Re: [PATCH, rs6000] Add additional support for vec_subc, vec_sube, vec_subec builtins.

Hi Carl, On Fri, Jun 30, 2017 at 01:32:01PM -0700, Carl Love wrote: > vector unsigned __int128 vec_sube (vector unsigned __int128, vector > unsigned __int128, vector unsigned __int128); Many of the changelog lines are much too long. > * gcc.target/powerpc/p8vector-builtin-8.c

Re: [PATCH, rs6000] Add support to __builtin_cpu_supports() for two new HWCAP2 bits

On Fri, Jun 30, 2017 at 11:53:48AM -0500, Peter Bergner wrote: > >> Not use an installed header, that's not what I'm asking. Share the > >> source file, i.e., just copy it over from the glibc source tree (it > >> should probably hold the master copy). Fewer typos, cannot forget to > >> update

Re: [PATCH,rs6000] PR80103: Fix typo in test case

On Fri, Jun 30, 2017 at 09:56:03AM -0600, Kelvin Nilsen wrote: > > While reviewing regression test results for a back port of the PR80103 > patch, I discovered a typographic error in the test case. This patch > corrects the error. A syntax error, even... I wonder why we haven't noticed that

Re: [PATCH] rs6000 branch probability changes

On Fri, Jun 30, 2017 at 10:36:27PM +0100, Ramana Radhakrishnan wrote: > On Fri, Jun 30, 2017 at 2:36 PM, David Edelsohn wrote: > > Convert the rs6000 port to use the new API for branch probabilities. > > > > Okay? > > > > Thanks, David > > > > * config/rs6000/rs6000.c

Re: Convert profile probabilities to new type

Also seeing a possibly related failure from my glibc bot, building for hppa-linux-gnu: /scratch/jmyers/glibc-bot/src/gcc/gcc/config/pa/pa.c: In function 'bool pa_expand_compare_and_swap_loop(rtx, rtx, rtx, rtx)': /scratch/jmyers/glibc-bot/src/gcc/gcc/config/pa/pa.c:10726:59: error: could not

Re: [C++ Patch] PR 65775 ("Late-specified return type bypasses return type checks (qualified, function, array)")

OK. On Sun, Jun 25, 2017 at 8:18 PM, Paolo Carlini wrote: > ... in fact, simply moving the checks forward, past the > splice_late_return_type call, appears to work fine. I'm finishing testing > the below. > > Thanks! > Paolo. > > /

Re: [PATCH] rs6000 branch probability changes

On Fri, Jun 30, 2017 at 2:36 PM, David Edelsohn wrote: > Convert the rs6000 port to use the new API for branch probabilities. > > Okay? > > Thanks, David > > * config/rs6000/rs6000.c (emit_unlikely_jump): Adjust to new branch > probability data type. > > Index: rs6000.c >

Fix ICE with -fno-guess-branch-probability

Hi, this patch fixes a problem when -fno-guess-branch-probability is inlined into expanded thunk. When inlining, the profile status needs to be dropped to lowest common one (well in theory we can avoid that when callee has only one BB or so). Index: ChangeLog

C++ PATCH for c++/81257, ICE with invalid ::template

The change for 54769 caused us to accept this invalid ::template A even though the scope is non-dependent, so we can do the lookup. Fixed by reverting part of the change to cp_parser_template_name, which isn't actually needed. Tested x86_64-pc-linux-gnu, applying to trunk. commit

Re: [PATCH] [PR 81245] Fix tree-if-conv calling of update_stmt after fold_stmt

On Fri, Jun 30, 2017 at 1:20 AM, Richard Biener wrote: > On Thu, Jun 29, 2017 at 10:12 PM, Andrew Pinski wrote: >> Hi, >> As described in the bug, tree-if-conv is calling update_stmt on an >> old stmt which might have been removed from the IR

[PATCH, rs6000] Add additional support for vec_subc, vec_sube, vec_subec builtins.

GCC Maintainers: The following patch adds support for additional the vec_subc, vec_sube, vec_subec builtins. It also adds a few missing tests for currently supported builtins. The patch has been tested on powerpc64le-unknown-linux-gnu (Power 8 LE) and powerpc64-unknown-linux-gnu(Power 8 BE).

Re: [C++ PATCH] conversion operator names

On Fri, Jun 30, 2017 at 3:53 PM, Nathan Sidwell wrote: > On 06/30/2017 03:11 PM, Jason Merrill wrote: >> >> On Fri, Jun 30, 2017 at 2:48 PM, Nathan Sidwell wrote: >>> >>> We used to give conversion operatos type-specific mangled names. Now we >>> just number them

Re: [PING] Re: [PATCH] c/c++: Add fix-it hints for suggested missing #includes

On Fri, 2017-06-30 at 09:40 -0600, Jeff Law wrote: > On 05/26/2017 01:54 PM, David Malcolm wrote: > > Ping: > > https://gcc.gnu.org/ml/gcc-patches/2017-05/msg00321.html > > > > On Thu, 2017-05-04 at 12:36 -0400, David Malcolm wrote: > > > As of r247522, fix-it-hints can suggest the insertion of

Re: [C++ PATCH] conversion operator names

On 06/30/2017 03:11 PM, Jason Merrill wrote: On Fri, Jun 30, 2017 at 2:48 PM, Nathan Sidwell wrote: We used to give conversion operatos type-specific mangled names. Now we just number them 'operator $N' for N in [1,...]. As such mangle_convop_name_for_type is misnamed and has

Re: [C++ PATCH] "decomposition declaration" -> "structured binding" in C++ diagnostics

On Fri, Jun 30, 2017 at 3:35 PM, Jakub Jelinek wrote: > On Fri, Jun 30, 2017 at 03:15:21PM -0400, Jason Merrill wrote: >> On Fri, Jun 30, 2017 at 1:26 PM, Jakub Jelinek wrote: >> > Hi! >> > >> > As C++17 decided to rename decompositions to structured bindings,

Re: [C++ PATCH] "decomposition declaration" -> "structured binding" in C++ diagnostics

On Fri, Jun 30, 2017 at 03:15:21PM -0400, Jason Merrill wrote: > On Fri, Jun 30, 2017 at 1:26 PM, Jakub Jelinek wrote: > > Hi! > > > > As C++17 decided to rename decompositions to structured bindings, this > > patch attempts to adjust the diagnostics by replacing > >

Re: [PATCH v2][RFC] Canonize names of attributes.

On Fri, Jun 30, 2017 at 5:23 AM, Martin Liška wrote: > This is v2 of the patch, where just names of attributes are canonicalized. > Patch can bootstrap on ppc64le-redhat-linux and survives regression tests. What is the purpose of the new "strict" parameter to cmp_attribs* ? I

Re: Convert profile probabilities to new type

> > After the change, edges that fail the predicate contribute REG_BR_PROB_BASE in > the upper hunk, but 0 in the lower hunk. Before the change, they contributed 0 > in both cases. > > The following patch should restore things: > > diff --git a/gcc/sel-sched-ir.c b/gcc/sel-sched-ir.c > index

Re: [C++ PATCH] "decomposition declaration" -> "structured binding" in C++ diagnostics

On Fri, Jun 30, 2017 at 1:26 PM, Jakub Jelinek wrote: > Hi! > > As C++17 decided to rename decompositions to structured bindings, this > patch attempts to adjust the diagnostics by replacing > "decomposition declaration" with "structured binding". > Or shall I use "structured

Re: [C++ PATCH] conversion operator names

On Fri, Jun 30, 2017 at 2:48 PM, Nathan Sidwell wrote: > We used to give conversion operatos type-specific mangled names. Now we > just number them 'operator $N' for N in [1,...]. As such > mangle_convop_name_for_type is misnamed and has no place in mangle.c Hmm, wouldn't

Re: [PATCH] [Aarch64] Variable shift count truncation issues

Okay I will take a look at this. Michael Collison > On Jun 30, 2017, at 11:04 AM, Andreas Schwab wrote: > >> On Jun 23 2017, Michael Collison wrote: >> >> diff --git a/gcc/testsuite/gcc.target/aarch64/var_shift_mask_1.c >>

Re: [C++ PATCH] conversion operator names

On 06/30/2017 02:48 PM, Nathan Sidwell wrote: We used to give conversion operatos type-specific mangled names. Now we just number them 'operator $N' for N in [1,...]. As such mangle_convop_name_for_type is misnamed and has no place in mangle.c This patch moves that machinery to lex.c, which

[C++ PATCH] conversion operator names

We used to give conversion operatos type-specific mangled names. Now we just number them 'operator $N' for N in [1,...]. As such mangle_convop_name_for_type is misnamed and has no place in mangle.c This patch moves that machinery to lex.c, which is a better place (best place?). Other than

Re: Convert profile probabilities to new type

on/torture/pr53505.c -O3 -fomit-frame-pointer > -funroll-loops -fpeel-loops -ftracer -finline-functions (test for excess > errors) > Excess errors: > during RTL pass: mach > /usr/local/gcc/gcc-20170630/gcc/testsuite/c-c++-common/torture/pr53505.c:26:1: > internal compiler err

Re: [PATCH][AArch64] Fix ILP32 memory access

ode); > break; > > case CONST: That breaks a lot of gnat tests in ilp32 mode: spawn -ignore SIGHUP /opt/gcc/gcc-20170630/Build/gcc/gnatmake --GCC=/opt/gcc/gcc-20170630/Build/gcc/xgcc --GNATBIND=/opt/gcc/gcc-20170630/Build/gcc/gnatbind --GNATLINK=/opt/gcc/gcc-20170630/Build/gcc/g

Re: [PATCH] [Aarch64] Variable shift count truncation issues

On Jun 23 2017, Michael Collison wrote: > diff --git a/gcc/testsuite/gcc.target/aarch64/var_shift_mask_1.c > b/gcc/testsuite/gcc.target/aarch64/var_shift_mask_1.c > new file mode 100644 > index 000..e2b020e > --- /dev/null > +++

Re: [C++ PATCH] "decomposition declaration" -> "structured binding" in C++ diagnostics

On 06/30/2017 01:26 PM, Jakub Jelinek wrote: Hi! As C++17 decided to rename decompositions to structured bindings, this patch attempts to adjust the diagnostics by replacing "decomposition declaration" with "structured binding". Or shall I use "structured binding declaration" instead (or is

Re: [C++] Fix decomp ICE with invalid initializer (PR c++/81258)

On 06/30/2017 01:24 PM, Jakub Jelinek wrote: The initializer for structured binding has to be one of: = assignment-expression ( assignment-expression ) { assignment-expression } but cp_parser_initializer can parse other forms, with fewer or more expressions in there. Some cases we caught with

[PATCH] Fix bb-reorder asm goto handling (PR sanitizer/81262)

Hi! The following testcases now ICE on the trunk. The problem is that fix_up_fall_thru_edges doesn't notice asm goto does have a fallthru edge when it has 3 edges and the EDGE_FALLTHRU is only 3rd. Fixed by using find_fallthru_edge if we didn't find it among the first 2 edges no matter what the

[C++ PATCH] "decomposition declaration" -> "structured binding" in C++ diagnostics

Hi! As C++17 decided to rename decompositions to structured bindings, this patch attempts to adjust the diagnostics by replacing "decomposition declaration" with "structured binding". Or shall I use "structured binding declaration" instead (or is that too longer/verbose), or something different?

[C++] Fix decomp ICE with invalid initializer (PR c++/81258)

Hi! The initializer for structured binding has to be one of: = assignment-expression ( assignment-expression ) { assignment-expression } but cp_parser_initializer can parse other forms, with fewer or more expressions in there. Some cases we caught with various cryptic errors or pedwarns, but

[PATCH] stringpool clean

The string pool provides 'empty_string', a zero-length string. But we only use it in 3 places, preferring the more obvious "" elsewhere. These days we have linkers that can do string table merging, so any size optimimization empty_string might have bought us is long gone. So nuking

Re: [PATCH, rs6000] Add support to __builtin_cpu_supports() for two new HWCAP2 bits

Segher, any response to my findings below about whether we should try and share header files with GLIBC? Peter On 6/27/17 1:06 PM, Peter Bergner wrote: > On 6/27/17 11:07 AM, Segher Boessenkool wrote: >> Not use an installed header, that's not what I'm asking. Share the >> source file, i.e.,

Re: [PATCH v2][ASAN] Implement dynamic allocas/VLAs sanitization.​

Hi, On 29/06/17 15:35, Jakub Jelinek wrote: Hi! Sorry for the review delay. On Mon, Jun 26, 2017 at 03:49:23PM +0300, Maxim Ostapenko wrote: (handle_builtin_stackrestore): Likewise. The function is called with _ between stack and restore. * match.pd: Add new pattern.

Re: Convert profile probabilities to new type

-ftracer -finline-functions (test for excess errors) Excess errors: during RTL pass: mach /usr/local/gcc/gcc-20170630/gcc/testsuite/c-c++-common/torture/pr53505.c:26:1: internal compiler error: in merge_expr, at sel-sched-ir.c:1886 0x410b897f merge_expr(_expr*, _expr*, rtx_insn

[gomp5] Fix ICE on clauses-1.c testcase

Hi! Another thing I've missed in the testing (or lack thereof) of the task reduction parsing patch. Reduction is a data sharing clause that is best located on the task construct for the taskloop IL sandwich, it will be of course added as argument to the GOMP_taskloop call when implemented later.

[GC PATCH] Clean ggc.h

ggc.h had a bit of funky formatting and widespread pre-C++ idiom 'static inline'. Fixed as obvious. template static inline T * just looks weird :) nathan -- Nathan Sidwell 2017-06-30 Nathan Sidwell * ggc.h: Replace all 'static inline' with plain 'inline'. Fix some

Re: [PATCH GCC][2/2]Refine CFG and bound information for split loops

On 06/14/2017 07:08 AM, Bin Cheng wrote: > Hi, > Loop split currently generates below control flow graph for split loops: > + > + .-- guard1 --. > + v v > + pre1(loop1).-->pre2(loop2) > + | |

Re: [PATCH,rs6000] PR80103: Fix typo in test case

On 6/30/17 10:56 AM, Kelvin Nilsen wrote: > While reviewing regression test results for a back port of the PR80103 > patch, I discovered a typographic error in the test case. This patch > corrects the error. > > I have tested this fix on powerpc64le-unknown-linux-gnu with no > regressions. Is

[PATCH, 6/4] Handle GOMP_OPENACC_NVPTX_JIT=-arch= in libgomp nvptx plugin

[ was: Re: [PATCH, 0/4] Handle GOMP_OPENACC_NVPTX_{DISASM,SAVE_TEMPS,JIT} in libgomp nvptx plugin ] On 06/26/2017 01:24 PM, Tom de Vries wrote: Hi, I've written a patch series to facilitate debugging libgomp openacc testcase failures on the nvptx accelerator. When running an openacc

Re: [PATCH] make find_taken_edge handle case with just default

On 6/30/17 3:27 AM, Richard Biener wrote: > On Thu, 29 Jun 2017, Peter Bergner wrote: >> On 6/29/17 8:58 AM, Richard Biener wrote: >>> We can also merge both loops, counting new_size upwards, storing >>> to label new_size if new_size != i ... >> >> Like this. I'm bootstrapping and regtesting the

Re: [PATCH, 5/4] Handle GOMP_OPENACC_NVPTX_PTXRW in libgomp nvptx plugin

On 06/27/2017 11:16 AM, Tom de Vries wrote: [ was: Re: [PATCH, 0/4] Handle GOMP_OPENACC_NVPTX_{DISASM,SAVE_TEMPS,JIT} in libgomp nvptx plugin ] On 06/26/2017 01:24 PM, Tom de Vries wrote: Hi, I've written a patch series to facilitate debugging libgomp openacc testcase failures on the nvptx

Re: [PATCH 1/2] Support C++-specific selftests

On 06/22/2017 06:20 PM, David Malcolm wrote: > Currently the "make selftest" target run during each stage of the > build just runs the selftests within cc1. > > As part of the fix for PR c++/81167 I want to be able to write > C++-specific selftests (to exercise the C++ implementation of > the

[PATCH,rs6000] PR80103: Fix typo in test case

While reviewing regression test results for a back port of the PR80103 patch, I discovered a typographic error in the test case. This patch corrects the error. I have tested this fix on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? gcc/testsuite/ChangeLog:

Re: [2/2] PR 80769: Incorrect strlen optimisation

On Tue, May 16, 2017 at 09:02:08AM +0100, Richard Sandiford wrote: > 2017-05-16 Richard Sandiford > > gcc/ > PR tree-optimization/80769 > * tree-ssa-strlen.c (strinfo): Document that "stmt" is also used > for malloc and calloc. Document the new

Re: [PATCH, 4/4] Handle GOMP_OPENACC_NVPTX_JIT=-ori in libgomp nvptx plugin

On 06/26/2017 01:44 PM, Tom de Vries wrote: On 06/26/2017 01:24 PM, Tom de Vries wrote: Hi, I've written a patch series to facilitate debugging libgomp openacc testcase failures on the nvptx accelerator. When running an openacc test-case on an nvptx accelerator, the following happens: -

Re: [PATCH, GCC/ARM, gcc-5-branch, ping2] Fix gcc.target/arm/fpscr.c

On 30/06/17 16:41, Thomas Preudhomme wrote: Ping? Best regards, Thomas On 28/06/17 12:35, Thomas Preudhomme wrote: Ping? Best regards, Thomas On 26/06/17 12:32, Thomas Preudhomme wrote: Hi, As raised by Christophe Lyon, fpscr.c FAILs because arm_fp_ok and arm_fp are not defined in GCC

Re: [PATCH, GCC/ARM, gcc-5-branch, ping2] Fix gcc.target/arm/fpscr.c

Ping? Best regards, Thomas On 28/06/17 12:35, Thomas Preudhomme wrote: Ping? Best regards, Thomas On 26/06/17 12:32, Thomas Preudhomme wrote: Hi, As raised by Christophe Lyon, fpscr.c FAILs because arm_fp_ok and arm_fp are not defined in GCC 5. This commit changes the test to use the

Re: [PING] Re: [PATCH] c/c++: Add fix-it hints for suggested missing #includes

On 05/26/2017 01:54 PM, David Malcolm wrote: > Ping: > https://gcc.gnu.org/ml/gcc-patches/2017-05/msg00321.html > > On Thu, 2017-05-04 at 12:36 -0400, David Malcolm wrote: >> As of r247522, fix-it-hints can suggest the insertion of new lines. >> >> This patch uses this to implement a new

Re: [ping * 2] PR78736: New C warning -Wenum-conversion

On 06/29/2017 05:01 PM, Joseph Myers wrote: > On Thu, 29 Jun 2017, Jeff Law wrote: > >> On 05/23/2017 07:54 AM, Prathamesh Kulkarni wrote: >>> Hi, >>> I would like to ping this patch for review: >>> https://gcc.gnu.org/ml/gcc-patches/2017-05/msg00775.html >> So was there any kind of resolution on

Re: [rtlanal] Do a better job of costing parallel sets containing flag-setting operations.

On 06/30/2017 03:03 AM, Richard Earnshaw (lists) wrote: > On 19/06/17 14:46, Richard Earnshaw (lists) wrote: >> Many parallel set insns are of the form of a single set that also sets >> the condition code flags. In this case the cost of such an insn is >> normally the cost of the part that

Re: [PATCH][DOC] Enhance target_clones documentation (PR other/78366).

On 06/30/2017 07:55 AM, Martin Liška wrote: > PING^1 > > On 06/22/2017 02:42 PM, Martin Liška wrote: >> Hi. >> >> As mentioned in the PR, we only generate a resolver function when there's a >> usage >> of a function with target_clones attribute. Let's document the behavior. >> >> Ready to be

[gomp4, nvptx, committed] Fix assert in nvptx_propagate_unified

Hi, with the openacc test-case in attached patch, I ran into an assert here: ... static void nvptx_propagate_unified (rtx_insn *unified) { rtx_insn *probe = unified; rtx cond_reg = SET_DEST (PATTERN (unified)); rtx pat; /* Find the comparison. (We could skip this and simply scan to

Re: [PATCH 0/3] Introduce internal_error_cont and exclude it from pot files

On 06/30/2017 04:31 AM, Martin Liška wrote: > On 04/10/2017 05:44 PM, Richard Biener wrote: >> On Mon, Apr 10, 2017 at 5:42 PM, Jeff Law wrote: >>> On 04/07/2017 04:30 AM, Martin Liška wrote: On 04/06/2017 06:44 PM, Jeff Law wrote: > > On 03/24/2017 03:29 AM,

[C++ PATCH] METHODVEC indexing

We store a class's overloaded methods on the METHODVEC, and provide one set of accessors that return an index (or -1), and another set of accessors that simply return the value of the slot (or NULL). This is confusing. The patch kills the former accessors, so everything outside of search.c

Re: [PATCH] Move static chain and non-local goto init after NOTE_INSN_FUNCTION_BEG (PR sanitize/81186).

On 06/30/2017 04:03 PM, Michael Matz wrote: > Hi, > > On Wed, 28 Jun 2017, Martin Liška wrote: > >> Thanks for the review. I'm not so familiar with RTL, but hopefully new >> version of the patch should do it properly. Idea is to come up with a >> new asan_stack_birth_insn that points after

Re: [PATCH] warn on mem calls modifying objects of non-trivial types (PR 80560)

On 06/30/2017 02:34 AM, Richard Biener wrote: On Thu, Jun 29, 2017 at 10:23 PM, Martin Sebor wrote: On 06/29/2017 10:15 AM, Jan Hubicka wrote: Hello, diff --git a/gcc/hash-table.h b/gcc/hash-table.h index 0f7e21a..443d16c 100644 --- a/gcc/hash-table.h +++

Re: [PATCH] Remove dead code in asan.c

> + /* Unpoison shadow memory that corresponds to a variable that was > + is subject of use-after-return sanitization. */ was is ? And shouldn't it be subject to instead of subject of? Otherwise LGTM. Jakub

Re: [PATCH] Fix PR81249

Hi Richard, On 29 June 2017 at 14:53, Richard Biener wrote: > > Bootstrap and regtest running on x86_64-unknown-linux-gnu. > > Richard. > > 2017-06-29 Richard Biener > > PR tree-optimization/81249 > * tree-vect-loop.c

Re: [PATCH] Move static chain and non-local goto init after NOTE_INSN_FUNCTION_BEG (PR sanitize/81186).

Hi, On Wed, 28 Jun 2017, Martin Liška wrote: > Thanks for the review. I'm not so familiar with RTL, but hopefully new > version of the patch should do it properly. Idea is to come up with a > new asan_stack_birth_insn that points after place where stack is ready > to use (when one uses

Re: [PATCH][DOC] Enhance target_clones documentation (PR other/78366).

PING^1 On 06/22/2017 02:42 PM, Martin Liška wrote: > Hi. > > As mentioned in the PR, we only generate a resolver function when there's a > usage > of a function with target_clones attribute. Let's document the behavior. > > Ready to be installed? > Martin > > gcc/ChangeLog: > > 2017-06-22

Re: [PATCH] Make inlining consistent in LTO and non-LTO mode (PR target/71991).

On 06/28/2017 05:18 PM, Jan Hubicka wrote: >> On 06/28/2017 04:24 PM, Jan Hubicka wrote: - /* If callee has no option attributes, then it is ok to inline. */ - if (!callee_tree) + /* If callee has no option attributes (or default), + then it is ok to inline. */ +

Re: [PATCH 4/N] Recover GOTO predictor.

On 06/22/2017 12:27 PM, Richard Biener wrote: > On Wed, Jun 21, 2017 at 3:06 PM, Martin Liška wrote: >> Hello. >> >> There's one additional predictor enhancement that is GOTO predict that >> used to working. Following patch adds expect statement for C and C++ family >> languages.

[PATCH] rs6000 branch probability changes

Convert the rs6000 port to use the new API for branch probabilities. Okay? Thanks, David * config/rs6000/rs6000.c (emit_unlikely_jump): Adjust to new branch probability data type. Index: rs6000.c === --- rs6000.c (revision 249839)

Re: [PATCH] Remove dead code in asan.c

On 06/30/2017 12:15 PM, Jakub Jelinek wrote: > On Fri, Jun 30, 2017 at 12:00:36PM +0200, Martin Liška wrote: >> Hi. >> >> Following crap code was added by me when I added use-after-scope. >> Actually decl always points to LASANPC, so asan_handled_variables->contains >> (decl) >> is always false.

[C++ PATCH] cdtor naming

This patch changes the way we name cdtors internally. Currently both in-charge ctor and dtor get the DECL_NAME naming the class they are a member of. That's the same DECL_NAME as the class's self-reference TYPE_DECL. (The specialized base/complete/deleting cdtors get special names already.)

[PATCH] Speed up SLP analysis

This removes the redundant analyzing of all scalar stmts in a SLP node (which also should cut down size of the dump file) as well as move the vector type computation for BBs into the SLP analysis routine (computing it only once). The transform phase later does essentially the same (only look at

Re: [PATCH] Fix vec_extract_lo_* patterns (PR target/81225)

Hello Jakub, On 29 Jun 18:51, Jakub Jelinek wrote: > Hi! > > This patch fixes various issues with the vec_extract_lo_* patterns. > There are splitters for these, but only for some cases (no mask, and > in one case also not xmm32+ reg) that change those into just a copy or load > of the low part

Re: C++ PATCH to remove WITH_CLEANUP_EXPR handling

On 06/29/2017 05:44 PM, Jason Merrill wrote: The C++ front end hasn't generated WITH_CLEANUP_EXPR in a very long time (20+ years?), so there's no need to handle it. I see dead code nathan -- Nathan Sidwell

Re: [PATCH] Fix removal of ifunc (PR ipa/81214).

> Hello. > > Following patch fixes the issue where we do not emit ifunc and resolver > for function that are not called in a compilation unit or and not > referenced. > > Patch can bootstrap on ppc64le-redhat-linux and survives regression tests. > i386.exp tests work on x86_64-linux-gnu. OK,

Re: [PATCH 4/N] Recover GOTO predictor.

> PING^1 > > Can you please Honza give a formal approval for the patch? OK, thanks! Honza > > Thanks, > Martin > > On 06/22/2017 01:47 PM, Richard Biener wrote: > > On Thu, Jun 22, 2017 at 12:57 PM, Martin Liška wrote: > >> On 06/22/2017 12:27 PM, Richard Biener wrote: > >>>

Re: [PATCH v9] add -fpatchable-function-entry=N,M option

Richard, Ping? On Tue, Jun 13, 2017 at 07:00:27PM +0200, Torsten Duwe wrote: > Changes since v8: > > * Documentation changes as requested by Sandra > * 3 functional test cases added > > Torsten > > > gcc/c-family/ChangeLog > 2017-06-13 Torsten Duwe > > *

[PR c++/81229] Dangling GC pointer

This fixes 81229, an ICE during GC. When converting the namespace symbol handling I was mystified by a SET_IDENTIFIER_TYPE call that seemingly had no effect. Turns out it is needed, because we've already set the IDENTIFIER_TYPE to the new decl's type. Whose name points back at the new

Re: [PATCH GCC][13/13]Distribute loop with loop versioning under runtime alias check

On Wed, Jun 28, 2017 at 2:09 PM, Bin.Cheng wrote: > On Wed, Jun 28, 2017 at 1:29 PM, Richard Biener > wrote: >> On Wed, Jun 28, 2017 at 1:46 PM, Bin.Cheng wrote: >>> On Wed, Jun 28, 2017 at 11:58 AM, Richard Biener >>>

Re: [PATCH GCC][01/13]Introduce internal function IFN_LOOP_DIST_ALIAS

On Wed, Jun 28, 2017 at 8:29 AM, Richard Biener wrote: > On Tue, Jun 27, 2017 at 6:46 PM, Bin.Cheng wrote: >> On Tue, Jun 27, 2017 at 3:59 PM, Richard Biener >> wrote: >>> On June 27, 2017 4:27:17 PM GMT+02:00,

Re: [PATCH 0/3] Introduce internal_error_cont and exclude it from pot files

On 04/10/2017 05:44 PM, Richard Biener wrote: > On Mon, Apr 10, 2017 at 5:42 PM, Jeff Law wrote: >> On 04/07/2017 04:30 AM, Martin Liška wrote: >>> >>> On 04/06/2017 06:44 PM, Jeff Law wrote: On 03/24/2017 03:29 AM, Martin Liška wrote: > > I would like to ping

Re: [PATCH] Remove dead code in asan.c

On Fri, Jun 30, 2017 at 12:00:36PM +0200, Martin Liška wrote: > Hi. > > Following crap code was added by me when I added use-after-scope. > Actually decl always points to LASANPC, so asan_handled_variables->contains > (decl) > is always false. > > Well, originally the idea was to not clear

[PATCH] Remove dead code in asan.c

Hi. Following crap code was added by me when I added use-after-scope. Actually decl always points to LASANPC, so asan_handled_variables->contains (decl) is always false. Well, originally the idea was to not clear content (place in shadow memory in between red zoner) of auto variables, but as

[PATCH] Fix dot-fn

Committed. 2017-06-30 Richard Biener * graph.c (draw_cfg_node_succ_edges): Fix broken dot syntax. Index: gcc/graph.c === --- gcc/graph.c (revision 249832) +++ gcc/graph.c (working copy) @@ -136,13

Re: [PATCH] ASAN: handle addressable params (PR sanitize/81040).

On Fri, Jun 30, 2017 at 11:21:48AM +0200, Martin Liška wrote: > @@ -858,6 +864,132 @@ sanitize_asan_mark_poison (void) > } > } > > +/* Rewrite all usages of tree OP which is a PARM_DECL with a VAR_DECL > + that is it's DECL_VALUE_EXPR. */ > + > +static tree > +rewrite_usage_of_param

Re: [PATCH 4/N] Recover GOTO predictor.

PING^1 Can you please Honza give a formal approval for the patch? Thanks, Martin On 06/22/2017 01:47 PM, Richard Biener wrote: > On Thu, Jun 22, 2017 at 12:57 PM, Martin Liška wrote: >> On 06/22/2017 12:27 PM, Richard Biener wrote: >>> On Wed, Jun 21, 2017 at 3:06 PM, Martin

[PATCH v2][RFC] Canonize names of attributes.

On 06/28/2017 09:01 PM, Jason Merrill wrote: > On Wed, Jun 28, 2017 at 12:05 PM, Joseph Myers > wrote: >> On Wed, 28 Jun 2017, Martin Liška wrote: >> >>> On 06/14/2017 07:24 PM, Jason Merrill wrote: On Tue, Jun 13, 2017 at 8:32 AM, Martin Liška

Re: [PATCH] ASAN: handle addressable params (PR sanitize/81040).

On 06/29/2017 01:17 PM, Jakub Jelinek wrote: > On Tue, Jun 20, 2017 at 03:06:56PM +0200, Martin Liška wrote: >> +/* Rewrite all usages of tree OP which is a PARM_DECL with a VAR_DECL >> + that is it's DECL_VALUE_EXPR. */ >> + >> +static tree >> +rewrite_usage_of_param (tree *op, int

Re: [rtlanal] Do a better job of costing parallel sets containing flag-setting operations.

On 19/06/17 14:46, Richard Earnshaw (lists) wrote: > Many parallel set insns are of the form of a single set that also sets > the condition code flags. In this case the cost of such an insn is > normally the cost of the part that doesn't set the flags, since updating > the condition flags is

[PATCH] Fix removal of ifunc (PR ipa/81214).

Hello. Following patch fixes the issue where we do not emit ifunc and resolver for function that are not called in a compilation unit or and not referenced. Patch can bootstrap on ppc64le-redhat-linux and survives regression tests. i386.exp tests work on x86_64-linux-gnu. Ready to be installed?

[PATCH] Fix ifunc and resolver (PR ipa/81213).

Hello. Following patch does refactoring of make_resolver_func where ifunc alias and resolver were probably confused. Patch can bootstrap on ppc64le-redhat-linux and survives regression tests. i386.exp tests work on x86_64-linux-gnu. Ready to be installed? Martin gcc/ChangeLog: 2017-06-29

Re: [PATCH] warn on mem calls modifying objects of non-trivial types (PR 80560)

On Thu, Jun 29, 2017 at 10:23 PM, Martin Sebor wrote: > On 06/29/2017 10:15 AM, Jan Hubicka wrote: >> >> Hello, >>> >>> diff --git a/gcc/hash-table.h b/gcc/hash-table.h >>> index 0f7e21a..443d16c 100644 >>> --- a/gcc/hash-table.h >>> +++ b/gcc/hash-table.h >>> @@ -803,7 +803,10

Re: PR80806

On Fri, Jun 30, 2017 at 10:25:51AM +0200, Richard Biener wrote: > Do not go down the route of -Wunreachable-code again please. Yeah, I don't think we want -Wunused-but-set* as a late warning, it is intentionally in the FE where is the only place where the false positive ratio of the warning can

Re: [PATCH] make find_taken_edge handle case with just default

On Thu, 29 Jun 2017, Peter Bergner wrote: > On 6/29/17 8:58 AM, Richard Biener wrote: > > On Thu, 29 Jun 2017, Peter Bergner wrote: > > > >> On 6/29/17 4:03 AM, Richard Biener wrote: > >>> > >>> This refactors things a bit to make CFG cleanup handle switches with > >>> just a default label. If

Re: PR80806

On Thu, 29 Jun 2017, Jeff Law wrote: > On 05/18/2017 12:55 PM, Prathamesh Kulkarni wrote: > > Hi, > > The attached patch tries to fix PR80806 by warning when a variable is > > set using memset (and friends) but not used. I chose to warn in dse > > pass since dse would detect if the variable

Re: [PATCH] [PR 81245] Fix tree-if-conv calling of update_stmt after fold_stmt

On Thu, Jun 29, 2017 at 10:12 PM, Andrew Pinski wrote: > Hi, > As described in the bug, tree-if-conv is calling update_stmt on an > old stmt which might have been removed from the IR already > (transforming of an assignment to a call in this case). This fixes > the problem

Re: [PATCH] Transform (m1 > m2) * d into m1> m2 ? d : 0

On Thu, Jun 29, 2017 at 5:10 PM, Wilco Dijkstra wrote: > Richard Biener wrote: > >> int f (int m, int c) >> { >> return (m & 1) * c; >> } > > This case (integer[0,1] rather than boolean input) should be transformed into > c & -(m & 1). The proposed patch handled both

[Committed] profile-count.h: Fix typos and whitespace issues.

I noticed a couple of typos and whitespace issues in profile-count.h. I've committed the following patch to fix them. gcc/ChangeLog: 2017-06-30 Andreas Krebbel * profile-count.h (enum profile_quality): Fix typos and whitespace issues. ---

[Committed] S/390: Adjust to the recent branch probability changes.

This fixes the bootstrap failure triggered by the recent changes wrt branch probabilities aka emit_cmp_and_jump_insns does not accept integers as branch probability anymore. Regressiontested on s390x. gcc/ChangeLog: 2017-06-30 Andreas Krebbel *

Re: gotools patch committed: Test runtime, misc/cgo/{test,testcarchive}

Hello! > This patch to the gotools Makefile adds tests to `make check`. We now > test the runtime package using the newly built go tool, and test that > cgo works by running the misc/cgo/test and misc/cgo/testcarchive > tests. Bootstrapped and ran Go testsuite on x86_64-pc-linux-gnu. >

Re: [PING 4] [PATCH] [AArch64] vec_pack_trunc_ should split after register allocator

Hi,  Please consider this as a personal reminder to review the patch at following link and let me know your comments on the same.  https://gcc.gnu.org/ml/gcc-patches/2017-04/msg01334.html Thanks, Naveen