| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
- Introduce Archi.ptr64 parameter.
- Define module Ptrofs of integers as wide as a pointer (64 if Archi.ptr64, 32 otherwise).
- Use Ptrofs.int as the offset type for Vptr values and anywhere pointer offsets are manipulated.
- Modify Val operations that handle pointers (e.g. Val.add, Val.sub, Val.cmpu) so that in 64-bit pointer mode it is the "long" operation (e.g. Val.addl, Val.subl, Val.cmplu) that handles pointers.
- Update the memory model accordingly.
- Modify C operations that handle pointers (e.g. addition, subtraction, comparisons) accordingly.
- Make it possible to turn off the splitting of 64-bit integers into pairs of 32-bit integers.
- Update the compiler front-end and back-end accordingly.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit changes the loc_arguments and loc_result functions that describe calling conventions so that each argument/result can be mapped either to a single location or (in the case of a 64-bit integer) to a pair of two 32-bit locations.
In the current CompCert, all arguments/results of type Tlong are systematically split in two 32-bit halves. We will need to change this in the future to support 64-bit processors. The alternative approach implemented by this commit enables the loc_arguments and loc_result functions to describe precisely which arguments need splitting. Eventually, the remainder of CompCert should not assume anything about splitting 64-bit types in two halves.
Summary of changes:
- AST: introduce the type "rpair A" of register pairs
- Conventions1, Conventions: use it when describing calling conventions
- LTL, Linear, Mach, Asm: honor the new calling conventions when observing external calls
- Events: suppress external_call', no longer useful
- All passes from Allocation to Asmgen: adapt accordingly.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
architectures
The original Stacking pass and its proof hard-wire assumptions about the processor and the register allocation, namely that integer registers are 32 bit wide and that all stack slots have natural alignment 4, which precludes having stack slots of type Tlong. Those assumptions become false if the target processor has 64-bit integer registers.
This commit makes minimal adjustments to the Stacking pass so as to lift these assumptions:
- Stack slots of type Tlong (or more generally of natural alignment 8) are supported. For slots produced by register allocation, the alignment is validated a posteriori in Lineartyping. For slots produced by the calling conventions, alignment is proved as part of the "loc_argument_acceptable" property in Conventions1.
- The code generated by Stacking to save and restore used callee-save registers no longer assumes 32-bit integer registers. Actually, it supports any combination of sizes for registers.
- To support the new save/restore code, Bounds was changed to record the set of all callee-save registers used, rather than just the max index of callee-save registers used.
On CompCert's current 32-bit target architectures, the new Stacking pass should generate pretty much the same code as the old one, modulo minor differences in the layout of the stack frame. (E.g. padding could be introduced at different places.)
The bulk of this big commit is related to the proof of the Stacking phase. The old proof strategy was painful and not obviously adaptable to the new Stacking phase, so I rewrote Stackingproof entirely, using an approach inspired by separation logic. The new library common/Separation.v defines assertions about memory states that can be composed using a separating conjunction, just like pre- and post-conditions in separation logic. Those assertions are used in Stackingproof to describe the contents of the stack frames during the execution of the generated Mach code, and relate them with the Linear location maps.
As a further simplification, the callee-save/caller-save distinction is now defined in Conventions1 by a function is_callee_save: mreg -> bool, instead of lists of registers of either kind as before. This eliminates many boring classification lemmas from Conventions1. LTL and Lineartyping were adapted accordingly.
Finally, this commit introduces a new library called Decidableplus to prove some propositions by reflection as Boolean computations. It is used to further simplify the proofs in Conventions1.
|
| |
|
|
|
|
|
|
|
|
| |
Use EF_debug instead of EF_annot for line number annotations.
Introduce PrintAsmaux.print_debug_info (very incomplete).
powerpc/Asmexpand: revise expand_memcpy_small.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Before, the back-end languages had distinct instructions
- Iannot for annotations, taking structured expressions (annot_arg)
as arguments, and producing no results'
- Ibuiltin for other builtins, using simple pseudoregs/locations/registers
as arguments and results.
This branch enriches Ibuiltin instructions so that they take structured
expressions (builtin_arg and builtin_res) as arguments and results.
This way,
- Annotations fit the general pattern of builtin functions,
so Iannot instructions are removed.
- EF_vload_global and EF_vstore_global become useless, as the
same optimization can be achieved by EF_vload/vstore taking
a structured argument of the "address of global" kind.
- Better code can be generated for builtin_memcpy between stack locations,
or volatile accesses to stack locations.
Finally, this commit also introduces a new kind of external function,
EF_debug, which is like EF_annot but produces no observable events.
It will be used later to transport debug info through the back-end,
without preventing optimizations.
|
|
|
|
|
|
|
|
|
|
| |
variables whose address is taken.
- CminorSel, RTL: add "annot" instructions.
- CminorSel to Asm: use type "annot_arg" for arguments of "annot" instructions.
- AST, Events: simplify EF_annot because constants are now part of the arguments.
Implementation is not complete yet.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- 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
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
1) Revised division of labor between RTLtyping and Lineartyping:
- RTLtyping no longer keeps track of single-precision floats,
switches from subtype-based inference to unification-based inference.
- Unityping: new library for unification-based inference.
- Locations: don't normalize at assignment in a stack slot
- Allocation, Allocproof: simplify accordingly.
- Lineartyping: add inference of locations that contain
a single-precision float.
- Stackingproof: adapted accordingly.
This addresses a defect report whereas RTLtyping was rejecting code that used a RTL pseudoreg to hold both double- and single-precision floats (see test/regression/singlefloats.c).
This corresponds to commits 2435+2436 plus improvements in Lineartyping.
2) Add -dtimings option to measure compilation times.
Moved call to C parser from Elab to Parse, to make it easier to
measure parsing time independently of elaboration time.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2449 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
| |
the new Lineartyping can't keep track of single floats that were spilled.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2438 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2437 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- RTLtyping no longer keeps track of single-precision floats,
switches from subtype-based inference to unification-based inference.
- Unityping: new library for unification-based inference.
- Locations: don't normalize at assignment in a stack slot
- Allocation, Allocproof: simplify accordingly.
- Lineartyping: add inference of machine registers that contain
a single-precision float.
- Stackingproof: adapted accordingly.
This addresses a defect report by D. Dickman whereas RTLtyping was rejecting code that used a RTL pseudoreg to hold both double- and single-precision floats.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2435 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
| |
- added RTL type "Tsingle"
- ABI-compatible passing of single-precision floats on ARM and x86
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2260 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
| |
1- new register allocator (+ live range splitting, spilling&reloading, etc)
based on a posteriori validation using the Rideau-Leroy algorithm
2- support for 64-bit integer arithmetic (type "long long").
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2200 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
| |
Update ARM port.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2085 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
| |
function definition, so that comparisons between function pointers
are correctly defined.
AST, Globalenvs, and many other files:
represent programs as a list of (function or variable) definitions
instead of two lists, one for functions and the other for variables.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2067 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
| |
Related general change: support for destroyed_at_moves.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1700 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
| |
functions, and more generally external functions
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1672 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- In pointer values "Vptr b ofs", interpret "ofs" as an unsigned int.
(Fixes issue with wrong comparison of pointers across 0x8000_0000)
- Revised Stacking pass to not use negative SP offsets.
- Add pointer validity checks to Cminor ... Mach
to support the use of memory injections in Stacking.
- Cleaned up Stacklayout modules.
- IA32: improved code generation for Mgetparam.
- ARM: improved code generation for op-immediate instructions.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1632 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
AST: add ef_inline flag to external functions.
Selection: recognize calls to inlined built-ins and inline them as Sbuiltin.
CminorSel to Asm: added Sbuiltin/Ibuiltin instruction.
PrintAsm: adapted expansion of builtins.
C2Clight: adapted detection of builtins.
Conventions: refactored in a machine-independent part (backend/Conventions)
and a machine-dependent part (ARCH/SYS/Conventions1).
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1356 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
| |
- Revised memory model with concrete representation of ints & floats,
and per-byte access permissions
- Revised Globalenvs implementation
- Matching changes in all languages and proofs.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1282 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
| |
accessing the table
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1172 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1171 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
| |
- In Cminor and below, removed pointer validity check in semantics of
comparisons, so that evaluation of expressions is independent of
memory state.
- In Cminor and below, removed "alloc" instruction.
- Cleaned up commented-away parts.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@945 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@925 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@605 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@604 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@489 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
| |
En particulier:
- Semantiques small-step depuis RTL jusqu'a PPC
- Cminor independant du processeur
- Ajout passes Selection et Reload
- Ajout des langages intermediaires CminorSel et LTLin correspondants
- Ajout des tailcalls depuis Cminor jusqu'a PPC
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@384 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
| |
que tous les arguments resident en registres
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@125 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@106 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Ajout de traces d'evenements d'E/S dans les semantiques
- Ajout constructions switch et allocation dynamique
- Initialisation des variables globales
- Portage Coq 8.1 beta
Debut d'integration du front-end C:
- Traduction Clight -> Csharpminor dans cfrontend/
- Modifications de Csharpminor et Globalenvs en consequence.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@72 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|