Tcwg-commits July 2015

tcwg-commits@lists.linaro.org

2 participants
1355 discussions

by git

This is an automated email from the git hooks/post-receive script. unknown user pushed a commit to branch hjl/pr66960/x86 in repository gcc. commit 386f403afc660ec59e3db58230e93dfa7ff5acb9 Author: H.J. Lu <hjl.tools(a)gmail.com> Date: Tue Jul 21 14:32:09 2015 -0700 Add __builtin_stack_top When __builtin_frame_address is used to retrieve the address of the function stack frame, the frame pointer register is required, which wastes one register and 2 instructions. For x86-32, one less register means significant negative impact on performance. This patch adds a new builtin function, __builtin_stack_top, to x86 backend. It returns the stack address when the function is called. gcc/ PR target/66960 * config/i386/i386.c (ix86_expand_prologue): Sorry if DRAP is used and the stack address has been taken. (ix86_builtins): Add IX86_BUILTIN_STACK_TOP. (ix86_init_mmx_sse_builtins): Add __builtin_stack_top. (ix86_expand_builtin): Handle IX86_BUILTIN_STACK_TOP. * config/i386/i386.h (machine_function): Add stack_top_taken. * doc/extend.texi: Document __builtin_stack_top. gcc/testsuite/ PR target/66960 * gcc.target/i386/pr66960-1.c: New test. * gcc.target/i386/pr66960-2.c: Likewise. * gcc.target/i386/pr66960-3.c: Likewise. * gcc.target/i386/pr66960-4.c: Likewise. * gcc.target/i386/pr66960-5.c: Likewise. --- gcc/config/i386/i386.c | 28 ++++++++++++++++++++++++++ gcc/config/i386/i386.h | 4 ++++ gcc/doc/extend.texi | 6 ++++++ gcc/testsuite/gcc.target/i386/pr66960-1.c | 33 +++++++++++++++++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-2.c | 33 +++++++++++++++++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-3.c | 17 ++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-4.c | 21 ++++++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-5.c | 21 ++++++++++++++++++++ 8 files changed, 163 insertions(+) diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c index ede8ea0..945bbf8 100644 --- a/gcc/config/i386/i386.c +++ b/gcc/config/i386/i386.c @@ -11575,6 +11575,12 @@ ix86_expand_prologue (void) { int align_bytes = crtl->stack_alignment_needed / BITS_PER_UNIT; + /* Can't use DRAP if the stack address has been taken. */ + if (cfun->machine->stack_top_taken) + sorry ("%<__builtin_stack_top%> not supported with stack" + " realignment. This may be worked around by adding" + " -maccumulate-outgoing-args."); + /* Only need to push parameter pointer reg if it is caller saved. */ if (!call_used_regs[REGNO (crtl->drap_reg)]) { @@ -30741,6 +30747,9 @@ enum ix86_builtins IX86_BUILTIN_READ_FLAGS, IX86_BUILTIN_WRITE_FLAGS, + /* Get the stack address when the function is called. */ + IX86_BUILTIN_STACK_TOP, + IX86_BUILTIN_MAX }; @@ -34353,6 +34362,10 @@ ix86_init_mmx_sse_builtins (void) def_builtin (OPTION_MASK_ISA_MWAITX, "__builtin_ia32_mwaitx", VOID_FTYPE_UNSIGNED_UNSIGNED_UNSIGNED, IX86_BUILTIN_MWAITX); + /* Get the stack address when the function is called. */ + def_builtin (0, "__builtin_stack_top", + PVOID_FTYPE_VOID, IX86_BUILTIN_STACK_TOP); + /* Add FMA4 multi-arg argument instructions */ for (i = 0, d = bdesc_multi_arg; i < ARRAY_SIZE (bdesc_multi_arg); i++, d++) { @@ -40291,6 +40304,21 @@ addcarryx: emit_insn (gen_xabort (op0)); return 0; + case IX86_BUILTIN_STACK_TOP: + cfun->machine->stack_top_taken = true; + + if (!target + || GET_MODE (target) != Pmode + || !register_operand (target, Pmode)) + target = gen_reg_rtx (Pmode); + + /* After the prologue, stack top is at -WORD(AP) in the current + frame. */ + emit_insn (gen_rtx_SET (target, + plus_constant (Pmode, arg_pointer_rtx, + -UNITS_PER_WORD))); + return target; + default: break; } diff --git a/gcc/config/i386/i386.h b/gcc/config/i386/i386.h index 7bd23ec..3781a22 100644 --- a/gcc/config/i386/i386.h +++ b/gcc/config/i386/i386.h @@ -2492,6 +2492,10 @@ struct GTY(()) machine_function { /* If true, it is safe to not save/restore DRAP register. */ BOOL_BITFIELD no_drap_save_restore : 1; + /* If true, the stack address of the current function has been + taken. */ + BOOL_BITFIELD stack_top_taken : 1; + /* If true, there is register available for argument passing. This is used only in ix86_function_ok_for_sibcall by 32-bit to determine if there is scratch register available for indirect sibcall. In diff --git a/gcc/doc/extend.texi b/gcc/doc/extend.texi index b18d8fb..971b584 100644 --- a/gcc/doc/extend.texi +++ b/gcc/doc/extend.texi @@ -16661,6 +16661,12 @@ The following built-in function is always available. @item void __builtin_ia32_pause (void) Generates the @code{pause} machine instruction with a compiler memory barrier. + +@item void *__builtin_stack_top (void) +This function is similar to calling @code{__builtin_frame_address} +with a value of @code{0}, but it returns the stack address when the +function is called. Unlike @code{__builtin_frame_address}, the frame +pointer register isn't required. @end table The following floating-point built-in functions are made available in the diff --git a/gcc/testsuite/gcc.target/i386/pr66960-1.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..aaab3cf --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-1.c @@ -0,0 +1,33 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -fomit-frame-pointer" { target { lp64 } } } */ +/* { dg-options "-O2 -fomit-frame-pointer -maddress-mode=short" { target { x32 } } } */ +/* { dg-options "-O2 -fomit-frame-pointer -miamcu" { target { ia32 } } } */ + +extern char **environ; +extern void exit (int status); +extern int main (long argc, char **argv, char **envp); + +void +_start (void) +{ + void *argc_p = __builtin_stack_top (); + char **argv = (char **) (argc_p + sizeof (void *)); + long argc = *(long *) argc_p; + int status; + + environ = argv + argc + 1; + + status = main (argc, argv, environ); + + exit (status); +} + +/* { dg-final { scan-assembler "movq\[ \t\]8\$%rsp\$, %rdi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]16\$%rsp\$, %rsi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]24\$%rsp,%rdi,8\$, %rdx" { target lp6 [...] +/* { dg-final { scan-assembler "movl\[ \t\]8\$%esp\$, %edi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]12\$%rsp\$, %esi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%rsi,%rdi,4\$, %edx" { target x32 [...] +/* { dg-final { scan-assembler "movl\[ \t\]\$%esp\$, %eax" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%esp\$, %edx" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]8\$%esp,%eax,4\$, %ecx" { target ia32 [...] diff --git a/gcc/testsuite/gcc.target/i386/pr66960-2.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..b9dbde2 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-2.c @@ -0,0 +1,33 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -fno-omit-frame-pointer" { target { lp64 } } } */ +/* { dg-options "-O2 -fno-omit-frame-pointer -maddress-mode=short" { target { x32 [...] +/* { dg-options "-O2 -fno-omit-frame-pointer -miamcu" { target { ia32 } } } */ + +extern char **environ; +extern void exit (int status); +extern int main (long argc, char **argv, char **envp); + +void +_start (void) +{ + void *argc_p = __builtin_stack_top (); + char **argv = (char **) (argc_p + sizeof (void *)); + long argc = *(long *) argc_p; + int status; + + environ = argv + argc + 1; + + status = main (argc, argv, environ); + + exit (status); +} + +/* { dg-final { scan-assembler "movq\[ \t\]8\$%rbp\$, %rdi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]16\$%rbp\$, %rsi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]24\$%rbp,%rdi,8\$, %rdx" { target lp6 [...] +/* { dg-final { scan-assembler "movl\[ \t\]8\$%ebp\$, %edi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]12\$%rbp\$, %esi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%rsi,%rdi,4\$, %edx" { target x32 [...] +/* { dg-final { scan-assembler "movl\[ \t\]4\$%ebp\$, %eax" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]8\$%ebp\$, %edx" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]12\$%ebp,%eax,4\$, %ecx" { target ia3 [...] diff --git a/gcc/testsuite/gcc.target/i386/pr66960-3.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..48cf25e --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-3.c @@ -0,0 +1,17 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -mno-accumulate-outgoing-args" { target { lp64 } } } */ +/* { dg-options "-O2 -mno-accumulate-outgoing-args -maddress-mode=short" { target [...] +/* { dg-options "-O2 -mno-accumulate-outgoing-args -miamcu" { target { ia32 } } } */ + +extern void abort (void); +extern int check_int (int *i, int align); +typedef int aligned __attribute__((aligned(64))); + +void * +foo (void) +{ + aligned j; + if (check_int (&j, __alignof__(j)) != j) + abort (); + return __builtin_stack_top (); +} /* { dg-message "sorry, unimplemented: .__builtin_stack_top. not supported" } */ diff --git a/gcc/testsuite/gcc.target/i386/pr66960-4.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..44c0b26 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-4.c @@ -0,0 +1,21 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -maccumulate-outgoing-args" { target { lp64 } } } */ +/* { dg-options "-O2 -maccumulate-outgoing-args -maddress-mode=short" { target { x [...] +/* { dg-options "-O2 -maccumulate-outgoing-args -miamcu" { target { ia32 } } } */ + +extern void abort (void); +extern int check_int (int *i, int align); +typedef int aligned __attribute__((aligned(64))); + +void * +foo (void) +{ + aligned j; + if (check_int (&j, __alignof__(j)) != j) + abort (); + return __builtin_stack_top (); +} + +/* { dg-final { scan-assembler "leaq\[ \t\]8\$%rbp\$, %rax" { target lp64 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]8\$%rbp\$, %eax" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%ebp\$, %eax" { target ia32 } } } */ diff --git a/gcc/testsuite/gcc.target/i386/pr66960-5.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..d449437 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-5.c @@ -0,0 +1,21 @@ +/* { dg-do link } */ +/* { dg-options "-O" } */ + +extern void link_error (void); + +__attribute__ ((noinline, noclone)) +void +foo (void) +{ + void **p = __builtin_stack_top (); + void *ra = __builtin_return_address (0); + if (*p != ra) + link_error (); +} + +int +main (void) +{ + foo (); + return 0; +} -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[dejagnu] 01/01: Merge branch 'master' into linaro

by git

This is an automated email from the git hooks/post-receive script. unknown user pushed a commit to branch linaro in repository dejagnu. commit 31aa24d3683d7150cae7b7c61b4d21af06eabd44 Merge: c465c48 e7d7a3e Author: Rob Savoye <rob(a)welcomehome.org> Date: Thu Jul 30 13:25:20 2015 -0600 Merge branch 'master' into linaro Conflicts: ChangeLog ChangeLog | 10 +++++-- lib/remote.exp | 86 ++++++++++++++++++++++++++++++++++++++++++---------------- 2 files changed, 70 insertions(+), 26 deletions(-) diff --cc ChangeLog index 9145142,0bdaeb9..62b5eff --- a/ChangeLog +++ b/ChangeLog @@@ -45,8 -46,7 +46,11 @@@ 2015-05-12 Steve Ellcey <sellcey(a)imgtec.com> * baseboards/generic-sim.exp: Check $DEJAGNU_SIM_OPTION. - >>>>>>> master + ++2015-04-17 Rob Savoye <rob.savoye(a)linaro.org> ++ ++ * lib/libgloss.exp (find_g++): Look for xg++ before g++. + 2015-03-30 Ben Elliston <bje(a)gnu.org> Revert the following change which causes breakage reported here: -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[dejagnu] branch linaro updated (c465c48 -> 31aa24d)

by git

This is an automated email from the git hooks/post-receive script. unknown user pushed a change to branch linaro in repository dejagnu. from c465c48 update from master since there has been another stable releas [...] adds e7d7a3e * lib/remote.exp (close_wait_program): New procedure. (loca [...] new 31aa24d Merge branch 'master' into linaro The 1 revisions listed above as "new" are entirely new to this repository and will be described in separate emails. The revisions listed as "adds" were already present in the repository and have only been added to this reference. Summary of changes: ChangeLog | 10 +++++-- lib/remote.exp | 86 ++++++++++++++++++++++++++++++++++++++++++---------------- 2 files changed, 70 insertions(+), 26 deletions(-) -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[gcc] branch hjl/pr66960/generic updated (f2f00c3 -> 03ff139)

by git

This is an automated email from the git hooks/post-receive script. unknown user pushed a change to branch hjl/pr66960/generic in repository gcc. discards f2f00c3 Add __builtin_stack_top adds e954729 Daily bump. adds 9b6b471 2015-07-27 Richard Biener <rguenther(a)suse.de> adds 6784a47 PR c++/66555 PR c/54979 * c-common.c (find_array_ref_with_ [...] adds ab2874b 2015-07-27 Matthew Wahab <matthew.wahab(a)arm.com> adds 1ed4ddb Fix logic error in Fortran OpenACC parsing adds eab7bd2 [PATCH][AArch64] Improve spill code - swap order in shl pattern adds 505e1f9 [ARM] PR/63870 Add qualifier to check lane bounds in expand adds 732205b [ARM] PR/63870 Add a __builtin_lane_check adds f4f2d7d [PATCH][AArch64] Improve spill code - swap order in shr patterns adds e5e363a Add more tests for PR target/66232 adds bccc47e PR debug/66468 * dwarf2out.c (gen_inlined_subroutine_die): [...] adds 37715b0 * constexpr.c (cxx_eval_call_expression): Don't add this cal [...] adds d9e4fe2 PR bootstrap/67030 * c-common.c (warn_tautological_cmp): Do [...] adds e5b7f5c * ipa-devirt.c (types_same_for_odr): Fix typo. adds 3b73c99 Update outer-4.c and uns-outer-4.c adds 81fbee0 Simplify gather_scalar_reductions adds 2bb2f6b * c-c++-common/Wtautological-compare-2.c: Remove unused line. adds dbbd4c6 gcc/c/ChangeLog adds b419f44 Daily bump. adds c14bac8 * config.sub, config.guess: Import from upstream. adds 60a4580 Honour DriverOnly for enum values in error messages adds ddc9c98 2015-07-28 Thomas Preud'homme <thomas.preudhomme(a)arm.com> adds 16db639 2015-07-28 Richard Biener <rguenther(a)suse.de> adds 95f4166 Handle double reduction in parloops adds 95328f3 * config/rl78/rl78.c (rl78_addsi3_internal): New function. [...] adds 66ed189 2015-07-28 Paolo Carlini <paolo.carlini(a)oracle.com> adds 9d8e046 [AArch64] Properly handle simple arith+extend ops in rtx costs adds 39e21ca 2015-07-28 Richard Biener <rguenther(a)suse.de> adds aa0bf07 2015-07-28 Richard Biener <rguenther(a)suse.de> adds 63593ac 2015-07-28 Eric Gallager <egall(a)gwmail.gwu.edu> adds 68aee03 gcc/ * config/fr30/fr30.md (indirect_jump): Use pmode_regist [...] adds de4e153 gcc/ * target-insns.def (indirect_jump): New targetm instruc [...] adds f027535 gcc/ * target-insns.def (eh_return): New targetm instruction [...] adds 4dd7c28 gcc/ * target-insns.def (can_extend, ptr_extend): New target [...] adds f37fec3 gcc/ * target-insns.def (atomic_test_and_set): New targetm i [...] adds 8a93a82 gcc/ * target-insns.def (reload_load_address): New targetm i [...] adds d145b68 gcc/ 2015-07-28 David Sherwood <david.sherwood(a)arm.com> adds 64ae101 Daily bump. adds afd2b95 2015-07-28 Bill Schmidt <wschmidt(a)linux.vnet.ibm.com> adds 940cb7d PR libstdc++/67015 * include/bits/regex_compiler.h (_Compil [...] adds 64a6962 Use fast unwinder for PowerPC adds 1ae7fdc 2015-07-29 Richard Biener <rguenther(a)suse.de> adds 951c31b 2015-07-29 Richard Biener <rguenther(a)suse.de> adds 5e0b3d2 2015-07-29 Richard Biener <rguenther(a)suse.de> adds 6f52065 [AArch64] Add basic FP16 support adds 87867af 2015-07-29 Ville Voutilainen <ville.voutilainen(a)gmail.com> adds b7c1cfe 2015-07-29 Ville Voutilainen <ville.voutilainen(a)gmail.com> adds ed5acf2 * testsuite/24_iterators/container_access.cc: New. adds afa3578 [ARM/AArch64 Testsuite] Add basic fp16 tests adds 236ea32 2015-07-29 Richard Biener <rguenther(a)suse.de> adds 042f02d [AArch64 Testsuite] vld1-vst1_1.c: Add missing float32x4_t case adds 1ee54bd [AArch64 Testsuite] vld1_lane.c: Remove unused test data adds a13f5fb PR libstdc++/66829 * testsuite/lib/libstdc++.exp (v3-build_ [...] adds df2106b PR libgfortran/66650 * libgfortran.h (GFC_DTYPE_SIZE_MASK): [...] adds d46c0c7 * MAINTAINERS (Write After Approval): Add myself. adds e4d0660 Fix whitespace in my MAINTAINERS change commit. adds 1fe39f9 Define DBX_REGISTER_NUMBER for IA MCU adds 0ba1028 [graphite] reduce the number of params in a scop to 3 adds 84a30a5 syscall: Fix to libgo/mksysinfo.sh adds dcb1277 Daily bump. adds b12954a PR target/66217 PR target/67045 * config/rs6000/rs6000.md [...] adds 081da05 DR 1558 PR c++/67021 * pt.c (tsubst_decl) [TYPE_DECL]: Cle [...] adds c4699ff 2015-07-30 Richard Biener <rguenther(a)suse.de> adds 18b3866 2015-07-30 Richard Biener <rguenther(a)suse.de> adds f58dd84 * c-common.c (warn_tautological_cmp): Bail for float types. adds 330d065 * include/experimental/any (any::operator=(const any&)): Che [...] adds 43d8404 2015-07-30 Paolo Carlini <paolo.carlini(a)oracle.com> adds 17af35c * include/bits/regex_automaton.h (_State_base, _State<>): R [...] adds 436dc21 2015-07-30 Richard Biener <rguenther(a)suse.de> adds 0879fd1 * testsuite/experimental/filesystem/operations/current_path. [...] adds 6e25659 gcc/ 2015-07-30 David Sherwood <david.sherwood(a)arm.com> adds de299e5 2015-07-30 Richard Biener <rguenther(a)suse.de> adds f4a09f4 gcc/ * builtins.c (HAVE_atomic_clear, gen_atomic_clear): Del [...] adds 7b42a29 [AArch64] Removed unused VRL2/3/4 iterator values adds 5638e47 * gcc.target/i386/pr66648.c: Move to correct place. new 03ff139 Add __builtin_stack_top This update added new revisions after undoing existing revisions. That is to say, some revisions that were in the old version of the branch are not in the new version. This situation occurs when a user --force pushes a change and generates a repository containing something like this: * -- * -- B -- O -- O -- O (f2f00c3) \ N -- N -- N refs/heads/hjl/pr66960/generic (03ff139) You should already have received notification emails for all of the O revisions, and so the following emails describe only the N revisions from the common base, B. Any revisions marked "omits" are not gone; other references still refer to them. Any revisions marked "discards" are gone forever. The 1 revisions listed above as "new" are entirely new to this repository and will be described in separate emails. The revisions listed as "adds" were already present in the repository and have only been added to this reference. Summary of changes: ChangeLog | 13 + MAINTAINERS | 1 + config.guess | 31 +- config.sub | 18 +- fixincludes/ChangeLog | 4 + fixincludes/check.tpl | 5 +- gcc/ChangeLog | 355 +++++++++++++++++++++ gcc/DATESTAMP | 2 +- gcc/builtins.c | 12 - gcc/c-family/ChangeLog | 18 ++ gcc/c-family/c-common.c | 76 +++++ gcc/c-family/c-common.h | 1 + gcc/c-family/c.opt | 4 + gcc/c/ChangeLog | 12 + gcc/c/c-decl.c | 20 +- gcc/c/c-typeck.c | 3 + gcc/config/aarch64/aarch64-builtins.c | 9 + gcc/config/aarch64/aarch64-modes.def | 4 + gcc/config/aarch64/aarch64-simd.md | 2 +- gcc/config/aarch64/aarch64.c | 60 +++- gcc/config/aarch64/aarch64.h | 4 +- gcc/config/aarch64/aarch64.md | 97 ++++-- gcc/config/aarch64/iterators.md | 18 +- gcc/config/arm/arm-arches.def | 3 +- gcc/config/arm/arm-builtins.c | 84 ++++- gcc/config/arm/arm-c.c | 5 + gcc/config/arm/arm-cores.def | 4 +- gcc/config/arm/arm-protos.h | 8 +- gcc/config/arm/arm-tables.opt | 31 +- gcc/config/arm/arm.c | 45 +-- gcc/config/arm/arm.h | 2 +- gcc/config/arm/arm_neon_builtins.def | 82 ++--- gcc/config/arm/driver-arm.c | 3 + gcc/config/arm/neon.md | 41 +-- gcc/config/fr30/fr30.md | 4 +- gcc/config/i386/i386.c | 31 +- gcc/config/i386/iamcu.h | 11 + gcc/config/i386/sse.md | 6 +- gcc/config/rl78/rl78-protos.h | 1 + gcc/config/rl78/rl78.c | 30 ++ gcc/config/rl78/rl78.md | 5 +- gcc/config/rs6000/rs6000.c | 2 +- gcc/config/rs6000/rs6000.md | 35 +- gcc/config/spu/spu.c | 10 +- gcc/cp/ChangeLog | 56 ++++ gcc/cp/call.c | 22 +- gcc/cp/class.c | 80 +++-- gcc/cp/constexpr.c | 5 +- gcc/cp/decl.c | 5 +- gcc/cp/except.c | 5 +- gcc/cp/friend.c | 6 +- gcc/cp/mangle.c | 18 +- gcc/cp/method.c | 22 +- gcc/cp/name-lookup.c | 36 ++- gcc/cp/parser.c | 6 +- gcc/cp/pt.c | 29 +- gcc/cp/search.c | 2 +- gcc/cp/semantics.c | 4 +- gcc/doc/invoke.texi | 19 +- gcc/dwarf2out.c | 4 + gcc/emit-rtl.c | 7 +- gcc/except.c | 6 +- gcc/expmed.c | 4 - gcc/fold-const.c | 26 +- gcc/fortran/ChangeLog | 5 + gcc/fortran/parse.c | 6 +- gcc/function.c | 2 - gcc/genmatch.c | 258 +++++++++------ gcc/genmodes.c | 4 +- gcc/gimple-fold.c | 216 +++++++------ gcc/go/gofrontend/MERGE | 2 +- gcc/ipa-devirt.c | 2 +- gcc/lto-streamer-out.c | 4 +- gcc/machmode.h | 20 +- gcc/match.pd | 97 ++++-- gcc/optabs.c | 32 +- gcc/opts-common.c | 2 + gcc/params.def | 2 +- gcc/reload1.c | 6 +- gcc/rtlanal.c | 15 +- gcc/simplify-rtx.c | 42 ++- gcc/stor-layout.c | 14 +- gcc/target-insns.def | 6 + gcc/testsuite/ChangeLog | 69 ++++ .../c-c++-common/Wtautological-compare-1.c | 70 ++++ .../c-c++-common/Wtautological-compare-2.c | 13 + .../c-c++-common/Wtautological-compare-3.c | 20 ++ gcc/testsuite/g++.dg/cpp0x/alias-decl-52.C | 24 ++ gcc/testsuite/g++.dg/cpp0x/decltype-54581.C | 2 +- gcc/testsuite/g++.dg/other/vector-compare.C | 2 +- gcc/testsuite/gcc.dg/autopar/outer-4.c | 19 +- gcc/testsuite/gcc.dg/autopar/uns-outer-4.c | 17 +- gcc/testsuite/gcc.dg/parm-incomplete-1.c | 1 - gcc/testsuite/gcc.dg/pr18809-1.c | 1 - gcc/testsuite/gcc.dg/pr27953.c | 1 - gcc/testsuite/gcc.dg/vla-11.c | 1 - gcc/testsuite/gcc.target/aarch64/f16_movs_1.c | 26 ++ .../gcc.target/aarch64/fp16/f16_convs_1.c | 34 ++ .../gcc.target/aarch64/fp16/f16_convs_2.c | 33 ++ .../{nvptx/nvptx.exp => aarch64/fp16/fp16.exp} | 13 +- gcc/testsuite/gcc.target/aarch64/vld1-vst1_1.c | 1 + gcc/testsuite/gcc.target/aarch64/vld1_lane.c | 29 +- gcc/testsuite/gcc.target/i386/pr66232-6.c | 13 + gcc/testsuite/gcc.target/i386/pr66232-7.c | 14 + gcc/testsuite/gcc.target/i386/pr66232-8.c | 13 + gcc/testsuite/gcc.target/i386/pr66232-9.c | 13 + gcc/testsuite/gcc.target/{ => i386}/pr66648.c | 0 gcc/tree-parloops.c | 90 +++++- gcc/tree-ssa-math-opts.c | 2 +- libgfortran/ChangeLog | 6 + libgfortran/libgfortran.h | 3 +- libgo/mksysinfo.sh | 6 +- libgomp/ChangeLog | 4 + libgomp/testsuite/libgomp.c/uns-outer-4.c | 36 +++ libsanitizer/ChangeLog | 14 + .../sanitizer_common/sanitizer_stacktrace.cc | 11 + .../sanitizer_common/sanitizer_stacktrace.h | 3 +- libstdc++-v3/ChangeLog | 56 ++++ libstdc++-v3/include/bits/functional_hash.h | 28 ++ libstdc++-v3/include/bits/range_access.h | 121 +++++-- libstdc++-v3/include/bits/regex_automaton.h | 87 ++++- libstdc++-v3/include/bits/regex_automaton.tcc | 18 +- libstdc++-v3/include/bits/regex_compiler.h | 9 +- libstdc++-v3/include/bits/regex_compiler.tcc | 35 +- libstdc++-v3/include/bits/regex_executor.h | 2 +- libstdc++-v3/include/bits/regex_executor.tcc | 8 +- libstdc++-v3/include/experimental/any | 3 + libstdc++-v3/include/std/regex | 1 + .../testsuite/20_util/hash/{52931.cc => 60970.cc} | 24 +- .../testsuite/24_iterators/container_access.cc | 89 ++++++ .../algorithms/regex_match/cstring_bracket_01.cc | 57 ++++ .../64646.cc => experimental/any/assign/self.cc} | 14 +- .../experimental/any/misc/any_cast_neg.cc | 2 +- .../filesystem/operations/current_path.cc | 6 +- libstdc++-v3/testsuite/lib/libstdc++.exp | 2 +- 135 files changed, 2615 insertions(+), 860 deletions(-) create mode 100644 gcc/testsuite/c-c++-common/Wtautological-compare-1.c create mode 100644 gcc/testsuite/c-c++-common/Wtautological-compare-2.c create mode 100644 gcc/testsuite/c-c++-common/Wtautological-compare-3.c create mode 100644 gcc/testsuite/g++.dg/cpp0x/alias-decl-52.C create mode 100644 gcc/testsuite/gcc.target/aarch64/f16_movs_1.c create mode 100644 gcc/testsuite/gcc.target/aarch64/fp16/f16_convs_1.c create mode 100644 gcc/testsuite/gcc.target/aarch64/fp16/f16_convs_2.c copy gcc/testsuite/gcc.target/{nvptx/nvptx.exp => aarch64/fp16/fp16.exp} (80%) create mode 100644 gcc/testsuite/gcc.target/i386/pr66232-6.c create mode 100644 gcc/testsuite/gcc.target/i386/pr66232-7.c create mode 100644 gcc/testsuite/gcc.target/i386/pr66232-8.c create mode 100644 gcc/testsuite/gcc.target/i386/pr66232-9.c rename gcc/testsuite/gcc.target/{ => i386}/pr66648.c (100%) create mode 100644 libgomp/testsuite/libgomp.c/uns-outer-4.c copy libstdc++-v3/testsuite/20_util/hash/{52931.cc => 60970.cc} (64%) create mode 100644 libstdc++-v3/testsuite/24_iterators/container_access.cc copy libstdc++-v3/testsuite/{25_algorithms/is_permutation/64646.cc => experimental [...] -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[gcc] 01/01: Add __builtin_stack_top

by git

This is an automated email from the git hooks/post-receive script. unknown user pushed a commit to branch hjl/pr66960/generic in repository gcc. commit 03ff13915082db6f527d3e46b0282c224c934668 Author: H.J. Lu <hjl.tools(a)gmail.com> Date: Tue Jul 21 14:32:09 2015 -0700 Add __builtin_stack_top When __builtin_frame_address is used to retrieve the address of the function stack frame, the frame pointer register is required, which wastes one register and 2 instructions. For x86-32, one less register means significant negative impact on performance. This patch adds a new builtin function, __builtin_stack_top. It returns the stack address when the function is called. This patch only enables __builtin_stack_top for x86 backend. Using __builtin_stack_top with other backends will lead to sorry, unimplemented: ‘__builtin_stack_top’ not supported on this target TARGET_STACK_TOP_RTX must be defined to enable __builtin_stack_top. default_stack_top_rtx may be extended to support more backends, including those with INITIAL_FRAME_ADDRESS_RTX. gcc/ PR target/66960 * builtin-types.def (BT_FN_PTR_VOID): New function type. * builtins.c (expand_builtin): Handle BUILT_IN_STACK_TOP. (is_simple_builtin): Likewise. * ipa-pure-const.c (special_builtin_state): Likewise. * builtins.def: Add BUILT_IN_STACK_TOP. * function.h (function): Add stack_top_taken. * target.def (stack_top_rtx): New target hook. * targhooks.c (default_stack_top_rtx): New. * targhooks.h (default_stack_top_rtx): Likewise. * config/i386/i386.c (ix86_expand_prologue): Sorry if DRAP is used and the stack address has been taken. (TARGET_STACK_TOP_RTX): New. * doc/extend.texi: Document __builtin_stack_top. * doc/tm.texi.in (TARGET_STACK_TOP_RTX): New. * doc/tm.texi: Regenerated. gcc/testsuite/ PR target/66960 * gcc.target/i386/pr66960-1.c: New test. * gcc.target/i386/pr66960-2.c: Likewise. * gcc.target/i386/pr66960-3.c: Likewise. * gcc.target/i386/pr66960-4.c: Likewise. * gcc.target/i386/pr66960-5.c: Likewise. --- gcc/builtin-types.def | 1 + gcc/builtins.c | 11 +++++++++++ gcc/builtins.def | 1 + gcc/config/i386/i386.c | 8 ++++++++ gcc/doc/extend.texi | 7 +++++++ gcc/doc/tm.texi | 5 +++++ gcc/doc/tm.texi.in | 2 ++ gcc/function.h | 3 +++ gcc/ipa-pure-const.c | 1 + gcc/target.def | 7 +++++++ gcc/targhooks.c | 9 +++++++++ gcc/targhooks.h | 3 +++ gcc/testsuite/gcc.target/i386/pr66960-1.c | 33 +++++++++++++++++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-2.c | 33 +++++++++++++++++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-3.c | 17 ++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-4.c | 21 ++++++++++++++++++++ gcc/testsuite/gcc.target/i386/pr66960-5.c | 21 ++++++++++++++++++++ 17 files changed, 183 insertions(+) diff --git a/gcc/builtin-types.def b/gcc/builtin-types.def index 0e34531..2b6b5ab 100644 --- a/gcc/builtin-types.def +++ b/gcc/builtin-types.def @@ -177,6 +177,7 @@ DEF_FUNCTION_TYPE_1 (BT_FN_COMPLEX_LONGDOUBLE_LONGDOUBLE, BT_COMPLEX_LONGDOUBLE, BT_LONGDOUBLE) DEF_FUNCTION_TYPE_1 (BT_FN_PTR_UINT, BT_PTR, BT_UINT) DEF_FUNCTION_TYPE_1 (BT_FN_PTR_SIZE, BT_PTR, BT_SIZE) +DEF_FUNCTION_TYPE_1 (BT_FN_PTR_VOID, BT_PTR, BT_VOID) DEF_FUNCTION_TYPE_1 (BT_FN_INT_INT, BT_INT, BT_INT) DEF_FUNCTION_TYPE_1 (BT_FN_INT_UINT, BT_INT, BT_UINT) DEF_FUNCTION_TYPE_1 (BT_FN_INT_LONG, BT_INT, BT_LONG) diff --git a/gcc/builtins.c b/gcc/builtins.c index d7eb65f..140df84 100644 --- a/gcc/builtins.c +++ b/gcc/builtins.c @@ -6206,6 +6206,16 @@ expand_builtin (tree exp, rtx target, rtx subtarget, machine [...] case BUILT_IN_CONSTANT_P: return const0_rtx; + case BUILT_IN_STACK_TOP: + if (targetm.calls.stack_top_rtx) + { + cfun->stack_top_taken = true; + return targetm.calls.stack_top_rtx (); + } + else + sorry ("%<__builtin_stack_top%> not supported on this target"); + break; + case BUILT_IN_FRAME_ADDRESS: case BUILT_IN_RETURN_ADDRESS: return expand_builtin_frame_address (fndecl, exp); @@ -12395,6 +12405,7 @@ is_simple_builtin (tree decl) case BUILT_IN_RETURN: case BUILT_IN_AGGREGATE_INCOMING_ADDRESS: case BUILT_IN_FRAME_ADDRESS: + case BUILT_IN_STACK_TOP: case BUILT_IN_VA_END: case BUILT_IN_STACK_SAVE: case BUILT_IN_STACK_RESTORE: diff --git a/gcc/builtins.def b/gcc/builtins.def index 80e4a9c..62f0523 100644 --- a/gcc/builtins.def +++ b/gcc/builtins.def @@ -778,6 +778,7 @@ DEF_EXT_LIB_BUILTIN (BUILT_IN_FFSL, "ffsl", BT_FN_INT_LONG, [...] DEF_EXT_LIB_BUILTIN (BUILT_IN_FFSLL, "ffsll", BT_FN_INT_LONGLONG, ATTR_CONST_NO [...] DEF_EXT_LIB_BUILTIN (BUILT_IN_FORK, "fork", BT_FN_PID, ATTR_NOTHROW_LIST) DEF_GCC_BUILTIN (BUILT_IN_FRAME_ADDRESS, "frame_address", BT_FN_PTR_UINT, A [...] +DEF_GCC_BUILTIN (BUILT_IN_STACK_TOP, "stack_top", BT_FN_PTR_VOID, ATTR_NULL) /* [trans-mem]: Adjust BUILT_IN_TM_FREE if BUILT_IN_FREE is changed. */ DEF_LIB_BUILTIN (BUILT_IN_FREE, "free", BT_FN_VOID_PTR, ATTR_NOTHROW_LEAF_LIST) DEF_GCC_BUILTIN (BUILT_IN_FROB_RETURN_ADDR, "frob_return_addr", BT_FN_PTR_P [...] diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c index ede8ea0..f806f16 100644 --- a/gcc/config/i386/i386.c +++ b/gcc/config/i386/i386.c @@ -11575,6 +11575,12 @@ ix86_expand_prologue (void) { int align_bytes = crtl->stack_alignment_needed / BITS_PER_UNIT; + /* Can't use DRAP if the stack address has been taken. */ + if (cfun->stack_top_taken) + sorry ("%<__builtin_stack_top%> not supported with stack" + " realignment. This may be worked around by adding" + " -maccumulate-outgoing-args."); + /* Only need to push parameter pointer reg if it is caller saved. */ if (!call_used_regs[REGNO (crtl->drap_reg)]) { @@ -52576,6 +52582,8 @@ ix86_operands_ok_for_move_multiple (rtx *operands, bool load, #define TARGET_UPDATE_STACK_BOUNDARY ix86_update_stack_boundary #undef TARGET_GET_DRAP_RTX #define TARGET_GET_DRAP_RTX ix86_get_drap_rtx +#undef TARGET_STACK_TOP_RTX +#define TARGET_STACK_TOP_RTX default_stack_top_rtx #undef TARGET_STRICT_ARGUMENT_NAMING #define TARGET_STRICT_ARGUMENT_NAMING hook_bool_CUMULATIVE_ARGS_true #undef TARGET_STATIC_CHAIN diff --git a/gcc/doc/extend.texi b/gcc/doc/extend.texi index b18d8fb..36af976 100644 --- a/gcc/doc/extend.texi +++ b/gcc/doc/extend.texi @@ -8695,6 +8695,13 @@ This function should only be used with a nonzero argument fo [...] purposes. @end deftypefn +@deftypefn {Built-in Function} {void *} __builtin_stack_top (void) +This function is similar to calling @code{__builtin_frame_address} +with a value of @code{0}, but it returns the stack address when the +function is called. Unlike @code{__builtin_frame_address}, the frame +pointer register isn't required. +@end deftypefn + @node Vector Extensions @section Using Vector Instructions through Built-in Functions diff --git a/gcc/doc/tm.texi b/gcc/doc/tm.texi index f95646c..e2cd480 100644 --- a/gcc/doc/tm.texi +++ b/gcc/doc/tm.texi @@ -11483,6 +11483,11 @@ argument list due to stack realignment. Return @code{NULL [...] is needed. @end deftypefn +@deftypefn {Target Hook} rtx TARGET_STACK_TOP_RTX (void) +This hook should return an rtx for the stack address when the function +is called. +@end deftypefn + @deftypefn {Target Hook} bool TARGET_ALLOCATE_STACK_SLOTS_FOR_ARGS (void) When optimization is disabled, this hook indicates whether or not arguments should be allocated to stack slots. Normally, GCC allocates diff --git a/gcc/doc/tm.texi.in b/gcc/doc/tm.texi.in index 2383fb9..9167069 100644 --- a/gcc/doc/tm.texi.in +++ b/gcc/doc/tm.texi.in @@ -8181,6 +8181,8 @@ and the associated definitions of those functions. @hook TARGET_GET_DRAP_RTX +@hook TARGET_STACK_TOP_RTX + @hook TARGET_ALLOCATE_STACK_SLOTS_FOR_ARGS @hook TARGET_CONST_ANCHOR diff --git a/gcc/function.h b/gcc/function.h index e92c17c..dd1c38a 100644 --- a/gcc/function.h +++ b/gcc/function.h @@ -378,6 +378,9 @@ struct GTY(()) function { /* Set when the tail call has been identified. */ unsigned int tail_call_marked : 1; + + /* Set when the address of the stack top has been taken. */ + unsigned int stack_top_taken : 1; }; /* Add the decl D to the local_decls list of FUN. */ diff --git a/gcc/ipa-pure-const.c b/gcc/ipa-pure-const.c index 8fd8c36..2405082 100644 --- a/gcc/ipa-pure-const.c +++ b/gcc/ipa-pure-const.c @@ -480,6 +480,7 @@ special_builtin_state (enum pure_const_state_e *state, bool *looping, case BUILT_IN_CXA_END_CLEANUP: case BUILT_IN_EH_COPY_VALUES: case BUILT_IN_FRAME_ADDRESS: + case BUILT_IN_STACK_TOP: case BUILT_IN_APPLY: case BUILT_IN_APPLY_ARGS: *looping = false; diff --git a/gcc/target.def b/gcc/target.def index 4edc209..7a30f39 100644 --- a/gcc/target.def +++ b/gcc/target.def @@ -4525,6 +4525,13 @@ argument list due to stack realignment. Return @code{NULL} [...] is needed.", rtx, (void), NULL) +/* Get the stack address when the function is called. */ +DEFHOOK +(stack_top_rtx, + "This hook should return an rtx for the stack address when the function\n\ +is called.", + rtx, (void), NULL) + /* Return true if all function parameters should be spilled to the stack. */ DEFHOOK diff --git a/gcc/targhooks.c b/gcc/targhooks.c index 3eca47e..f188272 100644 --- a/gcc/targhooks.c +++ b/gcc/targhooks.c @@ -1926,4 +1926,13 @@ can_use_doloop_if_innermost (const widest_int &, const wides [...] return loop_depth == 1; } +/* Get the stack address when the function is called. After the + prologue, stack top is at -WORD(AP) in the current frame. */ + +rtx +default_stack_top_rtx (void) +{ + return plus_constant (Pmode, arg_pointer_rtx, -UNITS_PER_WORD); +} + #include "gt-targhooks.h" diff --git a/gcc/targhooks.h b/gcc/targhooks.h index 5ae991d..094a589 100644 --- a/gcc/targhooks.h +++ b/gcc/targhooks.h @@ -240,4 +240,7 @@ extern void default_setup_incoming_vararg_bounds (cumulative_ar [...] tree type ATTRIBUTE_UNUSED, int *pretend_arg_size ATTRIBUTE_UNUSED, int second_time ATTRIBUTE_UNUSED); + +extern rtx default_stack_top_rtx (void); + #endif /* GCC_TARGHOOKS_H */ diff --git a/gcc/testsuite/gcc.target/i386/pr66960-1.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..aaab3cf --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-1.c @@ -0,0 +1,33 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -fomit-frame-pointer" { target { lp64 } } } */ +/* { dg-options "-O2 -fomit-frame-pointer -maddress-mode=short" { target { x32 } } } */ +/* { dg-options "-O2 -fomit-frame-pointer -miamcu" { target { ia32 } } } */ + +extern char **environ; +extern void exit (int status); +extern int main (long argc, char **argv, char **envp); + +void +_start (void) +{ + void *argc_p = __builtin_stack_top (); + char **argv = (char **) (argc_p + sizeof (void *)); + long argc = *(long *) argc_p; + int status; + + environ = argv + argc + 1; + + status = main (argc, argv, environ); + + exit (status); +} + +/* { dg-final { scan-assembler "movq\[ \t\]8\$%rsp\$, %rdi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]16\$%rsp\$, %rsi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]24\$%rsp,%rdi,8\$, %rdx" { target lp6 [...] +/* { dg-final { scan-assembler "movl\[ \t\]8\$%esp\$, %edi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]12\$%rsp\$, %esi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%rsi,%rdi,4\$, %edx" { target x32 [...] +/* { dg-final { scan-assembler "movl\[ \t\]\$%esp\$, %eax" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%esp\$, %edx" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]8\$%esp,%eax,4\$, %ecx" { target ia32 [...] diff --git a/gcc/testsuite/gcc.target/i386/pr66960-2.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..b9dbde2 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-2.c @@ -0,0 +1,33 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -fno-omit-frame-pointer" { target { lp64 } } } */ +/* { dg-options "-O2 -fno-omit-frame-pointer -maddress-mode=short" { target { x32 [...] +/* { dg-options "-O2 -fno-omit-frame-pointer -miamcu" { target { ia32 } } } */ + +extern char **environ; +extern void exit (int status); +extern int main (long argc, char **argv, char **envp); + +void +_start (void) +{ + void *argc_p = __builtin_stack_top (); + char **argv = (char **) (argc_p + sizeof (void *)); + long argc = *(long *) argc_p; + int status; + + environ = argv + argc + 1; + + status = main (argc, argv, environ); + + exit (status); +} + +/* { dg-final { scan-assembler "movq\[ \t\]8\$%rbp\$, %rdi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]16\$%rbp\$, %rsi" { target lp64 } } } */ +/* { dg-final { scan-assembler "leaq\[ \t\]24\$%rbp,%rdi,8\$, %rdx" { target lp6 [...] +/* { dg-final { scan-assembler "movl\[ \t\]8\$%ebp\$, %edi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]12\$%rbp\$, %esi" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%rsi,%rdi,4\$, %edx" { target x32 [...] +/* { dg-final { scan-assembler "movl\[ \t\]4\$%ebp\$, %eax" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]8\$%ebp\$, %edx" { target ia32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]12\$%ebp,%eax,4\$, %ecx" { target ia3 [...] diff --git a/gcc/testsuite/gcc.target/i386/pr66960-3.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..48cf25e --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-3.c @@ -0,0 +1,17 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -mno-accumulate-outgoing-args" { target { lp64 } } } */ +/* { dg-options "-O2 -mno-accumulate-outgoing-args -maddress-mode=short" { target [...] +/* { dg-options "-O2 -mno-accumulate-outgoing-args -miamcu" { target { ia32 } } } */ + +extern void abort (void); +extern int check_int (int *i, int align); +typedef int aligned __attribute__((aligned(64))); + +void * +foo (void) +{ + aligned j; + if (check_int (&j, __alignof__(j)) != j) + abort (); + return __builtin_stack_top (); +} /* { dg-message "sorry, unimplemented: .__builtin_stack_top. not supported" } */ diff --git a/gcc/testsuite/gcc.target/i386/pr66960-4.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..44c0b26 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-4.c @@ -0,0 +1,21 @@ +/* { dg-do compile { target *-*-linux* } } */ +/* { dg-options "-O2 -maccumulate-outgoing-args" { target { lp64 } } } */ +/* { dg-options "-O2 -maccumulate-outgoing-args -maddress-mode=short" { target { x [...] +/* { dg-options "-O2 -maccumulate-outgoing-args -miamcu" { target { ia32 } } } */ + +extern void abort (void); +extern int check_int (int *i, int align); +typedef int aligned __attribute__((aligned(64))); + +void * +foo (void) +{ + aligned j; + if (check_int (&j, __alignof__(j)) != j) + abort (); + return __builtin_stack_top (); +} + +/* { dg-final { scan-assembler "leaq\[ \t\]8\$%rbp\$, %rax" { target lp64 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]8\$%rbp\$, %eax" { target x32 } } } */ +/* { dg-final { scan-assembler "leal\[ \t\]4\$%ebp\$, %eax" { target ia32 } } } */ diff --git a/gcc/testsuite/gcc.target/i386/pr66960-5.c b/gcc/testsuite/gcc.target/i [...] new file mode 100644 index 0000000..d449437 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/pr66960-5.c @@ -0,0 +1,21 @@ +/* { dg-do link } */ +/* { dg-options "-O" } */ + +extern void link_error (void); + +__attribute__ ((noinline, noclone)) +void +foo (void) +{ + void **p = __builtin_stack_top (); + void *ra = __builtin_return_address (0); + if (*p != ra) + link_error (); +} + +int +main (void) +{ + foo (); + return 0; +} -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[gcc] branch aoliva/pr64164 updated (3004886 -> 2c8077e)

by git

This is an automated email from the git hooks/post-receive script. unknown user pushed a change to branch aoliva/pr64164 in repository gcc. discards 3004886 incremental fixes new 2c8077e incremental fixes This update added new revisions after undoing existing revisions. That is to say, some revisions that were in the old version of the branch are not in the new version. This situation occurs when a user --force pushes a change and generates a repository containing something like this: * -- * -- B -- O -- O -- O (3004886) \ N -- N -- N refs/heads/aoliva/pr64164 (2c8077e) You should already have received notification emails for all of the O revisions, and so the following emails describe only the N revisions from the common base, B. Any revisions marked "omits" are not gone; other references still refer to them. Any revisions marked "discards" are gone forever. The 1 revisions listed above as "new" are entirely new to this repository and will be described in separate emails. The revisions listed as "adds" were already present in the repository and have only been added to this reference. Summary of changes: gcc/function.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[gcc] 01/01: incremental fixes

by git

This is an automated email from the git hooks/post-receive script. unknown user pushed a commit to branch aoliva/pr64164 in repository gcc. commit 2c8077ed870fce546c29f88db71141ec3bbd2efa Author: Alexandre Oliva <aoliva(a)redhat.com> Date: Wed Jul 29 14:34:34 2015 -0300 incremental fixes PR bootstrap/66978 * function.c (expand_function_start): Convert static chain to Pmode if needed. From H.J. Lu <hongjiu.lu(a)intel.com>. PR middle-end/66983 PR middle-end/67035 * cfgexpand.c (align_local_variable, add_stack_var): Support anonymous SSA names. (defer_stack_allocation): Likewise. Declare earlier. (expand_one_ssa_partition): Record alignment before expanding stack vars. Support deferred allocation. (set_rtl): Do no record deferred-allocation marker in SA.partition_to_pseudo. (expand_stack_vars): Adjust check for the marker in it. (adjust_one_expanded_partition_var): Skip deferred-alloc vars. PR middle-end/67034 * function.c (assign_parm_setup_reg): Copy BLKmode byref args to RTL assigned to their default partition. PR rtl-optimization/67000 * expr.c (read_complex_part): Export. * expr.h (read_complex_part): Declare. * function.c (split_complex_args): Use it. Reset complex parm before fetching its default decl rtl. (assign_parms_unsplit_complex): Use the preexisting complex RTL if it matches the components. (assign_parm_setup_reg): Drop assert on from_expand mode. Adjust it to the promoted_mode, if not byref. --- gcc/cfgexpand.c | 77 +++++++++++++++++++++++++++++++++++++++++---------------- gcc/expr.c | 2 +- gcc/expr.h | 1 + gcc/function.c | 65 ++++++++++++++++++++++++++++++++++++------------ 4 files changed, 107 insertions(+), 38 deletions(-) diff --git a/gcc/cfgexpand.c b/gcc/cfgexpand.c index 0b19953..9577ad6d 100644 --- a/gcc/cfgexpand.c +++ b/gcc/cfgexpand.c @@ -97,6 +97,8 @@ gimple currently_expanding_gimple_stmt; static rtx expand_debug_expr (tree); +static bool defer_stack_allocation (tree, bool); + /* Return an expression tree corresponding to the RHS of GIMPLE statement STMT. */ @@ -253,7 +255,7 @@ set_rtl (tree t, rtx x) { if (SA.partition_to_pseudo[part]) gcc_assert (SA.partition_to_pseudo[part] == x); - else + else if (x != pc_rtx) SA.partition_to_pseudo[part] = x; } /* For the benefit of debug information at -O0 (where @@ -348,8 +350,15 @@ static bool has_short_buffer; static unsigned int align_local_variable (tree decl) { - unsigned int align = LOCAL_DECL_ALIGNMENT (decl); - DECL_ALIGN (decl) = align; + unsigned int align; + + if (TREE_CODE (decl) == SSA_NAME) + align = TYPE_ALIGN (TREE_TYPE (decl)); + else + { + align = LOCAL_DECL_ALIGNMENT (decl); + DECL_ALIGN (decl) = align; + } return align / BITS_PER_UNIT; } @@ -415,12 +424,15 @@ add_stack_var (tree decl) decl_to_stack_part->put (decl, stack_vars_num); v->decl = decl; - v->size = tree_to_uhwi (DECL_SIZE_UNIT (SSAVAR (decl))); + tree size = TREE_CODE (decl) == SSA_NAME + ? TYPE_SIZE_UNIT (TREE_TYPE (decl)) + : DECL_SIZE_UNIT (decl); + v->size = tree_to_uhwi (size); /* Ensure that all variables have size, so that &a != &b for any two variables that are simultaneously live. */ if (v->size == 0) v->size = 1; - v->alignb = align_local_variable (SSAVAR (decl)); + v->alignb = align_local_variable (decl); /* An alignment of zero can mightily confuse us later. */ gcc_assert (v->alignb != 0); @@ -1051,9 +1063,9 @@ expand_stack_vars (bool (*pred) (size_t), struct stack_vars_d [...] /* Skip variables that have already had rtl assigned. See also add_stack_var where we perpetrate this pc_rtx hack. */ decl = stack_vars[i].decl; - if ((TREE_CODE (decl) == SSA_NAME - ? SA.partition_to_pseudo[var_to_partition (SA.map, decl)] - : DECL_RTL (decl)) != pc_rtx) + if (TREE_CODE (decl) == SSA_NAME + ? SA.partition_to_pseudo[var_to_partition (SA.map, decl)] != NULL_RTX + : DECL_RTL (decl) != pc_rtx) continue; large_size += alignb - 1; @@ -1082,9 +1094,9 @@ expand_stack_vars (bool (*pred) (size_t), struct stack_vars_d [...] /* Skip variables that have already had rtl assigned. See also add_stack_var where we perpetrate this pc_rtx hack. */ decl = stack_vars[i].decl; - if ((TREE_CODE (decl) == SSA_NAME - ? SA.partition_to_pseudo[var_to_partition (SA.map, decl)] - : DECL_RTL (decl)) != pc_rtx) + if (TREE_CODE (decl) == SSA_NAME + ? SA.partition_to_pseudo[var_to_partition (SA.map, decl)] != NULL_RTX + : DECL_RTL (decl) != pc_rtx) continue; /* Check the predicate to see whether this variable should be @@ -1290,12 +1302,6 @@ expand_one_ssa_partition (tree var) if (SA.partition_to_pseudo[part]) return; - if (!use_register_for_decl (var)) - { - expand_one_stack_var_1 (var); - return; - } - unsigned int align = MINIMUM_ALIGNMENT (TREE_TYPE (var), TYPE_MODE (TREE_TYPE (var)), TYPE_ALIGN (TREE_TYPE (var))); @@ -1307,6 +1313,15 @@ expand_one_ssa_partition (tree var) record_alignment_for_reg_var (align); + if (!use_register_for_decl (var)) + { + if (defer_stack_allocation (var, true)) + add_stack_var (var); + else + expand_one_stack_var_1 (var); + return; + } + machine_mode reg_mode = promote_ssa_mode (var, NULL); rtx x = gen_reg_rtx (reg_mode); @@ -1331,6 +1346,13 @@ adjust_one_expanded_partition_var (tree var) rtx x = SA.partition_to_pseudo[part]; + if (!x) + { + /* This var will get a stack slot later. */ + gcc_assert (defer_stack_allocation (var, true)); + return; + } + set_rtl (var, x); if (!REG_P (x)) @@ -1409,10 +1431,14 @@ expand_one_error_var (tree var) static bool defer_stack_allocation (tree var, bool toplevel) { + tree size_unit = TREE_CODE (var) == SSA_NAME + ? TYPE_SIZE_UNIT (TREE_TYPE (var)) + : DECL_SIZE_UNIT (var); + /* Whether the variable is small enough for immediate allocation not to be a problem with regard to the frame size. */ bool smallish - = ((HOST_WIDE_INT) tree_to_uhwi (DECL_SIZE_UNIT (var)) + = ((HOST_WIDE_INT) tree_to_uhwi (size_unit) < PARAM_VALUE (PARAM_MIN_SIZE_FOR_STACK_SHARING)); /* If stack protection is enabled, *all* stack variables must be deferred, @@ -1421,16 +1447,24 @@ defer_stack_allocation (tree var, bool toplevel) if (flag_stack_protect || ((flag_sanitize & SANITIZE_ADDRESS) && ASAN_STACK)) return true; + unsigned int align = TREE_CODE (var) == SSA_NAME + ? TYPE_ALIGN (TREE_TYPE (var)) + : DECL_ALIGN (var); + /* We handle "large" alignment via dynamic allocation. We want to handle this extra complication in only one place, so defer them. */ - if (DECL_ALIGN (var) > MAX_SUPPORTED_STACK_ALIGNMENT) + if (align > MAX_SUPPORTED_STACK_ALIGNMENT) return true; + bool ignored = TREE_CODE (var) == SSA_NAME + ? !SSAVAR (var) || DECL_IGNORED_P (SSA_NAME_VAR (var)) + : DECL_IGNORED_P (var); + /* When optimization is enabled, DECL_IGNORED_P variables originally scoped might be detached from their block and appear at toplevel when we reach here. We want to coalesce them with variables from other blocks when the immediate contribution to the frame size would be noticeable. */ - if (toplevel && optimize > 0 && DECL_IGNORED_P (var) && !smallish) + if (toplevel && optimize > 0 && ignored && !smallish) return true; /* Variables declared in the outermost scope automatically conflict @@ -6135,7 +6169,8 @@ pass_expand::execute (function *fun) if (part == NO_PARTITION) continue; - gcc_assert (SA.partition_to_pseudo[part]); + gcc_assert (SA.partition_to_pseudo[part] + || defer_stack_allocation (name, true)); /* If this decl was marked as living in multiple places, reset this now to NULL. */ diff --git a/gcc/expr.c b/gcc/expr.c index d601129..fc49f92 100644 --- a/gcc/expr.c +++ b/gcc/expr.c @@ -3022,7 +3022,7 @@ write_complex_part (rtx cplx, rtx val, bool imag_p) /* Extract one of the components of the complex value CPLX. Extract the real part if IMAG_P is false, and the imaginary part if it's true. */ -static rtx +rtx read_complex_part (rtx cplx, bool imag_p) { machine_mode cmode, imode; diff --git a/gcc/expr.h b/gcc/expr.h index 32d1707..a2c8e1d 100644 --- a/gcc/expr.h +++ b/gcc/expr.h @@ -210,6 +210,7 @@ extern rtx_insn *emit_move_insn_1 (rtx, rtx); extern rtx_insn *emit_move_complex_push (machine_mode, rtx, rtx); extern rtx_insn *emit_move_complex_parts (rtx, rtx); +extern rtx read_complex_part (rtx, bool); extern void write_complex_part (rtx, rtx, bool); extern rtx emit_move_resolve_push (machine_mode, rtx); diff --git a/gcc/function.c b/gcc/function.c index c3d00cd..390b81e 100644 --- a/gcc/function.c +++ b/gcc/function.c @@ -152,6 +152,7 @@ static void prepare_function_start (void); static void do_clobber_return_reg (rtx, void *); static void do_use_return_reg (rtx, void *); static rtx rtl_for_parm (struct assign_parm_data_all *, tree); +static void maybe_reset_rtl_for_parm (tree); /* Stack of nested functions. */ @@ -2321,12 +2322,12 @@ split_complex_args (struct assign_parm_data_all *all, vec<t [...] from resetting their RTL. */ if (currently_expanding_to_rtl) { + maybe_reset_rtl_for_parm (cparm); rtx rtl = rtl_for_parm (all, cparm); - gcc_assert (!rtl || GET_CODE (rtl) == CONCAT); if (rtl) { - SET_DECL_RTL (p, XEXP (rtl, 0)); - SET_DECL_RTL (decl, XEXP (rtl, 1)); + SET_DECL_RTL (p, read_complex_part (rtl, false)); + SET_DECL_RTL (decl, read_complex_part (rtl, true)); DECL_CONTEXT (p) = cparm; DECL_CONTEXT (decl) = cparm; @@ -3104,10 +3105,11 @@ assign_parm_setup_reg (struct assign_parm_data_all *all, tr [...] rtx from_expand = rtl_for_parm (all, parm); - if (from_expand && !data->passed_pointer) + if (from_expand) { parmreg = from_expand; - gcc_assert (GET_MODE (parmreg) == promoted_nominal_mode); + if (!data->passed_pointer && GET_MODE (parmreg) != promoted_nominal_mode) + parmreg = gen_lowpart (promoted_nominal_mode, parmreg); } else { @@ -3118,7 +3120,7 @@ assign_parm_setup_reg (struct assign_parm_data_all *all, tree parm, /* If this was an item that we received a pointer to, set DECL_RTL appropriately. */ - if (data->passed_pointer) + if (!from_expand && data->passed_pointer) { rtx x = gen_rtx_MEM (TYPE_MODE (TREE_TYPE (data->passed_type)), parmreg); set_mem_attributes (x, parm, 1); @@ -3139,7 +3141,8 @@ assign_parm_setup_reg (struct assign_parm_data_all *all, tree parm, need_conversion = (data->nominal_mode != data->passed_mode || promoted_nominal_mode != data->promoted_mode); - moved = false; + gcc_assert (!(need_conversion && data->passed_pointer)); + moved = from_expand && data->passed_pointer; if (need_conversion && GET_MODE_CLASS (data->nominal_mode) == MODE_INT @@ -3278,12 +3281,15 @@ assign_parm_setup_reg (struct assign_parm_data_all *all, tr [...] if (data->passed_pointer && (from_expand || TYPE_MODE (TREE_TYPE (parm)) != BLKmode)) { + rtx src = DECL_RTL (parm); + /* We can't use nominal_mode, because it will have been set to Pmode above. We must use the actual mode of the parm. */ if (from_expand) { - parmreg = from_expand; gcc_assert (GET_MODE (parmreg) == TYPE_MODE (TREE_TYPE (parm))); + src = gen_rtx_MEM (GET_MODE (parmreg), validated_mem); + set_mem_attributes (src, parm, 1); } else if (use_register_for_decl (parm)) { @@ -3302,14 +3308,14 @@ assign_parm_setup_reg (struct assign_parm_data_all *all, tr [...] set_mem_attributes (parmreg, parm, 1); } - if (GET_MODE (parmreg) != GET_MODE (DECL_RTL (parm))) + if (GET_MODE (parmreg) != GET_MODE (src)) { - rtx tempreg = gen_reg_rtx (GET_MODE (DECL_RTL (parm))); + rtx tempreg = gen_reg_rtx (GET_MODE (src)); int unsigned_p = TYPE_UNSIGNED (TREE_TYPE (parm)); push_to_sequence2 (all->first_conversion_insn, all->last_conversion_insn); - emit_move_insn (tempreg, DECL_RTL (parm)); + emit_move_insn (tempreg, src); tempreg = convert_to_mode (GET_MODE (parmreg), tempreg, unsigned_p); emit_move_insn (parmreg, tempreg); all->first_conversion_insn = get_insns (); @@ -3318,8 +3324,21 @@ assign_parm_setup_reg (struct assign_parm_data_all *all, tree parm, did_conversion = true; } + else if (GET_MODE (parmreg) == BLKmode) + { + push_to_sequence2 (all->first_conversion_insn, + all->last_conversion_insn); + gcc_assert (TREE_CODE (data->passed_type) == POINTER_TYPE); + gcc_assert (TREE_TYPE (data->passed_type) == TREE_TYPE (parm)); + emit_block_move (parmreg, src, + GEN_INT (int_size_in_bytes (TREE_TYPE (parm))), + BLOCK_OP_NORMAL);; + all->first_conversion_insn = get_insns (); + all->last_conversion_insn = get_last_insn (); + end_sequence (); + } else - emit_move_insn (parmreg, DECL_RTL (parm)); + emit_move_insn (parmreg, src); SET_DECL_RTL (parm, parmreg); @@ -3495,11 +3514,21 @@ assign_parms_unsplit_complex (struct assign_parm_data_all *all, imag = DECL_RTL (fnargs[i + 1]); if (inner != GET_MODE (real)) { - real = gen_lowpart_SUBREG (inner, real); - imag = gen_lowpart_SUBREG (inner, imag); + real = simplify_gen_subreg (inner, real, GET_MODE (real), + subreg_lowpart_offset + (inner, GET_MODE (real))); + imag = simplify_gen_subreg (inner, imag, GET_MODE (imag), + subreg_lowpart_offset + (inner, GET_MODE (imag))); } - if (TREE_ADDRESSABLE (parm)) + if ((tmp = DECL_RTL_IF_SET (parm)) + && rtx_equal_p (real, + read_complex_part (tmp, false)) + && rtx_equal_p (imag, + read_complex_part (tmp, true))) + ; /* We have the right rtl in parm and tmp already. */ + else if (TREE_ADDRESSABLE (parm)) { rtx rmem, imem; HOST_WIDE_INT size = int_size_in_bytes (TREE_TYPE (parm)); @@ -3645,7 +3674,7 @@ assign_bounds (vec<bounds_parm_data> &bndargs, assign_parm_setup_block (&all, pbdata->bounds_parm, &pbdata->parm_data); else if (pbdata->parm_data.passed_pointer - || use_register_for_decl (pbdata->bounds_parm)) + || use_register_for_parm_decl (&all, pbdata->bounds_parm)) assign_parm_setup_reg (&all, pbdata->bounds_parm, &pbdata->parm_data); else @@ -5207,6 +5236,10 @@ expand_function_start (tree subr) SET_DECL_RTL (parm, local); mark_reg_pointer (local, TYPE_ALIGN (TREE_TYPE (TREE_TYPE (parm)))); + if (GET_MODE (local) != Pmode) + local = convert_to_mode (Pmode, local, + TYPE_UNSIGNED (TREE_TYPE (parm))); + insn = emit_move_insn (local, chain); /* Mark the register as eliminable, similar to parameters. */ -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[toolchain/abe] branch bernie/benchmarking updated: Use sleep rather than inotify to hold target session

by git

This is an automated email from the git hooks/post-receive script. bernie.ogden pushed a commit to branch bernie/benchmarking in repository toolchain/abe. The following commit(s) were added to refs/heads/bernie/benchmarking by this push: new 76c9b33 Use sleep rather than inotify to hold target session 76c9b33 is described below commit 76c9b33146667b881e2f6d95caeee241717bf664 Author: Bernard Ogden <bernie.ogden(a)linaro.org> Date: Thu Jul 30 19:43:10 2015 +0200 Use sleep rather than inotify to hold target session inotify is great for this, but inotify-tools is not available for jessie arm64. Change-Id: Iabc3ce5600acb36efb0d5caa23aee2849ced7c9c --- config/bench/lava/host-session | 2 +- config/bench/lava/target-session | 5 +---- config/bench/lava/target-session-kvm.yaml | 1 - config/bench/lava/target-session.yaml | 1 - 4 files changed, 2 insertions(+), 7 deletions(-) diff --git a/config/bench/lava/host-session b/config/bench/lava/host-session index cff154d..ad99b6e 100755 --- a/config/bench/lava/host-session +++ b/config/bench/lava/host-session @@ -41,7 +41,7 @@ fi #Non-parameters that we need to pass to Benchmark.job export ABE_DIR export post_build_cmd="rm ${HOME}/data/socket" -export post_target_cmd='touch ${HOME}/messages/release' +export post_target_cmd='kill `pidof sleep`' export maindir="/${HOME}/bench" mkdir "${maindir}" || exit 1 chmod 700 "${maindir}" || exit 1 diff --git a/config/bench/lava/target-session b/config/bench/lava/target-session index 899ce6d..f042d82 100755 --- a/config/bench/lava/target-session +++ b/config/bench/lava/target-session @@ -1,6 +1,3 @@ #!/bin/bash lava-network broadcast eth0 -mkdir "${HOME}/messages" -while ! test -e "${HOME}/messages/release"; do - inotifywait -e create "${HOME}/messages" -done +sleep infinity diff --git a/config/bench/lava/target-session-kvm.yaml b/config/bench/lava/target-s [...] index fd5d593..0e88369 100644 --- a/config/bench/lava/target-session-kvm.yaml +++ b/config/bench/lava/target-session-kvm.yaml @@ -15,7 +15,6 @@ install: - binutils - gawk - libgfortran3 - - inotify-tools run: steps: diff --git a/config/bench/lava/target-session.yaml b/config/bench/lava/target-session.yaml index 1185555..04215d8 100644 --- a/config/bench/lava/target-session.yaml +++ b/config/bench/lava/target-session.yaml @@ -12,7 +12,6 @@ install: - openssh-server - make - cpufrequtils - - inotify-tools run: steps: -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

[toolchain/abe] branch bernie/benchmarking updated: Stop creating a dir to store src

by git

9 years, 3 months

[toolchain/abe] branch bernie/benchmarking-agents deleted (was 123a21f)

by git

This is an automated email from the git hooks/post-receive script. bernie.ogden pushed a change to branch bernie/benchmarking-agents in repository toolchain/abe. was 123a21f Make Benchmark.job delete the socket after build phase This change permanently discards the following revisions: discards 123a21f Make Benchmark.job delete the socket after build phase discards 315ecde Split Benchmark.job into build and run phases, accept arbitra [...] -- To stop receiving notification emails like this one, please contact the administrator of this repository.

9 years, 3 months

← Newer
1
...
8
9
10
11
12
13
14
...
136
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

Tcwg-commits July 2015