Any comments? Thanks.
> Hi,
> This patch converts more intrinsics to use builtin functions instead of
> the
> previous inline assembly syntax.
> Passed the glorious testsuite of Christophe Lyon.
>
> Three testcases are added for the testing of intriniscs which are not
> covere
Hello,
libc could be installed in a directory prefix. This patch provides
a way to specify such a prefix for gcc at configuration time.
I have only tested the patch with glibc on amd64, x86 and arm.
It is logically straightforward. Please share your thoughts on it.
Cheers,
Benda
* con
Hello!
>> I also wonder if compare-elim ought to be helping here. Isn't that the
>> point here, to eliminate the comparison and instead get it for free as
>> part of the arithmetic? If so, is it the fact that we have memory
>> references that prevents compare-elim from kicking in?
>
> Yes, compa
On Wed, Dec 3, 2014 at 10:25 AM, Jonathan Wakely wrote:
> OK - thanks.
Committed. Does it need to be backported to 4.9? When do we usually do
the backporting?
--
Regards,
Tim Shen
This patch updates rs6000.md and rs6000.c to use std::swap instead of
open coding.
Bootstrapped on powerpc-linux.
Thanks, David
* config/rs6000/rs6000.md (floatsidf2_internal): Use std::swap.
(floatunssidf2_internal): Same.
* config/rs6000/rs6000.c (rs6000_emit_vector_com
On Wed, Dec 03, 2014 at 05:31:15PM +0800, Bin.Cheng wrote:
> Does your series of patches affect PR62151? I am going to revisit
> that one to see if I can answer Jeff's questions there.
I don't think any of my recent patches affects it. r215789 does,
however, so you'll need to revert that to show
On 4 December 2014 at 00:25, Tobias Burnus wrote:
> Manuel López-Ibáñez wrote:
>>
>> In any case, the current patch fixes these ICEs and I couldn't trigger
>> new ones. Bootstrapped and regression tested.
>> OK for Fortran?
>
>
> OK. Thanks.
It is still not clear to me if line_len is the length o
From: Trevor Saunders
Hi,
In this pr we have a section attribute being applied to something in a comdat
group. Since its ok for things to be in a comdat group and have an implicit
section it seems it should also be fine to have a section from the attribute.
bootstrapped + regtested x86_64-unkn
Hi,
this patch fixes fold-const folding if EQ/NE_EXPR of ADDR_EXPRS WRT aliases and
weaks. Similarly as my earlier nonzero_address_p patch, it moves the logic
whether two symbls can resolve to same location to symtab.
The existing code did not make much sense to me, so I tried to follow what
alias
Manuel López-Ibáñez wrote:
In any case, the current patch fixes these ICEs and I couldn't trigger
new ones. Bootstrapped and regression tested.
OK for Fortran?
OK. Thanks.
* * *
BTW: The output of the code changes with the common diagnostic. For a
different test case, I get (old)
On Mon, 10 Nov 2014, Martin Uecker wrote:
> + if (OPT_Wdiscarded_array_qualifiers
OPT_Wdiscarded_array_qualifiers is an enum constant, so it doesn't make
sense to test it in if conditions like this. You don't need a test in the
if condition for whether the warning is enabled - passing
Hello,
as discussed in the PR, this fixes a wrong-code reload bug in the
following situation in push_reload:
/* If this is an input reload and the operand contains a register that
dies in this insn and is used nowhere else, see if it is the right class
to be used for this reload. Use
Hi!
The following testcase ICEs, because base_checks vector contains
stale statements, and can_remove_asan_check relies on them not to be
there anymore (assumes that all statements in the vector dominate
the current statement, if that is not true, the loop going through immediate
dominators won't
> this patch is ready for commit now. Please apply. There have been no
> objections
> against doing dg-do compile only, since my last post in August.
I wasn’t back on active duty in August, so I didn’t follow that. But if the
only argument for compile-test over run-test is the runtime, then I do
On Wed, Dec 3, 2014 at 7:01 AM, H.J. Lu wrote:
> On Wed, Dec 3, 2014 at 5:47 AM, H.J. Lu wrote:
>> On Tue, Dec 2, 2014 at 12:01 PM, Uros Bizjak wrote:
>>> On Tue, Dec 2, 2014 at 8:40 PM, H.J. Lu wrote:
On Tue, Dec 2, 2014 at 11:19 AM, Uros Bizjak wrote:
> Hello!
>
>> Ping.
>>>
On 12/02/14 14:09, Jack Howarth wrote:
The attached patch fixes the regression in the
gcc/testsuite/gcc.dg/graphite/isl-codegen-loop-dumping.c testcase
caused by the accidental removal of -fgraphite-identity from
dg-options at r217315. Okay for gcc trunk?
OK.
jeff
On 12/01/14 16:45, Joseph Myers wrote:
On Mon, 1 Dec 2014, Jeff Law wrote:
diff --git a/gcc/c-family/c.opt b/gcc/c-family/c.opt
index 8f5d76c..283c632 100644
--- a/gcc/c-family/c.opt
+++ b/gcc/c-family/c.opt
@@ -1043,6 +1043,9 @@ Instrument only functions marked with bnd_instrument
attribute.
On 12/02/14 14:52, Michael Eager wrote:
As Cary said, the values are not fixed until the spec is released.
As Jason said, the values are unlikely to change.
Fair enough.
The standards process can be a bit messy at times, particularly
when we realize that something we added to a draft needs t
On Dec 3, 2014, at 6:13 AM, Kyrill Tkachov wrote:
>
> The check for the tiny memory model is something that is called frequently on
> aarch64 testing
> since we use it to prune relocation truncation errors in the testsuite.
> Therefore it makes sense to cached the result.
> I've done this in a s
On 12/03/14 07:13, Kyrill Tkachov wrote:
Hi all,
The check for the tiny memory model is something that is called
frequently on aarch64 testing
since we use it to prune relocation truncation errors in the testsuite.
Therefore it makes sense to cached the result.
I've done this in a similar way to
On 12/03/14 11:48, David Malcolm wrote:
On Mon, 2014-12-01 at 19:31 -0500, Ulrich Drepper wrote:
I think the jit patches introduced a problem when you rebuild within a
directory that contains an old build (i.e., no brand new build
directory). The gcc/Makefile creates a symlink for xgcc with th
On 12/03/14 12:10, Jack Howarth wrote:
The attached patch completes the removal of the cloog support in
gcc trunk by removing the use of -DCLOOG_INT_GMP on islinc in
config/isl.m4 and the top-level configure. Bootstrap tested against
isl 0.14 on x86_64-apple-darwin14. Okay for gcc trunk?
On 12/03/14 10:10, Jakub Jelinek wrote:
Hi!
While expand_mult correctly checks for multiplication by zero,
expand_widening_mult does not and as floor_log2 (0) returns -1,
it attempts to expand a double-word shift by -1 and on ppc32
happens to return 2 w* 0 = 1. Usually earlier optimizations
jus
On 12/03/14 10:11, Jakub Jelinek wrote:
Hi!
While looking at PR59708, I've noticed this. Ok for trunk?
2014-12-03 Jakub Jelinek
* expmed.c (expand_mult): Use std::swap.
Yes. In fact, I think this kind of stuff ought to fit under the obvious
rule.
jeff
The attached patch completes the removal of the cloog support in
gcc trunk by removing the use of -DCLOOG_INT_GMP on islinc in
config/isl.m4 and the top-level configure. Bootstrap tested against
isl 0.14 on x86_64-apple-darwin14. Okay for gcc trunk?
Jack
2014-12-03 Jack Howar
On Mon, 2014-12-01 at 19:31 -0500, Ulrich Drepper wrote:
> I think the jit patches introduced a problem when you rebuild within a
> directory that contains an old build (i.e., no brand new build
> directory). The gcc/Makefile creates a symlink for xgcc with the full
> driver name without first re
On 03/12/14 09:41 -0800, Tim Shen wrote:
Bootstrapped and tested.
Thanks!
OK - thanks.
Bootstrapped and tested.
Thanks!
--
Regards,
Tim Shen
commit 942a5f8e5f4fc4af4f1a18f15b529166e9bdd3ac
Author: timshen
Date: Wed Dec 3 09:36:04 2014 -0800
PR libstdc++/64140
* include/bits/regex.tcc (regex_iterator<>::operator++): Update
prefix.matched after modifying
On 11/28/2014 03:35 PM, Joseph Myers wrote:
If you find you need -fno-ms-extensions as well, that's not a bug in the
documentation or the testcase, it's a bug in GCC, failing to give a
diagnostic for certain code for certain targets even though the standard
requires it to do so. The testcase has
OK.
Jason
Hi!
As discussed in the PR, ideally this should be done in the middle-end,
but no fix materialized for it during the last 18 months and is unlikely
to happen for GCC 5 either, so this patch just restores what the FEs used
to produce. Bootstrapped/regtested on x86_64-linux and i686-linux, ok for
t
> Le 3 déc. 2014 à 18:08, Andre Vehreschild a écrit :
>
> Hi,
>
> this patch is ready for commit now. Please apply. There have been no
> objections
> against doing dg-do compile only, since my last post in August.
Not really true, I do have objections, but I won’t fight for them. I still
th
Hi!
While looking at PR59708, I've noticed this. Ok for trunk?
2014-12-03 Jakub Jelinek
* expmed.c (expand_mult): Use std::swap.
--- gcc/expmed.c.jj 2014-11-18 08:26:45.0 +0100
+++ gcc/expmed.c2014-12-03 11:12:37.519297971 +0100
@@ -3087,11 +3087,7 @@ expand_mult
Hi!
While expand_mult correctly checks for multiplication by zero,
expand_widening_mult does not and as floor_log2 (0) returns -1,
it attempts to expand a double-word shift by -1 and on ppc32
happens to return 2 w* 0 = 1. Usually earlier optimizations
just ensure that w* 0 isn't expanded using ex
Hi,
this patch is ready for commit now. Please apply. There have been no objections
against doing dg-do compile only, since my last post in August.
- Andre
On Tue, 26 Aug 2014 11:30:12 +0200
Andre Vehreschild wrote:
> Hi,
>
> On Sun, 17 Aug 2014 15:06:02 +0200
> Mikael Morin wrote:
>
> > L
On Wed, Dec 03, 2014 at 08:38:44AM -0500, Jason Merrill wrote:
> >(in f1 and f3 verifies it for _ZTI1V, in f2 and f4 verifies it for
> >_ZTI1S). Should I change it so that we get 2 instrumentations in f2 and f4
> >and one in f1/f3 (i.e. in f2/f4 check two vptrs, one _ZTI1S and one _ZTI1V),
> >or d
The two testcases rely on visibility support. This patch explicitly
adds the Dejagnu annotation to fix two more failures on systems
without visibility support.
Thanks, David
Tested on powerpc-ibm-aix7.1.0.0
* g++.dg/ext/visibility/anon[12].C: Require visibility support.
Index: anon1.C
* g++.dg/cpp0x/constexpr-shift1.C: New test.
> > * g++.dg/cpp1y/constexpr-shift1.C: New test.
> > * g++.dg/ubsan/pr63956.C: Add dg-errors.
>
> FAIL: 28_regex/algorithms/regex_match/ecma/wchar_t/anymatcher.cc (test for
> excess errors)
> Excess errors:
> /daten/ara
On Wed, Dec 03, 2014 at 04:03:37PM +, Jonathan Wakely wrote:
> On 03/12/14 16:47 +0100, Marek Polacek wrote:
> >@@ -417,7 +417,8 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
> > typedef typename std::is_same<_CharT, char>::type _UseCache;
> >
> > static constexpr size_t
> >- _S_cache_size
On 3 December 2014 at 15:22, Christophe Lyon wrote:
> Hi,
>
> Here is a fix for typos in the AdvSimd intrinsic tests, where vaddl
> and vaddw didn't actually execute the tests. (The function was
> declared in main, instead of called).
>
> This patch also fixes the expected output for these tests.
I've really got to start testing with -Wsystem-headers so I don't do
silly things like this.
Committed as obvious.
commit 88abdc37b00dbe64099f78a31163a299a023ec03
Author: Jonathan Wakely
Date: Wed Dec 3 16:08:08 2014 +
PR libstdc++/64168
* include/std/future (_Deferred_state::_M
On 03/12/14 16:47 +0100, Marek Polacek wrote:
@@ -417,7 +417,8 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
typedef typename std::is_same<_CharT, char>::type _UseCache;
static constexpr size_t
- _S_cache_size() { return 1ul << (sizeof(_CharT) * __CHAR_BIT__); }
+ _S_cache_size() { r
Hi,
in this bug Manuel noticed that a permerror emitted by
check_previous_goto_1 via identify_goto is followed by an explanatory
error instead of an inform, which then can't be suppressed with
-fpermissive -w. When consistently adjusting the testsuite I noticed
that check_goto too needs fixin
This fixes some testsuite fallout in 28_regex/. The problem was that
with -m32 we were triggering UB in a shift. Fixed by multiplying the
RHS of the shift-expression by 0 if we are not going to use the cache.
Regtested/bootstrapped on ppc64-linux, ok for trunk?
2014-12-03 Marek Polacek
Ping.
Alan Lawrence wrote:
vld1_lane intrinsics ICE at -O0 because they contain a call to the vset_lane
intrinsics, through which the lane index is not constant-propagated. (They are
fine at -O1 and higher!). This fixes the ICE by replacing said call by a macro.
Rather than defining many indi
Hi all,
This is a simple patch to change a few misused IDNET in AARCH64_CORE to
SCHED to reflect their original intentions.
The third argument of AARCH64_CORE macro is scheduler identity.
No functional change. aarch64-none-elf has been built and checked, no issue.
Okay for trunk?
Regards,
Re
Dear all,
the attached patches permit an in-tree build of GCC 4.9 using
CLooG with ISL-0.14 backend and ISL-0.14 - as wished by Richard.
The CLooG patches have been extracted from the CLooG git repository
and have been included as courtesy for Linux distribution people.
I do not intent to backp
On 02/12/14 21:45, Ramana Radhakrishnan wrote:
I've spent some time this evening pondering over your patch. Firstly
it appears that the current behaviour is going to cause more breakage
than originally expected. If this is to go in we'd have a number of
users having to add -mfloat-abi=soft to the
On Wed, Dec 3, 2014 at 5:47 AM, H.J. Lu wrote:
> On Tue, Dec 2, 2014 at 12:01 PM, Uros Bizjak wrote:
>> On Tue, Dec 2, 2014 at 8:40 PM, H.J. Lu wrote:
>>> On Tue, Dec 2, 2014 at 11:19 AM, Uros Bizjak wrote:
Hello!
> Ping.
>> Ping.
>>> Ping.
Ping.
It woul
On 02/12/14 23:58 +0100, François Dumont wrote:
DR libstdc++/13631
s/DR/PR/
* include/bits/codecvt.h (codecvt): friend class
std::messages.
(codecvt): friend class
std::messages.
* config/locale/gnu/messages_member.h
(messages::do_open): Specialized.
(messages::do_clo
Hi,
thank you very much for pointing out all the spelling mistakes. I
suppose you could tell I was writing this in a hurry :-) Nevertheless
I should probably make a habit of spell checking stuff before
sending it out.
As far as your question is concerned:
On Wed, Dec 03, 2014 at 10:53:54AM +0
2014-12-02 0:16 GMT+03:00 Jeff Law :
> On 11/24/14 07:06, Ilya Enkovich wrote:
>>>
>>> Normally GCC target libraries assigned to the FSF would use GPL+exception
>>> rather than LGPL (especially if the library might be linked in
>>> statically), to keep predictable what requirements are imposed by l
Hi,
Here is a fix for typos in the AdvSimd intrinsic tests, where vaddl
and vaddw didn't actually execute the tests. (The function was
declared in main, instead of called).
This patch also fixes the expected output for these tests.
OK?
Thanks
Christophe.
2014-12-03 Christophe Lyon
test
Also VQ_S to VDQ_BHSI! All these were in the ChangeLog, but apologies for a
not-very-good summary. So rebuilding that list:
* VDQ duplicates VDQ_I, use the latter
* VSDQ_I_BHSI has no uses, remove
* SDQ_I (duplicate of ALLI) has no uses, remove;
* VDQQHS (duplicate of VDQ_BHSI) has no
Hi all,
The check for the tiny memory model is something that is called
frequently on aarch64 testing
since we use it to prune relocation truncation errors in the testsuite.
Therefore it makes sense to cached the result.
I've done this in a similar way to how check_effective_target_vect_shift
On Wed, Dec 3, 2014 at 2:52 PM, H.J. Lu wrote:
>> As seen in Fedora gcc patch in pkgs.fedoraproject.org/cgit/gcc.git/ .
>>
>> libitm/ChangeLog:
>>
>> 2014-12-03 Uros Bizjak
>>
>> * configure.tgt (x86_64): Tune -m32 multilib to generic.
>>
>> libgomp/ChangeLog:
>>
>> 2014-12-03 Uros Bizjak
Hi,
On Mon, Dec 01, 2014 at 12:00:14PM +0100, Richard Biener wrote:
> On Fri, Nov 28, 2014 at 5:20 PM, Eric Botcazou wrote:
> > Hi,
> >
> > the attached Ada testcase triggers an assertion in the RTL expander for the
> > address operator because the operator has been applied to a non-byte-aligned
On Wed, Dec 3, 2014 at 5:49 AM, Uros Bizjak wrote:
> Hello!
>
> As seen in Fedora gcc patch in pkgs.fedoraproject.org/cgit/gcc.git/ .
>
> libitm/ChangeLog:
>
> 2014-12-03 Uros Bizjak
>
> * configure.tgt (x86_64): Tune -m32 multilib to generic.
>
> libgomp/ChangeLog:
>
> 2014-12-03 Uros Biz
On Wed, Dec 03, 2014 at 02:49:18PM +0100, Uros Bizjak wrote:
> Hello!
>
> As seen in Fedora gcc patch in pkgs.fedoraproject.org/cgit/gcc.git/ .
>
> libitm/ChangeLog:
>
> 2014-12-03 Uros Bizjak
>
> * configure.tgt (x86_64): Tune -m32 multilib to generic.
>
> libgomp/ChangeLog:
>
> 2014-
This moves fold_builtin_strncat and fold_builtin_printf.
Bootstrap and regtest pending on x86_64-unknown-linux-gnu.
Richard.
2014-12-03 Richard Biener
* builtins.c (target_newline): Export.
(target_percent_s_newline): Likewise.
(fold_builtin_1): Do not fold printf fu
On 1 December 2014 at 07:48, Gopalasubramanian, Ganesh
wrote:
> Please ignore the previous patch sent. The attachment was wrong.
>
>> There's no point in the buffer or the sprintf.
>> The text is short enough to repeat whole pattern in the array:
>
> Updated the patch for the above suggestions.
>
Hello!
As seen in Fedora gcc patch in pkgs.fedoraproject.org/cgit/gcc.git/ .
libitm/ChangeLog:
2014-12-03 Uros Bizjak
* configure.tgt (x86_64): Tune -m32 multilib to generic.
libgomp/ChangeLog:
2014-12-03 Uros Bizjak
* configure.tgt (x86_64-*-linux*): Tune -m32 multilib to gene
This moves sqrt simplification from GENERIC to patterns.
The testcase adjustment shows one of the side-effects I mentioned
in 1/n - simplification is delayed to SSA and for the testcase
we have to trick CCP not folding everything to a constant before
the expression simplification is done in forwp
On Tue, Dec 2, 2014 at 12:01 PM, Uros Bizjak wrote:
> On Tue, Dec 2, 2014 at 8:40 PM, H.J. Lu wrote:
>> On Tue, Dec 2, 2014 at 11:19 AM, Uros Bizjak wrote:
>>> Hello!
>>>
Ping.
> Ping.
>> Ping.
>>> Ping.
>>>
>>> It would probably help reviewers if you pointed to actual path
>>>
shift1.C: New test.
> * g++.dg/ubsan/pr63956.C: Add dg-errors.
FAIL: 28_regex/algorithms/regex_match/ecma/wchar_t/anymatcher.cc (test for
excess errors)
Excess errors:
/daten/aranym/gcc/gcc-20141203/Build/m68k-linux/libstdc++-v3/include/bits/regex_compiler.h:424:31:
in constexpr expans
While I was waiting for more testcases of broken mixed gimple/generic
folding to come in I completely forgot about this. Thus the following
patches finish moving strig function folding from builtins.c to
gimple-fold.c.
This is patch 1, moving fold_builtin_fprintf.
Bootstrapped on x86_64-unknown
On 12/03/2014 07:01 AM, Jakub Jelinek wrote:
One more question. My current version of the patch adds one ubsan vptr
instrumentation in each of the following functions:
struct S { int s; virtual void foo (); S(); virtual ~S(); };
struct T : S {};
struct U { int u; S s[4]; };
struct V { U v; virt
On Tue, Dec 2, 2014 at 5:55 AM, H.J. Lu wrote:
> On Tue, Dec 2, 2014 at 5:51 AM, H.J. Lu wrote:
>> On Tue, Dec 2, 2014 at 3:34 AM, Wilco Dijkstra wrote:
Jeff Law wrote:
OK with the appropropriate ChangeLog entires. THe original for
ira-costs.c was fine, so you just need the trivi
On Mon, Dec 01, 2014 at 09:45:30AM -0500, Jason Merrill wrote:
> On 11/28/2014 09:41 AM, Jakub Jelinek wrote:
> >>Why do you look through ARRAY_REF here? An element of an array is its own
> >>complete object.
> >
> >That had to do with only instrumenting dereferences surrounded by handled
> >compo
While working on PR57272 for unordered containers I was getting a
headache reading all the return types with nested-name-qualifiers
split over three or four lines.
This patch replaces them with trailing return types, so the names used
in the return type are in scope and don't need to be qualified
On Wed, 3 Dec 2014, Richard Biener wrote:
>
> This refactors fold_builtin_call_array to not build a CALL_EXPR if
> no simplification was possible (we have fold_build_call_array_loc
> for that). This requires to fix the single case where that was
> deemed necessary (fold_builtin_varargs calling f
Committed.
Prachi
-Original Message-
From: Matthew Fortune
Sent: Wednesday, December 3, 2014 4:18 PM
To: Prachi Godbole; gcc-patches@gcc.gnu.org
Subject: RE: [PATCH][MIPS] P5600 pipeline description fixes
> Changelog:
>
> 2014-12-03 Prachi Godbole
>
> * config/mips/p5600.md (
> Index: src/gcc/ipa-cp.c
> ===
> --- src.orig/gcc/ipa-cp.c
> +++ src/gcc/ipa-cp.c
> @@ -262,6 +262,9 @@ public:
>ipcp_lattice ctxlat;
>/* Lattices describing aggregate parts. */
>ipcp_agg_lattice *aggs;
> + /* Alignment
On Wed, Dec 3, 2014 at 11:45 AM, Jakub Jelinek wrote:
> Hi!
>
> I'd like to ping this patch to fix gcc/Makefile dependencies
> for gengtype objects as well as gcc-{ar,nm,ranlib} objects.
> https://gcc.gnu.org/ml/gcc-patches/2014-11/msg03092.html
Ok.
Thanks,
Richard.
> Jakub
> Changelog:
>
> 2014-12-03 Prachi Godbole
>
> * config/mips/p5600.md (define_automaton, define_cpu_unit): Replace
> p5600_agen_pipe and p5600_alu_pipe with p5600_agen_alq_pipe.
>
> (p5600_int_arith_1, p5600_int_arith_2, p5600_int_arith_4): Change
> reservation order.
Hi!
I'd like to ping this patch to fix gcc/Makefile dependencies
for gengtype objects as well as gcc-{ar,nm,ranlib} objects.
https://gcc.gnu.org/ml/gcc-patches/2014-11/msg03092.html
Jakub
Hi,
This patch merges automata p5600_agen_pipe and p5600_alu_pipe into one to
enable blocking of the cpu units in either-or reservations.
It also changes the order of the units in such reservations to benefit from
multi-issue scenarios.
Changelog:
2014-12-03 Prachi Godbole
* config
On Tue, Dec 02, 2014 at 03:43:43PM +0100, Richard Biener wrote:
> I'd like to get a little bit coverage of builtin simplification
> in the wild and wonder if the following is a piece that would
> be acceptable for trunk at this point (well, it fixes part of
> a bug which is fine in stage3). Eventu
This refactors fold_builtin_call_array to not build a CALL_EXPR if
no simplification was possible (we have fold_build_call_array_loc
for that). This requires to fix the single case where that was
deemed necessary (fold_builtin_varargs calling fold_builtin_fpclassify)
which is easily fixed to simp
On Wed, Nov 26, 2014 at 04:35:50PM +, James Greenhalgh wrote:
> Why do we want to turn off folding for the V4SF/V2SF/V2DF modes of these
> intrinsics? There should be no difference between the mid-end definition
> and the intrinsic definition of their behaviour.
Good point. Done.
> I also no
On Wed, Dec 03, 2014 at 11:00:06AM +0100, Tom de Vries wrote:
> As discussed*, this patch renames fuse-caller-save to fipa-ra.
>
> Bootstrapped and reg-tested on x86_64. Did minimal arm build.
>
> OK for stage3?
>
> Thanks,
> - Tom
>
> (*) https://gcc.gnu.org/ml/gcc-patches/2014-10/msg01943.htm
Jakub,
As discussed*, this patch renames fuse-caller-save to fipa-ra.
Bootstrapped and reg-tested on x86_64. Did minimal arm build.
OK for stage3?
Thanks,
- Tom
(*) https://gcc.gnu.org/ml/gcc-patches/2014-10/msg01943.html
2014-12-03 Tom de Vries
PR rtl-optimization/63957
* doc/invoke.te
On 18/11/14 11:58, Richard Earnshaw wrote:
On 18/11/14 11:30, Mantas Mikaitis wrote:
Incorrect predefinitions for certain target architectures. E.g. arm7-m
does not contain NEON but the defintion __ARM_NEON_FP was switched on.
Similarly with armv6 and even armv2.
This patch fixes the predefine
On Tue, Dec 2, 2014 at 9:32 PM, Martin Jambor wrote:
> Hi,
>
> apparently it is necessary to check that type sizes match before
> attempting to fold-V_C_E them in evaluate_conditions_for_known_args.
> So this patch does this.
>
> It passes bootstrap and testing on x86_64-linux and I have verified
On Tue, 2 Dec 2014, David Malcolm wrote:
> On Tue, 2014-12-02 at 08:49 +0100, Jakub Jelinek wrote:
> Hi!
> >
> > Here is an attempt to adjust gimple_build_assign* documentation.
> > Ok for trunk?
> >
> > Note, apparently the documentation has not been adjusted for
> > the gimple -> gassign * etc
On Wed, Dec 3, 2014 at 1:54 PM, Segher Boessenkool
wrote:
> On Tue, Dec 02, 2014 at 07:11:04PM -0600, Segher Boessenkool wrote:
>> Here is the testcase. I cannot actually test it on an m68k build, should
>> really do something about that (I build lots of cross compilers but they
>> cannot run the
On Wed, Dec 03, 2014 at 12:04:13PM +0300, Yury Gribov wrote:
> On 12/03/2014 11:36 AM, Jakub Jelinek wrote:
> >I thought I've agreed that we can optimize checks away for asan even when
> >recovering. The patch is ok for trunk anyway, we can perhaps make it
> >tunable later, or define asan_can_opti
On 12/03/2014 11:36 AM, Jakub Jelinek wrote:
I thought I've agreed that we can optimize checks away for asan even when
recovering. The patch is ok for trunk anyway, we can perhaps make it
tunable later, or define asan_can_optimize_checks to true unconditionally.
Snap, looks like my English fai
Manuel López-Ibáñez writes:
> libcpp uses diagnostic->override_column to give a custom column number
> to diagnostics. This is taken into account when building the prefix,
> but it was missing when placing the caret.
>
> Before:
>
> /home/manuel/override_column.c:1:4: warning: "/*" within comment
On Wed, Dec 03, 2014 at 11:09:41AM +0300, Yury Gribov wrote:
> >From 5a9ede4d120ba4e39ca47e212262af53de47eb5a Mon Sep 17 00:00:00 2001
> From: Yury Gribov
> Date: Tue, 25 Nov 2014 11:49:11 +0300
> Subject: [PATCH] 2014-12-02 Yury Gribov
>
> gcc/
> * asan.h (asan_can_optimize_checks): New
On 11/25/2014 08:06 PM, Yury Gribov wrote:
This patch improves current optimization of ASAN_CHECKS performed by
sanopt pass. In addition to searching the sanitized pointer in
asan_check_map, it also tries to search for definition of this pointer.
This allows more checks to be dropped when defi
91 matches
Mail list logo
