Hi Naveen,
On 15/03/17 05:32, Hurugalawadi, Naveen wrote:
Hi James,
My reason for asking is that the instruction fusion implemented in LLVM
( lib/Target/AArch64/AArch64MacroFusion.cpp::shouldScheduleAdjacent )
Sorry. There seems to be some confusion in the branch instructions.
The branch shou
Hi Martin,
> This is a small follow-up patch, where local.local flag should be also dropped
> for a default implementation.
>
> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
>
> Ready to be installed?
the testcase causes
WARNING: profopt.exp does not support dg-error
I
Huh, I forgot to attach the patch.
Martin
>From bb89456e6cecfa9497cf8e265d2083e762d5bc3e Mon Sep 17 00:00:00 2001
From: marxin
Date: Mon, 27 Feb 2017 14:07:03 +0100
Subject: [PATCH] Introduce IntegerRange for options (PR driver/79659).
gcc/ChangeLog:
2017-02-28 Martin Liska
PR driver/7965
Hi Kyrill,
>> I suggest you reword the whole comment and not talk about particular CPUs
>> but rather about the kinds of instructions you want to fuse
Modified as per the comments. Had modified the earlier version of patch
which had the vulcan reservation before James comments.
Please find attac
On 03/15/2017 10:30 AM, Rainer Orth wrote:
Hi Martin,
This is a small follow-up patch, where local.local flag should be also dropped
for a default implementation.
Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
Ready to be installed?
the testcase causes
WARNING: prof
On 03/15/2017 12:58 AM, Ilya Enkovich wrote:
2017-03-13 16:33 GMT+03:00 Martin Liška :
On 03/13/2017 02:07 PM, Richard Biener wrote:
No, that can't happen. I said that for example for
struct S { ... } s;
foo (s);
pass_by_reference may be true but on gimple you see a struct s as
actual argume
On 03/14/2017 11:27 PM, Ilya Enkovich wrote:
2017-03-10 16:15 GMT+03:00 Martin Liška :
Hello.
Currently, __builtin_ia32_bndret is used for all functions that have non-void
return type. I think the right fix is to return bounds just for a bounded type.
Patch can bootstrap on x86_64-linux-gnu an
On 03/14/2017 06:31 PM, Matthias Klose wrote:
On 14.03.2017 15:15, Richard Biener wrote:
On Tue, Mar 14, 2017 at 1:30 PM, Martin Liška wrote:
Tested on my local machine that's properly installed.
Ready for trunk?
Ok.
Richard.
shouldn't that go to the active branches as well?
Matthias
On 03/15/2017 04:12 AM, Jeff Law wrote:
> On 03/01/2017 04:36 AM, JonY wrote:
>> Patch OK?
>>
>> ChangeLog:
>> * unwind-seh.c: Suppress warnings for RtlUnwindEx() calls.
> You know this stuff better than anyone else working with GCC. If you
> think this is the right thing to do for the SEH cod
> First, I agree that the less formal language is becoming more
> acceptable. Some style guides explicitly allow contractions,
> but others advise against them. The technical specifications
> that significant parts of GCC aim to conform to, and those I
> happen to work with the most closely (the
We set ENUM_LOC, and then overwrite it without ever using it. Remove the line.
Bootstrapped/regtested on x86_64-linux, applying to trunk.
2017-03-15 Marek Polacek
* c-parser.c (c_parser_enum_specifier): Remove redundant line.
diff --git gcc/c/c-parser.c gcc/c/c-parser.c
index 46883e
2017-03-15 Uros Bizjak
PR target/80019
* config/i386/i386.c (ix86_vector_duplicate_value): Create
subreg of inner mode for values already in registers.
testsuite/ChangeLog:
2017-03-15 Uros Bizjak
PR target/80019
* gcc.target/i386/pr80019.c: New test.
Bootstrapped and
This fixes a failure in the testsuite. When we transform the doloop
pattern, the decrement of the old iteration register goes away, which is
a problem if it's used after the loop (where it should have the value zero).
Committed.
Bernd
* config/c6x/c6x.c (predicate_insn): Avoid rtl sharing fa
gcc/ChangeLog:
2017-03-13 Martin Liska
PR ipa/79857
* cgraph.c (cgraph_edge::verify_count_and_frequency):
Replace error with internal_error_cont.
(cgraph_node::verify_node): Likewise.
---
gcc/cgraph.c | 139 --
gcc/ChangeLog:
2017-03-13 Martin Liska
PR rtl-optimization/79856
* cfgrtl.c (try_redirect_by_replacing_jump): Fix GNU coding
style.
(fixup_partition_crossing): Likewise.
(fixup_new_cold_bb): Likewise.
(emit_barrier_after_bb): Likewise.
(f
On Tue, Mar 14, 2017 at 10:05 AM, Janne Blomqvist
wrote:
> On Tue, Mar 14, 2017 at 6:32 AM, NightStrike wrote:
>> On Mon, Mar 13, 2017 at 5:01 AM, Janne Blomqvist
>> wrote:
>>> On Sun, Mar 12, 2017 at 10:46 PM, Janne Blomqvist
>>> wrote:
On Sun, Mar 12, 2017 at 7:26 PM, NightStrike wrote:
On March 15, 2017 11:19:24 AM GMT+01:00, "Martin Liška" wrote:
>On 03/14/2017 06:31 PM, Matthias Klose wrote:
>> On 14.03.2017 15:15, Richard Biener wrote:
>>> On Tue, Mar 14, 2017 at 1:30 PM, Martin Liška
>wrote:
Tested on my local machine that's properly installed.
Ready for trun
Hello.
Following small series exclude some strings from pot files via
introduction of internal_error_cont function.
Patch can bootstrap and survives regression tests.
(Sorry for broken threading because my cover letter was rejected).
Martin
marxin (3):
Introduce internal_error_cont function
On 03/15/2017 12:09 AM, Bernd Schmidt wrote:
I'll retest with your
suggestion and with the bitmap creation conditional on i1 being nonnull.
Like this (also had to throw in a bitmap_empty_p). Retested as before. Ok?
Bernd
Index: gcc/combine.c
=
On 03/15/2017 04:00 PM, Bernd Schmidt wrote:
On 03/15/2017 12:09 AM, Bernd Schmidt wrote:
I'll retest with your
suggestion and with the bitmap creation conditional on i1 being nonnull.
Like this (also had to throw in a bitmap_empty_p). Retested as before. Ok?
Oops, that one also has dbg_cnt
Hi,
I think the patch isn't quite complete yet. You will also need changes in
generic code. Currently sched_macro_fuse_insns() does:
if (any_condjump_p (insn))
{
unsigned int condreg1, condreg2;
rtx cc_reg_1;
targetm.fixed_condition_code_regs (&condreg1, &condreg2);
It was XFAILed because there was a bug in glibc, but that bug was fixed
nine years ago. Nowadays everyone uses a version of glibc with the bug
fixed, so we should no longer XFAIL the test.
This fixes it. Testing on powerpc64-linux {-m32,-m64}, will commit if
that succeeds.
Segher
2017-03-15
On 10/03/17 15:26, Segher Boessenkool wrote:
On Fri, Mar 10, 2017 at 01:57:31PM +0100, Rainer Orth wrote:
I just noticed that nothing has happened at all in a month, so anything
is better than the tests XPASSing on a number of targets.
So the patch is ok for mainline with sparc*-*-* added to th
I'm committing this patch to fix a typo in /* Define if ... */ comment
that ends up in config.h
This was submitted as a pull request against the unofficial github
mirror: https://github.com/gcc-mirror/gcc/pull/10
* acinclude.m4 (GLIBCXX_CHECK_S_ISREG_OR_S_IFREG): Fix typo in
comm
Hi,
The gcc.dg/pic-2.c and gcc.dg/pie-2.c tests are failing for MIPS targets
because __PIC__ is always set to 1 for MIPS.
This patch makes the testsuite skip those two tests for all MIPS targets.
Tested with mips-mti-elf and mips-mti-linux-gnu.
Should I have fixed this in target-supports.exp in
Hi Catherine,
What is your opinion on this patch? I am OK with the temporary
workaround on the basis that the additional nop is successfully
eliminated so there is no code size increase. Also, I am happy
enough that the CFG is fixed very soon after the block move is
expanded so the 'bad' basic blo
Hi Jiong,
>> Subject: [PATCH] testsuite, 79356
>>
>> As stated in the PR (and elsewhere), this test now passes on aarch64,
>> ia64, mips, powerpc, sparc, and s390x. This patch disables the xfails
>> for those targets.
>>
>>
>> gcc/testsuite/
>> PR testsuite/79356
>> * gcc.dg/attr-alloc_
Some recent changes.
Index: htdocs/gcc-7/changes.html
===
RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-7/changes.html,v
retrieving revision 1.71
diff -u -r1.71 changes.html
--- htdocs/gcc-7/changes.html 14 Mar 2017 19:42:38 - 1.71
+++ ht
On 15/03/17 15:34, Rainer Orth wrote:
Hi Jiong,
Subject: [PATCH] testsuite, 79356
As stated in the PR (and elsewhere), this test now passes on aarch64,
ia64, mips, powerpc, sparc, and s390x. This patch disables the xfails
for those targets.
gcc/testsuite/
PR testsuite/79356
On 03/15/2017 05:00 AM, Richard Kenner wrote:
First, I agree that the less formal language is becoming more
acceptable. Some style guides explicitly allow contractions,
but others advise against them. The technical specifications
that significant parts of GCC aim to conform to, and those I
happ
Toma Tabacu writes:
> The gcc.dg/pic-2.c and gcc.dg/pie-2.c tests are failing for MIPS targets
> because __PIC__ is always set to 1 for MIPS.
>
> This patch makes the testsuite skip those two tests for all MIPS
> targets.
>
> Tested with mips-mti-elf and mips-mti-linux-gnu.
>
> Should I have fi
Hi All,
This fixes a bug in the scalar version of copysign where due to a subreg
were generating less than efficient code.
This patch replaces
return x * __builtin_copysignf (150.0f, y);
which used to generate
adrpx1, .LC1
mov x0, 2147483648
ins v3.d[0],
Hi All,
This patch fixes a bug in the SIMD mov pattern where we were doing an
insert and leaving the rest of the vector in an undetermined state. This
could cause a bug if something else expects the other lanes to be 0.
Semantically we wanted a mov here and not an insert. This is in line
with o
On 03/15/2017 09:40 AM, Martin Sebor wrote:
On 03/15/2017 05:00 AM, Richard Kenner wrote:
First, I agree that the less formal language is becoming more
acceptable. Some style guides explicitly allow contractions,
but others advise against them. The technical specifications
that significant par
On Wed, 15 Mar 2017, Martin Liška wrote:
> gcc/po/ChangeLog:
>
> 2017-03-15 Martin Liska
>
> * exgettext: Exclude internal_error_cont from location.
If you don't want a function's arguments to be extracted for translation,
you should not special-case it in exgettext. Instead:
* Ensu
On 03/15/2017 09:00 AM, Bernd Schmidt wrote:
On 03/15/2017 12:09 AM, Bernd Schmidt wrote:
I'll retest with your
suggestion and with the bitmap creation conditional on i1 being nonnull.
Like this (also had to throw in a bitmap_empty_p). Retested as before. Ok?
Yea, not surprised you needed th
On 03/15/2017 04:15 AM, Martin Liška wrote:
On 03/14/2017 11:27 PM, Ilya Enkovich wrote:
2017-03-10 16:15 GMT+03:00 Martin Liška :
Hello.
Currently, __builtin_ia32_bndret is used for all functions that have
non-void
return type. I think the right fix is to return bounds just for a
bounded type
On 03/15/2017 04:12 AM, Martin Liška wrote:
On 03/15/2017 12:58 AM, Ilya Enkovich wrote:
2017-03-13 16:33 GMT+03:00 Martin Liška :
On 03/13/2017 02:07 PM, Richard Biener wrote:
No, that can't happen. I said that for example for
struct S { ... } s;
foo (s);
pass_by_reference may be true but
> On Mar 14, 2017, at 9:25 AM, Richard Biener
> wrote:
>
> On Tue, Mar 14, 2017 at 3:20 PM, Bill Schmidt
> wrote:
>>
>>> On Mar 14, 2017, at 7:50 AM, Bill Schmidt
>>> wrote:
>>>
>>>
On Mar 14, 2017, at 3:57 AM, Richard Biener
wrote:
On Tue, Mar 14, 2017 at 1:04 AM,
On Wed, 2017-03-15 at 14:53 +0100, Martin Liška wrote:
> Hello.
>
> Following small series exclude some strings from pot files via
> introduction of internal_error_cont function.
>
> Patch can bootstrap and survives regression tests.
>
> (Sorry for broken threading because my cover letter was re
On Wed, Mar 15, 2017 at 04:09:05PM +, Tamar Christina wrote:
> This patch fixes a bug in the SIMD mov pattern where we were doing an
> insert and leaving the rest of the vector in an undetermined state. This
> could cause a bug if something else expects the other lanes to be 0.
>
> Semantically
2017-03-15 13:15 GMT+03:00 Martin Liška :
> On 03/14/2017 11:27 PM, Ilya Enkovich wrote:
>>
>> 2017-03-10 16:15 GMT+03:00 Martin Liška :
>>>
>>> Hello.
>>>
>>> Currently, __builtin_ia32_bndret is used for all functions that have
>>> non-void
>>> return type. I think the right fix is to return bound
2017-03-15 20:09 GMT+03:00 Jeff Law :
> On 03/15/2017 04:15 AM, Martin Liška wrote:
>>
>> On 03/14/2017 11:27 PM, Ilya Enkovich wrote:
>>>
>>> 2017-03-10 16:15 GMT+03:00 Martin Liška :
Hello.
Currently, __builtin_ia32_bndret is used for all functions that have
non-void
On 03/14/2017 09:53 PM, Richard Kenner wrote:
The GCC manual uses "cannot" in most places (280 lines) but there
are a few instances of "can't" (33 lines).
The attached patch replaces the informal "can't" with the former
for consistency.
In my opinion, this is the wrong direction. Contraction
> > The latter references other documents, which advocate for more use of
> > contractions even in formal writing.
>
> These are legal guides, not obviously relevant in the context
> of technical writing.
Yes and no. The argument for them is that legal writing is the most formal
of all and has b
On 10/03/17 11:54 +, Jonathan Wakely wrote:
On 10/03/17 19:36 +0800, Xi Ruoyao wrote:
Hi,
This was resent to be in both libstdc++ and gcc-patches list.
Thanks.
The ill-formed checking in binary_heap::push_heap has made it
O(n). Remove this checking.
The checking certainly looks redund
This looks very safe, but not critical enough for stage 4. So I think
wait for stage 1, and maybe also put it on the branch later.
Jason
On Tue, Mar 14, 2017 at 9:05 PM, David Malcolm wrote:
> PR c++/80016 reports an issue with bizarre underlined range
> for a complicated expression.
>
> The ro
PR 80047 - fixincludes/fixincl.c: PVS-Studio: Improper Release
of Memory Before Removing Last Reference (CWE-401) points out
that the fixincludes program calls getcwd with the first argument
set to NULL, apparently a Glibc extension, to have the function
allocate the memory to which it then return
I suggest we apply the following and close the PR as INVALID (not a
bug). Ok?
Bernd
Index: pr65693.c
===
--- pr65693.c (revision 245685)
+++ pr65693.c (working copy)
@@ -2,6 +2,11 @@
/* { dg-do compile } */
/* { dg-options "
On Tue, Mar 14, 2017 at 9:05 PM, David Malcolm wrote:
> OK for trunk now, or should this wait until stage 1?
Stage 1.
> + cp_token *close_paren = cp_parser_require (parser, CPP_CLOSE_PAREN,
> + RT_CLOSE_PAREN);
> + location_t end_loc =
Hello, gentle maintainer.
This is a message from the Translation Project robot.
A revised PO file for textual domain 'gcc' has been submitted
by the German team of translators. The file is available at:
http://translationproject.org/latest/gcc/de.po
(This file, 'gcc-7.1-b20170226.de.po', h
This testcase can only ever be built on x86 (it needs the "avx*"
attributes). This patch skips the test elsewhere.
Is this okay for trunk?
Segher
2017-03-15 Segher Boessenkool
gcc/testsuite/
* gcc.dg/tree-prof/pr66295.c: Skip unless on an x86 target.
---
gcc/testsuite/gcc.dg/tre
The issue here was that can_convert_bad doesn't allow this ill-formed
conversion, but instantiate_type does (with -fpermissive). Fixed by
recognizing that case. It might be nice to make can_convert_bad
smarter, but that doesn't seem worth the effort.
Tested x86_64-pc-linux-gnu, applying to trunk
Tested on Linux-x64.
2017-03-16 Ville Voutilainen
Implement LWG 2857, {variant,optional,any}::emplace should
return the constructed value.
* include/std/any (any_cast(any*)): Forward-declare.
(emplace(_Args&&...)): Change the return type and
return a reference to the constr
This patch is a simpler approach to fixing PR 71294, then I proposed last year.
The core of the problem is when we are creating V2DI/V2DF vectors via a splat
operation, the predicate for the second argument allows memory addresses, GPR
registers as well as VSX registers, but when we compile for po
On 16/03/17 00:21 +0200, Ville Voutilainen wrote:
Tested on Linux-x64.
2017-03-16 Ville Voutilainen
Implement LWG 2857, {variant,optional,any}::emplace should
return the constructed value.
* include/std/any (any_cast(any*)): Forward-declare.
(emplace(_Args&&...)): Change the retu
On 16 March 2017 at 00:31, Jonathan Wakely wrote:
>> emplace(_Args&&... __args)
>> {
>> __do_emplace<_Decay<_ValueType>>
>> (std::forward<_Args>(__args)...);
>> + return *(std::any_cast<_Decay<_ValueType>>(this));
>
>
> Can we avoid the branch in any_cast to che
In bug 52477 - Wrong initialization order? attribute constructor
vs static data access, the reporter expects C++ objects with static
storage duration to have their ctors called in the same order as
those of functions declared with attribute constructor interleaved
with them.
In his comment on th
On 16/03/17 00:43 +0200, Ville Voutilainen wrote:
On 16 March 2017 at 00:31, Jonathan Wakely wrote:
emplace(_Args&&... __args)
{
__do_emplace<_Decay<_ValueType>>
(std::forward<_Args>(__args)...);
+ return *(std::any_cast<_Decay<_ValueType>>(this));
Can we
The following patch fixes
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80017
The patch was successfully bootstrapped and tested on x86-64, ppc64, and
aarch64.
Committed as rev. 246181.
Index: ChangeLog
===
--- ChangeLog (revisio
The define_expand for copysign3 will call
gen_copysign3_fcpsgn if either TARGET_CMPB || VECTOR_UNIT_VSX_P
(mode) are true, but gen_copysign3_fcpsgn is missing the
check of VECTOR_UNIT_VSX_P (mode) which results in an
unrecognizable insn ICE. This is fixed with the following patch.
Bootstrap/regtes
So as noted in the PR, we're failing to discover an important jump
thread during threading from VRP. This is a case where the symbolic
range is actually more useful than constants in the range.
There's some talk of improving this in VRP for gcc-8, but in the mean
time, fixing this by improvin
As mentioned in the cover message, this is strictly a refactoring patch.
lookup_avail_expr should always have been a class member of the
available expression stack. That (of course) allows it to be used by
any instance of the class (which we'll take advantage of later from
within VRP's insta
Patch #3 will remove handle_dominating_asserts from the core of the jump
threading code and push it into VRP's callbacks where it should always
have been.
As a side effect it causes some code which was previously only used for
threading from VRP to be used unconditionally. It's actually a g
On 03/15/2017 09:17 PM, Jeff Law wrote:
Patch #3 will remove handle_dominating_asserts from the core of the jump
threading code and push it into VRP's callbacks where it should always
have been.
As a side effect it causes some code which was previously only used for
threading from VRP to be use
On Wed, Mar 15, 2017 at 6:59 PM, Martin Sebor wrote:
> In bug 52477 - Wrong initialization order? attribute constructor
> vs static data access, the reporter expects C++ objects with static
> storage duration to have their ctors called in the same order as
> those of functions declared with attri
66 matches
Mail list logo
