| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
Easier to type, and consistent with `-Os` (optimize for smaller code /
optimize for fewer conditional branches).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Extends the instruction selection pass with an if-conversion optimization:
some if/then/else statements are converted into "select" operations,
which in turn can be compiled down to branchless instruction sequences
if the target architecture supports them.
The statements that are converted are of the form
if (cond) { x = a1; } else { x = a2; }
if (cond) { x = a1; }
if (cond) { /*skip*/; } else { x = a2; }
where a1, a2 are "safe" expressions, containing no operations that can
fail at run-time, such as memory loads or integer divisions.
A heuristic in backend/Selectionaux.ml controls when the optimization occurs,
depending on command-line flags and the complexity of the "then" and "else"
branches.
|
|
|
|
|
|
|
|
|
|
| |
The option -fcommon controls whether uninitialized global
variables are placed in the COMMON section. If the option is given
in the negated form, -fno-common, variables are not placed in the
COMMON section. They are placed in the same sections as gcc does.
If the variables are not placed in the COMMON section merging of
tentative definitions is inhibited and multiple definitions lead
to a linker error, as it does for gcc.
|
|
|
|
|
|
|
|
|
| |
New inlining heuristic for static functions.
Static functions that are only called once can always be inlined,
since they can be removed safely after inlining and no call
prologue, epilogue, as well as register saving and needs to be
generated.
|
|
|
|
|
|
| |
The new option f(no-)inline controlls whether inlining is active
or not.
Bug 21343.
|
|
|
|
|
|
|
| |
The options controlling the generation of debugging information
are now moved into the Debug module. Futhermore the -gdepth
options are replaced in favor of a more gcc compatible version.
Bug 20193
|
| |
|
|
|
|
|
|
|
| |
Clightgen and CompCert share the code for preprocessing as well as
parsing C files. The code as well as command line switches is moved
in the new module Frontend.
Bug 18768
|
|
|
|
| |
The new option -dprepro allows it to keep the preprocessed source code files.
|
|
|
|
|
|
| |
- Rename '-fstruct-return' into '-fstruct-passing', because this emulation affects both function result passing and function argument passing. Keep '-fstruct-return' as a deprecated synonymous for '-fstruct-passing'
- Remove the ability to change the ABI for struct passing via the '-fstruct-passing=<abi>' and '-fstruct-return=<abi>' command-line flags. This was more confusing than useful.
- Produce an error if a struct/union is passed as function argument and '-fstruct-passing' is not set. This used to be supported, using CompCert's default ABI for passing struct arguments. However, this default ABI does not match any of the standard ABIs of our target platforms, so it is better to reject than to silently produce ABI-incompatible code.
|
|
|
|
|
|
|
|
|
|
|
| |
The new option gdepth subumes the gonly-globals. The option
allows it to control the level of information that is produced.
This option allows it to generate debugging inforation for:
-Only globals
-Global and local variables but without location information for
the local variable
-Full information
Bug 17638.
|
|
|
|
|
|
| |
The new option -gonly-global allows the generation of debuging
information for global variables only.
Bug 17566.
|
|
|
|
|
|
|
|
|
|
| |
The gcc produces DW_AT_ranges for non-contiguous address ranges, like
compilation units containing functions which are placed in different
ELF-sections or lexical scopes that are split up. With this commit
CompCert also uses this DWARF v3 feature for gnu backend based targets.
In order to ensure backward compability a flag is added which avoids
this and produces debug info in DWARF v2 format.
Bug 17392.
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
composites).
- Implement the "1/2/4/8" composite return policy, used by IA32/MacOS X and IA32/BSD.
- Move the default passing conventions from Machine.ml to compcert.ini, making it easier to test the various conventions.
- More comprehensive interoperability test in regression/interop1.c.
|
|
|
|
|
|
|
|
| |
configure: special ABI value for IA32/MacOSX and PowerPC/Linux
cparser/Machine: special config for PowerPC/Linux
cparser/StructReturn: generate better code for return-as-int
driver/Clflags, driver/Driver: add options -fstruct-return=<convention>
and -fstruct-passing=<convention> to simplify testing
|
|
|
|
| |
them off.
|
| |
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2572 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
| |
ARM: various tweaks, incl. support for SDIV and UDIV insns when available.
test/regression/funptr2.c: Thumb does weird things with <function ptr>+1.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2555 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2549 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
|
|
|
|
|
|
|
|
| |
Refactored compilation flags that affect the Coq part (module Compopts).
Added support for C99 for loops with declarations.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2410 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
| |
- Revised printing of intermediate RTL code.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2403 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
| |
(option -funprototyped, on by default).
Error if call to vararg function and -fvararg-calls is off.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2389 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- C function types and Cminor signatures annotated by calling conventions.
esp. vararg / not vararg
- Cshmgen: generate correct code for function call where there are
more arguments than listed in the function prototype. This is
still undefined behavior according to the formal semantics,
but correct code is generated.
- C2C, */PrintAsm.ml: remove "printf$iif" hack.
- powerpc/, checklink/: don't generate stubs for variadic functions.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2386 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2381 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2374 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
| |
with some static analysis tools).
*/PrintAsm.ml: don't cfa_adjust at Pfreeframe, as more code from the function can appear after (in case of tailcalls).
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2256 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2253 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
| |
test/c: added SHA3
cfrontend: support casts between long long and pointers, and comparisons between them.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2213 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
|
|
|
|
|
|
|
| |
Added options -falign-branch-targets and -falign-cond-branches (experimental).
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2113 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
| |
cparser: add primitive support for enum types.
bitfield emulation: for bitfields with enum type, choose signed/unsigned as appropriate
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2074 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2061 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1945 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
| |
- Revised memory model with Max and Cur permissions, but without bounds
- Constant propagation of 'const' globals
- Function inlining at RTL level
- (Unprovable) elimination of unreferenced static definitions
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1899 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
| |
cparser: renamed Errors to Cerrors; removed packing into Cparser.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1856 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
| |
(definitely not semantics-preserving; hard to justify).
CPragmas: make sure SDAs are not recognized on MacOSX.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1836 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1832 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
| |
- native treatment of volatile accesses in CompCert C's semantics
- translation of volatile accesses to built-ins in SimplExpr
- native treatment of struct assignment and passing struct parameter by value
- only passing struct result by value remains emulated
- in cparser, remove emulations that are no longer used
- added C99's type _Bool and used it to express || and && more efficiently.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1814 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
| |
Added -flongdouble option (to turn long double into double)
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1731 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
| |
test/*/Makefile: normalized 'bench' target
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1717 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
| |
test/regression: int main() so that interpretation works
Revised once more implementation of __builtin_memcpy (to check for PPC & ARM)
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1688 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
| |
cparser: added experimental emulation of packed structs (PackedStruct.ml)
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1650 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
| |
Added -dmach option and corresponding printer for Mach code.
CleanupLabelsproof.v: fixed for ARM
Driver.ml: -E sends output to stdout; support for .s and .S source files.
cparser/Elab.ml: spurious comment deleted.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1648 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1500 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Csyntax, Csem: source C language has side-effects within expressions,
performs implicit casts, and has nondeterministic reduction semantics
for expressions
- Cstrategy: deterministic red. sem. for the above
- Clight: the previous source C language, with pure expressions.
Added: temporary variables + implicit casts.
- New pass SimplExpr to pull side-effects out of expressions
(previously done in untrusted Caml code in cparser/)
- Csharpminor: added temporary variables to match Clight.
- Cminorgen: adapted, removed cast optimization (moved to back-end)
- CastOptim: RTL-level optimization of casts
- cparser: transformations Bitfields, StructByValue and StructAssign
now work on non-simplified expressions
- Added pretty-printers for several intermediate languages,
and matching -dxxx command-line flags.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1467 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|
|
|
|
| |
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1275 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
|