Re: [PATCH] Fix PR31096

Hi, >> I think we should handle at least INTEGER_CST and SSA_NAME >> with VRP, and it seems natural to add a VRP check The check should be added in the tree_single_nonzero_warnv_p for SSA_NAME case for tree_expr_nonzero_p. However, for tree_expr_nonnegative_p, its been handled in a different w

Re: [patch] fix an openacc reference-type reduction bug in PR70643

On Wed, Apr 13, 2016 at 03:56:22PM -0700, Cesar Philippidis wrote: > Thomas noticed that one of the openacc fortran tests involving > reductions inside fortran modules in gomp-4_0-branch had regressed after > I backported the recent reduction changes I made to trunk. That failure > occurred because

Re: PATCH] Fix PR 31531: A microoptimization of isnegative of signed integer

Hi Richard, >> What's the motivation of splitting this into a equal type >> and a non-equal type case? Simply only allow nop-conversions here >> (tree_nop_conversion_p) and unconditionally emit Done. Thanks for the review and Comments. Implemented the modifications as per you review comments.

Re: [PATCH] 69517 - [5/6 regression] SEGV on a VLA with excess initializer elements

On Wed, Apr 13, 2016 at 11:36 AM, Martin Sebor wrote: > On 04/12/2016 12:17 PM, Jason Merrill wrote: >> >> On 04/10/2016 07:14 PM, Martin Sebor wrote: >>> >>> The call to build_vla_check() in check_initializer() is to check >>> only explicitly initialized VLAs. The latter function may need >>> to

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On Wed, Apr 13, 2016 at 1:12 PM, Jason Merrill wrote: > On 04/13/2016 03:18 PM, Jakub Jelinek wrote: >> >> On Wed, Apr 13, 2016 at 03:11:34PM -0400, Jason Merrill wrote: >>> >>> commit 761983a023b5217ef831a43f423779940c788ecf >>> Author: Jason Merrill >>> Date: Tue Apr 12 13:16:50 2016 -0400 >>

[patch] fix an openacc reference-type reduction bug in PR70643

Thomas noticed that one of the openacc fortran tests involving reductions inside fortran modules in gomp-4_0-branch had regressed after I backported the recent reduction changes I made to trunk. That failure occurred because lower_oacc_reductions wasn't considering that reduction initializers can b

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On Wed, Apr 13, 2016 at 1:12 PM, Jason Merrill wrote: > On 04/13/2016 03:18 PM, Jakub Jelinek wrote: >> >> On Wed, Apr 13, 2016 at 03:11:34PM -0400, Jason Merrill wrote: >>> >>> commit 761983a023b5217ef831a43f423779940c788ecf >>> Author: Jason Merrill >>> Date: Tue Apr 12 13:16:50 2016 -0400 >>

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On 12/04/16 16:27 -0400, Jason Merrill wrote: A revision of the patch previously posted at https://gcc.gnu.org/ml/gcc-patches/2016-03/msg00841.html To recap quickly, the C++ compiler has used a different calling convention for passing empty classes, because C++ says they have size 1, while t

libgcc: Fix typos in comments for ARM FP emulation routines

A lifetime ago I contributed a patch that added CFI directives to ieee754-df.S, among other files. For unrelated reasons I looked at that file again and saw that some of the comments have extra '@' characters interwined; this is probably the result of splitting lines because they were too long. Thi

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On 04/13/2016 03:18 PM, Jakub Jelinek wrote: On Wed, Apr 13, 2016 at 03:11:34PM -0400, Jason Merrill wrote: commit 761983a023b5217ef831a43f423779940c788ecf Author: Jason Merrill Date: Tue Apr 12 13:16:50 2016 -0400 gcc/ * cfgexpand.c (pass_expand::execute): Handle attribute abi_

Re: [PATCH] c++/70594 debug info differences

On 04/13/2016 11:54 AM, Nathan Sidwell wrote: This patch builds from Jason's proto-patch in comment #26. As Jason discovered, that led to link problems with gt_ggc_mx (tree_node *&) when building cc1. Making 'hash_map' GC-able caused gengtype to put the GC-walkers for tree into constexpr.c. No id

Re: [PATCH] 69517 - [5/6 regression] SEGV on a VLA with excess initializer elements

On 04/13/2016 02:36 PM, Martin Sebor wrote: I don't see the benefit in doing the checking in build_vec_init and split_constant_init_1 when it can all be done just in check_initializer. I'm sure you have your reasons for going that route so I'd appreciate if you could let me know why you think th

Re: [PATCH] Don't include DECL_UIDs in tree-sra fancy names for -fdump-final-insns= (PR c++/70594)

On April 13, 2016 4:32:25 PM GMT+02:00, Jakub Jelinek wrote: >On Wed, Apr 13, 2016 at 09:25:56AM +0200, Richard Biener wrote: >> So if you think it's not acceptable to drop the relation between >> the artificial number used by SRA and the original UID then go >> with a hash-map unconditionally. Y

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On Wed, Apr 13, 2016 at 03:11:34PM -0400, Jason Merrill wrote: > commit 761983a023b5217ef831a43f423779940c788ecf > Author: Jason Merrill > Date: Tue Apr 12 13:16:50 2016 -0400 > > gcc/ > * cfgexpand.c (pass_expand::execute): Handle attribute abi_warning. > * expr.c (expand_expr_

Re: [PATCH] Fix ICE caused by nothrow pass (PR c++/70641)

On April 13, 2016 7:34:36 PM GMT+02:00, Jakub Jelinek wrote: >Hi! > >If the nothrow pass (which is GIMPLE, not IPA pass) determines that the >current function can't throw externally, it marks it nothrow. >But it fails to adjust self-recursive calls, which might change because >of >that from possib

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On 04/13/2016 11:54 AM, Jason Merrill wrote: On 04/13/2016 11:32 AM, Jakub Jelinek wrote: On Tue, Apr 12, 2016 at 04:27:48PM -0400, Jason Merrill wrote: Unfortunately, a drawback of doing this in the front end is that it's difficult to warn only about affected cases; the front end doesn't know

libgo patch committed: Update to 1.6.1 release

This patch updates libgo to the 1.6.1 release. This is a minor release that just fixes a couple of bugs with potential security implications (see https://github.com/golang/go/issues?utf8=%E2%9C%93&q=is%3Aopen+is%3Aissue+milestone%3AGo1.6.1 for more details). Bootstrapped and ran Go testsuite on x

[patch] removing aged ifdef 0 from boehm-gc/os_dep.c

This is my first effort at contributing to gcc, so I thought I would try with some of the easy stuff listed here: https://gcc.gnu.org/projects/beginner.html Attached is a patch removing a block which has been #if 0'd out since 2006. I tested the build afterward. I look forward to your feedback.

Re: [PATCH] 69517 - [5/6 regression] SEGV on a VLA with excess initializer elements

On 04/12/2016 12:17 PM, Jason Merrill wrote: On 04/10/2016 07:14 PM, Martin Sebor wrote: The call to build_vla_check() in check_initializer() is to check only explicitly initialized VLAs. The latter function may need to complete the VLA element type and reshape the initializer so the call canno

C++ PATCH for c++/70627 (ICE with opaque enum)

Here the redeclaration of H changed the underlying type from a typedef to a non-typedef, leading to confusion in the back end. We shouldn't change the underlying type from redeclarations. Tested x86_64-pc-linux-gnu, applying to trunk. commit bfabb2dff84a409cf856cda7959ff4395f073d8b Author: Jas

Re: Splitting up gcc/omp-low.c?

On 04/13/2016 07:56 PM, Thomas Schwinge wrote: Best way to present this might be to do diff -du old-omp-low.c . OK, I found Git "-C5%" produce something very similar to that; 0001-Split-up-gcc-omp-low.c-plain.patch.xz attached. That looks much better. However, the //OMPWHATEVER comments are

Re: [PATCH] Make distclean clean up more (PR70173)

On 04/11/2016 09:57 AM, Segher Boessenkool wrote: Currently, distclean does not remove the libcc1, gnattools, and gotools directories, and not the stage_final file either. Fix that. This finishes PR70173. Tested on powerpc64-linux, --enable-languages=all,ada,go,obj-c++ , followed by "make dist

Re: Splitting up gcc/omp-low.c?

On 04/13/2016 06:01 PM, Thomas Schwinge wrote: The attached 0001-Split-up-gcc-omp-low.c.patch.xz is a Git "--color --word-diff --ignore-space-change" patch, purely meant for manual review; I'm intentionally ;-) not attaching a "patch-applyable" patch at this point, to minimize the risk of other

[PATCH] Fix ICE caused by nothrow pass (PR c++/70641)

Hi! If the nothrow pass (which is GIMPLE, not IPA pass) determines that the current function can't throw externally, it marks it nothrow. But it fails to adjust self-recursive calls, which might change because of that from possibly throwing to not being able to throw, so we need to cleanup the EH

Re: Unreviewed patch

On 04/06/2016 04:58 AM, Rainer Orth wrote: The following patch has remainded unreviewed for a week: [testsuite, sparcv9] Fix gcc.dg/ifcvt-4.c on 64-bit SPARC (PR rtl-optimization/68749) https://gcc.gnu.org/ml/gcc-patches/2016-03/msg01631.html Although it's testsuite-only, I'm q

Re: [PATCH] avoid false positives in trailing operator rule in check_GNU_style.sh

On 04/04/2016 11:59 AM, Martin Sebor wrote: My recent change to check_GNU_style.sh (commit 1bbf98) introduces false positives for declarations of functions with a pointer return type. For example, in the following, the asterisk in the "void *" return type is incorrectly diagnosed as a trailing o

Re: [PATCH, rs6000] Add support for int versions of vec_adde

On Wed, Apr 13, 2016 at 10:47 AM, Bill Seurer wrote: > Here is an updated patch: > > > This patch adds support for the signed and unsigned int versions of the > vec_adde altivec builtins from the Power Architecture 64-Bit ELF V2 ABI > OpenPOWER ABI for Linux Supplement (16 July 2015 Version 1.1).

Re: PATCH for c++/70639 (ICE-on-valid with -Wmisleading-indentation and switch)

On 04/13/2016 06:11 PM, David Malcolm wrote: Your approach encapsulates the logic for rejecting this situation within should_warn_for_misleading_indentation, rather than at the callers, which is arguably better for modularity (similar to how we already call it for "do", which is then always rejec

[patch] fix an openacc reduction test for PR68242

As stated in PR68242, the OpenACC libgomp reduction test reduction-2.c has one test case involving a multiplication reduction which overflows and that results in a failure on i686-linux host targets. This patch adjusts the reduction size to prevent the overflow from happening. I've also adjusted re

Re: PATCH for c++/70639 (ICE-on-valid with -Wmisleading-indentation and switch)

On Wed, 2016-04-13 at 17:31 +0200, Marek Polacek wrote: > This is an ICE-on-valid-though-weirdo (TM) problem. We were trying > to warn > about misleading indentation for a switch statement, but > guard_tinfo_to_string > doesn't know what to do with RID_SWITCH and so a crash ensues. > Rather than

Re: Splitting up gcc/omp-low.c?

Hi! On Fri, 08 Apr 2016 11:36:03 +0200, I wrote: > On Thu, 10 Dec 2015 09:08:35 +0100, Jakub Jelinek wrote: > > On Wed, Dec 09, 2015 at 06:23:22PM +0100, Bernd Schmidt wrote: > > > On 12/09/2015 05:24 PM, Thomas Schwinge wrote: > > > >how about we split up gcc/omp-low.c into several > > > >files?

Re: [PATCH, rs6000] Add support for int versions of vec_adde

Segher pointed out a few formatting issues via IRC which I have fixed. On 04/13/16 09:47, Bill Seurer wrote: Here is an updated patch: This patch adds support for the signed and unsigned int versions of the vec_adde altivec builtins from the Power Architecture 64-Bit ELF V2 ABI OpenPOWER ABI f

Re: C PATCH to overhaul warning about dangling else (PR c/70436)

On Wed, Apr 13, 2016 at 05:00:00PM +0200, Bernd Schmidt wrote: > On 04/13/2016 04:14 PM, Marek Polacek wrote: > >I revamped the warning so that it follows what the C++ FE does (i.e. passing > >IF_P bools here and there) and it seems to work quite well. I didn't mean to > >tackle the OMP bits but

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On 04/13/2016 11:32 AM, Jakub Jelinek wrote: On Tue, Apr 12, 2016 at 04:27:48PM -0400, Jason Merrill wrote: Unfortunately, a drawback of doing this in the front end is that it's difficult to warn only about affected cases; the front end doesn't know what's actually going to be emitted, and has t

Re: [PATCH] c++/70594 debug info differences

This patch builds from Jason's proto-patch in comment #26. As Jason discovered, that led to link problems with gt_ggc_mx (tree_node *&) when building cc1. Making 'hash_map' GC-able caused gengtype to put the GC-walkers for tree into constexpr.c. No idea why. That's also what led me to add the u

Re: C PATCH to overhaul warning about dangling else (PR c/70436)

On Wed, Apr 13, 2016 at 05:43:25PM +0200, Marek Polacek wrote: > On Wed, Apr 13, 2016 at 05:09:58PM +0200, Jakub Jelinek wrote: > > On Wed, Apr 13, 2016 at 04:14:03PM +0200, Marek Polacek wrote: > > > PR c/70436 > > > * c-parser.c > > > (c_parser_statement_after_labels): Add IF_P argument and

Re: C PATCH to overhaul warning about dangling else (PR c/70436)

On Wed, Apr 13, 2016 at 05:09:58PM +0200, Jakub Jelinek wrote: > On Wed, Apr 13, 2016 at 04:14:03PM +0200, Marek Polacek wrote: > > PR c/70436 > > * c-parser.c > > (c_parser_statement_after_labels): Add IF_P argument and adjust > > callers. > > Can you put Were you going to say

Re: RFC: C++ PATCH to adjust empty class parameter passing ABI

On Tue, Apr 12, 2016 at 04:27:48PM -0400, Jason Merrill wrote: > Unfortunately, a drawback of doing this in the front end is that it's > difficult to warn only about affected cases; the front end doesn't know > what's actually going to be emitted, and has to warn conservatively, leading > to false

PATCH for c++/70639 (ICE-on-valid with -Wmisleading-indentation and switch)

This is an ICE-on-valid-though-weirdo (TM) problem. We were trying to warn about misleading indentation for a switch statement, but guard_tinfo_to_string doesn't know what to do with RID_SWITCH and so a crash ensues. Rather than teaching it about RID_SWITCH I think this warning can't usefully war

Re: [C++ Patch] PR 70635 ([4.9/5/6 Regression] ICE on (and rejects) valid code...)

On 04/13/2016 11:05 AM, Paolo Carlini wrote: As it happens, anyway, the below exceedingly trivial patch avoids the ICE in resolve_typename_type and appears to pass testing (is already past g++.dg/tm/tm.exp...). Should I apply it instead together with the amended testcase? Weird that nobody notice

[committed, Fortran] PR67039 and PR58000 minor documentation changes

I have committed the following patches to trunk after Jerry’s approval on bugzilla. Is it OK to back port them to the gcc-5 branch? Dominique Index: gcc/fortran/ChangeLog === --- gcc/fortran/ChangeLog (revision 234935) +++ gcc

Re: C/C++ PATCH to add -Wdangling-else option

On Wed, Apr 13, 2016 at 05:13:28PM +0200, Bernd Schmidt wrote: > On 04/13/2016 04:14 PM, Marek Polacek wrote: > >This patch is meant to be applied on top of the "Wparentheses overhaul" > >patch. > > > >I really think that warning about the dangling else problem isn't appropriate > >as a part of th

Re: C/C++ PATCH to add -Wdangling-else option

On 04/13/2016 04:14 PM, Marek Polacek wrote: This patch is meant to be applied on top of the "Wparentheses overhaul" patch. I really think that warning about the dangling else problem isn't appropriate as a part of the -Wparentheses warning, which I think should only deal with stuff like precede

Re: C/C++ PATCH to add -Wdangling-else option

On Wed, Apr 13, 2016 at 04:14:44PM +0200, Marek Polacek wrote: > This patch is meant to be applied on top of the "Wparentheses overhaul" patch. > > I really think that warning about the dangling else problem isn't appropriate > as a part of the -Wparentheses warning, which I think should only deal

Re: C PATCH to overhaul warning about dangling else (PR c/70436)

On Wed, Apr 13, 2016 at 04:14:03PM +0200, Marek Polacek wrote: > PR c/70436 > * c-parser.c > (c_parser_statement_after_labels): Add IF_P argument and adjust > callers. Can you put > @@ -5533,7 +5548,7 @@ c_parser_switch_statement (c_parser *parser) >c_start_case (swit

Re: [C++ Patch] PR 70635 ([4.9/5/6 Regression] ICE on (and rejects) valid code...)

Hi again, On 13/04/2016 16:42, Jason Merrill wrote: On 04/13/2016 03:36 AM, Paolo Carlini wrote: Hi, On 12/04/2016 15:50, Jason Merrill wrote: On 04/12/2016 09:25 AM, Paolo Carlini wrote: in this regression we have an infinite recursion affecting the same_type_p call at parser.c:25125 which

C++ PATCH for c++/70634 (ICE with self-referential initializer)

The compiler was getting into infinite recursion here because when looking at whether 'i>0' was value-dependent, we looked at the initializer for i, which contains sizeof(i), and we asked whether i is value-dependent again. The bug is that we shouldn't care whether the operand of sizeof is va

Re: C PATCH to overhaul warning about dangling else (PR c/70436)

On 04/13/2016 04:14 PM, Marek Polacek wrote: I revamped the warning so that it follows what the C++ FE does (i.e. passing IF_P bools here and there) and it seems to work quite well. I didn't mean to tackle the OMP bits but I bet it would be just a matter of passing IF_P somewhere. Bootstrapped

Re: [PATCH, rs6000] Add support for int versions of vec_adde

Here is an updated patch: This patch adds support for the signed and unsigned int versions of the vec_adde altivec builtins from the Power Architecture 64-Bit ELF V2 ABI OpenPOWER ABI for Linux Supplement (16 July 2015 Version 1.1). There are many of the builtins that are missing and this is the

Re: [C++ PATCH] Avoid code generation differences dependent on DECL_UID gaps (PR c++/70594)

OK. Jason

Re: [C++ Patch] PR 70635 ([4.9/5/6 Regression] ICE on (and rejects) valid code...)

On 04/13/2016 03:36 AM, Paolo Carlini wrote: Hi, On 12/04/2016 15:50, Jason Merrill wrote: On 04/12/2016 09:25 AM, Paolo Carlini wrote: in this regression we have an infinite recursion affecting the same_type_p call at parser.c:25125 which I added in the patch for c++/38313. The issue is that

[C++ PATCH] Avoid code generation differences dependent on DECL_UID gaps (PR c++/70594)

Hi! pop_labels adds LABEL_DECLs into BLOCK_VARS by traversing a hash table indexed by DECL_UID. This is something that is undesirable, if -g vs. -g0 has different DECL_UID values (but still the same order of them), which can happen even without the deletable constexpr tables, then the exact value

C++ PATCH for c++/70615 (ICE with pointer to member)

cp_gimplify_r doesn't always get to see the contents of a CONSTRUCTOR, so it can miss lowering PTRMEM_CSTs there. cp_genericize_r does look into CONSTRUCTORs, and lowering them earlier is fine. Tested x86_64-pc-linux-gnu, applying to trunk. commit fff4c8804414258331785f08435259ecfdb7311c Autho

Re: [PATCH] Don't include DECL_UIDs in tree-sra fancy names for -fdump-final-insns= (PR c++/70594)

On Wed, Apr 13, 2016 at 09:25:56AM +0200, Richard Biener wrote: > So if you think it's not acceptable to drop the relation between > the artificial number used by SRA and the original UID then go > with a hash-map unconditionally. You still have to dump and lookup > the actual relation though - th

Re: C PATCH to overhaul warning about dangling else (PR c/70436)

Looks like Joseph will be on vacation for the next few days so if we want to get this in sooner, someone else will need to review & approve the patch.

Re: C PATCH to overhaul warning about dangling else (PR c/70436)

On Wed, Apr 13, 2016 at 04:14:03PM +0200, Marek Polacek wrote: > I said I'd take care of the C FE part of this PR, so here it is. > > As this PR shows, our dangling else warning regressed in such a way that we > weren't warning about cases such as > > if (A) > for (;;) > if (B) > bar

C/C++ PATCH to add -Wdangling-else option

This patch is meant to be applied on top of the "Wparentheses overhaul" patch. I really think that warning about the dangling else problem isn't appropriate as a part of the -Wparentheses warning, which I think should only deal with stuff like precedence of operators, i.e. things where ()'s are mi

C PATCH to overhaul warning about dangling else (PR c/70436)

I said I'd take care of the C FE part of this PR, so here it is. As this PR shows, our dangling else warning regressed in such a way that we weren't warning about cases such as if (A) for (;;) if (B) bar (); else baz (); The problem was that we were setting FIRST_IF only when an "i

Unreviewed^2 patches

The following two patches have remained unreviewed for a week or (despite a reminder) two: [testsuite, sparcv9] Fix gcc.dg/ifcvt-4.c on 64-bit SPARC (PR rtl-optimization/68749) https://gcc.gnu.org/ml/gcc-patches/2016-03/msg01631.html Compile libcilkrts with -funwind-table

Re: [PATCH] Fix debug ICE on aarch64 due to bad rtl simplification (PR debug/70628)

On 04/12/2016 10:34 PM, Jakub Jelinek wrote: + return temp ? gen_rtx_fmt_ee (GET_CODE (x), to_mode, + temp, XEXP (x, 1)) + : temp; Wrap multi-line expressions in parens. No need for a full retest, just make sure the file still b

Re: [PATCH, i386] Fix operands order in kunpck* insns and corresponding expands

Hi Ilya, On 13 Apr 16:29, Ilya Enkovich wrote: > Hi, > > Current kunpck[hi|si|di] patterns emit operands in a wrong order. This > is compensated by a wrong operands order in vec_pack_trunc_[qi|hi|si] > expands and therefore we get correct code for vectorized loops. Code > using kunpck* intrinsics

[PATCH, i386] Fix operands order in kunpck* insns and corresponding expands

Hi, Current kunpck[hi|si|di] patterns emit operands in a wrong order. This is compensated by a wrong operands order in vec_pack_trunc_[qi|hi|si] expands and therefore we get correct code for vectorized loops. Code using kunpck* intrinsics would be wrong though. This patch fixes operands order an

Re: Fix for PR70498 in Libiberty Demangler

Hi Bernd, > -static long > +static int > d_compact_number (struct d_info *di) > { > - long num; > + int num; >if (d_peek_char (di) == '_') > num = 0; >else if (d_peek_char (di) == 'n') > @@ -2957,7 +2957,7 @@ d_compact_number (struct d_info *di) >else > num = d_number (di

Re: [Patch, testsuite] Require int32plus and scheduling support for some tests

Ping! Regards Senthil Senthil Kumar Selvaraj writes: > Hi, > > This patch add dg-require-effective-target directives to a few tests > that were failing unnecessarily for the AVR target. > > One of them invokes the compiler with -fschedule-insns2 - I've > required scheduling support for t

[PATCH] "Fix" PR70623

The following happens to fix the testcase in PR70623 while in reality I was only cleaning up code, adding comments when I figured that the worklist handling is a) obfuscated and b) visits nodes unnecessarily. b) may be the reason for this bug as the cruical difference appears in the second iterat

Re: [PATCH] Clear TREE_STATIC on vector ctors if needed (PR middle-end/70633)

On Wed, 13 Apr 2016, Jakub Jelinek wrote: > On Wed, Apr 13, 2016 at 09:50:29AM +0200, Richard Biener wrote: > > But I wonder why expansion doesn't simply "re-compute" TREE_STATIC > > here. That is, do we rely on TREE_STATIC constructors for correctness > > purposes as set by frontends? > > Clear

Re: [PATCH] Revert r234572 (aka PR testsuite/70577)

> On April 12, 2016 2:12:14 PM GMT+02:00, Jan Hubicka wrote: > >> Hello. > >> > >> As release managers agreed on IRC, following patch reverts r234572 > >> which introduced new PR testsuite/70577. > >> > >> I've been running bootstrap & regression tests on x86_64-linux-gnu. > >> Ready to be insta

[PATCH] Some tree-vect-data-refs.c cleanup

Hi, the attached patch is a start at cleaning up tree-vect-data-refs.c. I refactored some stuff I found strange and/or difficult to understand. Probably more to come. The misalignment calculation in vect_compute_data_ref_alignment seems quite coarse, is there any work planned for the future? No

[testsuite] gcc-dg: handle all return values when shouldfail is set

Hi, While investigating stability problems when running GCC validations, I fixed DejaGnu to properly handle cases where it cannot parse the testcase output: http://lists.gnu.org/archive/html/dejagnu/2016-04/msg8.html This means that such cases now return "unresolved" which confuses ${tool}_lo

Re: [PATCH] Remove dead code in omp-low.c

Hi! Ping Andi's patch; I just found the same thing: On Sun, 14 Dec 2014 11:34:18 -0800, Andi Kleen wrote: > Since the accelerator changes omp_expand_local is not used anymore. Specifically, and thereabouts. As

Re: [PATCH] Fix debug ICE on aarch64 due to bad rtl simplification (PR debug/70628)

Jakub Jelinek writes: > So here is what I'll throw at testing: > > 2016-04-12 Jakub Jelinek > > PR debug/70628 > * rtl.h (convert_memory_address_addr_space_1): New prototype. > * explow.c (convert_memory_address_addr_space_1): No longer static, > add NO_EMIT argument an

Re: [PATCH] Clear TREE_STATIC on vector ctors if needed (PR middle-end/70633)

On Wed, Apr 13, 2016 at 09:50:29AM +0200, Richard Biener wrote: > But I wonder why expansion doesn't simply "re-compute" TREE_STATIC > here. That is, do we rely on TREE_STATIC constructors for correctness > purposes as set by frontends? Clearly we do rely on it. If we did not rely on it, would t

Re: [PATCH] Clear TREE_STATIC on vector ctors if needed (PR middle-end/70633)

On Tue, 12 Apr 2016, Jakub Jelinek wrote: > Hi! > > As the testcase shows, gimplification of the VECTOR_TYPE CONSTRUCTOR > elements can turn previously valid initializer_constant_valid_p > into a temporary (thus no longer initializer_constant_valid_p). > > The following patch just clears TREE_ST

Re: [C++ Patch] PR 70635 ([4.9/5/6 Regression] ICE on (and rejects) valid code...)

Hi, On 12/04/2016 15:50, Jason Merrill wrote: On 04/12/2016 09:25 AM, Paolo Carlini wrote: in this regression we have an infinite recursion affecting the same_type_p call at parser.c:25125 which I added in the patch for c++/38313. The issue is that for, eg, the testcase at issue, we are passing

Re: [PATCH] Don't include DECL_UIDs in tree-sra fancy names for -fdump-final-insns= (PR c++/70594)

On Tue, 12 Apr 2016, Jakub Jelinek wrote: > On Tue, Apr 12, 2016 at 08:19:57PM +0200, Richard Biener wrote: > > >So you mean add another hash table that maps DECL_UIDs to these SRA > > >counters? Because if we dump there any number of say FIELD_DECL, it > > >would > > >be desirable to use the sam