loc_arguments (s: signature) : list loc := Fixpoint size_arguments_rec (tyl: list typ) (ir fr ofs: Z) {struct tyl} : Z := match tyl with | nil => ofs - | Tint :: tys => + | (Tint | Tany32) :: tys => match list_nth_z int_param_regs ir with | None => size_arguments_rec tys ir fr (ofs + 1) | Some ireg => size_arguments_rec tys (ir + 1) fr ofs end - | (Tfloat | Tsingle) :: tys => + | (Tfloat | Tsingle | Tany64) :: tys => match list_nth_z float_param_regs fr with | None => size_arguments_rec tys ir fr (align ofs 2 + 2) | Some freg => size_arguments_rec tys ir (fr + 1) ofs @@ -393,7 +403,7 @@ Opaque list_nth_z. destruct H. subst. split. omega. congruence. exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. intuition omega. assert (ofs <= align ofs 2) by (apply align_le; omega). - destruct H. subst. split. omega. congruence. + destruct H. subst. split. omega. congruence. destruct H. subst. split. omega. congruence. exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. intuition omega. - (* single *) @@ -404,6 +414,20 @@ Opaque list_nth_z. exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. assert (ofs <= align ofs 2) by (apply align_le; omega). intuition omega. +- (* any32 *) + destruct (list_nth_z int_param_regs ir) as [r|] eqn:E; destruct H. + subst. left. eapply list_nth_z_in; eauto. + eapply IHtyl; eauto. + subst. split. omega. congruence. + exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. intuition omega. +- (* any64 *) + destruct (list_nth_z float_param_regs fr) as [r|] eqn:E; destruct H. + subst. right. eapply list_nth_z_in; eauto. + eapply IHtyl; eauto. + subst. split. apply Zle_ge. apply align_le. omega. congruence. + exploit IHtyl; eauto. destruct l; auto. destruct sl; auto. + assert (ofs <= align ofs 2) by (apply align_le; omega). + intuition omega. Qed. Lemma loc_arguments_acceptable: @@ -441,6 +465,10 @@ Proof. destruct (list_nth_z float_param_regs fr); eauto. apply Zle_trans with (align ofs0 2). apply align_le; omega. apply Zle_trans with (align ofs0 2 + 2); auto; omega. + destruct (list_nth_z int_param_regs ir); eauto. apply Zle_trans with (ofs0 + 1); auto; omega. + destruct (list_nth_z float_param_regs fr); eauto. + apply Zle_trans with (align ofs0 2). apply align_le; omega. + apply Zle_trans with (align ofs0 2 + 2); auto; omega. Qed. Lemma size_arguments_above: @@ -490,6 +518,18 @@ Proof. transitivity (align ofs0 2 + 2). simpl; omega. eauto. apply size_arguments_rec_above. eauto. - (* single *) + destruct (list_nth_z float_param_regs fr); destruct H0. + congruence. + eauto. + inv H0. transitivity (align ofs0 2 + 2). simpl; omega. apply size_arguments_rec_above. + eauto. +- (* any32 *) + destruct (list_nth_z int_param_regs ir); destruct H0. + congruence. + eauto. + inv H0. apply size_arguments_rec_above. + eauto. +- (* any64 *) destruct (list_nth_z float_param_regs fr); destruct H0. congruence. eauto. diff --git a/test/c/fftsp.c b/test/c/fftsp.c index 42ae905e..3c7c23c3 100644 --- a/test/c/fftsp.c +++ b/test/c/fftsp.c @@ -52,17 +52,17 @@ int dfft(float x[], float y[], int np) for (k = 1; k <= m-1; k++ ) { n2 = n2 / 2; n4 = n2 / 4; - e = tpi / (double)n2; + e = tpi / (float)n2; a = 0.0; for (j = 1; j<= n4 ; j++) { - a3 = 3.0 * a; + a3 = 3.0f * a; cc1 = cosf(a); ss1 = sinf(a); cc3 = cosf(a3); ss3 = sinf(a3); - a = e * (double)j; + a = e * (float)j; is = j; id = 2 * n2; @@ -162,7 +162,7 @@ int main(int argc, char ** argv) xi = calloc(np, sizeof(float)); pxr = xr; pxi = xi; - *pxr = (enp - 1.0) * 0.5; + *pxr = (enp - 1.0) * 0.5f; *pxi = 0.0; n2 = np / 2; *(pxr+n2) = -0.5; @@ -171,8 +171,8 @@ int main(int argc, char ** argv) j = np - i; *(pxr+i) = -0.5; *(pxr+j) = -0.5; - z = t * (double)i; - y = -0.5*(cosf(z)/sinf(z)); + z = t * (float)i; + y = -0.5f*(cosf(z)/sinf(z)); *(pxi+i) = y; *(pxi+j) = -y; } diff --git a/test/regression/NaNs.c b/test/regression/NaNs.c index 45f1e7fb..618ce295 100644 --- a/test/regression/NaNs.c +++ b/test/regression/NaNs.c @@ -32,17 +32,17 @@ inline float single_of_bits(u32 i) u.i = i; return u.f; } -volatile double val[8]; - char * valname[8] = { "+0", "-0", "+inf", "-inf", "snan(5)", "qnan(6)", "snan(-9)", "qnan(-1)" }; -int main() +void test64(void) { + volatile double val[8]; int i, j; + printf("--- Double precision\n"); val[0] = 0.0; val[1] = - val[0]; val[2] = double_of_bits(0x7FF0000000000000); @@ -56,10 +56,7 @@ int main() for (i = 0; i < 8; i++) { printf("opp(%s) = 0x%016llx\n", valname[i], bits_of_double(- val[i])); printf("single(%s) = 0x%08x\n", valname[i], bits_of_single((float)(val[i]))); -#if 0 - /* The reference interpreter doesn't support __builtin_fabs */ printf("abs(%s) = 0x%016llx\n", valname[i], bits_of_double(__builtin_fabs(val[i]))); -#endif for (j = 0; j < 8; j++) { printf("%s + %s = 0x%016llx\n", valname[i], valname[j], bits_of_double(val[i] + val[j])); @@ -71,5 +68,44 @@ int main() valname[i], valname[j], bits_of_double(val[i] / val[j])); } } +} + +void test32(void) +{ + volatile float val[8]; + int i, j; + + printf("--- Single precision\n"); + val[0] = 0.0f; + val[1] = - val[0]; + val[2] = single_of_bits(0x7F800000); + val[3] = - val[2]; + + val[4] = single_of_bits(0x7F800005); + val[5] = single_of_bits(0x7FC00006); + val[6] = single_of_bits(0xFF800009); + val[7] = single_of_bits(0xFFC00001); + + for (i = 0; i < 8; i++) { + printf("opp(%s) = 0x%08x\n", valname[i], bits_of_single(- val[i])); + printf("double(%s) = 0x%016llx\n", valname[i], bits_of_double((double)(val[i]))); + printf("abs(%s) = 0x%08x\n", valname[i], bits_of_single(__builtin_fabs(val[i]))); + for (j = 0; j < 8; j++) { + printf("%s + %s = 0x%08x\n", + valname[i], valname[j], bits_of_single(val[i] + val[j])); + printf("%s - %s = 0x%08x\n", + valname[i], valname[j], bits_of_single(val[i] - val[j])); + printf("%s * %s = 0x%08x\n", + valname[i], valname[j], bits_of_single(val[i] * val[j])); + printf("%s / %s = 0x%08x\n", + valname[i], valname[j], bits_of_single(val[i] / val[j])); + } + } +} + +int main(void) +{ + test64(); + test32(); return 0; } diff --git a/test/spass/Makefile.bak b/test/spass/Makefile.bak deleted file mode 100644 index 320f6221..00000000 --- a/test/spass/Makefile.bak +++ /dev/null @@ -1,13 +0,0 @@ -LEVEL = ../../.. -PROG = SPASS - -CPPFLAGS = -DCLOCK_NO_TIMING -fno-strict-aliasing -w -LDFLAGS = -lm - -ifdef SMALL_PROBLEM_SIZE -RUN_OPTIONS="$(PROJ_SRC_DIR)/small_problem.dfg" -else -RUN_OPTIONS="$(PROJ_SRC_DIR)/problem.dfg" -endif - -include ../../Makefile.multisrc -- cgit