Re: [PATCH 1/9] gensupport: Fix define_subst operand renumbering.

On 01/14/2016 05:33 PM, Andreas Krebbel wrote: When processing substitutions the operands are renumbered. To find a free operand number the array used_operands_numbers is used to record the operand numbers already in use. Currently this array is used to assign new numbers *before* all the

Re: [PATCH PR68542]

Richard, Here is the second part of patch which really preforms mask stores and all statements related to it to new basic block guarded by test on zero mask. Hew test is also added. Is it OK for trunk? Thanks. Yuri. 2016-01-18 Yuri Rumyantsev PR middle-end/68542 *

Re: [PATCH][ARM] Add movv4hf/v8hf expanders & later insns; disable VnHF immediates.

On 18 January 2016 at 14:12, Kyrill Tkachov wrote: > Hi Alan, > > > On 18/01/16 12:14, Alan Lawrence wrote: >> >> This fixes ICEs on armeb for float16x[48]_t vectors, e.g. in >> check_effective_target_arm_neon_fp_16_ok. >> >> At present, without the expander, moving

C++ PATCH for c++/68586 (rejects-valid with enum in C++11)

In this PR, we find ourselves in a curious situation. When parsing this enum: enum E { x = 1, y = x << 1 }; we process the LSHIFT_EXPR in cp_build_binary_op and call fold_non_dependent_expr on each of the operands. Then fold_non_dependent_expr calls maybe_constant_value which, for

Re: C++ PATCH for c++/68586 (rejects-valid with enum in C++11)

This wouldn't cover cases where this change affects the type or value of more complicated expressions, so my preference would be to clear the caches when we finish_enum_value_list. Jason

Re: genattrab.c generate switch

On Mon, Jan 18, 2016 at 03:15:08PM +0100, Bernd Schmidt wrote: > Secondly, we're currently in a development phase where we only accept bug > fixes for gcc-6. You should resubmit/ping the patch once stage1 opens again. I think this is a bug fix, it is a workaround for a broken compiler that some

[PING^2][PATCH, 3/16] Ignore reduction clause on kernels directive

On 24/11/15 13:21, Tom de Vries wrote: On 09/11/15 16:50, Tom de Vries wrote: On 09/11/15 16:35, Tom de Vries wrote: Hi, this patch series for stage1 trunk adds support to: - parallelize oacc kernels regions using parloops, and - map the loops onto the oacc gang dimension. The patch series

[PATCH] Fix PR69297

The following patch fixes us miscounting the number of scalar instructions for BB vectorization leading to vectorizations that are not profitable. A simple fix is to count each scalar stmt at most once. Bootstrapped and tested on x86_64-unknown-linux-gnu, applied. Richard. 2016-01-18 Richard

Re: [PING^2][PATCH, 3/16] Ignore reduction clause on kernels directive

On Mon, Jan 18, 2016 at 03:24:21PM +0100, Tom de Vries wrote: > >>As discussed here ( > >>https://gcc.gnu.org/ml/gcc-patches/2015-11/msg00785.html ), the kernels > >>directive does not allow the reduction clause. This patch fixes that. > >> > > > > Ping^2. Ok. Jakub

[PING^2][PATCH, 12/16] Handle acc loop directive

On 24/11/15 13:26, Tom de Vries wrote: On 09/11/15 21:06, Tom de Vries wrote: On 09/11/15 16:35, Tom de Vries wrote: Hi, this patch series for stage1 trunk adds support to: - parallelize oacc kernels regions using parloops, and - map the loops onto the oacc gang dimension. The patch series

Re: genattrab.c generate switch

On 18/01/16 15:15, Bernd Schmidt wrote: On 01/13/2016 01:53 AM, Jesper Broge Jørgensen wrote: genattrab.c can generate if statements that have very deep bracket nesting causing clang to produce errors (when target=arm-none-eabi) as explained at https://gcc.gnu.org/ml/gcc/2014-05/msg00032.html

[Committed] Allow pass_parallelize_loops to be run outside the loop pipeline

[ was: Re: [PIING][PATCH, 9/16] Add pass_parallelize_loops_oacc_kernels ] On 14/12/15 16:22, Richard Biener wrote: Can the pass not just use a pass parameter to switch between oacc/non-oacc? It can, and that means that parloops is run outside the loops pipeline. This patch enables that.

[committed] Add oacc_kernels_p argument to pass_parallelize_loops

[was: Re: [PIING][PATCH, 9/16] Add pass_parallelize_loops_oacc_kernels ] On 14/12/15 16:22, Richard Biener wrote: On Sun, Dec 13, 2015 at 5:58 PM, Tom de Vries wrote: On 24/11/15 13:24, Tom de Vries wrote: On 16/11/15 12:59, Tom de Vries wrote: On 09/11/15 20:52,

Re: [PATCH][ARM] Add movv4hf/v8hf expanders & later insns; disable VnHF immediates.

Hi Alan, On 18/01/16 12:14, Alan Lawrence wrote: This fixes ICEs on armeb for float16x[48]_t vectors, e.g. in check_effective_target_arm_neon_fp_16_ok. At present, without the expander, moving v4hf/v8hf values around is done via subregs. On armeb, this ICEs because REG_CANNOT_CHANGE_MODE_P.

Re: [PATCH PR68542]

Thanks Richard. I changed the check on type as you proposed. What about the second back-end part of patch (it has been sent 08.12.15). Thanks. Yuri. 2016-01-18 15:44 GMT+03:00 Richard Biener : > On Mon, Jan 11, 2016 at 11:06 AM, Yuri Rumyantsev

[PING] genattrab.c generate switch

Ping patch: https://gcc.gnu.org/ml/gcc-patches/2016-01/msg00784.html thanks

[committed] Add pass_parallelize_loops to pass_oacc_kernels

[ was: Re: [committed] Add oacc_kernels_p argument to pass_parallelize_loops ] On 18/01/16 14:07, Tom de Vries wrote: [was: Re: [PIING][PATCH, 9/16] Add pass_parallelize_loops_oacc_kernels ] On 14/12/15 16:22, Richard Biener wrote: On Sun, Dec 13, 2015 at 5:58 PM, Tom de Vries

[committed] Add oacc kernels tests in goacc

[ was: Re: [PATCH, 13/16] Add c-c++-common/goacc/kernels-*.c ] On 09/11/15 21:07, Tom de Vries wrote: On 09/11/15 16:35, Tom de Vries wrote: Hi, this patch series for stage1 trunk adds support to: - parallelize oacc kernels regions using parloops, and - map the loops onto the oacc gang

[comitted] Add oacc kernels test in libgomp

[ was: Re: [PATCH, 15/16] Add libgomp.oacc-c-c++-common/kernels-*.c ] On 09/11/15 21:10, Tom de Vries wrote: On 09/11/15 16:35, Tom de Vries wrote: Hi, this patch series for stage1 trunk adds support to: - parallelize oacc kernels regions using parloops, and - map the loops onto the oacc gang

Re: [PATCH PR68542]

On Mon, Jan 18, 2016 at 3:02 PM, Yuri Rumyantsev wrote: > Thanks Richard. > > I changed the check on type as you proposed. > > What about the second back-end part of patch (it has been sent 08.12.15). Can't see it in my inbox - can you reply to the mail with a ping? Thanks,

Re: [PATCH, i386, AVX512] PR target/67895: Fix position of embedded rounding/SAE mode in AVX512 vrangep* and vcvt?si2s* instructions.

Hello, On 15 Jan 15:39, Alexander Fomin wrote: > I've bootstrapped and regtested it against GCC v5 on x86_64-gnu-linux. > OK for 5-branch? Yes, it is ok for gcc-5-branch -- Thanks, K > > -- > Thanks, > Alexander > > On Fri, Oct 09, 2015 at 05:24:56PM +0300, Kirill Yukhin wrote: > > Hello, > >

[PATCH] Fix PR69308

This fixes missing handling of GIMPLE_COND in gimple_could_trap_p[_1]. Bootstrapped and tested on x86_64-unknown-linux-gnu, applied. Richard. 2016-01-18 Richard Biener PR middle-end/69308 * gimple.c (gimple_could_trap_p_1): Handle GIMPLE_COND. Index:

Re: [Patch, fortran] (4/5-regression) PR61831 side-effect deallocation of variable components)

The failures with -m32 are Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Without the closing brace, I get UNRESOLVED: gfortran.dg/derived_constructor_comps_6.f90 -O0 scan-tree-dump-times original "__builtin_free » 33 Dominique > Le 18 janv. 2016 à 13:48,

Re: [PATCH v2] libstdc++: Make certain exceptions transaction_safe.

On Sun, 2016-01-17 at 18:30 -0500, David Edelsohn wrote: > On Sun, Jan 17, 2016 at 3:21 PM, Torvald Riegel wrote: > > On Sat, 2016-01-16 at 15:38 -0500, David Edelsohn wrote: > >> On Sat, Jan 16, 2016 at 8:35 AM, Jakub Jelinek wrote: > >> > On Sat, Jan 16,

Re: genattrab.c generate switch

On 01/13/2016 01:53 AM, Jesper Broge Jørgensen wrote: genattrab.c can generate if statements that have very deep bracket nesting causing clang to produce errors (when target=arm-none-eabi) as explained at https://gcc.gnu.org/ml/gcc/2014-05/msg00032.html At the above link it was suggested that

Re: [hsa merge 08/10] HSAIL BRIG description header file

Hi, On Sat, Jan 16, 2016 at 12:43:07PM +0100, Jakub Jelinek wrote: > On Fri, Jan 15, 2016 at 06:23:05PM +0100, Martin Jambor wrote: > > BRIG_KIND_OPERAND_REGISTER = 0x300a, > > BRIG_KIND_OPERAND_STRING = 0x300b, > > BRIG_KIND_OPERAND_WAVESIZE = 0x3009c, > > BRIG_KIND_OPERAND_END = 0x300d

Re: [hsa merge 09/10] Majority of the HSA back-end

Hi, On Sat, Jan 16, 2016 at 09:58:51AM +0100, Jakub Jelinek wrote: > On Sat, Jan 16, 2016 at 12:49:12AM +0100, Martin Jambor wrote: > > bootstrapping on i686-linux revealed the need for the following simple > > patch. I've run into two types of compilation errors on > > powerpc-ibm-aix (no

Re: [PATCH] Fix the remaining PR c++/24666 blockers (arrays decay to pointers too early)

On 12/25/2015 12:37 PM, Patrick Palka wrote: That alone would not be sufficient because more_specialized_fn() doesn't call maybe_adjust_types_for_deduction() beforehand, yet we have to do the decaying there too (and on both types, not just one of them). And maybe_adjust_types_for_deduction()

Re: [hsa merge 09/10] Majority of the HSA back-end

Hi! PDP endian is gcc_assert (!BYTES_BIG_ENDIAN); gcc_assert (WORDS_BIG_ENDIAN); and 16-bit words, thus within uint16_t it is little endian, and the 16-bit words are ordered in larger units in big endian order. > +#else > + val = ((val & 0xff00ff00) >> 8) | ((val & 0xff00ff) << 8);

Re: [PATCH] Fix RTL DSE (PR rtl-optimization/68955)

On Mon, Jan 18, 2016 at 10:06:44AM +0100, Eric Botcazou wrote: > > The following testcase is miscompiled on i686-linux at -O3. > > The bug is in DSE record_store, which for group_id < 0 uses mem_addr > > set to result of get_addr (base->val_rtx) (plus optional offset), > > which is fine for

Re: [PATCH] Fix a warning in mpx wrappers

2016-01-17 20:53 GMT+03:00 Jakub Jelinek : > Hi! > > The following patch fixes a warning in libmpx: > ../../../../libmpx/mpxwrap/mpx_wrappers.c:492:8: warning: assignment discards > 'const' qualifier from pointer target type [-Wdiscarded-qualifiers] > *d = *s; > ^ >

RE: [PATCH] [ARC] Add basic support for double load and store instructions

> >if (n_pieces >= (unsigned int) (optimize_size ? 3 : 15)) > > return false; > > - if (piece > 4) > > + if (TARGET_LL64 && (piece != 8) && (align >= 4)) > > +piece = 8; > > + else if (piece > 4) > > piece = 4; > >dst_addr = force_offsettable (XEXP (operands[0],

Re: [PATCH] DWARF: add abstract origin links on lexical blocks DIEs

On Sun, Jan 17, 2016 at 9:09 PM, Eric Botcazou wrote: >> Sounds like a good excuse to add a guality for Ada (which has unique >> needs for dwarf). > > Well, the guality testsuite is a pain to maintain so I'd rather not. > The GDB testsuite is clearly the right place for

Re: [PATCH] DWARF: add abstract origin links on lexical blocks DIEs

> But that tests GDB and not GCCs generation of DWARF ... But GDB only consumes the DWARF generated by GCC, it cannot synthetize it. ;-) > which means take the other option of writing a scan-assembler testcase > looking for the previously missing DWARF. Fine with me (either Ada or C as far as

Re: [PING][PATCH] Fix line number that is expected to generate an error.

On Mon, Jan 11, 2016 at 12:28:40PM +0100, Dominik Vogt wrote: > The attached patch fixes a test failure caused by expecting the > error message for the wrong line. Can this be committed? > gcc/testsuite/ChangeLog > > * g++.dg/cpp0x/constexpr-reinterpret1.C: Fix line number that is >

Re: [PING][PATCH] Fix line number that is expected to generate an error.

On Mon, Jan 18, 2016 at 10:53:51AM +0100, Dominik Vogt wrote: > On Mon, Jan 11, 2016 at 12:28:40PM +0100, Dominik Vogt wrote: > > The attached patch fixes a test failure caused by expecting the > > error message for the wrong line. > > Can this be committed? > > > gcc/testsuite/ChangeLog > >

RE: [Patch,tree-optimization]: Add new path Splitting pass on tree ssa representation

-Original Message- From: Jeff Law [mailto:l...@redhat.com] Sent: Saturday, January 16, 2016 12:03 PM To: Ajit Kumar Agarwal; Richard Biener Cc: GCC Patches; Vinod Kathail; Shail Aditya Gupta; Vidhumouli Hunsigida; Nagaraju Mekala Subject: Re: [Patch,tree-optimization]: Add new path

Re: [aarch64] Fix target/69176

> +(define_constraint "Upl" > + "A constraint that matches two uses of add instructions." That's not a particularly helpful description for external users of the compiler. I think that either needs to be sufficiently precise that people who understand the ISA but not the guts of GCC can use it,

Re: [PATCH] Fix RTL DSE (PR rtl-optimization/68955)

> The following testcase is miscompiled on i686-linux at -O3. > The bug is in DSE record_store, which for group_id < 0 uses mem_addr > set to result of get_addr (base->val_rtx) (plus optional offset), > which is fine for canon_true_dependence with other MEMs in that function, > but we also store

Re: Thoughts on memcmp expansion (PR43052)

On Fri, Jan 15, 2016 at 5:58 PM, Bernd Schmidt wrote: > PR43052 is a PR complaining about how the rep cmpsb expansion that gcc uses > for memcmp is slower than the library function. As is so often the case, if > you investigate a bit, you can find a lot of issues with the

Fix PR ada/69219

This is a spurious error on nested subprograms with pragma Inline_Always and Intrinsic, which comes from a thinko in check_inlining_for_nested_subprog, so it's a regression present on the mainline and 5 branch. Tested on x86_64-suse-linux, applied on the mainline and 5 branch. 2016-01-18

Re: [PATCH] Fix RTL DSE (PR rtl-optimization/68955)

> So, do you suggest to tweak get_addr like the patch below, and remove the > mem_addr = get_addr (mem_addr); > line above and the comment? Yes, exactly. And if that doesn't easily work, then go for your solution and add a blurb to the comment explaining why get_addr cannot be easily changed.

[PATCH PR66796]Obvious, revise check condition in test case

Hi, Turns out the check on number of iv_uses is still too large on target hppa. It only supports small offset in REG+offset addressing mode for floating point load/store. Even with this restriction, the grouped version is better than before, so I am going to further relax the check condition

Re: [PATCH] DWARF: add abstract origin links on lexical blocks DIEs

On 01/18/2016 10:45 AM, Eric Botcazou wrote: which means take the other option of writing a scan-assembler testcase looking for the previously missing DWARF. Fine with me (either Ada or C as far as I'm concerned). Thank you for your inputs! I’m going to try that, then. I hope this test will

Re: [PATCH] S/390: Reduce accuracy of bessel_6.f90.

On Mon, Jan 11, 2016 at 03:40:56PM +0100, Dominik Vogt wrote: > Another patch reducing the accuracy required in the bessel_6 test. Can this be committed? > gcc/testsuite/ChangeLog > > * gfortran.dg/bessel_6.f90: Reduce accuracy for S/390. > >From 70a35dd6f6bf906d8e5907667ad0f04f981a61ac

Re: [PATCH 0/2][AArch64] Implement AAPCS64 updates for alignment attribute

> Similarly to ARM, I note that Ada is affected. Indeed, with a gcc 4.9 host > compiler, I saw a bootstrap miscompare iff including Ada; however, I was > able to bootstrap Ada successfully, if I first built a GCC including this > patch with --disable-bootstrap, and then used that as host compiler.

[PATCH] Add fopt-info-oacc

Hi, This patch introduces an option fopt-info-oacc. When using the option like this with a kernels region in kernels-loop.c that parloops does not manage to parallelize: ... $ gcc kernels-loop.c -S -O2 -fopenacc -fopt-info-oacc-all ... we get a message: ... kernels-loop.c:23:9: note: kernels

Re: genattrab.c generate switch

On 18/01/16 14:39, Jesper Broge Jørgensen wrote: No i have not gone through copyright assignment. This is my first time trying to contribute to a GNU project so i have tried following the "Contributing to GCC"@ https://gcc.gnu.org/contribute.html There i followed the advice to run the patch

Re: C++ PATCH for c++/68586 (rejects-valid with enum in C++11)

On 01/18/2016 11:57 AM, Marek Polacek wrote: On Mon, Jan 18, 2016 at 10:04:12AM -0500, Jason Merrill wrote: This wouldn't cover cases where this change affects the type or value of more complicated expressions, so my preference would be to clear the caches when we finish_enum_value_list. So

RE: [Patch,tree-optimization]: Add new path Splitting pass on tree ssa representation

-Original Message- From: Jeff Law [mailto:l...@redhat.com] Sent: Saturday, January 16, 2016 4:33 AM To: Ajit Kumar Agarwal; Richard Biener Cc: GCC Patches; Vinod Kathail; Shail Aditya Gupta; Vidhumouli Hunsigida; Nagaraju Mekala Subject: Re: [Patch,tree-optimization]: Add new path

Re: [PATCH] Add fopt-info-oacc

On 01/18/2016 10:26 AM, Tom de Vries wrote: Hi, This patch introduces an option fopt-info-oacc. When using the option like this with a kernels region in kernels-loop.c that parloops does not manage to parallelize: ... $ gcc kernels-loop.c -S -O2 -fopenacc -fopt-info-oacc-all ... we get a

Re: [PING] genattrab.c generate switch

On 01/18/2016 07:09 AM, Jesper Broge Jørgensen wrote: Ping patch: https://gcc.gnu.org/ml/gcc-patches/2016-01/msg00784.html I'd put it in my gcc-7 queue. But if Richard, Bernd, Richi or someone else wants to work though the changes as a bugfix for bootstrapping on platforms with crippled

Re: [PATCH] PR testsuite/69181: ensure expected multiline outputs is cleared per-test (v2)

On Jan 18, 2016, at 8:14 AM, David Malcolm wrote: > I assumed that these differences were unintentional, so the patch > consolidates things to make the cleanup identical between (A) and (B). I also think this is the right path forward.

Re: [PATCH PR66796]Obvious, revise check condition in test case

On Mon, Jan 18, 2016 at 9:28 AM, Bin Cheng wrote: > Hi, > Turns out the check on number of iv_uses is still too large on target hppa. > It only supports small offset in REG+offset addressing mode for floating > point load/store. Even with this restriction, the grouped

Re: [Patch, fortran] (4/5-regression) PR61831 side-effect deallocation of variable components)

Dear Paul, Sorry for the late feedback. There is a missing right brace in gfortran.dg/derived_constructor_comps_6.f90. This is fixed by the obvious patch: --- ../5_clean/gcc/testsuite/gfortran.dg/derived_constructor_comps_6.f90 2016-01-17 19:27:04.0 +0100 +++

[patch] libstdc++/69293 Fix construction of std::function from null pointer-to-member

The wrong overload of _M_not_empty_function gets chosen and we treat a null pointer-to-member as a valid target. Tested powerpc64le-linux, comitted to trunk. commit c0f055172fb4ceda0257a1a4ccd5f244609a0f37 Author: Jonathan Wakely Date: Mon Jan 18 11:25:43 2016 +

Re: [PATCH][PR tree-optimization/69270] Exploit VRP information in DOM

On Mon, Jan 18, 2016 at 11:38:37AM +, Kyrill Tkachov wrote: > On 18/01/16 11:31, Andreas Schwab wrote: > >Jeff Law writes: > > > >>commit 1384b36abcd52a7ac72ca6538afa2aed2e04f8e0 > >>Author: Jeff Law > >>Date: Fri Jan 15 17:15:24 2016 -0500 >

Re: [PATCH PR68542]

On Mon, Jan 11, 2016 at 11:06 AM, Yuri Rumyantsev wrote: > Hi Richard, > > Did you have anu chance to look at updated patch? diff --git a/gcc/tree-vrp.c b/gcc/tree-vrp.c index acbb70b..208a752 100644 --- a/gcc/tree-vrp.c +++ b/gcc/tree-vrp.c @@ -5771,6 +5771,10 @@

Re: [PATCH 1/4] Make SRA scalarize constant-pool loads

On Fri, Jan 15, 2016 at 11:27 AM, Alan Lawrence wrote: > On 24/12/15 11:53, Alan Lawrence wrote: >> >> Here's a new version that fixes the gcc.dg/guality/pr54970.c failures seen >> on >> aarch64 and powerpc64. Prior to SRA handling constant pool decls, >>

[PATCH][GCC-5] Fix "#pragma GCC pop_options" warning.

Hi there, Can we have the "#pragma GCC pop_options" fix backported to GCC-5? Patch found in https://gcc.gnu.org/ml/gcc-patches/2015-10/msg01261.html and was committed in r228794. The same patch applies cleanly to gcc-5, which would otherwise not be able to use this pragma even though the

[Ada] Housekeeping work in gigi (bis)

In preparation for the implementation of the 'char' compatibility fix. Tested on x86_64-suse-linux, applied on the mainline. 2016-01-18 Eric Botcazou * gcc-interface/gigi.h (build_call_raise_column): Adjust prototype. (build_call_raise_range):

Re: [PATCH][PR tree-optimization/69270] Exploit VRP information in DOM

On 18/01/16 11:49, Jakub Jelinek wrote: On Mon, Jan 18, 2016 at 11:38:37AM +, Kyrill Tkachov wrote: On 18/01/16 11:31, Andreas Schwab wrote: Jeff Law writes: commit 1384b36abcd52a7ac72ca6538afa2aed2e04f8e0 Author: Jeff Law Date: Fri Jan

[PATCH][ARM] Remove neon_reinterpret, use casts

This cleans up the neon_reinterpret code on ARM in a similar way to AArch64. Rather than a builtin backing onto an expander that emits a mov insn, we can just use a cast, because GCC defines casts of vector types as keeping the same bit pattern. On armeb, this fixes previously-failing test:

Re: [patch] libstdc++/69293 Fix construction of std::function from null pointer-to-member

On 18/01/16 11:43 +, Jonathan Wakely wrote: Fix construction of std::function from null pointer-to-member PR libstdc++/69293 I've done it again, that was for 69243 not 69293. I have too many similar PR numbers flying around :-(

[Ada] Housekeeping work in gigi

In preparation for the implementation of the 'char' compatibility fix. Tested on x86_64-suse-linux, applied on the mainline. 2016-01-18 Eric Botcazou * gcc-interface/ada-tree.h (TYPE_IMPLEMENTS_PACKED_ARRAY_P): Rename to (TYPE_IMPL_PACKED_ARRAY_P):

Re: [PATCH 1/2] DWARF: process all TYPE_DECL nodes when iterating on scopes

On Tue, Jan 5, 2016 at 3:32 PM, Pierre-Marie de Rodat wrote: > Hello, > > In Ada, it is possible to have nested subprograms in the following > configuration: > > procedure Parent is >type T; >[...] >procedure Child (Value : T) is >begin >

Re: [PATCH][ARM,AARCH64] target/PR68674: relayout vector_types in expand_expr

On Fri, Jan 8, 2016 at 2:29 PM, Christian Bruel wrote: > When compiling code with attribute targets on arm or aarch64, > vector_type_mode returns different results (eg Vmode or BLKmode) depending > on the current simd flags that are not set between functions. > > for

[PATCH][ARM] Add movv4hf/v8hf expanders & later insns; disable VnHF immediates.

This fixes ICEs on armeb for float16x[48]_t vectors, e.g. in check_effective_target_arm_neon_fp_16_ok. At present, without the expander, moving v4hf/v8hf values around is done via subregs. On armeb, this ICEs because REG_CANNOT_CHANGE_MODE_P. (On arm-*, moving via two subregs is less efficient

Re: Thoughts on memcmp expansion (PR43052)

Hi Bernd, + rtx op0 = force_reg (direct_mode, arg1_rtx); + rtx op1 = force_reg (direct_mode, arg2_rtx); + rtx tem = emit_store_flag (target, NE, op0, op1, +direct_mode, true, false); This is me being ignorant here... wouldn't it be easier to have a

Re: [PATCH][PR tree-optimization/69270] Exploit VRP information in DOM

Jeff Law writes: > commit 1384b36abcd52a7ac72ca6538afa2aed2e04f8e0 > Author: Jeff Law > Date: Fri Jan 15 17:15:24 2016 -0500 > > PR tree-optimization/69270 > * tree-ssanames.c (ssa_name_has_boolean_range): Moved here from >

Re: [PATCH][PR tree-optimization/69270] Exploit VRP information in DOM

On 18/01/16 11:31, Andreas Schwab wrote: Jeff Law writes: commit 1384b36abcd52a7ac72ca6538afa2aed2e04f8e0 Author: Jeff Law Date: Fri Jan 15 17:15:24 2016 -0500 PR tree-optimization/69270 * tree-ssanames.c

[PATCH] Fix PR69337

This fixes ICEs when trying to merge functions and variables. As this is a fatal error in the end no need to try doing sth fancy. Applied as obvious. Richard. 2016-01-18 Richard Biener PR lto/69337 * lto-symtab.c (lto_symtab_merge): Return early for

Re: [Patch, fortran] (4/5-regression) PR61831 side-effect deallocation of variable components)

Hi Dominique, Late or not, thanks for the feedback. I'll fix the right brace. More worrying is the failure with -m32. I presume that the failure with -O0/O1 is at runtime? If not, the correction of the missing right brace is a mysterious trigger for a fault that is optimization dependent. Cheers

Re: [PATCH] c++/58109 - alignas() fails to compile with constant expression

On 01/12/2016 01:11 PM, Martin Sebor wrote: On 01/11/2016 10:20 PM, Jason Merrill wrote: On 12/22/2015 09:32 PM, Martin Sebor wrote: + if (is_attribute_p ("aligned", name) + || is_attribute_p ("vector_size", name)) +{ + /* Attribute argument may be a dependent indentifier. */ +

Re: C++ PATCH to suppress bogus -Wunused warning for parameter packs (PR c++/68965)

On 01/13/2016 05:52 PM, Marek Polacek wrote: On Wed, Jan 13, 2016 at 03:33:27PM -0500, Jason Merrill wrote: On 01/13/2016 11:50 AM, Marek Polacek wrote: So to quash that -Wunused-parameter warning, I decided to set TREE_USED at the place where we create those #xs parameters. Won't that cause

[patch] libstdc++/60637 Fix C++98 std::signbit

This fixes PR60637 by using the appropriate built-in for the size of the argument type. In Bugzilla Marc asked why we don't just use the same code as for C++11, but I want to make this less intrusive change on the branches (trunk is already OK anyway). The new test is only run on x86 because it

Re: Fix pasto in lto-partition.c

On January 18, 2016 4:59:30 PM GMT+01:00, Jan Hubicka wrote: >Hi, >this patch fixes pasto that leads to undefined symbol in the testcase >in PR69003. >Unfortunately i am not sure how to do incremental linking tests in the >testsuite. Only with custom .exp files I guess. >The

[PATCH] libcc1: rerun configure when gcc/BASE-VER changes

* configure.ac (CONFIG_STATUS_DEPENDENCIES): Substitute. * configure: Regenerate. * Makefile.in: Regenerate. diff --git a/libcc1/configure.ac b/libcc1/configure.ac index 6c97afd..e2e3fda 100644 --- a/libcc1/configure.ac +++ b/libcc1/configure.ac @@ -50,6 +50,7 @@

C++/c-common PATCH for c++/68767 (warning regression with ?:)

In this testcase, we weren't getting the benefits of fold's cleverness in handling COND_EXPR because we were only calling fold_for_warn on the condition itself. This patch changes check_function_arguments_recurse to fold the entire COND_EXPR, and also fixes cp_fold to actually fold COND_EXPR.

[PATCH] PR testsuite/69181: ensure expected multiline outputs is cleared per-test (v2)

On Tue, 2016-01-12 at 23:21 -0700, Jeff Law wrote: On 01/12/2016 12:34 PM, David Malcolm wrote: > >> I looked at this code, and there are two near-identical blocks which > >> reset all these variables. You are modifying only one of them, leaving > >> the one inside the if { catch } thing unchanged

Re: C++/c-common PATCH for c++/68767 (warning regression with ?:)

On 01/18/2016 11:06 AM, Jakub Jelinek wrote: On Mon, Jan 18, 2016 at 10:53:41AM -0500, Jason Merrill wrote: In this testcase, we weren't getting the benefits of fold's cleverness in handling COND_EXPR because we were only calling fold_for_warn on the condition itself. This patch changes

Re: [PATCH v2] libstdc++: Make certain exceptions transaction_safe.

On Mon, 2016-01-18 at 14:54 +0100, Torvald Riegel wrote: > On Sun, 2016-01-17 at 18:30 -0500, David Edelsohn wrote: > > On Sun, Jan 17, 2016 at 3:21 PM, Torvald Riegel wrote: > > > On Sat, 2016-01-16 at 15:38 -0500, David Edelsohn wrote: > > >> On Sat, Jan 16, 2016 at 8:35 AM,

Fix pasto in lto-partition.c

Hi, this patch fixes pasto that leads to undefined symbol in the testcase in PR69003. Unfortunately i am not sure how to do incremental linking tests in the testsuite. The patch should work for the release branches, too, Honza Index: ChangeLog

Re: C++/c-common PATCH for c++/68767 (warning regression with ?:)

On Mon, Jan 18, 2016 at 10:53:41AM -0500, Jason Merrill wrote: > In this testcase, we weren't getting the benefits of fold's cleverness in > handling COND_EXPR because we were only calling fold_for_warn on the > condition itself. This patch changes check_function_arguments_recurse to > fold the

Re: [PATCH] PR testsuite/69181: ensure expected multiline outputs is cleared per-test (v2)

So we do need error cleanup for both blocks (A) and (B). gcc/testsuite/ChangeLog: PR testsuite/69181 * gcc.dg/pr69181-1.c: New test file. * gcc.dg/pr69181-2.c: New test file. * lib/gcc-dg.exp (dg-test): Consolidate post-test cleanup of globals by moving

Re: C++ PATCH for c++/68586 (rejects-valid with enum in C++11)

On Mon, Jan 18, 2016 at 10:04:12AM -0500, Jason Merrill wrote: > This wouldn't cover cases where this change affects the type or value of > more complicated expressions, so my preference would be to clear the caches > when we finish_enum_value_list. So like this? Bootstrapped/regtested on

[patch, fortran] PR65996 [5/6 Regression] gfortran ICE with -dH

This patch follows the suggestion Jakub made in the PR and is very straightforward. With the patch, an abort is given on actual errors, in agreement with the documentation for -dH. (Yes, not very useful, but we can clear this PR) Buffered errors bypass this abort by saving and restoring the

Re: [PATCH] ARM PR68620 (ICE with FP16 on armeb)

Thanks for working on this, Christophe, and sorry I missed the PR. You got further in fixing more things than I did though :). A couple of comments: > For the vec_set_internal and neon_vld1_dup patterns, I > switched to an existing iterator which already had the needed > V4HF/V8HF (so I switched

Re: [PATCH] Fix the remaining PR c++/24666 blockers (arrays decay to pointers too early)

On Mon, Jan 18, 2016 at 10:34 AM, Jason Merrill wrote: > On 12/25/2015 12:37 PM, Patrick Palka wrote: >> >> That alone would not be sufficient because more_specialized_fn() >> doesn't call maybe_adjust_types_for_deduction() beforehand, yet we >> have to do the decaying there too

[PATCH] gcc/configure test for AIX DWARF

AIX7 has added support for DWARF to XCOFF, but complete and correct support did not occur with a single update and the initial release of AIX7. The initial support defined a subset of common DWARF debug sections. A later update added most of the remaining sections for location lists and frames,

[PATCH] fix #69253 - [6 Regression] ICE in cxx_incomplete_type_diagnostic initializing a flexible array member with empty string

The attached patch fixes the ICE reported for the test case below: struct str { int a; char s[]; }; void fn1() { (struct str){1, ""}; } While I don't think the patch is incorrect as far as it goes, it's not the last word on the subject of initializing flexible array members. I

Re: [patch, fortran] Inline MATMUL(A,TRANSPOSE(B)), PR 66094

On 01/17/2016 01:44 PM, Thomas Koenig wrote: So... comments? Toon, would this help you? Could yo maybe give this a spin? Thanks, the nightly test at my home computer will build with your patch. 2016-01-17 Thomas Koenig PR fortran/66094 *

[PATCH] Fix debug info handling in prepare_shrink_wrap (PR debug/65779)

Hi! On the following testcase with -mrelocatable on ppc32 we get assembly that contains undefined reference to a local .LC* symbol. The problem is that prepare_shrink_wrap attempts to schedule some instructions from the entry block to later basic blocks, if they set a register that is only used

[PATCH] Fix RTL DSE (PR rtl-optimization/68955, take 2)

On Mon, Jan 18, 2016 at 11:40:23AM +0100, Eric Botcazou wrote: > > So, do you suggest to tweak get_addr like the patch below, and remove the > > mem_addr = get_addr (mem_addr); > > line above and the comment? > > Yes, exactly. And if that doesn't easily work, then go for your solution and >

[PATCH] fix #69251 - [6 Regression] ICE in unify_array_domain on a flexible array member

The attached is a minimal patch to avoid the ICE. The patch doesn't fix the type substitution of flexible array members as that seems more involved and is, strictly speaking, outside the scope of this bug. Type substitution of flexible array is wrong in 5.3.0 (which treats flexible array

Re: [PATCH 1/2] fix memory chunk corruption for opts_obstack (PR jit/68446)

On Fri, Jan 15, 2016 at 03:04:33PM -0500, David Malcolm wrote: > OK for trunk? (assuming it bootstraps by itself) > > gcc/ChangeLog: > PR jit/68446 > * gcc.c (driver::decode_argv): Add call to > init_opts_obstack before init_options_struct. > * opts.c (init_opts_obstack):

-z bndplt documentation in GCC manual

I think the documentation relating to '-z bndplt' in the GCC manual description of -fcheck-pointer-bounds is incorrect. It looks like, as of r225862, the GCC driver is supposed to emit an error message if GCC was configured with a linker that doesn't support this option and you pass -mmpx

Re: reject decl with incomplete struct/union type in check_global_declaration()

On Sat, 16 Jan 2016, Prathamesh Kulkarni wrote: > > There's a GNU C extension allowing forward declarations of enums, and it > > seems that > > > > static enum e x; > > > > doesn't get diagnosed either with -fsyntax-only. Thus I think you should > > cover that case as well. > Done in the

Re: [PATCH 2/4] Equate MEM_REFs and ARRAY_REFs in tree-ssa-scopedtables.c

On Thu, Dec 24, 2015 at 3:55 AM, Alan Lawrence wrote: > This version changes the test cases to fix failures on some platforms, by > rewriting the initializers so that they aren't pushed out to the constant > pool. > > gcc/ChangeLog: > > * tree-ssa-scopedtables.c

[PR66726] Fixe regression caused by Factor conversion out of COND_EXPR

Hi, This is an updated version of https://gcc.gnu.org/ml/gcc-patches/2015-07/msg02196.html. Patch to fix PR66726 missed optimization, factor conversion out of COND_EXPR caused a regression for targets with branch cost greater than i.e., testcase gcc.dg/pr46309.c failed for these targets. I

  1   2   >