The following is the only way I currently see to fix PR67438, register
pressure increase over a conditional due to a match pattern applying
to ops with multiple uses.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied.
Richard.
2015-12-04 Richard Biener
PR middle-end/67438
When we compute excessive byte alignment get_pointer_alignment_1
may end up returning an alignment of zero due to overflow when
multiplying with BITS_PER_UNIT. This in turn causes get_object_alignment
to return a too conservative (byte) alignment.
Fixed as follows.
Bootstrapped and tested on x8
On Thu, 3 Dec 2015, Jakub Jelinek wrote:
> Hi!
>
> As mentioned in the PR, maybe_optimize_range_tests considers basic blocks
> with not just the final GIMPLE_COND (or for last_bb store feeding into PHI),
> but also assign stmts that don't trap, don't have side-effects and where
> the SSA_NAMEs th
On Fri, Dec 04, 2015 at 09:15:25AM +0100, Richard Biener wrote:
> > +modified one, up to and including last_bb, to be executed even if
> > +they would not be in the original program. If the value ranges of
> > +assignment lhs' in those bbs were dependent on the conditions
> > +guar
On Fri, 4 Dec 2015, Jakub Jelinek wrote:
> On Fri, Dec 04, 2015 at 09:15:25AM +0100, Richard Biener wrote:
> > > + modified one, up to and including last_bb, to be executed even if
> > > + they would not be in the original program. If the value ranges of
> > > + assignment lhs' in those bbs we
Hello,
Patch in the bottom fixes spec2k6/437.leslie3d illigal insn generation.
The problem is that for AVX-512F broacasts are allowed to 512b registers only.
[x|y]mm variants are AVX-512VL.
Bootstrapped and regtested.
I'll commit it into GCC main trunk on Monday if no objections.
gcc/
PR
Hi,
> If you haven't set up write-access to the repository, please go ahead and get
> that process started:
>
> https://www.gnu.org/software/gcc/svnwrite.html
>
> You can list me as your sponsor on the form.
>
> Once your account is set up, you can commit patches which have been
> approved.
>
On Thu, 3 Dec 2015, Jan Hubicka wrote:
> > >may lead to wrong code.
> >
> > Can you try generating a testcase?
> > Because with equal vptr and voffset I can't see how that can happen
> > unless some pass extracts information from the pointer types without
> > sanity checking with the pointers
On 4 December 2015 at 03:34, Bin Cheng wrote:
> 2015-12-01 Bin Cheng
>
> * config/aarch64/atomics.md (atomic_store): Use predicate
> aarch64_sync_memory_operand.
>
OK /Marcus
Hi Evandro,
On 03/12/15 20:58, Evandro Menezes wrote:
On 11/20/2015 11:17 AM, James Greenhalgh wrote:
On Tue, Nov 10, 2015 at 11:54:00AM -0600, Evandro Menezes wrote:
2015-11-10 Evandro Menezes
gcc/
* config/aarch64/aarch64-cores.def: Use the Exynos M1 sched model.
Hi all,
The wrong-code in this PR occurs for pre-ARMv5 architectures with Thumb
interworking when trying
to use a static chain. Our output_call_mem function that outputs the assembly
for the call explicitly
clobbers the IP register, which is also used as the static chain register.
Richard sugg
> Indeed we don't do code hoisting yet. Maybe one could trick PPRE
> into doing it.
>
> Note that for OBJ_TYPE_REFs in calls you probably should better use
> gimple_call_fntype instead of the type of the OBJ_TYPE_REF anyway
> (well, fntype will be the method-type, not pointer-to-method-type).
>
Hi all,
We don't handle properly the patterns for the [us]bfiz and [us]bfx instructions
when they
have an extend+ashift form. For example, the
*_ashl pattern.
This leads to rtx costs recuring into the extend and assigning a cost to these
patterns that is too
large.
This patch fixes that overs
On Thu, 3 Dec 2015, Jakub Jelinek wrote:
> Hi!
>
> As mentioned in the PR, GCC 4.7+ seems to have regressed for
> -fstack-protector*, functions containing VLAs and no other arrays are not
> protected anymore. Before 4.7, VLAs were gimplified as __builtin_alloca
> call, which sets ECF_MAY_BE_ALLO
On 3 December 2015 at 12:17, Eric Botcazou wrote:
>> I can understand this restriction, but...
>>
>> > + /* See the same assertion on PROBE_INTERVAL above. */
>> > + gcc_assert ((first % 4096) == 0);
>>
>> ... why isn't this a test that FIRST is aligned to PROBE_INTERVAL?
>
> Because that isn't
On Fri, 4 Dec 2015, Jan Hubicka wrote:
> > Indeed we don't do code hoisting yet. Maybe one could trick PPRE
> > into doing it.
> >
> > Note that for OBJ_TYPE_REFs in calls you probably should better use
> > gimple_call_fntype instead of the type of the OBJ_TYPE_REF anyway
> > (well, fntype will
On Fri, Dec 4, 2015 at 9:27 AM, Kyrill Tkachov wrote:
> Hi all,
>
> The wrong-code in this PR occurs for pre-ARMv5 architectures with Thumb
> interworking when trying
> to use a static chain. Our output_call_mem function that outputs the
> assembly for the call explicitly
> clobbers the IP registe
Hello.
I noticed that Builtins section of documentation does not mention clog10{,f,l}
functions.
I've tried to write a patch, however I'm not sure how should be these functions
described.
Thanks,
Martin
>From 2cb8dfd30ac14e1de00f3788a0b6e55c6a7fa8b9 Mon Sep 17 00:00:00 2001
From: marxin
Date:
If there are no objections, I intent to commit the following patch
tomorrow as obvious.
Tobias
On Tue, Oct 27, 2015 at 11:27:41AM +0100, Tobias Burnus wrote:
> recently, support for ISL 0.15 was added to GCC and also
> ftp://gcc.gnu.org/pub/gcc/infrastructure/ now contains ISL 0.15.
>
> Hence, t
Writing reliable vectorizer testcases is hard - the following factors
out target dependent defaults for tree-reassoc-width for
gcc.dg/vect/pr45752.c
Committed.
Richard.
2015-12-04 Richard Biener
PR testsuite/68681
* gcc.dg/vect/pr45752.c: Add --param tree-reassoc-width=1.
I
A few warning options lack the 'Warning' flag, which since r228094 (2015-09-24)
has the effect that -W(no-)error= doesn't work for them. Additionally,
--help=warnings doesn't work for them either.
Successfully bootstrapped with c,c++,fortran,lto,go,objc,obj-c++,java on
x86-64-gnu-linux & checked w
And now with attached patch ...
On Fri, Dec 04, 2015 at 10:59:29AM +0100, Tobias Burnus wrote:
> A few warning options lack the 'Warning' flag, which since r228094
> (2015-09-24)
> has the effect that -W(no-)error= doesn't work for them. Additionally,
> --help=warnings doesn't work for them eithe
On Thu, Dec 3, 2015 at 9:29 PM, Jeff Law wrote:
> On 12/02/2015 08:35 AM, Richard Biener wrote:
>
>>>
>>> The most interesting side effect, and one I haven't fully analyzed yet is
>>> an
>>> unexpected jump thread -- which I've traced back to differences in what
>>> the
>>> alias oracle is able to
On Thu, Dec 3, 2015 at 10:14 PM, Sebastian Pop wrote:
> Richard Biener wrote:
>> On Wed, Dec 2, 2015 at 10:36 PM, Sebastian Paul Pop
>> wrote:
>> > Do you recommend that we add a gcc_assert that min is always lower than
>> > max?
>>
>> No, min can be one less than max if the array has size zero
> Does the patch mean we inlined __builtin_alloca_with_align ()
> functions? We might run into the issue Eric fixed lately with
> mixing alloca and VLAs (don't see the patch being committed though).
But I'm about to do it (I was waiting for the approval of the aarch64
specific part). Note that
On Thu, Dec 3, 2015 at 6:54 PM, Jeff Law wrote:
> This is something I noticed while working on fixing 67816.
>
> Essentially I was seeing trivially true or trivially false conditionals left
> in the IL for DOM to clean up.
>
> While DOM can and will clean that crud up, but a trivially true or triv
On Fri, Dec 04, 2015 at 10:30:38AM +0100, Richard Biener wrote:
> > The following patch has been bootstrapped/regtested on x86_64-linux and
> > i686-linux.
>
> The patch is ok - it looks like you could have removed the
> __builtin_alloca strcmp with it though.
Ok, will remove the strcmp then.
>
On Fri, 4 Dec 2015, Jakub Jelinek wrote:
> On Fri, Dec 04, 2015 at 10:30:38AM +0100, Richard Biener wrote:
> > > The following patch has been bootstrapped/regtested on x86_64-linux and
> > > i686-linux.
> >
> > The patch is ok - it looks like you could have removed the
> > __builtin_alloca strcmp
On Wed, 2 Dec 2015, Jakub Jelinek wrote:
> On Wed, Dec 02, 2015 at 09:16:10PM +0100, Jan Hubicka wrote:
> > * cfgexpand.c: Include alias.h
>
> Missing full stop at the end.
>
> > (expand_call_stmt, expand_debug_expr): Set no_new_alias_sets;
> > do not fiddle with flag_strict_aliasing
On 12/03/2015 09:33 PM, David Malcolm wrote:
The attached patch updates the handling of %q+D, simplifying
the implementation, and ensuring that it retains the range
information of the decl, giving:
diagnostic-ranges-1.c:6:7: warning: unused variable ‘redundant’
[-Wunused-variable]
int redun
On Fri, 4 Dec 2015, Jan Hubicka wrote:
> Hi,
> this is the last patch of the series. It makes operand_equal_p to compare
> alias sets even in !flag_strict_aliasing before inlining so inlining
> !flag_strict_aliasing to flag_strict_aliasing is possible when callee is
> merged comdat. I tried to
> The patch should bring C++ support for flexible array members closer
> to C (most of the same constructs should be accepted and rejected).
> The only C change in this patch is to include the size of excessively
> large types in diagnostics (I found knowing the size helpful when
> adding tests an
This properly guards the bitmap_mem_desc.get_descriptor_for_instance
call with GATHER_STATISTICS - the hashmap is globally initialized
and thus a query isn't very well inline optimized even if the map
is empty.
Committed as obvious.
Richard.
2015-12-04 Richard Biener
* bitmap.c (bit
> Looks ok to me. OK /Marcus
Thanks. Testing was successful so I have installed it with a small change
(s/ARITH_BASE/ARITH_FACTOR/, it's a bit more mathematically correct).
--
Eric Botcazou
On Mon, Nov 30, 2015 at 2:11 PM, Yuri Rumyantsev wrote:
> Hi All,
>
> Here is a patch for 481.wrf preformance regression for avx2 which is
> sligthly modified mask store optimization. This transformation allows
> perform unpredication for semi-hammock containing masked stores, other
> words if we
As described in pr other/67627, the all-multi target can be
built in parallel with the %_.lo targets which generate make
dependencies that are parsed during the build of all-multi.
gcc -MD does not generate the makefile dependencies in an
atomic way so make can fail if it concurrently parses thos
This patch permits
interface
subroutine sub (x)
real x(10)[*]
end subroutine
end interface
real :: x(100)[*]
call sub (x(10))
end
where one passes an array element ("x(10)") of a contiguous array to a
coarray dummy argument. That's permitted per interpretation
Ping?
On 27 November 2015 at 14:00, Christophe Lyon
wrote:
> Hi,
>
> After the recent commits from Christian adding target attributes
> support for ARM FPU settings, I've noticed that some of the tests
> were failing because of incorrect assumptions wrt to the default
> cpu/fpu/float-abi of the
On Fri, Mar 27, 2015 at 03:04:05PM -0300, Alexandre Oliva wrote:
> This patch reworks the out-of-ssa expander to enable coalescing of SSA
> partitions that don't share the same base name. This is done only when
> optimizing.
>
> The test we use to tell whether two partitions can be merged no long
PING.
-- Forwarded message --
From: H.J. Lu
Date: Mon, Oct 19, 2015 at 1:04 PM
Subject: PING: [PATCH] X86: Optimize access to globals in PIE with copy reloc
To: GCC Patches , Richard Biener
, Jakub Jelinek ,
Richard Henderson
PING.
-- Forwarded message --
Fro
On 05/11/15 21:43, Sebastian Pop wrote:
* graphite-optimize-isl.c (optimize_isl): Call
isl_options_set_schedule_maximize_band_depth.
* gcc.dg/graphite/fuse-1.c: New.
* gcc.dg/graphite/fuse-2.c: New.
* gcc.dg/graphite/interchange-13.c: Remove bogus check.
I pressed "Send" too early - as the testsuite fails unless the following
patch is applied. I think I will just use this test case (with patch)
instead of adding a new test-suite file. Required patch:
--- a/gcc/testsuite/gfortran.dg/coarray_args_2.f90
+++ b/gcc/testsuite/gfortran.dg/coarray_args_2.
Status
==
The GCC 5 branch is open again for regression and documentation fixes.
If nothing unusual happens you can expect GCC 5.4 somewhen closely
before GCC 6 is released.
Quality Data
Priority # Change from last report
--- --
Hi!
While this has been posted after stage1 closed and I'm not really happy
that it missed the deadline, I'm willing to grant an exception, the patch
is small enough that it is ok at this point of stage3. That said, next time
please try to submit new features in time.
Are there any plans for GCC
Hello.
Following patch handles just floating pointer internal functions and
for the rest if displays warning message.
Patch has been installed to the HSA branch.
Thanks,
Martin
>From 5c6581d913c754e4e0a197e073bfa3c17c20b31f Mon Sep 17 00:00:00 2001
From: marxin
Date: Fri, 4 Dec 2015 09:56:11 +0
Hello.
The patch handles builtins mentioned in the email subject and was installed
to the HSA branch.
Thanks,
Martin
>From b56ba10d46c03cadfda16c6658dd3134f5da09f8 Mon Sep 17 00:00:00 2001
From: marxin
Date: Thu, 3 Dec 2015 13:31:28 +0100
Subject: [PATCH 2/3] HSA: implement __builtin_bzero
gcc/
On 12/04/2015 04:41 PM, Jakub Jelinek wrote:
Hi!
While this has been posted after stage1 closed and I'm not really happy
that it missed the deadline, I'm willing to grant an exception, the patch
is small enough that it is ok at this point of stage3. That said, next time
please try to submit new
Hi Ilya,
On 02 Dec 16:51, Ilya Enkovich wrote:
> Hi,
>
> This patch fixes wrong alignment check in _store_mask
> pattern. Currently we check a register operand instead of a memory
> one. This fixes segfault on 481.wrf compiled at -O3 for KNL target.
> I bootstrapped and tested this patch on x86_
The recently added test graphite/id-28.c requires pthreads due to its use of
-fcilkplus. Committed as obvious.
nathan
2015-12-04 Nathan Sidwell
* gcc.dg/graphite/id-28.c: Requires pthreads.
Index: testsuite/gcc.dg/graphite/id-28.c
==
> + (unspec_volatile:PTR [(match_operand:PTR 1 "register_operand" "0")
> + (match_operand:PTR 2 "register_operand" "r")]
> +UNSPEC_PROBE_STACK_RANGE))]
Minor nit. Since this is used in an unspec_volatile, the name should be
UNSPECV_ and de
The PTX backend has to emit PTX function prototypes and prologues in addition to
the regular argument marshalling machinery. And of course these all have to
agree with each other. The first two are done by two different pieces of code,
This patch changes things so that prototype and prologue
Hello,
Patch in the bottom fixes miscompare issue on Spec2k6/434.zeus.
Bootstrapped & reg-tested.
If no objections - I'll commit it into GCC main trunk on Monday.
gcc/
PR target/68633
* config/i386/sse.md (define_insn "kunpckhi"): Fix operands order.
(define_insn "kunpcksi
Next version of the patch with the changes discussed internally.
(Sorry for the version number; I now use my internam branch
numbers for sake of clarity.)
Ciao
Dominik ^_^ ^_^
--
Dominik Vogt
IBM Germany
gcc/ChangeLog
* config/s390/s390.opt (s390_arch_string): Remove.
(s390_t
On Thu, 2015-12-03 at 17:17 -0500, Jason Merrill wrote:
> On 12/03/2015 04:43 PM, David Malcolm wrote:
> > On Thu, 2015-12-03 at 15:33 -0500, Jason Merrill wrote:
> >> On 12/03/2015 09:55 AM, David Malcolm wrote:
> >>> This patch adds bulletproofing to detect purged tokens, and avoid using
> >>> th
"register_operand" was a bit confusing. Also some other minor cleanups.
Tested on powerpc64-linux; okay for mainline?
Segher
2015-12-04 Segher Boessenkool
* (cstore4_unsigned): Use gpc_reg_operand instead of
register_operand. Remove empty constraints. Use std::swap.
On 12/04/2015 03:14 PM, Dominik Vogt wrote:
> Next version of the patch with the changes discussed internally.
> (Sorry for the version number; I now use my internam branch
> numbers for sake of clarity.)
Applied. Thanks!
-Andreas-
Ping.
This almost fell through the cracks.
https://gcc.gnu.org/ml/gcc-patches/2015-10/msg00055.html
Thanks,
Kyrill
On 01/10/15 14:00, Kyrill Tkachov wrote:
Hi all,
As part of the SWITCHABLE_TARGET work I inadvertently changed the behaviour of
-mgeneral-regs-only with respect to the .arch dire
On 09/25/2015 04:02 PM, Dominik Vogt wrote:
> On Fri, Sep 25, 2015 at 02:59:41PM +0100, Dominik Vogt wrote:
>> The following set of two patches implements the function
>> __attribute__ ((target("..."))) and the corresponding #pragma GCC
>> target("...") on S/390. It comes with certain limitations:
On Fri, Dec 4, 2015 at 9:34 AM, Segher Boessenkool
wrote:
> "register_operand" was a bit confusing. Also some other minor cleanups.
>
> Tested on powerpc64-linux; okay for mainline?
>
>
> Segher
>
>
> 2015-12-04 Segher Boessenkool
>
> * (cstore4_unsigned): Use gpc_reg_operand instead o
On 12/04/2015 02:23 AM, Dominik Vogt wrote:
gcc/testsuite/ChangeLog
* gcc.dg/Wframe-address.c: S/390 requires the -mbackchain option to
access arbitrary stack frames.
* gcc.dg/Wno-frame-address.c: Likewise.
Applied. Thanks!
-Andreas-
This implements cstore for the last case we do not yet handle, using
the superopt algo from the venerable CWG. The only integer cases we
do still not handle after this are for -m32 -mpowerpc64. Those case
now generate a branch sequence, which is better than what we had
before.
Tested on powerpc6
On Fri, Dec 4, 2015 at 9:34 AM, Segher Boessenkool
wrote:
> This implements cstore for the last case we do not yet handle, using
> the superopt algo from the venerable CWG. The only integer cases we
> do still not handle after this are for -m32 -mpowerpc64. Those case
> now generate a branch seq
On 04/12/15 14:40, Kyrill Tkachov wrote:
Ping.
This almost fell through the cracks.
https://gcc.gnu.org/ml/gcc-patches/2015-10/msg00055.html
Thanks,
Kyrill
On 01/10/15 14:00, Kyrill Tkachov wrote:
Hi all,
As part of the SWITCHABLE_TARGET work I inadvertently changed the
behaviour of -mgeneral
On 02 Dec 16:27, Richard Biener wrote:
> On Wed, Dec 2, 2015 at 4:24 PM, Ilya Enkovich wrote:
> >
> > The problem is that conversion is supposed to be handled by
> > vectorizable_conversion,
> > but it fails to because it is not actually a conversion. I suppose it
> > may be handled
> > in vectori
Hi Richard.
Thanks a lot for your review.
Below are my answers.
You asked why I inserted additional check to
++ b/gcc/tree-ssa-forwprop.c
@@ -373,6 +373,11 @@ combine_cond_expr_cond (gimple *stmt, enum
tree_code code, tree type,
gcc_assert (TREE_CODE_CLASS (code) == tcc_comparison);
+ /* Do
On 17/11/15 16:30, Andre Vieira wrote:
On 17/11/15 12:29, Bernd Schmidt wrote:
On 11/16/2015 04:48 PM, Andre Vieira wrote:
On 16/11/15 15:34, Joern Wolfgang Rennecke wrote:
I just happened to stumble on this problem with another port.
The volatile & test solution doesn't work, though.
What do
On 27/11/15 08:30, Richard Biener wrote:
This is part 1 of a fix for PR68533 which shows that some targets
cannot can_vec_perm_p on an identity permutation. I chose to fix
this in the vectorizer by detecting the identity itself but with
the current structure of vect_transform_slp_perm_load this
On 12/03/2015 05:08 PM, David Malcolm wrote:
On Thu, 2015-12-03 at 15:38 -0500, Jason Merrill wrote:
On 12/03/2015 09:55 AM, David Malcolm wrote:
Testcase g++.dg/template/ref3.C:
1// PR c++/28341
2
3template struct A {};
4
5template st
On December 4, 2015 4:32:33 PM GMT+01:00, Alan Lawrence
wrote:
>On 27/11/15 08:30, Richard Biener wrote:
>>
>> This is part 1 of a fix for PR68533 which shows that some targets
>> cannot can_vec_perm_p on an identity permutation. I chose to fix
>> this in the vectorizer by detecting the identity
On Thu, 3 Dec 2015, Martin Sebor wrote:
> The only C change in this patch is to include the size of excessively
> large types in diagnostics (I found knowing the size helpful when
> adding tests and I think it might be helpful to others as well).
I don't see what that C change has to do with flex
On 12/04/2015 03:19 AM, Richard Biener wrote:
On Thu, Dec 3, 2015 at 6:54 PM, Jeff Law wrote:
This is something I noticed while working on fixing 67816.
Essentially I was seeing trivially true or trivially false conditionals left
in the IL for DOM to clean up.
While DOM can and will clean tha
The PowerPC port provides reciprocal sqrt but doesn't implement the
extra incantation to utilize it for sqrtf.
The current implementation re-associates terms in the N-R iteration to
utilize one constant instead of two, but does not provide a
pre-computed estimate multiplied by the source, which re
Hi!
GCC changed recently to disallow -Werror= on W options that don't have
Warning keyword set on them, the following patch fixes 3 spots where
Warning has been unintentionally omitted. The PR also mentions -Wpsabi,
but I believe -Werror=psabi is not appropriate, because -Wpsabi is not
really a w
I've committed the obvious patch after confirmation from
the original author that it is correct.
2015-12-04 Steven G. Kargl
PR fortran/68684
* resolve.c (resolve_lock_unlock_event): Fix logic faux pas.
Index: resolve.c
==
On Fri, 2015-12-04 at 11:01 -0500, Jason Merrill wrote:
> On 12/03/2015 05:08 PM, David Malcolm wrote:
> > On Thu, 2015-12-03 at 15:38 -0500, Jason Merrill wrote:
> >> On 12/03/2015 09:55 AM, David Malcolm wrote:
> >>> Testcase g++.dg/template/ref3.C:
> >>>
> >>>1 // PR c++/28341
> >>>
Hi!
As mentioned in the PR, the process_options diagnostics is about errors
on the command line (incompatible options, unsupported options etc.),
which aren't really related to any source code location.
This patch fixes it to use explicit UNKNOWN_LOCATION, instead of
explicit or implicit input_loc
>
> I wonder if you can split out the re-naming at this stage. Further
> comments below.
OK, I will commit the renaming and ipa-icf fix separately.
>
> > Bootstrapped/regtested x86_64-linux, OK?
> >
> > I will work on some testcases for the ICF and fold-const that would lead
> > to wrong code
> On Fri, 4 Dec 2015, Jan Hubicka wrote:
>
> > > Indeed we don't do code hoisting yet. Maybe one could trick PPRE
> > > into doing it.
> > >
> > > Note that for OBJ_TYPE_REFs in calls you probably should better use
> > > gimple_call_fntype instead of the type of the OBJ_TYPE_REF anyway
> > > (we
I think marking stuff with Warning as appropriate qualifies as obvious.
On 12/04/2015 05:37 PM, Jakub Jelinek wrote:
+ /* If the switch takes an integer, convert it. */
+ if (arg && cl_options[opt_index].cl_uinteger)
+ {
+ value = integral_argument (arg);
+
On 12/03/2015 09:55 AM, David Malcolm wrote:
@@ -362,10 +362,11 @@ convert_to_real_1 (tree type, tree expr, bool fold_p)
case REAL_TYPE:
/* Ignore the conversion if we don't need to store intermediate
results and neither type is a decimal float. */
- return build1 ((f
On 12/04/2015 11:45 AM, David Malcolm wrote:
On Fri, 2015-12-04 at 11:01 -0500, Jason Merrill wrote:
On 12/03/2015 05:08 PM, David Malcolm wrote:
On Thu, 2015-12-03 at 15:38 -0500, Jason Merrill wrote:
On 12/03/2015 09:55 AM, David Malcolm wrote:
Testcase g++.dg/template/ref3.C:
1
On 12/04/2015 05:45 PM, Jakub Jelinek wrote:
This patch fixes it to use explicit UNKNOWN_LOCATION, instead of
explicit or implicit input_location, which for most of process_options
is somewhere on line 1 of the main source file.
Ok.
Bernd
Version 6 with another fix. This should work now.
Ciao
Dominik ^_^ ^_^
--
Dominik Vogt
IBM Germany
gcc/ChangeLog
* config/s390/s390.md ("movstr", "*movstr"): Fix warning.
("movstr"): New indirect expanders used by "movstr".
gcc/testsuite/ChangeLog
* gcc.target/s390/
Version 5 with the latest requested changes. Seems to work now.
I've dropped the extra patch and rather marked the failing tests
as "xfail".
Ciao
Dominik ^_^ ^_^
--
Dominik Vogt
IBM Germany
gcc/ChangeLog
* config/s390/s390.c (s390_expand_setmem): Use new expanders.
* config/
On 12/04/2015 03:12 AM, Richard Biener wrote:
On Thu, Dec 3, 2015 at 9:29 PM, Jeff Law wrote:
On 12/02/2015 08:35 AM, Richard Biener wrote:
The most interesting side effect, and one I haven't fully analyzed yet is
an
unexpected jump thread -- which I've traced back to differences in what
the
On December 4, 2015 5:20:52 PM GMT+01:00, Jeff Law wrote:
>On 12/04/2015 03:19 AM, Richard Biener wrote:
>> On Thu, Dec 3, 2015 at 6:54 PM, Jeff Law wrote:
>>> This is something I noticed while working on fixing 67816.
>>>
>>> Essentially I was seeing trivially true or trivially false
>conditiona
Hi Guys,
I am applying the attached patch to reduce the number of multilibs for
the MSP430 target. This is at the request of TI, on behalf of their
customers, who complained that the toolchain was too large. The patch
only affects MSP430 specific files, and parts of files. It does not
On Fri, Dec 4, 2015 at 2:41 PM, Jakub Jelinek wrote:
> Hi!
>
> While this has been posted after stage1 closed and I'm not really happy
> that it missed the deadline, I'm willing to grant an exception, the patch
> is small enough that it is ok at this point of stage3. That said, next time
> please
On Fri, Dec 4, 2015 at 2:45 PM, Yury Gribov wrote:
> On 12/04/2015 04:41 PM, Jakub Jelinek wrote:
>>
>> Hi!
>>
>> While this has been posted after stage1 closed and I'm not really happy
>> that it missed the deadline, I'm willing to grant an exception, the patch
>> is small enough that it is ok at
On 12/04/2015 04:18 PM, Andre Vieira wrote:
Reworked following Joern's suggestion.
Is this OK?
Yes.
Bernd
This patch ensures that this_thread::get_id() returns a value that is
distinct from the "not a thread" value, and avoids undefined behaviour
in pthread_equal.
Previously programs using glibc but not linking to libpthread would
get the "not a thread" value for std::this_thread::get_id() in main.
W
On Fri, Dec 04, 2015 at 06:32:38PM +0100, Dmitry Vyukov wrote:
> +2015-12-04 Dmitry Vyukov
> +
> + * sancov.c: New file.
> + * Makefile.in (OBJS): Add sancov.o.
> + * invoke.texi (-fsanitize-coverage=trace-pc): Describe.
> + * passes.def (sancov_pass): Add.
> + * tree-pass.h
On 04/12/15 16:04, Richard Biener wrote:
> On December 4, 2015 4:32:33 PM GMT+01:00, Alan Lawrence
> wrote:
>> On 27/11/15 08:30, Richard Biener wrote:
>>>
>>> This is part 1 of a fix for PR68533 which shows that some targets
>>> cannot can_vec_perm_p on an identity permutation. I chose to fix
On 12/03/2015 11:42 PM, Martin Sebor wrote:
+ if (next && TREE_CODE (next) == FIELD_DECL)
This will break if there's a non-field between the array and the next field.
@@ -4114,7 +4115,10 @@ walk_subobject_offsets (tree type,
/* Avoid recursing into objects that are not interes
On 25/11/2015 14:55, Paolo Bonzini wrote:
> Left shifts into the sign bit is a kind of overflow, and the
> standard chooses to treat left shifts of negative values the
> same way.
>
> However, the -fwrapv option modifies the language to one where
> integers are defined as two's complement---whic
On Fri, Dec 04, 2015 at 06:01:58PM +0100, Bernd Schmidt wrote:
> I think marking stuff with Warning as appropriate qualifies as obvious.
>
> On 12/04/2015 05:37 PM, Jakub Jelinek wrote:
> >+ /* If the switch takes an integer, convert it. */
> >+ if (arg && cl_options[opt_index].cl_uinte
The Cilk+ code bizarrely parses function-trailing attributes in a
different place from normal operation, and tells the compiler that they
were C++11 attributes when they were actually specified with the GNU
syntax, which changes what they apply to. This patch fixes the second
issue, which is e
On Dec 4, 2015, at 5:13 AM, Alan Lawrence wrote:
> On 05/11/15 21:43, Sebastian Pop wrote:
>>* graphite-optimize-isl.c (optimize_isl): Call
>>isl_options_set_schedule_maximize_band_depth.
>>
>>* gcc.dg/graphite/fuse-1.c: New.
>>* gcc.dg/graphite/fuse-2.c: New.
>>
Hi,
this is the patch implementing renaming and fixing logic in ipa-icf/lto-symtab
WRT compuattion of this flag. I don't seem to be able to construct testcase for
this: the merged flag is currently used in inliner only to decide whether to
ingore optimize_size/optimize levels which should not lead
Hi,
this is the patch for fold-const.c. Can you think of some testcase for the
MR_DEPENDENCE_CLIQUE comparsion? I am not that familiar with the code to
be able to construct it :(
Bootstrapped/regtested x86_64-linux, OK?
Honza
* fold-const.c (operand_equal_p): Do not use flag_strict_alias
1 - 100 of 135 matches
Mail list logo
