Various optimizations. Committed.
2013-09-14 DJ Delorie
Nick Clifton
* config/rl78/mulsi3.S: Remove a few unneeded moves and branches.
* config/rl78/vregs.h: New.
* config/rl78/signbit.S: New file. Implements signbit function.
* config/rl78/divm
tree if conversion is an enabler pass for vectorization, so by
default, it is only turned on when vectorization is on, but may also
depend on the optimization level. Currently, the logic to handle this
is in the gate function which become hard to understand and extend.
The proposed patch move the
Applied to trunk.
-- Gaby
Index: MAINTAINERS
===
--- MAINTAINERS (revision 202586)
+++ MAINTAINERS (working copy)
@@ -195,7 +195,6 @@
dwarf debugging code Jason Merrill ja...@redhat.com
dwarf debugging code Cary Cout
This patch updates the maintainership of the c++-concepts and
var-template branches.
-- Gaby
Index: svn.html
===
RCS file: /cvs/gcc/wwwdocs/htdocs/svn.html,v
retrieving revision 1.188
diff -p -r1.188 svn.html
*** svn.html18 Aug
Wei Mi writes:
>> Checking corei7/corei7-avx explicitly isn't a good idea.
>> It is also useful for Ivy Bridge and Haswell. I think you
>> should use a variable to control it, similar to
>> TARGET_FUSE_CMP_AND_BRANCH.
>>
>>
>> --
>> H.J.
>
> Different x86 microarchitectures support macro-fusion
On Fri, Sep 13, 2013 at 4:07 PM, Wei Mi wrote:
> On Fri, Sep 13, 2013 at 1:45 PM, Wei Mi wrote:
>> On Fri, Sep 13, 2013 at 12:09 PM, H.J. Lu wrote:
>>> On Fri, Sep 13, 2013 at 11:28 AM, Wei Mi wrote:
> Checking corei7/corei7-avx explicitly isn't a good idea.
> It is also useful for Ivy
On Fri, Sep 13, 2013 at 1:45 PM, Wei Mi wrote:
> On Fri, Sep 13, 2013 at 12:09 PM, H.J. Lu wrote:
>> On Fri, Sep 13, 2013 at 11:28 AM, Wei Mi wrote:
Checking corei7/corei7-avx explicitly isn't a good idea.
It is also useful for Ivy Bridge and Haswell. I think you
should use a var
On 09/12/2013 06:20 PM, Tim Shen wrote:
On Thu, Sep 12, 2013 at 12:00 PM, Jakub Jelinek wrote:
That is not the right test, because you need to count all tabs (at least
the leading ones, you probably shouldn't have other tabs) as 8 columns
rather than just one. And there are 11 lines that are o
> I’ve attached fix for this issue:
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57737
> There are fix, two tests and change log message.
>
> Is it ok?
>
> Cary, can you commit it for me?
>
> Ping this patch, http://gcc.gnu.org/ml/gcc-patches/2013-07/msg00053.html
Thanks for the fix, and sorry fo
On Fri, Sep 13, 2013 at 11:57 AM, James Greenhalgh
wrote:
> On Fri, Sep 13, 2013 at 07:39:08PM +0100, Andrew Pinski wrote:
>> I don't think this works for big-endian due to the way ARM decided the
>> lanes don't match up with array entry there.
>
> Hi Andrew,
>
> Certainly for the testcase I've ad
On Sep 13, 2013, at 23:35 , Olivier Hainque wrote:
>> # Darwin (Mac OS X)
>> -ifeq ($(strip $(filter-out darwin%,$(target_cpu))),)
>> +ifeq ($(strip $(filter-out darwin%,$(target_os))),)
>> 2) I don't have write access.
>
> I'll take care of committing. Thanks!
rev 202578. Thanks again Domin
Hello Dominique,
On Sep 13, 2013, at 20:03 , domi...@lps.ens.fr (Dominique Dhumieres) wrote:
> The bootstrap failure is fixed with the following patch:
>
> [macbook] gcc/work% diff -up ../_clean/gcc/ada/gcc-interface/Makefile.in
> gcc/ada/gcc-interface/Makefile.in
> --- ../_clean/gcc/ada/gcc-in
On Sep 13, 2013, at 1:08 AM, Richard Sandiford
wrote:
>>> We don't really model that properly yet. Partial modes are just defined
>>> using something like:
>>>
>>> PARTIAL_INT_MODE (SI);
True, but, all ports are trivial to fix so that the precision is included and
for the interface to change
On Fri, Sep 13, 2013 at 12:09 PM, H.J. Lu wrote:
> On Fri, Sep 13, 2013 at 11:28 AM, Wei Mi wrote:
>>> Checking corei7/corei7-avx explicitly isn't a good idea.
>>> It is also useful for Ivy Bridge and Haswell. I think you
>>> should use a variable to control it, similar to
>>> TARGET_FUSE_CMP_AN
On Fri, Sep 13, 2013 at 07:39:08PM +0100, Andrew Pinski wrote:
> I don't think this works for big-endian due to the way ARM decided the
> lanes don't match up with array entry there.
Hi Andrew,
Certainly for the testcase I've added in this patch there are no issues.
Vector indexing should work c
The bootstrap failure is fixed with the following patch:
[macbook] gcc/work% diff -up ../_clean/gcc/ada/gcc-interface/Makefile.in
gcc/ada/gcc-interface/Makefile.in
--- ../_clean/gcc/ada/gcc-interface/Makefile.in 2013-09-12 13:18:34.0
+0200
+++ gcc/ada/gcc-interface/Makefile.in 2013-09-
This is kind of fugly, but don't have anything better at the moment.
The thing is that we were always instrumenting shift expressions, even
when both operands were INTEGER_CSTs and we could prove at compile
time that the expression is well defined. This causes problems in the
C FE, mainly at place
On Fri, 13 Sep 2013, Marek Polacek wrote:
> This is kind of fugly, but don't have anything better at the moment.
> 2013-09-13 Marek Polacek
>
> PR sanitizer/58413
> c-family/
> * c-ubsan.c (ubsan_instrument_shift): Don't instrument
> an expression if we can prove it is correc
On 09/13/2013 11:11 AM, Andrew MacLeod wrote:
On 09/13/2013 03:54 AM, Richard Biener wrote:
On Thu, Sep 12, 2013 at 11:09 PM, Andrew MacLeod
wrote:
There are 2 parts of tre-ssa-ter.c to address.
is_replaceable_p() is also used in expr.c, It has a flag to indicate
where
its being called from,
On Fri, Sep 13, 2013 at 11:28 AM, Wei Mi wrote:
>> Checking corei7/corei7-avx explicitly isn't a good idea.
>> It is also useful for Ivy Bridge and Haswell. I think you
>> should use a variable to control it, similar to
>> TARGET_FUSE_CMP_AND_BRANCH.
>>
>>
>> --
>> H.J.
>
> Different x86 microarc
A simple cut-and-paste error. When I copied
any_value_dependent_elements_p to any_type_dependent_elements_p, I
should have changed the body of the function as well...
Tested x86_64-pc-linux-gnu, applying to trunk, 4.8, 4.7.
commit 0515f1a9424ba750c345826d5504aecb6e23b8e0
Author: Jason Merrill
A new test case is added to testsuite/gcc.dg/vect, which will fail
without this patch and pass with it. Bootstrap also get passed. No
additional test failure is introduced.
The new test case includes a dot product on two arrays with short and
int types. The loop will still be vectorized (using pun
On Fri, Sep 13, 2013 at 10:28 AM, Wei Mi wrote:
>> Thanks. At this point you need feedback from x86 and scheduler maintainers.
>> I would recommend you to resubmit the patch with a Changelog text, and with
>> the text of the patch inline in the email (your last mail has the patch as a
>> binary a
On Fri, Sep 13, 2013 at 11:35 AM, James Greenhalgh
wrote:
>
> Hi,
>
> The vset_lane_<8,16,32,64> intrinsics are currently
> written useing assembler, but can be easily expressed
> in C.
>
> As I expect we will want to efficiently compose these intrinsics
> I've added them as macros, just as was do
Hi,
This patch adds intrinsics for vcopy_lane_<8,16,32,64>.
These are implemented in an optimal way using the vget_lane and vset_lane
intrinsics and a combine pattern.
I've added a testcase and run a full regression run for aarch64-none-elf.
OK?
Thanks,
James
---
gcc/
2013-09-13 James Gree
Hi,
fixed mainline and 4_8-branch. Tested x86_64-linux.
Thanks,
Paolo.
2013-09-13 Paolo Carlini
PR libstdc++/58415
* include/ext/sso_string_base.h (__sso_string_base<>::
__sso_string_base(__sso_string_base&&)): Fix thinkos about
_M_le
Hi,
This patch converts the vmul_lane_<16,32,64> intrinsics
in arm_neon.h to a C implementation.
To support this, we add some patterns for the combiner to pick
up. We need a few patterns for this.
mul3_elt covers vmul_lane, vmulq_laneq variants, where the number
of lanes selected from matches t
Hi,
The vset_lane_<8,16,32,64> intrinsics are currently
written useing assembler, but can be easily expressed
in C.
As I expect we will want to efficiently compose these intrinsics
I've added them as macros, just as was done with the vget_lane
intrinsics.
Regression tested for aarch64-none-elf
Applied at revision 202573.
Thanks,
Kai
> Checking corei7/corei7-avx explicitly isn't a good idea.
> It is also useful for Ivy Bridge and Haswell. I think you
> should use a variable to control it, similar to
> TARGET_FUSE_CMP_AND_BRANCH.
>
>
> --
> H.J.
Different x86 microarchitectures support macro-fusion for different
compare and br
Updated patch implementing the logic that more specific option wins.
Ok for trunk?
thanks,
David
On Fri, Sep 13, 2013 at 9:48 AM, Xinliang David Li wrote:
> Ok -- then my updated patch is wrong then. The implementation in the
> first version matches the requirement.
>
> thanks,
>
> David
>
>
>
> Thanks. At this point you need feedback from x86 and scheduler maintainers.
> I would recommend you to resubmit the patch with a Changelog text, and with
> the text of the patch inline in the email (your last mail has the patch as a
> binary attachment, which makes it harder to review and respon
On Fri, 13 Sep 2013, Marek Polacek wrote:
> On Thu, Sep 12, 2013 at 04:05:48PM +, Joseph S. Myers wrote:
> > cause stack overflow that doesn't get detected by the kernel. So maybe
> > ubsan should imply -fstack-check or similar.
>
> Well, I have a patch for that, but I no longer think that
This patch implements semantics for constrained friend templates and
classes. The only significant changes are in determine_specializaiton
and check_constrained_friend.
Unless a friend function is defined, a constraints on friend
declarations are never actually checked. The checking happens during
Ok -- then my updated patch is wrong then. The implementation in the
first version matches the requirement.
thanks,
David
On Fri, Sep 13, 2013 at 9:45 AM, Joseph S. Myers
wrote:
> On Fri, 13 Sep 2013, Richard Biener wrote:
>
>> @@ -1691,6 +1695,12 @@ common_handle_option (struct gcc_options
>>
r202570 | abutcher | 2013-09-13 17:14:15 +0100 (Fri, 13 Sep 2013) | 7
lines
Revert r202554, r202540 and r202539.
- r202554: Fix uninitialized variables causing breakage with -Werror.
- r202540: Support using 'auto' in a
On Fri, 13 Sep 2013, Richard Biener wrote:
> @@ -1691,6 +1695,12 @@ common_handle_option (struct gcc_options
> opts->x_flag_ipa_reference = false;
>break;
>
> +case OPT_ftree_vectorize:
> + if (!opts_set->x_flag_tree_loop_vectorize)
> + opts->x_flag_tree_loop_vectorize =
New patch attached.
1) the peeling part is removed
2) the new patch implements the last-one-wins logic. -ftree-vectorize
behaves like a true alias. -fno-tree-vectorize can override previous
-ftree-xxx-vectorize.
Ok for trunk after testing?
thanks,
David
On Fri, Sep 13, 2013 at 8:16 AM, Xinlian
On Fri, 13 Sep 2013, Kai Tietz wrote:
> ChangeLog
>
> 2013-09-13 Kai Tietz
>
> PR target/57484
> * c/c-decl.c (c_builtin_function_ext_scope): Remove
> wrong assumption that it is never called on prexisting
> symbol.
c/ has its own ChangeLog so the entry should go there withou
On Fri, Sep 13, 2013 at 1:30 AM, Richard Biener
wrote:
> On Thu, Sep 12, 2013 at 10:31 PM, Xinliang David Li
> wrote:
>> Currently -ftree-vectorize turns on both loop and slp vectorizations,
>> but there is no simple way to turn on loop vectorization alone. The
>> logic for default O3 setting is
Rap Diego's knuckles for that :-) Bad Diego.
Andrew
On 09/13/2013 11:07 AM, Jakub Jelinek wrote:
On Thu, Sep 05, 2013 at 11:11:43AM +0200, Richard Biener wrote:
On Thu, Sep 5, 2013 at 2:06 AM, Mike Stump wrote:
On Aug 30, 2013, at 4:22 PM, Diego Novillo wrote:
Thanks for the suggestions
On Fri, Sep 13, 2013 at 05:35:27PM +0200, Jakub Jelinek wrote:
> On Fri, Sep 13, 2013 at 03:15:56PM +0200, Jakub Jelinek wrote:
> > On Fri, Sep 13, 2013 at 05:11:09PM +0400, Michael V. Zolotukhin wrote:
> > > > FYI, I'm attaching a WIP patch with the splay tree stuff.
> > > Thanks, I'll take a look
On Fri, Sep 13, 2013 at 03:15:56PM +0200, Jakub Jelinek wrote:
> On Fri, Sep 13, 2013 at 05:11:09PM +0400, Michael V. Zolotukhin wrote:
> > > FYI, I'm attaching a WIP patch with the splay tree stuff.
> > Thanks, I'll take a look. By the way, isn't it better to move splay-tree
> > implementation to
Chung-Ju Wu writes:
> It has been a while since last v2 patch.
> I create a new v3 patch to fix some typo and indentation.
I had a read through out of curiosity, and FWIW, it looks very clean and
well-commented to me. See below for a few questions and comments.
This isn't an official review thou
Hi all,
gcc.target/arm/minmax_minus.c is really only valid when we have
conditional execution available, that is non Thumb1-only targets. I've
added an effective target check for that and used it in the test so that
it does not get run and give a false negative when testing Thumb1 targets.
O
On Fri, Sep 13, 2013 at 11:07 AM, Jakub Jelinek wrote:
> E.g. today I've noticed you've lost OMP_CLAUSE_LINEAR_NO_COPYIN
> comment that has been added to tree.h recently, but you haven't
> actually moved it into tree-core.h.
Sorry about that. I remember an update conflict, but I thought I had
i
Hi!
I've merged today gomp-4_0-branch from the trunk, but I'm seeing various
regressions.
The ICEs look like (e.g. on libgomp.c/simd-1.c (and other simd tests):
0xadf61a crash_signal
../../gcc/toplev.c:335
0x571a0e tree_check2(tree_node*, char const*, int, char const*, tree_code,
tree_c
On 09/13/2013 03:54 AM, Richard Biener wrote:
On Thu, Sep 12, 2013 at 11:09 PM, Andrew MacLeod wrote:
There are 2 parts of tre-ssa-ter.c to address.
is_replaceable_p() is also used in expr.c, It has a flag to indicate where
its being called from, and we do different checks for each one. There
Ok.
David
On Fri, Sep 13, 2013 at 7:21 AM, Teresa Johnson wrote:
> Testing passes, is the below patch ok for google/4_8?
> Thanks, Teresa
>
> On Thu, Sep 12, 2013 at 10:18 PM, Teresa Johnson wrote:
>> On Thu, Sep 12, 2013 at 2:32 PM, Xinliang David Li
>> wrote:
>>> When absolute path is speci
On Thu, Sep 05, 2013 at 11:11:43AM +0200, Richard Biener wrote:
> On Thu, Sep 5, 2013 at 2:06 AM, Mike Stump wrote:
> > On Aug 30, 2013, at 4:22 PM, Diego Novillo wrote:
> >> Thanks for the suggestions. I've incorporated them into the patch.
> >> It now adds tree-core.h with all the structures,
Hi,
in Ada we can have misaligned array components in record types, that is to say
object with BLKmode whose alignment is lower than that of their type. In this
case IVOPTS can generate misaligned TARGET_MEM_REFs, which will lead to an
unaligned access on strict-alignment platforms.
Tested on
Hi,
with improvements on devirtualization side, the thunks are getting more
problematic
for optimizers. Inliner won't inline them, profiling won't instrument them and
there are other places where they bock optimization.
This patch solves part of the problem by making gimple thunks to go same way
Testing passes, is the below patch ok for google/4_8?
Thanks, Teresa
On Thu, Sep 12, 2013 at 10:18 PM, Teresa Johnson wrote:
> On Thu, Sep 12, 2013 at 2:32 PM, Xinliang David Li wrote:
>> When absolute path is specified for the object file, no prefix will be
>> prepended to the gcda path. If you
On Fri, Sep 13, 2013 at 3:18 AM, Ilya Enkovich wrote:
>> Did you check the above with x32, where Pmode != word_mode on x86_64?
>> The inner UNSPEC will be generated in SImode, but the matching pattern
>>
>> +(define_insn "*_mk"
>> + [(set (match_operand:BND 0 "register_operand" "=B")
>> +(uns
Hi,
this patch fixes detection of recursion in do_estimate_growth to work well with
aliases.
Bootstrapped/regtested x86_64-linux, comitted.
Honza
Index: ChangeLog
===
--- ChangeLog (revision 202565)
+++ ChangeLog (working copy)
Hi,
in Ada parameters can be passed by reference: in this case, the address of the
argument is directly passed to the callee, which dereferences it to access the
argument; now Ada also enables -fexceptions -fnon-call-exceptions, which means
that any pointer dereference is seen as the potential
On 13.09.2013 13:26, Andreas Schwab wrote:
FAIL: g++.dg/cpp0x/auto3.C -std=c++11 (test for errors, line 24)
FAIL: g++.dg/cpp0x/auto3.C -std=c++11 (test for excess errors)
Excess errors:
: error: use of 'auto' in parameter declaration only
available with -std=c++1y or -std=gnu++1y
I'm on it.
On Sep 13, 2013, at 4:21 AM, Jan-Benedict Glaw wrote:
> On Wed, 2013-07-31 18:34:26 +0200, Jan-Benedict Glaw
> wrote:
>> We've seen ICEs while outputting an operand (not even the excessive
>> CISC of a VAX could do that), which should be fixed by this patch:
>>
>> 2013-07-31 Jan-Benedict Gla
FAIL: g++.dg/cpp0x/lambda/lambda-conv4.C (internal compiler error)
FAIL: g++.dg/cpp0x/lambda/lambda-conv4.C (test for excess errors)
Excess errors:
../gcc/testsuite/g++.dg/cpp0x/lambda/lambda-conv4.C: In function ‘void func()’:
../gcc/testsuite/g++.dg/cpp0x/lambda/lambda-conv4.C:12:30: internal com
FAIL: g++.dg/cpp0x/auto3.C -std=c++11 (test for errors, line 24)
FAIL: g++.dg/cpp0x/auto3.C -std=c++11 (test for excess errors)
Excess errors:
: error: use of 'auto' in parameter declaration only available with -\
std=c++1y or -std=gnu++1y
Andreas.
--
Andreas Schwab, sch...@linux-m68k.org
GPG K
The following patch makes loop-distribution able to distribute loops
that have control flow. It does so by adding control dependence
edges to the RDG (removing the need for special-casing the loop
exit condition and its dependencies). With this we can now
distribute the loop of the testcase
f
Kai Tietz writes:
> Hello,
>
> this patch fixes a wrong assumption in c_builtin_function_ext_scope.
> The check for never being called on preexisting symbols (with
> meaning), isn't correct as the turning on of builtins via pragmas
> wasn't considered.
>
> The following sample demonstrate
Hello,
On 12 Sep 08:19, Caroline Tice wrote:
> Yes, that patch is ok.
>
> -- Caroline Tice
> cmt...@google.com
>
> On Thu, Sep 12, 2013 at 3:56 AM, Alexander Ivchenko
> wrote:
> >
> > Is following patch OK?
Checked into main trunk: http://gcc.gnu.org/ml/gcc-cvs/2013-09/msg00425.html
--
Thanks
On Wed, 2013-07-31 18:34:26 +0200, Jan-Benedict Glaw wrote:
> We've seen ICEs while outputting an operand (not even the excessive
> CISC of a VAX could do that), which should be fixed by this patch:
>
> 2013-07-31 Jan-Benedict Glaw
>
> * config/vax/constraints.md (T): Add missing CONSTA
> Well, I have a patch for that, but I no longer think that ubsan should
> imply -fstack-check, since e.g.
>
> int
> main (void)
> {
> int x = -1;
> int b[x - 4];
> /* ... */
> return 0;
> }
>
> segfaults at runtime on int b[x - 4]; line when -fstack-check is used
> (even without sanitizi
On Fri, 13 Sep 2013, Richard Biener wrote:
>
> This is the last cleanup before I start adding new features.
>
> Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
>
> Richard.
>
> 2013-09-13 Richard Biener
>
> * tree-data-ref.h (known_dependences_p): Move here ...
This is the last cleanup before I start adding new features.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
Richard.
2013-09-13 Richard Biener
* tree-data-ref.h (known_dependences_p): Move here ...
* tree-loop-distribution.c (known_dependences_p): ...
Hello,
On 09 Sep 15:11, Kirill Yukhin wrote:
> Hello,
> On 06 Sep 17:41, Kirill Yukhin wrote:
> > Hello,
> >
> > PING.
> PING.
PING.
--
Thanks, K
2013/9/11 Uros Bizjak :
> On Tue, Sep 10, 2013 at 1:38 PM, Ilya Enkovich wrote:
>> Ping^4
>>
>> Could please someone look at this patch? It is mostly i386 target
>> specific and is basic for further MPX based features.
>>
>> Thanks,
>> Ilya
>>
>> 2013/9/2 Ilya Enkovich :
>>> Ping^3
>>>
>>> Attache
Hi,
On 09/13/2013 02:01 AM, Paul Pluzhnikov wrote:
On Wed, Sep 4, 2013 at 9:55 PM, Daniel Krügler
wrote:
Did you mean "pessimises code size", or something else?
Yes.
Daniel's idea proved a good one, and I now have a patch that I am
happy with, and that will be easy to extend to string::at()
On Thu, Sep 12, 2013 at 04:05:48PM +, Joseph S. Myers wrote:
> cause stack overflow that doesn't get detected by the kernel. So maybe
> ubsan should imply -fstack-check or similar.
Well, I have a patch for that, but I no longer think that ubsan should
imply -fstack-check, since e.g.
int
ma
Yes, you are right. Patch rejected due it is pretty useless. Now it
makes me wonder why I have this in my tree as this default got set
already at rev 70194.
Thanks,
Kai
On Thu, 12 Sep 2013, Wei Mi wrote:
> Thanks, fixed. New patch attached.
Thanks. At this point you need feedback from x86 and scheduler maintainers.
I would recommend you to resubmit the patch with a Changelog text, and with
the text of the patch inline in the email (your last mail has the patch
> This is a followup on Thomas' proposal at
> ...
This commit (r202518) breaks bootstrap on x86_64-apple-darwin10
(I have bootstrapped r202538 with the commit reverted):
/opt/gcc/build_w/./gcc/xgcc -B/opt/gcc/build_w/./gcc/
-B/opt/gcc/gcc4.9w/x86_64-apple-darwin10.8.0/bin/
-B/opt/gcc/gcc4.9w/x
>> While it is straight forward to remove the movmisalign path in 4.9 and 4.8,
>> this is not so simple in the 4.7 branch. The reason is that 4.7 uses
>> "to_rtx = expand_normal (tem);" while 4.8 and 4.9 use
>> "to_rtx = expand_expr (tem, NULL_RTX, VOIDmode, EXPAND_WRITE);"
>> which does almost the
On 09/13/2013 11:27 AM, Kai Tietz wrote:
The check for this is broken for some Windows targets, due printf
doesn't support in all cases the %ll width modifier. So why probing,
if we know it works.
Which check? I don't see any, this is my point.
Paolo.
Hi,
the assert in try_make_edge_direct_simple_call does not really do the
right thing when we have actually removed a speculation rather than
just a previously indirect edge direct because the cs->callee might
have been inlined or cloned and thus be different from the constant in
the jump function
2013/9/13 Paolo Carlini :
> Hi,
>
>
> On 09/13/2013 08:51 AM, Kai Tietz wrote:
>>
>> Hello,
>>
>> this patch enables the 'long long' use in libstdc++ for Windows native
>> targets.
>>
>> ChangeLog
>>
>> 2013-09-13 Kai Tietz
>>
>> * config/os/mingw32/os_defines.h (_GLIBCXX_USE_LONG_LONG):
>>
Hi,
On 09/13/2013 08:51 AM, Kai Tietz wrote:
Hello,
this patch enables the 'long long' use in libstdc++ for Windows native targets.
ChangeLog
2013-09-13 Kai Tietz
* config/os/mingw32/os_defines.h (_GLIBCXX_USE_LONG_LONG):
Enable feature.
* config/os/mingw-w64/os_defines.h (
> While it is straight forward to remove the movmisalign path in 4.9 and 4.8,
> this is not so simple in the 4.7 branch. The reason is that 4.7 uses
> "to_rtx = expand_normal (tem);" while 4.8 and 4.9 use
> "to_rtx = expand_expr (tem, NULL_RTX, VOIDmode, EXPAND_WRITE);"
> which does almost the same
Ok for trunk?
Thanks,
Kyrill
2013-09-13 Kyrylo Tkachov
* config/arm/arm.md (arm_cmpsi_insn): Split rI alternative.
Set type attribute correctly. Set predicable_short_it attribute.
(cmpsi_shiftsi): Remove %? from output template.
Ok.
Ramana
On Fri, Sep 13, 2013 at 10:33 AM, Richard Sandiford
wrote:
> Richard Biener writes:
>>> Yeah. I don't think it makes sense to canonise PSI to 32 bits when we
>>> know it has fewer than 32 bits. It's still going to be wrong, and will
>>> still defeat one of the main purposes of canonical constan
Hi all,
This patch splits the r/rI alternative in arm_cmpsi_insn so as to
better specify the "type" attribute. Also, the predicable_short_it
attribute is set properly to conform to the -mrestrict-it rules.
A nearby pattern has a redundant "%?" removed.
Tested arm-none-eabi on a model and bootst
For 4.8 and 4.9
2013-09-13 Christian Bruel
PR target/58314
* config/sh/sh.md (mov_reg_reg): Allow memory reloads.
2013-09-13 Christian Bruel
PR target/58314
* gcc.target/sh/torture/pr58314.c: New test.
Index: gcc/config/sh/sh.md
===
Richard Biener writes:
>> Yeah. I don't think it makes sense to canonise PSI to 32 bits when we
>> know it has fewer than 32 bits. It's still going to be wrong, and will
>> still defeat one of the main purposes of canonical constants, which is
>> to make equality obvious. E.g. we'd still treat
On Thu, Sep 12, 2013 at 10:31 PM, Xinliang David Li wrote:
> Currently -ftree-vectorize turns on both loop and slp vectorizations,
> but there is no simple way to turn on loop vectorization alone. The
> logic for default O3 setting is also complicated.
>
> In this patch, two new options are introd
On Mon, Sep 9, 2013 at 9:01 PM, Teresa Johnson wrote:
> I've attached a patch that implements the cleanup of newline emission
> by the new dump framework as discussed here:
>
> http://gcc.gnu.org/ml/gcc-patches/2013-08/msg01779.html
>
> Essentially, I have removed the leading newline emission from
This removes what loop-distribution calls "components" (SCCs of
the RDG). They are not in any way useful and just an unnecessary
intermediate step.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied.
Richard.
2013-09-13 Richard Biener
* tree-loop-distribution.c (struct rd
Ping
http://gcc.gnu.org/ml/gcc-patches/2013-08/msg01381.html
On Fri, 23 Aug 2013, Marc Glisse wrote:
On Sun, 4 Aug 2013, Gerald Pfeifer wrote:
On Sat, 13 Jul 2013, Marc Glisse wrote:
2013-07-14 Marc Glisse
gcc/cp/
* call.c (build_conditional_expr_1): Handle the case with 1 vector
On Fri, Sep 13, 2013 at 10:08 AM, Richard Sandiford
wrote:
> Richard Biener writes:
(what about partial integer modes which have weird precision (none)?)
>>>
>>> We don't really model that properly yet. Partial modes are just defined
>>> using something like:
>>>
>>> PARTIAL_INT_MODE (SI);
Richard Biener writes:
>>> (what about partial integer modes which have weird precision (none)?)
>>
>> We don't really model that properly yet. Partial modes are just defined
>> using something like:
>>
>> PARTIAL_INT_MODE (SI);
>>
>> i.e. without the partial precision. But trunc_int_for_mode is
On Thu, Sep 12, 2013 at 11:09 PM, Andrew MacLeod wrote:
> There are 2 parts of tre-ssa-ter.c to address.
>
> is_replaceable_p() is also used in expr.c, It has a flag to indicate where
> its being called from, and we do different checks for each one. There is a
> wrapper function stmt_is_replaceab
Hello,
this patch fixes a wrong assumption in c_builtin_function_ext_scope.
The check for never being called on preexisting symbols (with
meaning), isn't correct as the turning on of builtins via pragmas
wasn't considered.
The following sample demonstrate this issue pretty well (it is reduced
tes
On Thu, Sep 12, 2013 at 6:19 PM, Richard Sandiford
wrote:
> Richard Biener writes:
>> On Wed, Sep 11, 2013 at 8:39 PM, Richard Sandiford
>> wrote:
>>> Graham Stott writes:
Hi Richard,
There is some minor testsuite fallout with these patches on MIPS a
couple of tests (see bel
On Thu, 12 Sep 2013, Jakub Jelinek wrote:
> Hi!
>
> This patch fixes PR58392 and some related issues I've discovered.
> move_sese_region_to_fn wasn't remapping loop->simduid VAR_DECL,
> and wasn't setting dest_cfun->has_{simduid,force_vect}_loops
> when needed, the inliner wasn't copying/remappin
95 matches
Mail list logo
