2013/6/25 Kirill Yukhin :
> Hello,
>>
>> Looks good to me; just say "the new Intel...".
> Checked int www CVS.
>
> Thanks, K
Hi, Kirill,
By reviewing webpage of gcc-4.9 changes:
http://gcc.gnu.org/gcc-4.9/changes.html
I noticed there is a duplicate subsection 'IA-32/x86-64'
under "New Targets an
On 7/2/2013 11:31 AM, Chung-Ju Wu wrote:
> I think that would be better to merge the content into single
> 'IA-32/x86-64' subsection. But I have no idea about the order of those
> two items. Perhaps you or Sriraman can make the decision for the
> change? :-)
Hi, thanks for pointing!
I think order o
Since there obviously is no interest in this patch I'm simply going to
attach it to the ticket mentioned below (Bug 56044) in case someone
decides to look into it in the future. If anyone should want to comment
on this please do so at the bug tracker since I probably won't be
following this li
I've find out that compiler may crash with SEGFAULT on some openmp-enabled code
(attached as debug-1.c) called with the following flags:
-g -fopenmp -gdwarf-2 -femit-struct-debug-reduced
Also, can be reproduced with -femit-struct-debug-baseonly
The problem is in dwarf2out.c:350:should_emit_stru
Dear Tobias,
The patch itself seems obvious enough. I have just a nit pick
with the grammar of
"Don't warn for unused dummy arguments when it is used in namelists"
I think it should be either "a dummy argument" or "when they are".
Probably something such as
"Don't warn as unused for dummy argume
Hello,
As mentioned by Uros [1], we have few intrinsics which described in
Intel Spec, but absent in bmiintin.h
Attached patch resolves this.
Bootstrap passing. Updated tests passing on BMI-featured HW.
ChangeLog:
2013-07-02 Kirill Yukhin
* config/i386/bmiintrin.h (_blsi_u32): New.
Hello,
As they stand the ld1 and st1 intrinsics are broken. The constraints
used in their assembler definitions are not strong enough to prevent
GCC from performing some exciting reorders moving around loads and
stores.
Rewriting these intrinsics as RTL builtins settles GCC's errant
behaviour an
On 20/06/13 14:50, Kyrylo Tkachov wrote:
Hi all,
This patch adjusts the andsi3, iorsi3, xorsi3 patterns for -mrestrict-it. It
is done by adding 16-bit alternatives that are fit be cond_exec'd when
arm_restrict_it is on.
Tested arm-none-eabi on model and qemu with ARMv7 and ARMv8. Also tested as
On 27 June 2013 17:09, Ian Bolton wrote:
> 2013-06-27 Ian Bolton
>
> gcc/
> * config/aarch64/aarch64.md (*extr_insv_reg): New pattern.
>
> testsuite/
> * gcc.target/aarch64/bfxil_1.c: New test.
> * gcc.target/aarch64/bfxil_2.c: Likewise.
>
OK
/Marcus
On 24 June 2013 16:57, Ian Bolton wrote:
> 2013-06-24 Ian Bolton
>
> * gcc.target/config/aarch64/insv_1.c: Update to show it doesn't work
> on big endian.
> * gcc.target/config/aarch64/insv_2.c: New test for big endian.
> * lib/target-supports.exp: Define aarch6
On 25 June 2013 17:48, Tejas Belagod wrote:
> 2013-06-25 Tejas Belagod
>
> gcc/
> * config/aarch64/aarch64-protos.h (cpu_vector_cost): New.
> (tune_params): New member 'const vec_costs'.
> * config/aarch64/aarch64.c (generic_vector_cost): New.
> (generic_tunings
On 25 June 2013 18:04, Ian Bolton wrote:
> 2013-06-25 Ian Bolton
>
> gcc/
> * config/aarch64/aarch64-simd.md (absdi2): Support abs for
> DI mode.
>
> testsuite/
> * gcc.target/aarch64/abs_1.c: New test.
OK
/Marcus
On 21/06/13 11:53, Kyrylo Tkachov wrote:
Hi all,
This patch adjusts the patterns in thumb2.md to conform with -mrestrict-it.
16-bit alternatives are added, patterns that explicitly generate IT blocks
that don't conform are disabled or replaced with appropriate patterns for
arm_restrict_it. A new
On 21/06/13 11:53, Kyrylo Tkachov wrote:
Hi all,
This is the final patch in the series. It adjusts patterns in arm.md that do
arithmetic operations and comparisons. We add alternatives for 16-bit
encodings to the patterns. Some alternatives and patterns are disabled for
arm_restrict_it or their
On 01/07/13 12:05, Kyrylo Tkachov wrote:
Hi Richard,
This hurts code size.
Therefore I've disabled the new peephole2 for optimize_insn_for_size_p
so that
the original peephole before r200197 is used when optimising for size.
I've also added a test to confirm that the new peephole2 for the no
Jakub Jelinek wrote:
>Hi!
>
>My recent patch introduced a regression as shown by the attached
>testcases, there was nothing actually checking the type of the
>induction,
>but because we allowed only INTEGER_CST steps that implied
>integral/pointer
>types only. With -ffast-math I don't see a reas
On Tue, Jul 2, 2013 at 1:16 AM, Marc Glisse wrote:
> On Mon, 1 Jul 2013, Gabriel Dos Reis wrote:
>
>> On Mon, Jul 1, 2013 at 10:36 AM, David Malcolm
>> wrote:
>>>
>>> My plan for removal of global variables in gcc 4.9 [1] calls for several
>>> hundred new classes, which will be singletons in a cl
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 Russian team of translators. The file is available at:
http://translationproject.org/latest/gcc/ru.po
(This file, 'gcc-4.8.0.ru.po', has just
This patch removes unused types and variables claimed to handle abi
types in aarch64.
OK for the trunk?
Thanks,
Yufeng
gcc/
* config/aarch64/aarch64.h (enum arm_abi_type): Remove.
(ARM_ABI_AAPCS64): Ditto.
(arm_abi): Ditto.
(ARM_DEFAULT_ABI): Ditto.diff --git a
On 07/01/2013 08:12 PM, Peter Bergner wrote:
> libitm/
> * acinclude.m4 (LIBITM_CHECK_AS_HTM): New.
> * configure.ac: Use it.
> (AC_CHECK_HEADERS): Check for sys/auxv.h.
> (AC_CHECK_FUNCS): Check for getauxval.
> * config.h.in, configure: Rebuild.
> * confi
Hi guys,
Thanks for the review and for your responses - please find my answers
below.
> Yes, I looked on the patch in detail this week (I am currently on a travel
> with
> sporadic internet access and my days before leaving was extremely hectic).
> The
> patch is OK except for the expr.c bits I
Hi Richard,
Thanks for the review.
> (define_insn_and_split "*arm_cmpdi_unsigned"
> [(set (reg:CC_CZ CC_REGNUM)
> - (compare:CC_CZ (match_operand:DI 0 "s_register_operand" "r")
> -(match_operand:DI 1 "arm_di_operand" "rDi")))]
> + (compare:CC_CZ (match_operand
As discussed in http://gcc.gnu.org/ml/gcc/2013-07/msg8.html there
are four macros that create implicit uses of the input_location global.
I'm attaching a patch which removes these macros.
The patch converts most uses of the macros to a simple expansion of the
former definition, making the use
Hi,
I have just committed backports of the following revisions from trunk
to linaro/gcc-4_8-branch:
r199810 as r200607
r200019 as r200608
r200062 as r200613
r200096 as r200614
Thanks,
Christophe.
> "David" == David Malcolm writes:
David> gcc/java/
David> 2013-07-02 David Malcolm
David> * class.c (maybe_layout_super_class): Update comment.
David> * decl.c (java_add_stmt): Remove use of input_filename macro.
David> * jcf-parse.c (set_source_filename): Remove use of
David> input_f
On 02/07/13 16:13, Kyrylo Tkachov wrote:
Hi Richard,
Thanks for the review.
(define_insn_and_split "*arm_cmpdi_unsigned"
[(set (reg:CC_CZ CC_REGNUM)
- (compare:CC_CZ (match_operand:DI 0 "s_register_operand" "r")
- (match_operand:DI 1 "arm_di_operand" "rDi"))
On Tue, 2013-07-02 at 06:37 -0500, Gabriel Dos Reis wrote:
> On Tue, Jul 2, 2013 at 1:16 AM, Marc Glisse wrote:
> > On Mon, 1 Jul 2013, Gabriel Dos Reis wrote:
> >
> >> On Mon, Jul 1, 2013 at 10:36 AM, David Malcolm
> >> wrote:
> >>>
> >>> My plan for removal of global variables in gcc 4.9 [1] ca
Hello everyone,
Is this Patch OK for trunk?
-Balaji V. Iyer.
> -Original Message-
> From: Iyer, Balaji V
> Sent: Monday, June 17, 2013 8:10 AM
> To: gcc-patches@gcc.gnu.org
> Subject: [PING][PATCH] for for c/PR57541
>
> Hello Everyone,
> Is this patch OK for trunk?
>
> Tha
On 30 June 2013 08:55, Tim Shen wrote:
> ChangeLog entry fixed.
Some comments on the latest patch (it's a long-ish email but most of
my comments are very minor things regarding the _RegexMask type) ...
Defining struct _RegexMask in the typedef declaration makes it public,
I think I would prefer i
As already proposed in the comments to PR52267, this patch restricts the
vrp66.c test to targets with int sizes other (and if standards conformant,
that means larger) than 16 bit.
It also adds a test for 16 bit int targets.
I found that when running the original test using the sim target of avr-
> The choice appears to be to continue to have broken volatile bitfields
> on ARM with no way for users to make them conform to the ABI, or to
> change things so that they conform to the ABI if you specify
> -fstrict-volatile-bitfields explicitly and to the C/C++ standard by
> default, without
Hi Richard,
This is another small microMIPS cleanup, this time for the testsuite. Okay?
Thanks,
Catherine
jalrs.cl
Description: jalrs.cl
jalrs.patch
Description: jalrs.patch
On Tue, 2013-07-02 at 08:16 +0200, Marc Glisse wrote:
> On Mon, 1 Jul 2013, Gabriel Dos Reis wrote:
>
> > On Mon, Jul 1, 2013 at 10:36 AM, David Malcolm wrote:
> >> My plan for removal of global variables in gcc 4.9 [1] calls for several
> >> hundred new classes, which will be singletons in a cla
Hi Andrew,
Please find the updated patch in the attachment that addresses your
comments.
It now builds both ilp32 and lp64 multilibs by default, with the
--with-multilib-list support remaining to provide options to turn off
one of them.
-mabi=ilp32 and -mabi=lp64 are now the command line o
"Moore, Catherine" writes:
> This is another small microMIPS cleanup, this time for the testsuite. Okay?
OK, thanks. I assume this is really a fix rather than a cleanup though?
"Cleanup" to me means "make things look prettier", but it sounds like
these tests were failing as-is.
Richard
On Wed, Jul 3, 2013 at 1:11 AM, Jonathan Wakely wrote:
> Defining struct _RegexMask in the typedef declaration makes it public,
> I think I would prefer if the struct definition was a private member
> of regex_traits and only the char_class_type typedef was public.
> Either that or just rename the
> -Original Message-
> From: Richard Sandiford [mailto:rdsandif...@googlemail.com]
> Sent: Tuesday, July 02, 2013 3:07 PM
> To: Moore, Catherine
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [Patch] microMIPS cleanup for testsuite
>
> "Moore, Catherine" writes:
> > This is another small
On Tue, 2013-07-02 at 10:33 -0600, Tom Tromey wrote:
> > "David" == David Malcolm writes:
>
> David> gcc/java/
> David> 2013-07-02 David Malcolm
> David>* class.c (maybe_layout_super_class): Update comment.
> David>* decl.c (java_add_stmt): Remove use of input_filename macr
Forgot the patch file >.<
--
Tim Shen
regex_traits.patch
Description: Binary data
Hi,
>Second, making it `const` enables more potential compiler
>optimizations.
Stylistically, we already discussed this in the past and the outcome was that
normally we don't use const by value anywhere. But personally I don't have a
strong opinion and certainly I would not object if it makes
> Echoing Tom's question: given that last 2011 commit for improved macro
> diagnostics which appears to assume a diagnostic->location, is it
> reasonable to expect every call to diagnostic_report_current_function
> call to have the right filename for diagnostic->location for the lang
> hook?
Not o
This patch is for the google/gcc-4_7 and google/gcc-4_8 branches.
When using -gsplit-dwarf, we use die_checksum to calculate a hash
over the compilation unit for the DW_AT_dwo_id attribute. When
hashing location expressions, the hash may include pointers that
can vary from run to run of GCC. Thi
On 2 July 2013 20:27, Paolo Carlini wrote:
>
>
> Hi,
>
>>Second, making it `const` enables more potential compiler
>>optimizations.
>
> Stylistically, we already discussed this in the past and the outcome was that
> normally we don't use const by value anywhere. But personally I don't have a
> st
On Tue, 2 Jul 2013, David Malcolm wrote:
> So I envisage a "--enable-shared" configuration switch to opt-in to the
> shared library code, but I want as minimize the difference between the
> two cases.
--enable-host-shared or --enable-shared-gcc or some other such option
name, please, not plain -
> I've find out that compiler may crash with SEGFAULT on some openmp-enabled
> code (attached as debug-1.c) called with the following flags:
> -g -fopenmp -gdwarf-2 -femit-struct-debug-reduced
> Also, can be reproduced with -femit-struct-debug-baseonly
>
> The problem is in dwarf2out.c:350:should
On Mon, Jul 1, 2013 at 8:36 AM, David Malcolm wrote:
> My plan for removal of global variables in gcc 4.9 [1] calls for several
> hundred new classes, which will be singletons in a classic monolithic
> build, but have multiple instances in a shared-library build.
>
> In order to avoid the register
On Tue, Jul 2, 2013 at 11:53 AM, Yufeng Zhang wrote:
> Hi Andrew,
>
> Please find the updated patch in the attachment that addresses your
> comments.
>
> It now builds both ilp32 and lp64 multilibs by default, with the
> --with-multilib-list support remaining to provide options to turn off one of
>>> This patch replaces expand_location_to_spelling_point with
>>> expand_location so that macros will not be expanded.
OK for google/gcc-4_8.
-cary
On Tue, 2013-07-02 at 13:54 -0700, Andrew Pinski wrote:
> On Mon, Jul 1, 2013 at 8:36 AM, David Malcolm wrote:
> > My plan for removal of global variables in gcc 4.9 [1] calls for several
> > hundred new classes, which will be singletons in a classic monolithic
> > build, but have multiple instanc
On 2 July 2013 20:10, Tim Shen wrote:
> On Wed, Jul 3, 2013 at 1:11 AM, Jonathan Wakely wrote:
>> I think the _RegexMask constructor should be explicit.
>
> I intentionally make _RegexMask act like an unsigned number. Since
> lookup_collatename is required to return a value compares equal to 0,
O
On Tue, Jul 2, 2013 at 1:37 PM, Cary Coutant wrote:
>
> This patch is for the google/gcc-4_7 and google/gcc-4_8 branches.
>
> When using -gsplit-dwarf, we use die_checksum to calculate a hash
> over the compilation unit for the DW_AT_dwo_id attribute. When
> hashing location expressions, the hash
On Tue, 2013-07-02 at 20:49 +, Joseph S. Myers wrote:
> On Tue, 2 Jul 2013, David Malcolm wrote:
>
> > So I envisage a "--enable-shared" configuration switch to opt-in to the
> > shared library code, but I want as minimize the difference between the
> > two cases.
>
> --enable-host-shared or
When using -gsplit-dwarf, we use die_checksum to calculate a hash
over the compilation unit for the DW_AT_dwo_id attribute. When
hashing location expressions, the hash may include pointers that
can vary from run to run of GCC. This patch fixes loc_checksum
to use the (already existing) hash_loc_o
On Tue, Jul 2, 2013 at 2:09 PM, David Malcolm wrote:
> On Tue, 2013-07-02 at 13:54 -0700, Andrew Pinski wrote:
>> On Mon, Jul 1, 2013 at 8:36 AM, David Malcolm wrote:
>> > My plan for removal of global variables in gcc 4.9 [1] calls for several
>> > hundred new classes, which will be singletons i
Ping~
Can I get an OK please if there is no objection?
Regards,
Yufeng
On 06/26/13 23:39, Yufeng Zhang wrote:
This patch updates assign_parm_find_data_types to assign passed_mode and
nominal_mode with the mode of the built pointer type instead of the
hard-coded Pmode in the case of pass-by-ref
On Wed, 26 Jun 2013, Iyer, Balaji V wrote:
> Hello Everyone,
> This patch will fix a FAIL in one of the test cases for array
> notations. The reason for fail is that the array sizes were huge and thus it
> was causing a stack overflow. This patch should fix the issue. I am
> committing thi
On Mon, Jul 1, 2013 at 2:12 PM, Uros Bizjak wrote:
> On Mon, Jul 1, 2013 at 8:01 PM, Sriraman Tallam wrote:
>
>>So, something like the patch attached?
>>
>> * config/i386/i386.c (ix86_option_override_internal): Turn
>> on all -mavx target flags by default as they are dependent
This patch fixes the discriminator assignments for stmts. It used to
apply a signle locus for everything, which is not right (will
invalidate locus for macro expansion).
Bootstrapped and passed regression test.
OK for google-4_8?
Thanks,
Dehao
Index: gcc/input.c
Two weeks since the last comment... Anyone?
http://gcc.gnu.org/ml/gcc-patches/2013-06/msg01089.html
All fixed. I use `_RegexMask` rather than `const _RegexMask&` just
because, IMO, const reference is for something very large and doesn't
need to be every-byte-accessed.
Add _RegexMask::_S_valid_mask = 0x3 instead of using the literal everywhere.
--
Tim Shen
regex_traits.patch
Description: Bina
> -Original Message-
> From: Hans-Peter Nilsson [mailto:h...@bitrange.com]
> Sent: Tuesday, July 02, 2013 7:24 PM
> To: Iyer, Balaji V
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH] Fix for PR 57692
>
> On Wed, 26 Jun 2013, Iyer, Balaji V wrote:
> > Hello Everyone,
> > This pat
On Tue, Jul 2, 2013 at 10:32 AM, Kirill Yukhin wrote:
> Hello,
> As mentioned by Uros [1], we have few intrinsics which described in
> Intel Spec, but absent in bmiintin.h
> Attached patch resolves this.
>
> Bootstrap passing. Updated tests passing on BMI-featured HW.
>
> ChangeLog:
> 2013-07-02
On Wed, Jul 03, 2013 at 08:14:25AM +0200, Uros Bizjak wrote:
> On Tue, Jul 2, 2013 at 10:32 AM, Kirill Yukhin
> wrote:
> BTW: Do we want to backport this patch (and your previous) to 4.8 branch?
I'd say yes to both.
Jakub
> I'd say yes to both.
I'll prepare a patch for 4.8 then
Thanks, K
64 matches
Mail list logo
