[PATCH] Fix for mklog

Hi all, Current mklog in some cases do not checks whether function have real changes or not. I fixed it by removing $doubtfunc and adding protection against EOF. Example: diff --git a/gcc/cfg.c b/gcc/cfg.c index 6070d9e..cb3dfd9 100644 --- a/gcc/cfg.c +++ b/gcc/cfg.c @@ -555,10 +555,10 @@ al

Re: [PATCH, ifcvt] Allow CC mode if HAVE_cbranchcc4 (fix s390 build)

On 11/06/2014 08:44 AM, Zhenqiang Chen wrote: > Hi, > > The patch add runtime check to fix s390 build fail > (https://gcc.gnu.org/ml/gcc-patches/2014-11/msg00050.html). > > And there is additional code to workaround s390 cstorecc4 issue. > > Bootstrap and no make check regression on X86-64. > Bu

Re: [PATCH, PR63742][ARM] Fix arm *movhi_insn_arch4 pattern for big-endian

> > The idea is simple: Use movw for certain const source operand instead > > of > ldrh. And exclude the const values which cannot be handled by > mov/mvn/movw. > > I am doing regression test for this patch. Assuming no issue pops up, > OK for trunk? > > So, doesn't that makes the bug

RE: [Ping] [PATCH, 9/10] aarch64: generate conditional compare instructions

> -Original Message- > From: Richard Henderson [mailto:r...@redhat.com] > Sent: Wednesday, November 05, 2014 9:42 PM > To: Zhenqiang Chen > Cc: gcc-patches@gcc.gnu.org > Subject: Re: [Ping] [PATCH, 9/10] aarch64: generate conditional compare > instructions > > On 11/05/2014 10:05 AM, Zhe

[PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

This merges patterns implementing the bitwise patterns from tree-ssa-forwprop.c. I've removed duplicate functionality from fold-const.c as I found them, some may be still lurking in the depths. This also fixes a bug in genmatch which made user-defined predicates matching anything, thus (match

Re: [PATCH 3/5] IPA ICF pass

> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63747 is likely caused by > this patch. compare_gimple_switch does not check CASE_LOW and > CASE_HIGH, resulting merging functions not identical. > > Interestingly in the first a few versions of this patch CASE_LOW/HIGH > were checked. But last versio

[PATCH, i386, MPX] Add target filter for MPX tests

Hi, This patch adds target filter for recently added MPX tests. I put check function into a new file because I'm going to also add there some MPX runtime tests related functions and use in it another exp file. Since there is no change in compiler I just checked I get no fails for modified tes

Re: [PATCH] Fix for mklog

On 11/06/2014 11:00 AM, Marat Zakirov wrote: - if ($doubtfunc) { - $idx = $line_idx; + $idx = $line_idx; # Skip line info in context diffs. - while ($is_context_diff && $diff_lines[$idx + 1] =~ /^[-\*]{3} [0-9]/) { - ++$idx; -

Re: [gomp4] Move libgomp plugins into subdirectory

Hi Julian! On Wed, 5 Nov 2014 17:57:10 +, Julian Brown wrote: > This patch moves plugin-nvptx.c and plugin-host.c (from oacc-host.c) > into a new "plugin" subdirectory, as requested by Jakub, and to match > more closely the layout of the Intel MIC pieces. This also moves the > autotools bits

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

> On Nov 6, 2014, at 12:55 AM, Richard Biener wrote: > > > This merges patterns implementing the bitwise patterns from > tree-ssa-forwprop.c. I've removed duplicate functionality from > fold-const.c as I found them, some may be still lurking in the > depths. > > This also fixes a bug in g

Re: [Ping] [PATCH, 9/10] aarch64: generate conditional compare instructions

> + /* Generate insn to match ccmp_and/ccmp_ior. */ > + target = gen_rtx_REG (mode, CC_REGNUM); > + emit_insn (gen_rtx_SET (VOIDmode, target, > + gen_rtx_fmt_ee (COMPARE, VOIDmode, > + bit_op, const0_rtx))); Invalid mode for the

Re: [AArch64] [BE] Fix vector load/stores to not use ld1/st1

> On 10 October 2014 16:19, Alan Hayward wrote: > > This patch is dependant on "[AArch64] [BE] [1/2] Make large opaque > > integer modes endianness-safe.” > > > > It fixes up movoi/ci/xi for Big Endian, so that we end up with the lsb > > of a big-endian integer to be in the low byte of the highest

Re: [PATCH] Fix for mklog

On 11/06/2014 12:07 PM, Yury Gribov wrote: On 11/06/2014 11:00 AM, Marat Zakirov wrote: -if ($doubtfunc) { -$idx = $line_idx; +$idx = $line_idx; # Skip line info in context diffs. -while ($is_context_diff && $diff_lines[$idx + 1] =~ /^[-\*]{3} [0-9]/) { -

Re: [AArch64] [BE] Fix vector load/stores to not use ld1/st1

On 06/11/2014 09:09, "Yangfei (Felix)" wrote: >> On 10 October 2014 16:19, Alan Hayward wrote: >> > This patch is dependant on "[AArch64] [BE] [1/2] Make large opaque >> > integer modes endianness-safe.” >> > >> > It fixes up movoi/ci/xi for Big Endian, so that we end up with the lsb >> > of a

Re: [PATCH, i386, MPX] Add target filter for MPX tests

Hello! > 2014-11-06 Ilya Enkovich > > * lib/mpx-dg.exp: New. > * gcc.target/i386/i386.exp: Load mpx-dg.exp. > * gcc.target/i386/chkp-builtins-1.c: Require mpx target. > * gcc.target/i386/chkp-builtins-2.c: Likewise. > * gcc.target/i386/chkp-builtins-3.c: Likewise. > * gcc.target/i386/chkp-built

Re: [AArch64] [BE] Fix vector load/stores to not use ld1/st1

> > > > > >Hello, > > > > It seems that this patch breaks the compile of some testcases under > >big-endian. > > On example: testsuite/gcc.target/aarch64/advsimd-intrinsics/ vldX.c > > Any thing I missed? Please confirm. Thanks. > > > >$ aarch64_be-linux-gnu-gcc vldX.c > > > >vldX.c: In function

Re: [gomp4] Move libgomp plugins into subdirectory

On Thu, 6 Nov 2014 10:06:00 +0100 Thomas Schwinge wrote: > Hi Julian! > > On Wed, 5 Nov 2014 17:57:10 +, Julian Brown > wrote: > > This patch moves plugin-nvptx.c and plugin-host.c (from oacc-host.c) > > into a new "plugin" subdirectory, as requested by Jakub, and to > > match more closely

Re: [PATCH] Limit removal of UBSAN_NULL checks

On Wed, Nov 05, 2014 at 11:08:27PM +0100, Jakub Jelinek wrote: > I wonder if for the case where you don't remove, gimple_bb (g) == gimple_bb > (stmt) and tree_int_cst_compare (cur_align, align) == 0 it wouldn't be > worthwhile to v.pop (); so that the vector doesn't grow up unnecessarily too Good

Re: [gomp4] acc reductions with multiple variables

Hi! On Fri, 31 Oct 2014 16:35:27 -0700, Cesar Philippidis wrote: > This patch does the following: [...] > gcc/ > * omp-low.c (oacc_max_threads): New function. > (lower_reduction_clauses): Use the GOACC thread builtin functions to > determine the thread count. Handle mul

[PATCH, testsuite, obvious] Do not produce optimized dump in devirt-4[45].C

Hi, g++.dg/ipa/devirt-44.C and g++.dg/ipa/devirt-45.C leave behind an optimized dump which they don't even use. Fixed thusly, tested by runing make -k check RUNTESTFLAGS="dg.exp=ipa/*.C", will commit shortly as obvious. Thanks, Martin 2014-11-06 Martin Jambor * g++.dg/ipa/devirt-4

Re: [PATCH] Limit removal of UBSAN_NULL checks

On Thu, Nov 06, 2014 at 10:56:17AM +0100, Marek Polacek wrote: > --- gcc/sanopt.c > +++ gcc/sanopt.c > @@ -130,7 +130,30 @@ sanopt_optimize_walker (basic_block bb, struct > sanopt_ctx *ctx) > /* At this point we shouldn't have any statements >tha

Re: [gomp4] Move libgomp plugins into subdirectory

On Thu, Nov 06, 2014 at 10:06:00AM +0100, Thomas Schwinge wrote: > Hi Julian! > > On Wed, 5 Nov 2014 17:57:10 +, Julian Brown > wrote: > > This patch moves plugin-nvptx.c and plugin-host.c (from oacc-host.c) > > into a new "plugin" subdirectory, as requested by Jakub, and to match > > more c

Re: [AArch64] [BE] Fix vector load/stores to not use ld1/st1

On 6 November 2014 10:28, Alan Hayward wrote: > > > On 06/11/2014 09:09, "Yangfei (Felix)" wrote: > >>> On 10 October 2014 16:19, Alan Hayward wrote: >>> > This patch is dependant on "[AArch64] [BE] [1/2] Make large opaque >>> > integer modes endianness-safe.” >>> > >>> > It fixes up movoi/ci/xi

[PATCH][AArch64] Add bounds checking to vqdm*_lane intrinsics via a qualifier that also flips endianness

This generates out-of-range errors at compile- (rather than assemble-)time for the vqdm*_lane intrinsics, and also provides a single place to do bigendian lane-swapping for all those intrinsics (and others to follow in later patches). This allows us to remove many define_expands that just do a r

[PATCHv5][PING^3] Vimrc config with GNU formatting

On 10/21/2014 07:24 PM, Yury Gribov wrote: On 10/13/2014 02:26 PM, Yury Gribov wrote: On 10/02/2014 09:14 PM, Yury Gribov wrote: On 09/17/2014 09:08 PM, Yury Gribov wrote: > On 09/16/2014 08:38 PM, Yury Gribov wrote: >> Hi all, >> >> This is the third version of the patch. A list of changes

Fix ICE with thunks taking argument passed by reference

Hi, PR63573 is about ICE when expanding thunk call for function taking as a parameter structure passed by reference. This structure in fact contains only one integer and thus it is promoted to register by argument setup in function.c (as an optimization). This is an sensible optimization, but when

Re: [PATCH, libfortran] PR 47007, 61847 Locale failures in libgfortran

On Wed, Nov 5, 2014 at 12:48 PM, Janne Blomqvist wrote: > Hi, > > the attached patch fixes a few locale related failures in libgfortran, > in the case where the POSIX 2008 extended locale functionality and > extensions strto{f,d,ld}_l are present. > > These failures typically occur when libgfortra

Re: [PATCH, testsuite, obvious] Do not produce optimized dump in devirt-4[45].C

Hello! > g++.dg/ipa/devirt-44.C and g++.dg/ipa/devirt-45.C leave behind an > optimized dump which they don't even use. Fixed thusly, tested by > runing make -k check RUNTESTFLAGS="dg.exp=ipa/*.C", will commit > shortly as obvious. > > 2014-11-06 Martin Jambor > > * g++.dg/ipa/devirt-44.C: Do n

Re: [PATCH] Limit removal of UBSAN_NULL checks

On Thu, Nov 06, 2014 at 11:10:37AM +0100, Jakub Jelinek wrote: > Sorry for being slow on this, but in the current algorithm you are using > the latest vector element from each bb only anyway, so the > && tree_int_cst_compare (cur_align, align) == 0 > is unnecessary, even if the condition isn't true

Re: Fix ICE with thunks taking argument passed by reference

On Thu, 6 Nov 2014, Jan Hubicka wrote: > Hi, > PR63573 is about ICE when expanding thunk call for function taking as a > parameter structure passed by reference. This structure in fact contains only > one integer and thus it is promoted to register by argument setup in > function.c > (as an opti

RE: [PATCH, C++] Fix PR63366: __complex not equivalent to __complex double in C++

> From: Jason Merrill [mailto:ja...@redhat.com] > Sent: Wednesday, October 29, 2014 4:43 PM > > Since this is a GNU extension, I think we want to be compatible with the > C front end, which has > > > else if (specs->complex_p) > > { > > specs->typespec_word = cts_double; >

Re: [libcc1, build] Enable libcc1 on Solaris

Hi Phil, > The configure change is fine. Also the include. (From a libcc > authorship point of view). I am not aware of the history of AF_UNIX > over AF_LOCAL so I have no comment on that. Please await permission the glibc manual states AF_UNIX This is a synonym for AF_LOCAL. Although A

Re: [libcc1, build] Enable libcc1 on Solaris

On 06/11/14 11:33, Rainer Orth wrote: > Hi Phil, > >> The configure change is fine. Also the include. (From a libcc >> authorship point of view). I am not aware of the history of AF_UNIX >> over AF_LOCAL so I have no comment on that. Please await permission > > the glibc manual states > > AF_UN

[PATCH, Pointer Bounds Checker, Builtins instrumentation 1/5] Builtin codes and decls

Hi, This is the first patch in a series to enable builtin function calls instrumentation by Poitner Bounds Checker. Previously builtins instrumentation was disabled in checker because it allowed two function calls with the same function code have different set of arguments. Here I try another

[Patchv3] Control SRA and IPA-SRA by a param rather than MOVE_RATIO

On Fri, Oct 31, 2014 at 10:46:12AM +, Richard Biener wrote: > On Wed, Oct 29, 2014 at 3:39 PM, James Greenhalgh wrote: > >> I suppose I could port any target with a definition of MOVE_RATIO to > >> override the default parameter value in their option overriding code, > >> but that makes this a

Re: [gofrontend-dev] [PATCH 4/4] Gccgo port to s390[x] -- part II

On Tue, Nov 04, 2014 at 08:16:51PM -0800, Ian Taylor wrote: > The way to do it is not by > copying the test. If the test needs to be customized, add additional > files that use // +build lines to pick which files is built. Move > them into a directory, like method4.go or other tests that use > "r

[PATCH, Pointer Bounds Checker, Builtins instrumentation 2/5] Instrument builtin calls

Hi, This patch enables instrumentation of chosen builtin calls. Thanks, Ilya -- 2014-11-06 Ilya Enkovich * ipa-chkp.c (chkp_versioning): Clone builtin functions. (chkp_instrument_normal_builtin): New. (chkp_add_bounds_to_call_stmt): Support builtin functions. (

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

On Thu, 6 Nov 2014, Richard Biener wrote: +/* Simplify (A & B) OP0 (C & B) to (A OP0 C) & B. */ +/* (x | CST1) & CST2 -> (x & CST2) | (CST1 & CST2) */ As soon as there are 2+ operations in the result, this comment in fwprop_ssa_val becomes outdated, no? /* We continue matching along SS

[PATCH, Pointer Bounds Checker, Builtins instrumentation 3/5] Expand instrumented builtin calls

Hi, This patch adds support of instrumented builtin calls in expand. Calls are mostly expanded as calls. But some of them reuse existing string function calls expand functions (memcpy expand was slightly refactored for that). This is the last enabling patch in this series. Remaining two patc

Re: [PATCH] Correctly check dg-require-effective-target in avx512 tests.

On 05 Nov 17:17, Uros Bizjak wrote: > On Wed, Nov 5, 2014 at 5:14 PM, Ilya Tocar wrote: > > Hi, > > > > Currently we only check for dg-require-effective-target avx512vl in > > avx512vl tests. We should also check for avx512dq/avx512bw. > > Patch bwllow does this. > > Ok for trunk? > > > > OK, if

Re: [PING^3][PATCH] Warn about unclosed pragma omp declare target.

Ping. On 30 Oct 18:31, Ilya Tocar wrote: > Ping. > On 20 Oct 19:26, Ilya Tocar wrote: > > Ping. > > > > On 02 Oct 17:38, Ilya Tocar wrote: > > > Ping. > > > On 15 Aug 16:26, Ilya Tocar wrote: > > > > Ping. > > > > > > > > On 29 Jul 18:45, Ilya Tocar wrote: > > > > > Hi, > > > > > > > > > > As di

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

On Thu, 6 Nov 2014, Richard Biener wrote: +/* Try simple folding for X op !X, and X op X with the help + of the truth_valued_p and logical_inverted_value predicates. */ +(match truth_valued_p + @0 + (if (INTEGRAL_TYPE_P (type) && TYPE_PRECISION (type) == 1))) +(for op (lt le eq ne ge gt truth

[PATCH, Pointer Bounds Checker, Builtins instrumentation 4/5] Use alternative instrumented string function calls

Hi, This patch adds instrumentation optimization to use string function with no checks and/or bounds copy. Thanks, Ilya -- gcc/ 2014-11-06 Ilya Enkovich * tree-chkp-opt.c (chkp_get_nobnd_fndecl): New. (chkp_get_nochk_fndecl): New. (chkp_optimize_string_function_calls

Re: [PATCH] PR36312

It looks like -soname, and libgnat-5.0.so both end up in the infiles array. I'm not sure if that's supposed to happen or not. On Wed, Nov 5, 2014 at 9:57 PM, Eric Botcazou wrote: >> 2014-10-25 Anthony Brandon >> >> PR driver/36312 >> * diagnostic-core.h: Add prototype for fatal_error.

[PATCH, Pointer Bounds Checker, Builtins instrumentation 5/5] Support instrumented calls in strlen pass

Hi, This patch adds support of instrumented function calls into strlen pass. Whole series pass bootstrap and check on linux-x86_64. OK for trunk? Thanks, Ilya -- gcc/ 2014-11-06 Ilya Enkovich * tree-ssa-strlen.c: include ipa-chkp.h. (get_string_length): Handle calls with bo

Re: [PATCH 00/13] Go closures, libffi, and the static chain

On Thu, Nov 06, 2014 at 07:59:16AM +0100, Richard Henderson wrote: > I haven't done powerpc yet. If you'd like to help, I'd be delighted. I was going to say that it doesn't look too difficult, but then I noticed we have a problem. PowerPC uses r11 as the static chain, a register that is allowed

Re: [PATCH 4/n] OpenMP 4.0 offloading infrastructure: lto-wrapper

On 16 Oct 13:22, Jakub Jelinek wrote: > Can you please extract the configure{,.ac}, langhooks.c and varasm.c > bits into a separate patch? That is preapproved for trunk right now, that > isn't dependent on anything else. Done. > The rest LGTM, but please run it through LTO review (Richard/Honza)

Re: [PARCH 1/2, x86, PR63534] Fix darwin bootstrap

Now I see that equiv reload could be special for PIC register. Let's apply more conservative patch. Darwin bootstrap passed with the patch applied on r216304 (along with already committed to trunk patches from PR63618 and PR63620). 2014-11-06 Evgeny Stupachenko PR target/63534

Re: [PATCH 00/13] Go closures, libffi, and the static chain

On 11/06/2014 01:48 PM, Alan Modra wrote: > On Thu, Nov 06, 2014 at 07:59:16AM +0100, Richard Henderson wrote: >> I haven't done powerpc yet. If you'd like to help, I'd be delighted. > > I was going to say that it doesn't look too difficult, but then I > noticed we have a problem. PowerPC uses r

Re: Add uniform_inside_sphere_distribution

On 11/06/2014 02:23 AM, Marc Glisse wrote: On Wed, 5 Nov 2014, Ed Smith-Rowland wrote: On 11/05/2014 04:25 PM, Marc Glisse wrote: On Wed, 5 Nov 2014, Ed Smith-Rowland wrote: Like the uniform_on_sphere_distribution which is used inside, the 2-dimensional case uses rejection Could you point

Re: [PATCH 00/13] Go closures, libffi, and the static chain

Aren't there cases where the static chain register is needed? How does that work if it could be trashed on a plt call? On 11/06/2014 06:48 AM, Alan Modra wrote: On Thu, Nov 06, 2014 at 07:59:16AM +0100, Richard Henderson wrote: I haven't done powerpc yet. If you'd like to help, I'd be deligh

Re: FW: [AArch64] [BE] [1/2] Make large opaque integer modes endianness-safe.

On 10 October 2014 16:58, David Sherwood wrote: > Hi, > > I forgot to mention that this patch needs was tested in combination with > another > patch by Alan Hayward (coming soon) and observed the following: > > x86_64: No regressions. > aarch64: No regressions. > aarch64_be: Lots of vector tests

Re: [PATCH 00/13] Go closures, libffi, and the static chain

On 11/06/2014 02:10 PM, Lynn A. Boger wrote: > Aren't there cases where the static chain register is needed? How does that > work if it could be trashed on a plt call? No. At the moment the static chain is only used for nested functions, which are local to the translation unit, and are therefore

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

On Thu, 6 Nov 2014, Marc Glisse wrote: > On Thu, 6 Nov 2014, Richard Biener wrote: > > > +/* Simplify (A & B) OP0 (C & B) to (A OP0 C) & B. */ > > > +/* (x | CST1) & CST2 -> (x & CST2) | (CST1 & CST2) */ > > As soon as there are 2+ operations in the result, this comment in > fwprop_ssa_val beco

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

On Thu, 6 Nov 2014, Marc Glisse wrote: > On Thu, 6 Nov 2014, Richard Biener wrote: > > > +/* Try simple folding for X op !X, and X op X with the help > > + of the truth_valued_p and logical_inverted_value predicates. */ > > +(match truth_valued_p > > + @0 > > + (if (INTEGRAL_TYPE_P (type) && T

[patch] Fix ACATS c65003a

Hi, the compiler started to segfault on this ACATS test very recently, probably when the large Pointer Bounds Checker patch went in, although the issue had been latent. For the attached reduced testcase, the backtrace is: Starting program: /home/eric/build/gcc/native/stage1-gcc/gnat1 -quiet o

Re: [gofrontend-dev] [PATCH 1/4] Gccgo port to s390[x] -- part II

Ian Taylor writes: > On Tue, Nov 4, 2014 at 4:15 AM, Dominik Vogt wrote: >> See commit comment and ChangeLog for details. > > Committed patch 0001 with various formatting fixes, as attached. Unfortunately, the mksysinfo.sh part broke Solaris bootstrap: there's no type _ucred in gen-sysinfo.go,

Re: [gomp4] acc enter/exit data

Hi! On Thu, 30 Oct 2014 17:11:04 -0700, Cesar Philippidis wrote: > This patch add support for OpenACC's enter/exit data directive. > gcc/ > * gimple.h (enum gf_mask): Add GF_OMP_TARGET_KIND_OACC_ENTER_EXIT_DATA. Applied to gomp-4_0-branch in r217190: commit 81c45b54c802bd76efc9417

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

On Thu, 6 Nov 2014, Richard Biener wrote: On Thu, 6 Nov 2014, Marc Glisse wrote: On Thu, 6 Nov 2014, Richard Biener wrote: +/* Try simple folding for X op !X, and X op X with the help + of the truth_valued_p and logical_inverted_value predicates. */ +(match truth_valued_p + @0 + (if (INTE

Re: [Patchv3] Control SRA and IPA-SRA by a param rather than MOVE_RATIO

On Thu, Nov 6, 2014 at 12:53 PM, James Greenhalgh wrote: > > On Fri, Oct 31, 2014 at 10:46:12AM +, Richard Biener wrote: >> On Wed, Oct 29, 2014 at 3:39 PM, James Greenhalgh wrote: >> >> I suppose I could port any target with a definition of MOVE_RATIO to >> >> override the default parameter v

Re: [patch] Fix ACATS c65003a

On Thu, Nov 6, 2014 at 2:58 PM, Eric Botcazou wrote: > Hi, > > the compiler started to segfault on this ACATS test very recently, probably > when the large Pointer Bounds Checker patch went in, although the issue had > been latent. For the attached reduced testcase, the backtrace is: > > Starting

Re: [PATCH] OpenACC for C++ front end

Hi! On Wed, 5 Nov 2014 21:41:02 +, Joseph Myers wrote: > I think the TODO: XXX FIX -2 and TODO XXX: FIX -1 comments need, at least, > more explanation of what the issue is and where the constants come from, > even if something is left until later to be fixed. Such constants should go into

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

On Thu, 6 Nov 2014, Marc Glisse wrote: > On Thu, 6 Nov 2014, Richard Biener wrote: > > > On Thu, 6 Nov 2014, Marc Glisse wrote: > > > > > On Thu, 6 Nov 2014, Richard Biener wrote: > > > > > > > +/* Try simple folding for X op !X, and X op X with the help > > > > + of the truth_valued_p and lo

[gimple-classes, committed 06/44] Yet more gcall *

gcc/ChangeLog.gimple-classes: * tree-ssa-forwprop.c (simplify_builtin_call): Strengthen local "stmt2" from gimple to gcall * via a checked cast. Replace is_gimple_call with a dyn_cast, introducing local "call_stmt1", and using the latter in place of "stmt1" for type

[gimple-classes, committed 02/44] Make replace_call_with_call_and_fold work on a gcall *

gcc/ChangeLog.gimple-classes: * gimple-fold.c (replace_call_with_call_and_fold): Strengthen param "repl" from gimple to gcall *. (gimple_fold_builtin_strcpy): Likewise for local "repl". (gimple_fold_builtin_strncpy): Likewise. (gimple_fold_builtin_strcat): Li

[gimple-classes, committed 00/44] Yet more accessor typesafety

I've pushed the following 44 patches to the git branch "dmalcolm/gimple-classes". This is a work-in-progress towards making the rest of the gimple_assign_* accessors require a gassign * (or a const gassign *). Successfully bootstrapped®rtested the combination of the 44 patches upon the branch on

[gimple-classes, committed 03/44] gimple-low.c: Use gcall * in a few places

gcc/ChangeLog.gimple-classes: * gimple-low.c (lower_builtin_setjmp): Strengthen local "stmt" from gimple to gcall *, via a checked cast. (lower_builtin_posix_memalign): Likewise for local "call". Strengthen local "cond" from gimple to gcond *. Split local gim

[gimple-classes, committed 07/44] Add casting variants of various gimple_assign_*_p functions

Use them to provide various gimple_assign_ accessors with gassign * rather than plain gimple. gcc/ChangeLog.gimple-classes: * gimple.h (is_a_helper ::test): New. (gimple_assign_copy_p): Update return type from bool to gassign *, returning GS if the predicate holds.

[gimple-classes, committed 01/44] Use gassign * rather than gimple in various places

Use gassign * everywhere that a stmt is built from a gimple_build_assign_* call, to ease making the gimple_assign_ accessors typesafe (and to better document the code). gcc/ChangeLog.gimple-classes: * asan.c (build_shadow_mem_access): Strengthen local "g" from gimple to gassign *.

[gimple-classes, committed 12/44] gimple-fold.c: Use gassign

gcc/ChangeLog.gimple-classes: * gimple-fold.c (fold_gimple_assign): Strengthen local "stmt" from gimple to gassign * via a checked cast. (gsi_replace_with_seq_vops): Add a checked cast within region guarded by gimple_assign_single_p. (get_maxval_strlen): With

[gimple-classes, committed 08/44] tree-ssa-forwprop.c: Use gassign * in various places

gcc/ChangeLog.gimple-classes: * tree-ssa-forwprop.c (get_prop_source_stmt): Strengthen return type and local "def_stmt" from gimple to gassign *, using a dyn_cast in place of an is_gimple_assign. (can_propagate_from): Strengthen param from gimple to gassign *.

[gimple-classes, committed 05/44] More gcall *

gcc/ChangeLog.gimple-classes: * gimple.c (gimple_set_lhs): Add checked casts to gassign * and gcall * as appropriate. * tree-cfg.c (execute_fixup_cfg): Introduce local "call_stmt" via a dyn_cast and use it in place of "stmt" where appropriate * tree-cfgcleanu

[gimple-classes, committed 21/44] tree-eh.c: Use gassign

gcc/ChangeLog.gimple-classes: * tree-eh.c (stmt_could_throw_1_p): Add checked casts. (sink_clobbers): Add checked cast. --- gcc/ChangeLog.gimple-classes | 5 + gcc/tree-eh.c| 8 +--- 2 files changed, 10 insertions(+), 3 deletions(-) diff --git a/gcc/ChangeL

[gimple-classes, committed 10/44] gimplify-me.c: Use gassign

gcc/ChangeLog.gimple-classes: * gimplify-me.c (gimple_regimplify_operands): Replace a couple of is_gimple_assign calls with dyn_cast, introducing locals "assign_stmt" and using them in place of "stmt" for typesafety. --- gcc/ChangeLog.gimple-classes | 6 ++ gcc/gimplif

[gimple-classes, committed 04/44] Use gcall * in more places

gcc/ChangeLog.gimple-classes: * config/i386/i386.c (add_condition_to_bb): Strengthen local "call_cond_stmt" from gimple to gcall *. * gimple-fold.c (gimple_build): Likewise for local "stmt". * gimple-match-head.c (maybe_push_res_to_seq): Likewise for local "n

[gimple-classes, committed 25/44] tree-ssa-coalesce.c: Use gassign

gcc/ChangeLog.gimple-classes: * tree-ssa-coalesce.c (build_ssa_conflict_graph): Replace is_gimple_assign with dyn_cast, introducing local "assign_stmt", using it in place of "stmt" for typesafety. (create_outofssa_var_map): Within case GIMPLE_ASSIGN introduce

[gimple-classes, committed 24/44] tree-ssa-ccp.c: Use gassign

gcc/ChangeLog.gimple-classes: * tree-ssa-ccp.c (get_default_value): Replace is_gimple_assign with a dyn_cast, introducing local "assign_stmt", using it in place of "stmt" for typesafety. (likely_value): Add a checked cast. (evaluate_stmt): Replace checks agai

[gimple-classes, committed 26/44] tree-ssa-copy.c: Use gassign

gcc/ChangeLog.gimple-classes: * tree-ssa-copy.c (stmt_may_generate_copy): Replace check against GIMPLE_ASSIGN with a dyn_cast, introducing local gassign * "assign_stmt", using it in place of "stmt" for typesafety. (copy_prop_visit_assignment): Strengthen param "stmt"

[gimple-classes, committed 14/44] gimple-match-head.c: Use gassign

gcc/ChangeLog.gimple-classes: * gimple-match-head.c (gimple_simplify): Within case GIMPLE_ASSIGN, introduce local "assign_stmt" via a checked cast and use it in place of "stmt" for typesafety. --- gcc/ChangeLog.gimple-classes | 6 ++ gcc/gimple-match-head.c | 25 +

[gimple-classes, committed 09/44] auto-profile.c: Use gassign

gcc/ChangeLog.gimple-classes: * auto-profile.c (afdo_propagate_circuit): Replace a check for GIMPLE_ASSIGN within the while loop with a dyn_cast, introducing a local "def_assign", using it in place of "stmt" for typesafety. This involves replacing the series of while

[gimple-classes, committed 28/44] tree-ssa-dom.c: Use gassign

gcc/ChangeLog.gimple-classes: * tree-ssa-dom.c (initialize_hash_element): Replace check for GIMPLE_ASSIGN with a dyn_cast, introducing local "assign_stmt", using it in place of "stmt" for typesafety. (record_equivalences_from_incoming_edge): Replace is_gimple_assign

[gimple-classes, committed 31/44] ipa-polymorphic-call.c: Use gassign

gcc/ChangeLog.gimple-classes: * ipa-polymorphic-call.c (walk_ssa_copies): Add checked cast. (ipa_polymorphic_call_context::ipa_polymorphic_call_context): Introduce local gassign * "def_assign" from result of gimple_assign_single_p, using it in place of SSA_NA

[gimple-classes, committed 35/44] gimple_assign_rhs_to_tree takes a const gassign *

gcc/ChangeLog.gimple-classes: * cfgexpand.c (gimple_assign_rhs_to_tree): Strengthen param from gimple to const gassign *. (expand_debug_expr): Add checked cast. (expand_gimple_basic_block): Likewise. * cfgexpand.h (gimple_assign_rhs_to_tree): Strengthen param

[gimple-classes, committed 40/44] ipa-icf-gimple.c: Use gassign

gcc/ChangeLog.gimple-classes: * ipa-icf-gimple.c (func_checker::compare_bb): Add checked casts. (func_checker::compare_gimple_assign): Strengthen params from gimple to const gassign *. * ipa-icf-gimple.h (func_checker::compare_gimple_assign): Likewise. --- g

[gimple-classes, committed 43/44] sese.c: Use gassign

gcc/ChangeLog.gimple-classes: * sese.c (rename_uses): Add checked cast to gassign within region guarded by is_gimple_assign. (graphite_copy_stmts_from_block): Likewise. --- gcc/ChangeLog.gimple-classes | 6 ++ gcc/sese.c | 4 ++-- 2 files changed, 8 in

[gimple-classes, committed 39/44] ipa-prop.c: Use gassign

gcc/ChangeLog.gimple-classes: * ipa-prop.c (stmt_may_be_vtbl_ptr_store): Replace is_gimple_assign with dyn_cast, introducing local gassign * "assign_stmt" and using it in place of "stmt" for typesafety. (load_from_unmodified_param): Use "assign" rather than "stmt" fo

[gimple-classes, committed 44/44] trans-mem.c: Use gassign

gcc/ChangeLog.gimple-classes: * trans-mem.c (thread_private_new_memory): Replace is_gimple_assign with a dyn_cast, introducing local gassign * "assign_stmt", using it in place of "stmt" for typesafety. (examine_assign_tm): Strengthen local "stmt" from gimple to

[gimple-classes, committed 42/44] predict.c: Use gassign

gcc/ChangeLog.gimple-classes: * predict.c (expr_expected_value_1): Replace is_gimple_assign with a dyn_cast, introducing local "def_assign" and using it in place of "def" for typesafety. --- gcc/ChangeLog.gimple-classes | 6 ++ gcc/predict.c| 13 +++

[gimple-classes, committed 37/44] ipa-split.c: Use gassign

gcc/ChangeLog.gimple-classes: * ipa-split.c (find_return_bb): Eliminate check for GIMPLE_ASSIGN since this is covered by gimple_assign_single_p, using the latter to introduce local gassign * "assign_stmt", using it in place of "stmt" for typesafety. (find_ret

[gimple-classes, committed 32/44] Fix issues with gimple_assign_single_p and gimple_assign_load_p

gcc/ChangeLog.gimple-classes: * asan.c (get_mem_ref_of_assignment): Weaken param "assignment" from const gassign * to gassign *. * gimple.h (gimple_assign_single_p): Eliminate const_gimple variant, since this could call itself, leading to an infinite recursion.

[gimple-classes, committed 38/44] ipa-inline-analysis.c: Use gassign

gcc/ChangeLog.gimple-classes: * ipa-inline-analysis.c (unmodified_parm): Capture result of gimple_assign_single_p as a new local gassign * "def_assign" and use in place of SSA_NAME_DEF_STMT (op) for typesafety. (unmodified_parm_or_parm_agg_item): Likewise. (e

[gimple-classes, committed 41/44] omp-low.c: Use gassign

gcc/ChangeLog.gimple-classes: * omp-low.c (expand_omp_taskreg): Strengthen locals "parcopy_stmt" and "stmt" from gimple to gassign *, replacing a check against GIMPLE_ASSIGN with a dyn_cast. (expand_omp_atomic_fetch_op): Likewise for local "stmt". (expand_omp

[gimple-classes, committed 34/44] build_assign and build_type_cast take const gassign *

gcc/ChangeLog.gimple-classes: * asan.c (asan_expand_check_ifn): Strengthen local "shadow_test" from gimple to gassign *. Introduce gassign * locals "add_7", "cast", "t_ge_shadow", "and_expr", using them in place of gimple_seq_last_stmt (seq) for typesafety.

[gimple-classes, committed 30/44] gimple-walk.c: Use gassign

gcc/ChangeLog.gimple-classes: * gimple-walk.c (walk_gimple_op): Within case GIMPLE_ASSIGN, introduce local gassign * "assign_stmt" and use it in place of "stmt" for typesafety. --- gcc/ChangeLog.gimple-classes | 6 gcc/gimple-walk.c| 74 +++

[gimple-classes, committed 36/44] cfgexpand.c: Use gassign *

gcc/ChangeLog.gimple-classes: * cfgexpand.c (add_scope_conflicts_1): Add checked cast. (expand_gimple_cond): Replace check for GIMPLE_ASSIGN with a dyn_cast, introducing local "second_assign", using it in place of "assign" for typesafety. (expand_call_stmt):

Re: [PATCH][11/n] Merge from match-and-simplify, bit patterns from forwprop

On Thu, 6 Nov 2014, Richard Biener wrote: On Thu, 6 Nov 2014, Marc Glisse wrote: On Thu, 6 Nov 2014, Richard Biener wrote: On Thu, 6 Nov 2014, Marc Glisse wrote: On Thu, 6 Nov 2014, Richard Biener wrote: +/* Try simple folding for X op !X, and X op X with the help + of the truth_valued

[gimple-classes, committed 11/44] gimple-ssa-isolate-paths.c: Use gassign

gcc/ChangeLog.gimple-classes: * gimple-ssa-isolate-paths.c (insert_trap_and_remove_trailing_statements): Replace a call to is_gimple_assign with a dyn_cast, introducing a local "assign_stmt", and using it in place of "stmt" for typesafety. --- gcc/ChangeLog.gimple-c

[gimple-classes, committed 23/44] tree-ssa-alias.c: Use gassign

gcc/ChangeLog.gimple-classes: * tree-ssa-alias.c (stmt_may_clobber_global_p): Within case GIMPLE_ASSIGN, add a checked cast. (ao_ref_init_from_ptr_and_size): Both the gimple_assign_single_p and is_gimple_assign conditionals require a GIMPLE_ASSIGN, so begin w

[gimple-classes, committed 15/44] graphite-sese-to-poly.c: Use gasssign

gcc/ChangeLog.gimple-classes: * graphite-sese-to-poly.c (rewrite_cross_bb_scalar_deps): Add checked cast. (is_reduction_operation_p): Strengthen param from gimple to gassign *. (follow_ssa_with_commutative_ops): Replace is_gimple_assign with a dyn_cas

[gimple-classes, committed 27/44] tree-ssa-dce.c: Use gassign

gcc/ChangeLog.gimple-classes: * tree-ssa-dce.c (mark_stmt_if_obviously_necessary): Within case GIMPLE_ASSIGN, introduce local "assign_stmt" and use in place of "stmt" for typesafety. (mark_all_reaching_defs_necessary_1): Add a checked cast. (propagate_necessi

  1   2   >