Hi Andre,
> Because this patch is obvious I plan to commit it tomorrow if no one objects?!
The patch itself is obviously OK.
About the test case: In general, it is better not to change existing
test cases unless absolutely necessary (e.g. adjust an error message).
This makes it easier to track
> Hi,
> this patch makes it possible for non-LTO alias oracle to TBAA disambiguate
> pointer types. It makes void * conflicting with all of them and does not put
> it
> to alias set 0. It also preserves the property that qualifiers of pointer-to
> type should not matter to determine the alias set
Ping?
> -Original Message-
> From: Thomas Preud'homme [mailto:thomas.preudho...@arm.com]
> Sent: Thursday, April 30, 2015 3:19 PM
> To: Thomas Preud'homme; Richard Earnshaw; 'gcc-patches@gcc.gnu.org';
> Marcus Shawcroft; Ramana Radhakrishnan
> (ramana.radhakrish...@arm.com)
> Subject: RE:
> > >
> > > I am not sure if TYPE_MAIN_VARIANT is really needed here. What I know is
> > > that
> > > complete types may have incomplete variants.
> >
> > How can that be? TYPE_FIELDS is shared across variants and all variants
> > should be layed out.
>
> Because TYPE_FILEDS are not always sh
Hi,
this patch makes it possible for non-LTO alias oracle to TBAA disambiguate
pointer types. It makes void * conflicting with all of them and does not put it
to alias set 0. It also preserves the property that qualifiers of pointer-to
type should not matter to determine the alias set and that poin
> >
> > I am not sure if TYPE_MAIN_VARIANT is really needed here. What I know is
> > that
> > complete types may have incomplete variants.
>
> How can that be? TYPE_FIELDS is shared across variants and all variants
> should be layed out.
Because TYPE_FILEDS are not always shared across varian
Segher suggested that the change to favor (plus (ashift (...))) as a
split point be conditionalized on the shift count being a constant --
seems quite reasonable to me.
Bootstrapped and regression tested on x86_64-linux-gnu as well as
verified the 300+ test files generate the same code befor
Jakub pointed out a formatting issue with one of my recent memory leak
patches. It was a tabs vs spaces issue.
This trivial patch replaces spaces with tabs where appropriate in
tree-ssa-threadedge.c. It has no functional changes.
Bootstrapped and regression tested (as part of a different
Hi,
The first patch backports the fix of PR
target/64011(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64011) to the 4.9
branch from trunk r219717,
and the second patch backports the fix of PR
target/61749(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61749) to the 4.9
branch from trunk r215046.
On Thu, Apr 30, 2015 at 12:09 PM, Jonathan Wakely wrote:
> On 30/04/15 11:19 -0700, Doug Evans wrote:
>>
>> Hi.
>>
>> This patch adds operator-> xmethod support for std::unique_ptr.
>>
>> Regression tested on amd64-linux.
>>
>> Ok to commit?
>
>
> OK, thanks.
Committed, thanks.
On Tue, May 26, 2015 at 3:11 PM, Matthew Fortune
wrote:
> The change for MIPS looks fine by visual inspection and I've built both
> a default pie and default no-pie compiler. The default pie won't build
> glibc but I am pretty sure it is not down to this patch. I haven't had
> time to look into wh
Hi,
I've committed the attached patches to fix PR target/65979
which is a 5/6 regression in behalf of Oleg. The problematic
peephole can clobber a register before its use. See PR65979 for
details. The first patch is for trunk and the second one is
a slightly different version against gcc-5-bran
The change for MIPS looks fine by visual inspection and I've built both
a default pie and default no-pie compiler. The default pie won't build
glibc but I am pretty sure it is not down to this patch. I haven't had
time to look into why it won't build though, something related to
selecting the CRT f
/gcc/gcc-mainline-test/gcc/../libcpp/include \
> > -DBASEVER="\"6.0.0\"" -DDATESTAMP="\" 20150526\"" \
> > -DREVISION="\" [trunk revision 223719]\"" \
> > -DDEVPHASE="\" (experimental)\""
build
> -I/home/wschmidt/gcc/gcc-mainline-test/gcc
> /../include -I/home/wschmidt/gcc/gcc-mainline-test/gcc/../libcpp/include \
> -DBASEVER="\"6.0.0\"" -DDATESTAMP="\" 20150526\"" \
> -DREVISION="\" [trunk revision 223719]
On Wed, Apr 29, 2015 at 11:14 AM, Jonathan Wakely wrote:
> On 29/04/15 10:57 -0700, Doug Evans wrote:
>>
>> On Tue, Apr 28, 2015 at 5:16 AM, Jonathan Wakely
>> wrote:
>>>
>>> On 27/04/15 15:44 -0700, Doug Evans wrote:
PR libstdc++/65839
* python/libstdcxx/v6/xm
We don't need the C99-like treatment if we don't expect NaN's.
Tested on x86_64-suse-linux, applied on the mainline.
2015-05-26 Eric Botcazou
* gcc-interface/trans.c (Attribute_to_gnu) : Do not
bother about NaN's if Machine_Overflows is true.
--
Eric BotcazouIndex: gcc-int
When an object is declared with an initializing expression and has its
elaboration delayed, the following code in gnat_to_gnu is invoked:
/* If this object has its elaboration delayed, we must force
evaluation of GNU_EXPR right now and save it for when the object
> This change ensures on all targets that A**B = A**C when B is a small
> static constant in the range 0 .. 4 and C is a variable with the same
> value. Previously for Float and Long_Float on some targets, this was
> not the case. The results given were both within the error bounds that
> are allow
> An arbitrary restriction was imposed on renaming in conjunction with the new
> Aspect/Pragma Volatile_Full_Access for implementation reasons: the compiler
> was rejecting renamings of components of Volatile_Full_Access objects.
This is the second part of the overhaul, tested on x86_64-suse-linux
> An arbitrary restriction was imposed on renaming in conjunction with the new
> Aspect/Pragma Volatile_Full_Access for implementation reasons: the compiler
> was rejecting renamings of components of Volatile_Full_Access objects.
The reason was that supporting Volatile_Full_Access requires an over
Chung-Lin posted this patch last year but it seems never to have been
reviewed:
https://gcc.gnu.org/ml/gcc-patches/2014-06/msg00714.html
I've just re-applied and re-tested it and it still seems to be good.
Can somebody please take a look at it?
-Sandra
Hi!
This patch adds changes similar to those that have been added
for OpenMP 4.1 to the C FE so far. Note, taskloop with C++ class
iterators isn't yet supported and the testsuite coverage is limited too.
2015-05-26 Jakub Jelinek
gcc/c/
* c-parser.c (c_parser_pragma): Return result of
On 05/25/15 21:18, Jason Merrill wrote:
On 05/25/2015 04:55 PM, Nathan Sidwell wrote:
else if (TYPE_CANONICAL (to_type) != to_type)
TYPE_CANONICAL (t)
= build_reference_type_for_mode (TYPE_CANONICAL (to_type),
mode, false);
But we're passing 'false' in a
On 26 May 2015 at 18:25, Alan Lawrence wrote:
> Christophe Lyon wrote:
>>
>> On 22 April 2015 at 19:36, Alan Lawrence wrote:
>>>
>>> In the first revision of Christophe Lyon's advsimd-intrinsics tests,
>>> https://gcc.gnu.org/ml/gcc-patches/2014-06/msg00532.html , both
>>> gcc-dg-runtest (to asse
> On Fri, 22 May 2015, Jan Hubicka wrote:
>
> > >
> > > And no, I'm hesitant to change operand_equal_p too much. It's
> > > very much deep-rooted into GENERIC.
> >
> > OK, as another option, i can bring relevant logic from operand_equal_p
> > to ipa-icf and separate it into the compare_operand
> Hi,
>
> On Fri, 22 May 2015, Jan Hubicka wrote:
>
> > Index: tree-streamer-out.c
> > ===
> > --- tree-streamer-out.c (revision 223508)
> > +++ tree-streamer-out.c (working copy)
> > @@ -346,6 +346,7 @@ pack_ts_type_common_v
> > Will do if we agree on having this.
> >
> > I know you would like ipa-icf to keep original bodies and use them for
> > inlining declaring alias sets to be function local. This is wrong plan.
> > Consder:
> >
> > void t(int *ptr)
> > {
> > *ptr=1;
> > }
> >
> > int a(int *ptr1, int *ptr
> > Now the change does not really translate to great increase of
> > disambiguations
> > for Firefox (it seems more in noise). The reason is the pointer_type
> > globbing
> > in alias.c.
>
> Yeah, we only get the improvement because of some "hack" in the tree
> alias oracle which also uses the
Hi all,
unfortunately introduced the latest optimization for this patch an ICE in
pFUnit again. The issue occurs when the -fbounds-check is used while
allocating an array of non-polymorphic or intrinsic type. It is caused by the
location of the gfc_expr of e3rhs not set which is needed to generate
On Tue, May 26, 2015 at 9:17 AM, Lynn A. Boger
wrote:
>
> On 05/22/2015 05:50 PM, Ian Lance Taylor wrote:
>>
>> On Fri, May 22, 2015 at 3:11 PM, Andrew Chambers
>> wrote:
>>>
>>> I'm not suggesting breaking go conventions, I just think the default if
>>> no
>>> GOARCH is specified then it should
w.r.t. the PR48052, here is the patch which finds out if scev would wrap or not.
The patch symbolically evaluates if valid_niter >= loop->nb_iterations is true.
In that case the scev would not wrap.
Currently, we only look for two special 'patterns', which are sufficient to
analyze the test cases.
On 05/06/2015 12:28 PM, Jason Merrill wrote:
On 05/05/2015 04:33 PM, Aldy Hernandez wrote:
On 05/05/2015 02:08 PM, Jason Merrill wrote:
On 05/04/2015 09:29 PM, Aldy Hernandez wrote:
The code handling parameter DIEs needed a little tweaking for variable
length template arguments. I've relaxed
Christophe Lyon wrote:
On 22 April 2015 at 19:36, Alan Lawrence wrote:
In the first revision of Christophe Lyon's advsimd-intrinsics tests,
https://gcc.gnu.org/ml/gcc-patches/2014-06/msg00532.html , both
gcc-dg-runtest (to assemble only) and c-torture-execute were used. In review
the gcc-dg-run
On 05/22/2015 05:50 PM, Ian Lance Taylor wrote:
On Fri, May 22, 2015 at 3:11 PM, Andrew Chambers
wrote:
I'm not suggesting breaking go conventions, I just think the default if no
GOARCH is specified then it should match --target.
Sounds good to me.
Perhaps we could check the symlink name f
Hi,
On Fri, 22 May 2015, Richard Biener wrote:
> >It's currently regstrapping on x86_64-linux, okay for trunk if that
> >passes?
>
> OK.
r223704 now. I've tried to also add the fortran runtime testcase from the
PR that now exists, but failed. The gfortran.dg/vect testsuite is strange
and c
Thanks James
Committed as https://gcc.gnu.org/viewcvs/gcc?view=revision&revision=223703
Regards,
Venkat.
> -Original Message-
> From: James Greenhalgh [mailto:james.greenha...@arm.com]
> Sent: Tuesday, May 26, 2015 2:09 PM
> To: Kumar, Venkataramanan
> Cc: gcc-patches@gcc.gnu.org; Marcu
David Edelsohn writes:
> On Tue, May 26, 2015 at 10:51 AM, Richard Sandiford
> wrote:
>> Bill Schmidt writes:
>>> Hi Richard,
>>>
>>> Unfortunately this broke the Power builds:
>>>
>>> /home/wschmidt/gcc/gcc-mainline-base/gcc/config/rs6000/constraints.md:211:
>>> reference to unknown predicate '
On Tue, May 26, 2015 at 10:51 AM, Richard Sandiford
wrote:
> Bill Schmidt writes:
>> Hi Richard,
>>
>> Unfortunately this broke the Power builds:
>>
>> /home/wschmidt/gcc/gcc-mainline-base/gcc/config/rs6000/constraints.md:211:
>> reference to unknown predicate 'mem_operand_gpr'
>> /home/wschmidt/
On 05/22/15 17:44:23, Jan Hubicka wrote:
> PR 66181 is about ICE in verify_type that complains that type and its variant
> differs
> by TYPE_NO_FORCE_BLK.
Also PR 66283, on IA64. Has a small test case.
Bill Schmidt writes:
> Hi Richard,
>
> Unfortunately this broke the Power builds:
>
> /home/wschmidt/gcc/gcc-mainline-base/gcc/config/rs6000/constraints.md:211:
> reference to unknown predicate 'mem_operand_gpr'
> /home/wschmidt/gcc/gcc-mainline-base/gcc/config/rs6000/constraints.md:242:
> referen
On 22/05/15 18:48 +0100, Jonathan Wakely wrote:
On 22/05/15 16:21 +0100, Jonathan Wakely wrote:
On 22/05/15 17:13 +0200, Jakub Jelinek wrote:
On Fri, May 22, 2015 at 03:59:47PM +0100, Jonathan Wakely wrote:
+ alignas(alignof(_Tp2)) unsigned char _M_storage[sizeof(_Tp)];
Is alignof(_Tp2)
I don't have commit access. I would appreciate if someone does that for me.
Thanks,
-Aditya
> Date: Fri, 22 May 2015 14:52:29 -0600
> From: l...@redhat.com
> To: hiradi...@msn.com; richard.guent...@gmail.com; gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH]
I don't have commit access. I would appreciate if someone does that for me.
Thanks,
-Aditya
> Date: Tue, 26 May 2015 08:14:41 -0600
> From: l...@redhat.com
> To: hiradi...@msn.com; gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH] Break when has_sample is tr
Hi Richard,
Unfortunately this broke the Power builds:
/home/wschmidt/gcc/gcc-mainline-base/gcc/config/rs6000/constraints.md:211:
reference to unknown predicate 'mem_operand_gpr'
/home/wschmidt/gcc/gcc-mainline-base/gcc/config/rs6000/constraints.md:242:
reference to unknown predicate 'small_dat
On 05/23/2015 04:46 AM, Manuel López-Ibáñez wrote:
This patch adds descriptive names to various constants in line-map.c.
There were some differences between the constants used a various
places, but my understanding is that these differences are arbitrary
and it is easier to understand the code i
Some utility subprograms used in GNATprove were generally useful, and
thus are better located in GNAT frontend than in GNATprove code.
Tested on x86_64-pc-linux-gnu, committed on trunk
2015-05-26 Yannick Moy
* inline.adb (Has_Initialized_Type): Adapt to new names.
* sem_aux.ad
On 05/26/2015 08:07 AM, Aditya Kumar wrote:
gcc/ChangeLog:
2015-05-26 Aditya Kumar
* auto-profile.c (afdo_calculate_branch_prob): Break once has_sample
is true.
OK. Please install onto the trunk.
Thanks,
jeff
On Sun, May 24, 2015 at 8:47 AM, Bin.Cheng wrote:
> On Fri, May 22, 2015 at 7:45 PM, Richard Biener
> wrote:
>> On Wed, May 20, 2015 at 11:41 AM, Bin Cheng wrote:
>>> Hi,
>>> As we know, GCC is too conservative when checking overflow behavior in SCEV
>>> and loop related optimizers. Result is s
Hello.
Following patch enhanced dump output for LTO balanced map.
Sample output:
Partition sizes:
partition 0 contains 2413 (13.33%) symbols and 56646 (3.62%) insns
partition 1 contains 2006 (11.08%) symbols and 55901 (3.57%) insns
partition 2 contains 1954 (10.79%) symbols and 61054 (3.90%) ins
On 05/25/2015 01:52 PM, Andreas Tobler wrote:
On 25.05.15 14:42, Dominique Dhumieres wrote:
Wouldn't it be better to remove the target selector and instead add:
/* { dg-require-effective-target pie } */
...
This allows the tests to be run on Darwin and fail because the code
generated does not
On 05/25/2015 06:42 AM, Dominique Dhumieres wrote:
Wouldn't it be better to remove the target selector and instead add:
/* { dg-require-effective-target pie } */
...
This allows the tests to be run on Darwin and fail because the code generated
does not
match the scan-assembler:
FAIL: gcc.tar
gcc/ChangeLog:
2015-05-26 Aditya Kumar
* auto-profile.c (afdo_calculate_branch_prob): Break once has_sample is
true.
---
gcc/auto-profile.c | 9 +++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/gcc/auto-profile.c b/gcc/auto-profile.c
index 55dd8d1..1936487 10
On 05/20/2015 10:01 AM, Marat Zakirov wrote:
On 05/12/2015 02:16 PM, Marat Zakirov wrote:
On 04/07/2015 03:22 PM, Jakub Jelinek wrote:
How are the automatic misaligned variables different from say heap
allocated ones, or global vars etc.?
No difference you are right Jakub. Shadow memory in
On Thu, May 21, 2015 at 04:57:00PM +0100, Matthew Wahab wrote:
> On Aarch64, the __sync builtins are implemented using the __atomic operations
> and barriers. This makes the the __sync builtins inconsistent with their
> documentation which requires stronger barriers than those for the __atomic
> bu
The following fixes the testcase in PR66142
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
Richard.
2015-05-26 Richard Biener
PR tree-optimization/66142
* tree-ssa-sccvn.c (vn_reference_lookup_3): Manually compare
MEM_REFs for the same base ad
Hi,
On Fri, 22 May 2015, Jan Hubicka wrote:
> Index: tree-streamer-out.c
> ===
> --- tree-streamer-out.c (revision 223508)
> +++ tree-streamer-out.c (working copy)
> @@ -346,6 +346,7 @@ pack_ts_type_common_value_fields (s
On Wed, May 13, 2015 at 9:49 AM, Christian Bruel wrote:
> 2 parts for maintainers
>
> - c-family: machine descriptions export macro definitions into c
> implementation : need to export 'builtin_define_with_int_value' and '
> builtin_define_type_sizeof'
>
>Could a global reviewer check this ?
On Tue, May 26, 2015 at 2:42 PM, Christian Bruel wrote:
>
>
> On 05/26/2015 12:11 PM, Ramana Radhakrishnan wrote:
>>
>>
>>
>> On 13/05/15 11:16, Christian Bruel wrote:
>>>
>>>
>>> > - [2.2/6]: Redefine TARGET_MACROS for #pragma GCC target without
>>> "thumbness" the glue.
>>>
>>> Here it is,
An arbitrary restriction was imposed on renaming in conjunction with the new
Aspect/Pragma Volatile_Full_Access for implementation reasons: the compiler
was rejecting renamings of components of Volatile_Full_Access objects.
It is lifted by this change and the following package must now be accepted
On Tue, May 26, 2015 at 02:21:16PM +0200, Kai Tietz wrote:
> Hi,
>
> I am not sure if suggested fix is the thing we want here. But infact
> having here a VAR_DECL can be caused by prior errors introducing them.
> So we simply might want to treat VAR_DECL in unify via
> unify_template_argument_mis
On 05/25/2015 08:29 PM, Andreas Tobler wrote:
> Ok for trunk?
OK, thanks.
Andrew.
Hi, Richard
Thanks for review starting.
Do you see any major issues with this patch (i.e. algorithms and ideas
that should be completely replaced, effectively causing the re-write
of most code)?
To decide if there are major issues in the patch, perhaps, you need
additional clarifications from me
On 26/05/15 09:58, Richard Biener wrote:
On Fri, May 22, 2015 at 5:44 PM, Jan Hubicka wrote:
Hi,
PR 66181 is about ICE in verify_type that complains that type and its variant
differs
by TYPE_NO_FORCE_BLK. This flag is kind-of internal to stor-layout.c, so the
divergence
may not matter (I a
This patch extends the machinery which detects dangerous order
dependencies caused by out-mode parameters of Ada 2012 functions
(AI-0144) to detect the error in array aggregates that have a
nonstatic range (RM 6.20/3).
The compiler now catches the error in the following sources:
pragma Ada_2012;
GCC built with latest binutils and the patch give the following
performance improve:
spec2000INT +3% at "-O2 -m32", +1,5% at "-O2 -m64".
Some other benchmark scores at "-O2" were also improved up to 6%.
The patch is very efficient for PIE mode.
Thanks,
Evgeny
On Tue, May 5, 2015 at 6:30 PM, H.J
This patch fixes a visibility error on a subtype of a derived private type,
when the enclosing unit is both the parent of an instantiated generic child
unit, and contains an expression function that is automatically inlined.
The following must compile and execute quietly:
gnatmake -q main
This change ensures on all targets that A**B = A**C when B is a small
static constant in the range 0 .. 4 and C is a variable with the same
value. Previously for Float and Long_Float on some targets, this was
not the case. The results given were both within the error bounds that
are allowed by the
On Thu, 14 May 2015, Tom de Vries wrote:
> On 20-04-15 14:25, Richard Biener wrote:
> > On Wed, 15 Apr 2015, Tom de Vries wrote:
> >
> > > On 03-04-15 14:39, Tom de Vries wrote:
> > > > On 27-03-15 15:10, Tom de Vries wrote:
> > > > > Hi,
> > > > >
> > > > > this patch fixes PR65443, a todo in t
Hi,
My first part patch improving how we handle overflow in scev is posted at
https://gcc.gnu.org/ml/gcc-patches/2015-05/msg01795.html . Here comes the
second part patch.
This patch does below improvements:
1) Computes and records control iv for each loop's exit edge. This
provides a way to co
On Fri, May 22, 2015 at 5:44 PM, Jan Hubicka wrote:
> Hi,
> PR 66181 is about ICE in verify_type that complains that type and its variant
> differs
> by TYPE_NO_FORCE_BLK. This flag is kind-of internal to stor-layout.c, so the
> divergence
> may not matter (I am not sure about it as C++ FE fina
The new debug switch -gnatd.k suppresses occurrences of line numbers
within error messages referring to a location in an internal file.
The following test normally compiles as follows (with -gnatj55)
1. with Ada.Text_IO; use Ada.Text_IO;
2. with Ada.Containers.Vectors;
3. procedure
Previously, A**B could give different results for float and
long float depending on whether B was a small static constant
in the range 2..4 or a variable with the same value. Although
both values are valid, this discrepancy is undesirable, both
for general use and in particular for formal analysis
Hi,
I am not sure if suggested fix is the thing we want here. But infact
having here a VAR_DECL can be caused by prior errors introducing them.
So we simply might want to treat VAR_DECL in unify via
unify_template_argument_mismatch.
Tested for 5.x branch, and trunk. Ok for apply?
Kai
ChangeLo
This update allows the use of the following restrictions pragmas
in package System:
No_Specification_Of_Aspect
No_Use_Of_Attribute
No_Use_Of_Pragma
Given a system.ads file containing the lines:
pragma Restrictions (No_Use_Of_Pragma => Attach_Handler);
pragma Restrictions (No_Specification_
On Sun, May 24, 2015 at 3:17 PM, Marc Glisse wrote:
>
> I forgot to mention I optimistically tried to write something like this:
>
> (match
> (negated_value_for_comparison @0)
> (negate @0))
> (match
> (negated_value_for_comparison (negate @0))
> @0)
> (match
> (negated_value_for_comparison (
Ping
2015-05-19 12:39 GMT+03:00 Ilya Enkovich :
> Ping
>
> 2015-05-05 11:05 GMT+03:00 Ilya Enkovich :
>> Ping
>>
>> 2015-04-14 17:35 GMT+03:00 Ilya Enkovich :
>>> On 10 Apr 03:27, Jan Hubicka wrote:
>
> + /* We might propagate instrumented function pointer into
> + not instrume
Ping
2015-05-19 12:40 GMT+03:00 Ilya Enkovich :
> Ping
>
> 2015-05-05 11:06 GMT+03:00 Ilya Enkovich :
>> Ping
>>
>> 2015-04-14 12:14 GMT+03:00 Ilya Enkovich :
>>> On 10 Apr 03:15, Jan Hubicka wrote:
>
> References are not streamed out for nodes which are referenced in a
> partition
This implements a new aspect Disable_Controlled which can only be
used for controlled record types. It causes suppression of related
calls to Initialize, Adjust, Finalize (for conditional compilation
purposes).
The following test:
1. with Ada.Finalization; use Ada.Finalization;
2. with
This patch ensures that assertion policy Ignore interacts properly with aspect
Type_Invariant'Class.
-- Source --
-- pack_type_invariant_class.ads
pragma Assertion_Policy (Type_Invariant'Class => Ignore);
package Pack_Type_Invariant_Class is
type Priv_Typ is tagged
On Fri, May 8, 2015 at 11:07 PM, Evgeniya Maenkova
wrote:
> Hi,
>
> Could you please review my patch for predicated lim?
>
> Let me note some details about it:
>
>
>
> 1) Phi statements are still moved only if they have 1 or 2
> arguments. However, phi statements could be move under condition
This refactors code to arrange that for loop SLP we pass in the SLP
node and instance to the vectorizable_* functions also at analysis
phase (not only at transform phase).
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
Richard.
2015-05-26 Richard Biener
* tre
One of my dev trees had this.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
Richard.
2015-05-26 Richard Biener
* fold-const.c (fold_binary_loc): Move X % -Y -> X % Y and
X % C -> X & (C - 1) for C being a power-of two to ...
* match.pd: ... p
On 13/05/15 11:16, Christian Bruel wrote:
> - [2.2/6]: Redefine TARGET_MACROS for #pragma GCC target without
"thumbness" the glue.
Here it is, no regression for
arm-sim/
arm-sim//-march=armv7-a
arm-sim//-mthumb
arm-sim//-mthumb/-march=armv7-a
Obviously, [4/6],[5/6] an
On 26/05/15 09:45 +0100, Pedro Alves wrote:
On 05/25/2015 07:41 PM, Jonathan Wakely wrote:
Index: include/debug/functions.tcc
===
--- include/debug/functions.tcc (revision 0)
+++ include/debug/functions.tcc (working copy)
@@ -0,0 +
On Tue, May 19, 2015 at 11:12:55AM +0100, Kyrill Tkachov wrote:
> Hi all,
>
> This patch adds a mention of the new native cpu detection feature in aarch64
> GNU/Linux. Gerald, this is a patch against htdocs/gcc-6/changes.html and I
> thought I had seen the 'changes' link in gcc.gnu.org earlier bu
On 06 Apr 09:28, Jeff Law wrote:
> On 04/06/2015 09:17 AM, Ilya Enkovich wrote:
> >>
> >>To tell the truth, I can't figure out what this means from a user
> >>perspective. How does a user know whether the linker option is
> >>being ignored, or if they have a new enough linker? If the linker
> >>a
A few files I added earlier this year are missing 2015 in the list of
copyright years, because I forgot to update them when moving them from
local branches to the FSF repo.
Committed to trunk.
commit 378bbf405af6b83dd44c9d92b6e819a0af3e7697
Author: Jonathan Wakely
Date: Tue May 26 10:25:24 201
On Fri, May 22, 2015 at 09:30:18AM +0100, Matthew Wahab wrote:
> [Added PR number and updated patches]
>
> This patch adds tests for the code generated by the Aarch64 backend for the
> __sync builtins.
>
> Tested aarch64-none-linux-gnu with check-gcc.
>
> Ok for trunk?
> Matthew
This is OK once
On Fri, 22 May 2015, Jan Hubicka wrote:
> > > + case OBJ_TYPE_REF:
> > > + {
> > > + if (!operand_equal_p (OBJ_TYPE_REF_EXPR (arg0),
> > > + OBJ_TYPE_REF_EXPR (arg1), flags))
> > > + return false;
> > > + if (flag_devirtualize && virtual_method_call_p (arg
Hi,
The first part patch improving how we handle overflow in scev is
posted at https://gcc.gnu.org/ml/gcc-patches/2015-05/msg01795.html .
Here comes the second part patch.
This patch does below improvements:
1) Computes and records control iv for each loop's exit edge. This
provides a way to co
On Sun, May 24, 2015 at 3:33 PM, Marc Glisse wrote:
> Hello,
>
> I noticed we were only doing this transformation for floats and not for
> integers, so I took the chance to move it to match.pd. Regtested on
> ppc64le-redhat-linux.
Ok.
Thanks,
Richard.
> 2015-05-25 Marc Glisse
>
> * m
On Tue, May 19, 2015 at 07:05:14AM +0100, Dominik Vogt wrote:
> The attached patch activates the -mtune=native and -march=native
> command line options on S390 and System z. Most of the code is a
> modified copy of gcc/config/mips/driver-native.c. ChangeLog
> attached.
Version 2 of the patch to
Ping.
On Tue, May 19, 2015 at 04:07:53PM +0200, Marek Polacek wrote:
> This PR points out that we output same -Wformat warning twice when using
> __attribute__ ((format)). The problem was that attribute_value_equal
> (called when processing merge_attributes) got two lists:
> "format printf, 1, 2"
Hi,
My first part patch improving how we handle overflow in scev is posted at
https://gcc.gnu.org/ml/gcc-patches/2015-05/msg01795.html . Here comes the
second part patch.
This patch does below improvements:
1) Computes and records control iv for each loop's exit edge. This
provides a way to co
Since the move to C++14 is now imminent, I regtested/bootstrapped
the patch with -Wshift-overflow always enabled for C++ and that
went well.
Marek
On Fri, May 01, 2015 at 09:24:20AM +0100, Kyrill Tkachov wrote:
> Hi all,
>
> It occurs to me that in the IOR-of-shifts form of the rev16 operation we
> should be costing the operand properly. For that we'd want to reuse the
> aarch_rev16_p function that does all the heavy lifting and get it to w
On 05/25/2015 07:41 PM, Jonathan Wakely wrote:
>> Index: include/debug/functions.tcc
>> ===
>> --- include/debug/functions.tcc (revision 0)
>> +++ include/debug/functions.tcc (working copy)
>> @@ -0,0 +1,159 @@
>> +// Debugg
On Mon, May 25, 2015 at 06:39:36AM +0100, Kumar, Venkataramanan wrote:
> Ping!
>
> -Original Message-
> From: gcc-patches-ow...@gcc.gnu.org [mailto:gcc-patches-ow...@gcc.gnu.org] On
> Behalf Of Kumar, Venkataramanan
> Sent: Tuesday, May 19, 2015 9:07 PM
> To: James Greenhalgh (james.gree
On Mon, 25 May 2015, Dharmakan Rohit Arul Raj wrote:
> > -Original Message-
> > From: Richard Biener [mailto:rguent...@suse.de]
> > Sent: Monday, May 18, 2015 1:06 PM
> > To: Jeff Law
> > Cc: Dharmakan Rohit-B30502; gcc-patches@gcc.gnu.org; Jakub Jelinek; Alan
> > Modra; David Edelsohn; Wi
1 - 100 of 107 matches
Mail list logo
