Hi!
When memcpy has constant length, we leave probable_max_size uninitialized
and later on create a CONST_INT out of it.
Fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux,
committed to trunk as obvious.
2014-01-15 Jakub Jelinek
PR other/58712
* builtins.c (d
Ping ? Ok for trunk ?
On 01/09/14 15:17, Laurent Alfonsi wrote:
On 01/09/14 06:02, Jeff Law wrote:
On 01/08/14 02:05, Laurent Alfonsi wrote:
All,
I was looking at PR49718. I have enclosed a simple fix for this bug
report.
2014-01-07 Laurent Alfonsi
* c-family/c-common.c
(ha
On 15 January 2014 00:09, Andrew Pinski
wrote:
> ChangeLog:
> * config/aarch64/aarch64.c (aarch64_register_move_cost): Correct cost
> of moving from/to the STACK_REG register class.
+ /* Moving between GPR and stack cost is the same as GP2GP. */
Don't forget the double space after period.
Oth
On Tue, Jan 14, 2014 at 06:52:00PM +0100, Florian Weimer wrote:
> On 01/13/2014 09:48 PM, Marek Polacek wrote:
> >+bool
> >+pointer_to_zero_sized_aggr_p (tree t)
> >+{
> >+ t = strip_pointer_operator (t);
> >+ if (RECORD_OR_UNION_TYPE_P (t)
> >+ && TYPE_SIZE (t)
> >+ && integer_zerop (T
Hi there,
With trunk enhancement at
http://gcc.gnu.org/ml/gcc-patches/2013-11/msg00533.html, gcc can properly
handle PLD rtx. Otherwise the PLD rtx will be treated as SET rtx and gcc
will end up with ICE. The attached patch intends to back port this
enhancement to 4.8 branch. Tested with gcc regre
Hi,
we have run into optimization regressions in Ada caused by the interaction
between the GIMPLE clobbers and SRA: on the one hand GIMPLE clobbers create
artificial EH handlers for aggregate variables, on the other hand SRA refuses
to scalarize objects that appear in a statement that can throw
Hi,
fix for BZ59803. For more details please see the comment in the
patch.
Bootstrapped and regtested on s390x with 4.8 and mainline.
Committed to 4.8 and mainline.
Bye,
-Andreas-
2014-01-15 Andreas Krebbel
* config/s390/s390.c (s390_preferred_reload_class): Don't return
On 15/01/14 09:23, Terry Guo wrote:
> Hi there,
>
> With trunk enhancement at
> http://gcc.gnu.org/ml/gcc-patches/2013-11/msg00533.html, gcc can properly
> handle PLD rtx. Otherwise the PLD rtx will be treated as SET rtx and gcc
> will end up with ICE. The attached patch intends to back port this
Hi,
The patch fixes ICE when building kernel raid6 neon code.
lib/raid6/neon4.c: In function 'raid6_
neon4_gen_syndrome_real':
lib/raid6/neon4.c:113:1: internal compiler error: in
dwarf2out_frame_debug_adjust_cfa, at dwarf2cfi.c:1090
}
https://bugs.launchpad.net/gcc-linaro/+bug/1268893
Root c
On Tue, Jan 14, 2014 at 09:42:37PM +, Joseph S. Myers wrote:
> On Mon, 13 Jan 2014, Marek Polacek wrote:
>
> > +/* Return true if T is a pointer to a zero-sized struct/union. */
> > +
> > +bool
> > +pointer_to_zero_sized_aggr_p (tree t)
> > +{
> > + t = strip_pointer_operator (t);
> > + if
On Wed, Jan 15, 2014 at 11:27:37AM +0100, Marek Polacek wrote:
> Perhaps, but I don't think we can do it easily. Consider
>
> int
> foo (int *p, int *q)
> {
> return p - q;
> }
That is not a difference of pointers to zero sized arrays though,
that is pointers to int, and there is no division b
On 13/01/14 21:05, Richard Earnshaw wrote:
> On 11/01/14 23:42, Kugan wrote:
>> Hi,
>>
>> aarch64_build_constant incorrectly truncates the immediate when
>> constants are generated with MOVN. This causes coinor-osi tests to fail
>> (tracked also in https://bugs.launchpad.net/gcc-linaro/+bug/1263576
> -Original Message-
> From: Richard Earnshaw
> Sent: Wednesday, January 15, 2014 5:54 PM
> To: Terry Guo
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [GCC, ARM] Backport trunk fix to 4.8 branch to properly
handle
> rtx of ARM PLD instruction
>
> On 15/01/14 09:23, Terry Guo wrote:
> > H
This improves compile time of the testcase in PR39326 at -O3 from
loop invariant motion : 24.41 (32%) usr 0.03 ( 4%) sys 24.43 (32%)
wall 148 kB ( 0%) ggc
loop unswitching: 15.16 (20%) usr 0.02 ( 3%) sys 15.30 (20%)
wall 0 kB ( 0%) ggc
TOTAL : 76.
On Tue, Jan 14, 2014 at 3:49 PM, Richard Earnshaw wrote:
> On 14/01/14 14:32, Jakub Jelinek wrote:
>> Anyway, the above is really a simple case, and I'd call it a
>> backend bug if it isn't able to generate good code out of that.
>
> Exactly which back-end pass are you expecting to simplify
>
> (s
On Wed, Jan 15, 2014 at 10:24 AM, Eric Botcazou wrote:
> Hi,
>
> we have run into optimization regressions in Ada caused by the interaction
> between the GIMPLE clobbers and SRA: on the one hand GIMPLE clobbers create
> artificial EH handlers for aggregate variables, on the other hand SRA refuses
On 15/01/14 10:45, Terry Guo wrote:
>
>
>> -Original Message-
>> From: Richard Earnshaw
>> Sent: Wednesday, January 15, 2014 5:54 PM
>> To: Terry Guo
>> Cc: gcc-patches@gcc.gnu.org
>> Subject: Re: [GCC, ARM] Backport trunk fix to 4.8 branch to properly
> handle
>> rtx of ARM PLD instructi
On 15/01/14 10:38, Kugan wrote:
> On 13/01/14 21:05, Richard Earnshaw wrote:
>> On 11/01/14 23:42, Kugan wrote:
>>> Hi,
>>>
>>> aarch64_build_constant incorrectly truncates the immediate when
>>> constants are generated with MOVN. This causes coinor-osi tests to fail
>>> (tracked also in https://bu
On 14 January 2014 14:44, Jonathan Wakely wrote:
> On 9 January 2014 21:55, François Dumont wrote:
>> Hi
>>
>> Here is a patch for this small problem with clang. It is not a blocking
>> issue for the 4.9 release but at the same time it is a rather safe fix so
>> just tell me if I can commit it.
On 15/01/14 10:04, Zhenqiang Chen wrote:
> Hi,
>
> The patch fixes ICE when building kernel raid6 neon code.
>
> lib/raid6/neon4.c: In function 'raid6_
>
> neon4_gen_syndrome_real':
> lib/raid6/neon4.c:113:1: internal compiler error: in
> dwarf2out_frame_debug_adjust_cfa, at dwarf2cfi.c:1090
>
> What you want is tree-eh.c:optimize_clobbers, right? Can't we
> do this optimization during EH refactoring / lowering as well?
We need the SSA form for sink_clobbers.
> Also why does SRA refuse to scalarize here?
Because of the EH handler and the following predicate:
/* Disqualify LHS and RH
>
> Preferably, particularly since you haven't supplied a testcase.
>
> R.
Bug is reported at http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59826. I
shall update the patch to include the test case.
BR,
Terry
> -Original Message-
> From: Terry Guo [mailto:terry@arm.com]
> Sent: Wednesday, January 15, 2014 8:21 PM
> To: Richard Earnshaw
> Cc: gcc-patches@gcc.gnu.org
> Subject: RE: [GCC, ARM] Backport trunk fix to 4.8 branch to properly
handle
> rtx of ARM PLD instruction
>
> >
> > Preferab
On Wed, Jan 15, 2014 at 11:52 AM, Richard Biener wrote:
> I can split off a df_analyze_1 and have a df_analyze () doing what
> it does now and a df_analyze_loop () combining df_set_blocks ()
> and df_analyze () for example.
df_analyze_loop sounds good to me.
But shouldn't it be possible to avoid
On Wed, 15 Jan 2014, Steven Bosscher wrote:
> On Wed, Jan 15, 2014 at 11:52 AM, Richard Biener wrote:
> > I can split off a df_analyze_1 and have a df_analyze () doing what
> > it does now and a df_analyze_loop () combining df_set_blocks ()
> > and df_analyze () for example.
>
> df_analyze_loop s
Hi,
this is a 4.9 Regression, but just an ICE on invalid. It's a little
tricky to fix, because, eg, in mainline (vs 4_8-branch) we want to
produce somewhat more concise diagnostic in this area (eg, a single
error for init/array26.C). It seems Ok to me, and passes testing, to
return NULL_TREE
This should fix PR59822, us creating in valid SSA form when
hoisting a[i_2] out of a loop where i_2 is defined inside the loop.
Bootstrap / regtest running on x86_64-unknown-linux-gnu.
Richard.
2014-01-15 Richard Biener
PR tree-optimization/59822
* tree-vect-stmts.c (hoist_d
Oops, forgot to CC the x86 maintainers. Is the i386.c change ok for 4.8?
Peter
Forwarded Message
From: Peter Bergner
To: gcc-patches@gcc.gnu.org
Cc: Richard Biener , Jakub Jelinek
, Vladimir Makarov
Subject: [PATCH, reginfo.c, i386.c] Backport fix for PR58139 to 4.8
Date:
On Mon, Jan 13, 2014 at 11:11 AM, Jakub Jelinek wrote:
> Hi!
>
> This patch fixes the following testcase by preevaluating rhs if it has
> (can have) side-effects in lhs op= rhs expressions. Bootstrapped/regtested
> on x86_64-linux and i686-linux, ok for trunk?
> C++ already does a similar thing (
On Wed, Jan 15, 2014 at 11:35:40AM +0100, Jakub Jelinek wrote:
> On Wed, Jan 15, 2014 at 11:27:37AM +0100, Marek Polacek wrote:
> > Perhaps, but I don't think we can do it easily. Consider
> >
> > int
> > foo (int *p, int *q)
> > {
> > return p - q;
> > }
>
> That is not a difference of pointe
On Wed, 15 Jan 2014, Steven Bosscher wrote:
> On Wed, Jan 15, 2014 at 11:52 AM, Richard Biener wrote:
> > I can split off a df_analyze_1 and have a df_analyze () doing what
> > it does now and a df_analyze_loop () combining df_set_blocks ()
> > and df_analyze () for example.
>
> df_analyze_loop s
Am 12/18/2013 12:37 PM, schrieb Richard Biener:
On Tue, Dec 17, 2013 at 2:05 PM, Georg-Johann Lay wrote:
Am 12/05/2013 04:09 PM, schrieb Richard Biener:
On Thu, Dec 5, 2013 at 3:53 PM, Georg-Johann Lay wrote:
This is a fix of a wrong warning for a bas ISR name. The assumption was
that if
On 15/01/14 12:37, Terry Guo wrote:
>
>
>> -Original Message-
>> From: Terry Guo [mailto:terry@arm.com]
>> Sent: Wednesday, January 15, 2014 8:21 PM
>> To: Richard Earnshaw
>> Cc: gcc-patches@gcc.gnu.org
>> Subject: RE: [GCC, ARM] Backport trunk fix to 4.8 branch to properly
> handle
On Wed, Jan 15, 2014 at 1:06 PM, Eric Botcazou wrote:
>> What you want is tree-eh.c:optimize_clobbers, right? Can't we
>> do this optimization during EH refactoring / lowering as well?
>
> We need the SSA form for sink_clobbers.
I know, I'm saying it may be possible to implement an equivalent
op
Ping.
Am 06.01.2014 20:24, schrieb Ian Lance Taylor:
> This libgo patch from Michael Hudson-Doyle recognizes arm64 as the Go
> name for the AArch64 architecture. Bootstrapped and ran Go testsuite on
> x86_64-unknown-linux-gnu. Committed to mainline.
this missed the one reference to the goarch in the g
On Wed, Jan 15, 2014 at 7:24 AM, Matthias Klose wrote:
> Am 06.01.2014 20:24, schrieb Ian Lance Taylor:
>> This libgo patch from Michael Hudson-Doyle recognizes arm64 as the Go
>> name for the AArch64 architecture. Bootstrapped and ran Go testsuite on
>> x86_64-unknown-linux-gnu. Committed to ma
We ICE on the following (invalid) testcase, because the DECL_ARG_TYPE
is error_mark_node and useless_type_conversion_p doesn't check that
it operates on a type.
Regtested/bootstrapped on x86_64-linux, ok for trunk/4.8/4.7?
2014-01-15 Marek Polacek
PR middle-end/59827
* cgraph.
> -Original Message-
> From: Richard Earnshaw
> Sent: Wednesday, January 15, 2014 10:30 PM
> To: Terry Guo
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [GCC, ARM] Backport trunk fix to 4.8 branch to properly
handle
> rtx of ARM PLD instruction
>
> On 15/01/14 12:37, Terry Guo wrote:
> >
Hi!
Since started, the VAX backend doesn't compile with config-list.mk
because one of VAX's macros doesn't use its argument:
g++ -c -g -O2 -DIN_GCC -DCROSS_DIRECTORY_STRUCTURE -fno-exceptions
-fno-rtti -fasynchronous-unwind-tables -W -Wall -Wno-narrowing -Wwrite-strings
-Wcast-qual -Wmissin
On Wed, Jan 15, 2014 at 08:23:45AM -0800, Richard Henderson wrote:
> --- a/gcc/reginfo.c
> +++ b/gcc/reginfo.c
> @@ -86,7 +86,7 @@ static const char initial_call_really_used_regs[] =
> CALL_REALLY_USED_REGISTERS;
> char global_regs[FIRST_PSEUDO_REGISTER];
>
> /* Declaration for the global regi
IMO the test case is invalid, but as always an ICE is unacceptable.
I can imagine there are some legitimate uses for absconding with
the hard frame pointer, under -fomit-frame-pointer plus knowledge
that the code is simple enough.
Previously when we got into this situation, we'd silently smash th
On Wed, Jan 15, 2014 at 8:23 AM, Richard Henderson wrote:
> IMO the test case is invalid, but as always an ICE is unacceptable.
>
> I can imagine there are some legitimate uses for absconding with
> the hard frame pointer, under -fomit-frame-pointer plus knowledge
> that the code is simple enough.
Hi,
On Wed, Jan 15, 2014 at 03:39:09PM +0100, Richard Biener wrote:
> On Wed, Jan 15, 2014 at 1:06 PM, Eric Botcazou wrote:
> > Yes, it's scalarized during SRA but not ESRA because there is an ehcleanup
> > pass in-between. It used to be scalarized during ESRA up to 4.6.x.
>
> I'm saying even E
On 15/01/14 15:59, Terry Guo wrote:
>
>
>> -Original Message-
>> From: Richard Earnshaw
>> Sent: Wednesday, January 15, 2014 10:30 PM
>> To: Terry Guo
>> Cc: gcc-patches@gcc.gnu.org
>> Subject: Re: [GCC, ARM] Backport trunk fix to 4.8 branch to properly
> handle
>> rtx of ARM PLD instruct
Eric,
This patch adds scanning of clobbers in CALL_INSN_FUNCTION_USAGE to
find_all_hard_reg_sets.
For MIPS, calls are split at some point. After the split, one of the resulting
insns may clobber $6. But before the split, that's not explicit in the rtl
representation of the unsplit call.
Fo
Hi
Here is a patch to partially fix 2 string_view tests. It looks like
we can't use 2 dg-options into the same source, one override the over,
the dg-options directives have been merged into 1. I also update a
script to make the experimental folder part of the folders to look for
tests. I
Hi
Here is a patch to fix profile mode compilation errors. It makes
tree based containers C++11 allocator aware in profile mode as they are
in normal mode. I also try to use default implementation as much as
possible to benefit from the normal mode noexcept qualifications on the
defaulted
On Tue, Jan 14, 2014 at 11:09 PM, Jakub Jelinek wrote:
> On Tue, Jan 14, 2014 at 06:18:22AM -0800, H.J. Lu wrote:
>> 2014-01-14 H.J. Lu
>>
>> PR target/59794
>> * config/i386/i386.c (type_natural_mode): Add a bool parameter
>> to indicate if type is used for function return va
Handle NULL vdef for call in the case where we have a matching vnresult
that has a vdef (it already handles the NULL vdef case when !vnresult). This
can happen for promoted indirect calls if the fallback indirect call
(which has a vdef) can be proven equivalent to the promoted direct call
(which mi
Regression tested, ok for google/4_8?
Thanks,
Teresa
2014-01-15 Teresa Johnson
* loop-unroll.c (code_size_limit_factor): Handle int overflow.
Index: loop-unroll.c
===
--- loop-unroll.c (revision 206601)
+++ loop-un
On 15 January 2014 16:59, François Dumont wrote:
> Hi
>
> Here is a patch to fix profile mode compilation errors. It makes tree
> based containers C++11 allocator aware in profile mode as they are in normal
> mode. I also try to use default implementation as much as possible to
> benefit from t
ok.
David
On Wed, Jan 15, 2014 at 9:10 AM, Teresa Johnson wrote:
> Regression tested, ok for google/4_8?
>
> Thanks,
> Teresa
>
> 2014-01-15 Teresa Johnson
>
> * loop-unroll.c (code_size_limit_factor): Handle int overflow.
>
> Index: loop-unroll.c
> ===
On 15 January 2014 16:56, François Dumont wrote:
> Hi
>
> Here is a patch to partially fix 2 string_view tests. It looks like we
> can't use 2 dg-options into the same source, one override the over, the
> dg-options directives have been merged into 1. I also update a script to
> make the experi
On 01/15/2014 06:17 PM, Jonathan Wakely wrote:
I think we decided we want functions to be constexpr in debug mode if
they are constexpr in normal mode.
I think std::array has solved the same problem without losing the
constexpr qualifier.
Yes, I think the only complete solution we know of is tha
On Wed, Jan 15, 2014 at 02:06:24PM +0100, Richard Biener wrote:
>
> This should fix PR59822, us creating in valid SSA form when
> hoisting a[i_2] out of a loop where i_2 is defined inside the loop.
>
> Bootstrap / regtest running on x86_64-unknown-linux-gnu.
Unfortunately with my patch on top of
The following patch fixes:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59511
Register move cost calculations were too conservative for some cases
and it resulted into overflows. The patch fixes also some wrong
frequencies usage I found.
The patch has very small affect on SPEC2000. Only 7 prog
> I know, I'm saying it may be possible to implement an equivalent
> optimization without SSA form.
Sure, but what would be the point of this duplication exactly?
> I'm saying even ESRA should be able to scalarize it just fine. It just
> needs to be careful where to insert the loads from the agg
On 15 January 2014 17:22, Paolo Carlini wrote:
> On 01/15/2014 06:17 PM, Jonathan Wakely wrote:
>>
>> I think we decided we want functions to be constexpr in debug mode if
>> they are constexpr in normal mode.
>>
>> I think std::array has solved the same problem without losing the
>> constexpr qual
This patch is now committed.
--joel
On 1/7/2014 10:58 AM, Sebastian Huber wrote:
> v2: Do not override an existing tmake_file in (arm*-*-uclinux*eabi*).
>
> The ARM changes are already approved:
>
> http://gcc.gnu.org/ml/gcc-patches/2014-01/msg00254.html
>
> The rest is RTEMS specific.
>
> gcc/Cha
Hello Everyone,
I noticed that there was an error in the testcode in
"check_libcilkrts_available" target-supports function for C++. It was working
fine in C but needed an extern "C" along with a prototype for C++. This patch
below should fix that. This patch is committed as obvious (..m
On 01/13/14 00:34, Jakub Jelinek wrote:
On Sun, Jan 12, 2014 at 11:21:59PM -0700, Jeff Law wrote:
--- a/gcc/ree.c
+++ b/gcc/ree.c
@@ -297,6 +297,13 @@ combine_set_extension (ext_cand *cand, rtx curr_insn, rtx
*orig_set)
else
new_reg = gen_rtx_REG (cand->mode, REGNO (SET_DEST (*orig_se
On Wed, Jan 15, 2014 at 10:56:35AM -0700, Jeff Law wrote:
> On 01/13/14 00:34, Jakub Jelinek wrote:
> >On Sun, Jan 12, 2014 at 11:21:59PM -0700, Jeff Law wrote:
> >>--- a/gcc/ree.c
> >>+++ b/gcc/ree.c
> >>@@ -297,6 +297,13 @@ combine_set_extension (ext_cand *cand, rtx curr_insn,
> >>rtx *orig_set)
Hi,
On Wed, Jan 15, 2014 at 06:37:30PM +0100, Eric Botcazou wrote:
> > I know, I'm saying it may be possible to implement an equivalent
> > optimization without SSA form.
>
> Sure, but what would be the point of this duplication exactly?
>
> > I'm saying even ESRA should be able to scalarize it
On 01/14/2014 08:37 AM, Nick Clifton wrote:
> gcc/ChangeLog
> 2014-01-13 Nick Clifton
>
> PR middle-end/28865
> * varasm.c (output_constant): Return the number of bytes actually
> emitted.
> (output_constructor_array_range): Update the field size with the
> number
In libgcov-driver.c,
1) there are couple of places with trailing white spaces (e.g, in
gcov_sort_n_vals body), please remove
2) gcov_exit_write_gcda in trunk takes eof_pos as an arg, and check it
before writing the header. I think this is more correct than in your
patch
3) It would be better to ke
Tom de Vries writes:
> This patch adds scanning of clobbers in CALL_INSN_FUNCTION_USAGE to
> find_all_hard_reg_sets.
>
> For MIPS, calls are split at some point. After the split, one of the
> resulting insns may clobber $6. But before the split, that's not
> explicit in the rtl representation of
On 01/15/14 10:07, Teresa Johnson wrote:
Handle NULL vdef for call in the case where we have a matching vnresult
that has a vdef (it already handles the NULL vdef case when !vnresult). This
can happen for promoted indirect calls if the fallback indirect call
(which has a vdef) can be proven equiv
The test c-c++-common/cilk-plus/AN/builtin_func_double2.c is failing on MIPS
when compiled with C++ but not when compiled with C because GCC is generating
an illegal assembly language instruction during the C++ compilation. The
reason this works in C but fails for C++ is because the
__sec_reduce_a
On Wed, Jan 15, 2014 at 10:40 AM, Xinliang David Li wrote:
> In libgcov-driver.c,
>
> 1) there are couple of places with trailing white spaces (e.g, in
> gcov_sort_n_vals body), please remove
They are from the existing code. But I'll fix them.
> 2) gcov_exit_write_gcda in trunk takes eof_pos as a
In the original testcase, the user found that a large std::array of
std::atomics was taking forever to compile. There were two issues
involved here:
1) Because atomics do have user-provided constructors, we were going
down the constructor path and not realizing that it just gave us
zero-init
On 01/09/14 07:17, Laurent Alfonsi wrote:
On 01/09/14 06:02, Jeff Law wrote:
On 01/08/14 02:05, Laurent Alfonsi wrote:
All,
I was looking at PR49718. I have enclosed a simple fix for this bug
report.
2014-01-07 Laurent Alfonsi
* c-family/c-common.c
(handle_no_instrument_functio
Two patches in a week about dbr_schedule and redundant insns. This one
fixes a miscompilation of libgcov-driver.c for n32 and n64 MIPS.
It's independent of (and predates) the 59137 patch.
We had:
barrier
L1:
C: $r1 = ...
D: $r2 = ...
...
A: if ... goto L1
B: if
umips-branch-4.c relies on direct JALs being possible, so add
addressing=absolute to force that.
Tested on mips64-linux-gnu and applied.
Thanks,
Richard
gcc/testsuite/
* gcc.target/mips/umips-branch-4.c: Add addressing=absolute.
Index: gcc/testsuite/gcc.target/mips/umips-branch-4.c
===
On 01/15/2014 07:54 AM, Paolo Carlini wrote:
It seems Ok to me, and passes testing, to
return NULL_TREE from build_value_init. Alternately, more
conservatively, we could change build_value_init_noctor to not call
build_value_init at all when ftype == error_mark_node.
I think I would prefer that
On Tue, Jan 14, 2014 at 7:13 AM, Kirill Yukhin wrote:
> Hello,
> This patch introduces missing AVX-512PF intrinsics and tests.
> It also renames store/load intrinsics according to EAS.
>
> gcc/
> * config/i386/avx512fintrin.h (_mm512_loadu_si512): Rename.
> (_mm512_storeu_si512): D
On Wed, Jan 15, 2014 at 10:46 AM, Jeff Law wrote:
> On 01/15/14 10:07, Teresa Johnson wrote:
>>
>> Handle NULL vdef for call in the case where we have a matching vnresult
>> that has a vdef (it already handles the NULL vdef case when !vnresult).
>> This
>> can happen for promoted indirect calls if
Hi James,
This commit (SVN r206045) seems to have introduced a problem when
compiling multiple source files if a -mcpu option is also present on
the command line.
This can be reproduced in a arm-unknown-linux-gnueabihf build with any
source file which use floating point arguments/results. For exa
Hi,
On 01/15/2014 08:45 PM, Jason Merrill wrote:
On 01/15/2014 07:54 AM, Paolo Carlini wrote:
It seems Ok to me, and passes testing, to
return NULL_TREE from build_value_init. Alternately, more
conservatively, we could change build_value_init_noctor to not call
build_value_init at all when ftyp
On Wed, Jan 15, 2014 at 12:00 PM, Teresa Johnson wrote:
> On Wed, Jan 15, 2014 at 10:46 AM, Jeff Law wrote:
>> On 01/15/14 10:07, Teresa Johnson wrote:
>>>
>>> Handle NULL vdef for call in the case where we have a matching vnresult
>>> that has a vdef (it already handles the NULL vdef case when !
Hi,
this patch fixes ICE in ipa-devirt that is caused by get_binfo_at_offset
reporting NULL for a valid query. This is because how virtual inheritance
is represented.
Here we have chain A<-B<-C where A is a virtual base. We look for A within
C that is there at offset 64. On this query get_binfo
There are options you can use to control passes explicitly:
-fdisable-... -fenable-
To disable early inline:
-fdisable-tree-einline
David
On Wed, Jan 15, 2014 at 1:17 PM, Teresa Johnson wrote:
> On Wed, Jan 15, 2014 at 12:00 PM, Teresa Johnson wrote:
>> On Wed, Jan 15, 2014 at 10:46 AM, J
On Wed, 15 Jan 2014, Marek Polacek wrote:
> +/* Return true if T is a pointer to a zero-sized struct/union. */
> +
> +bool
> +pointer_to_zero_sized_aggr_p (tree t)
> +{
> + t = strip_pointer_operator (t);
> + return ((RECORD_OR_UNION_TYPE_P (t) || TREE_CODE (t) == ARRAY_TYPE)
> + && TYPE_
> Hi,
> this patch fixes ICE in ipa-devirt that is caused by get_binfo_at_offset
> reporting NULL for a valid query. This is because how virtual inheritance
> is represented.
>
> Here we have chain A<-B<-C where A is a virtual base. We look for A within
> C that is there at offset 64. On this q
On 01/15/14 12:27, Richard Sandiford wrote:
[ snip ]
barrier
L1:
C'': (use ($r1 = ...))
L2:
D: $r2 = ...
L3:
...
A: if ... goto L2
C': $r1 = ...
B: if ... goto L3
D': $r2 = ...
So far so good. The problem is that redirecting B to L3 makes
I did try disabling early inlining and some other things but still
didn't have luck. The indirect call promotion is also in a different
phase in trunk and google/4_8, which was adding another complication.
Teresa
On Wed, Jan 15, 2014 at 1:23 PM, Xinliang David Li wrote:
> There are options you ca
Our SSA copy-prop passes do a pretty pathetic job at cleaning up after
themselves when const/copy propagation exposes new trivial copies and
constant initializations.
This can be seen in the code for pr57904 after copyprop2 for bb2
_2 = 344;
ubound.0_3 = _2 & 7;
size.1_4 = ubound.0_3 +
On 01/15/2014 08:28 AM, Jakub Jelinek wrote:
>> -static tree GTY(()) global_regs_decl[FIRST_PSEUDO_REGISTER];
>> > +tree GTY(()) global_regs_decl[FIRST_PSEUDO_REGISTER];
> Do you need the GTY(()) marker here when there is GTY(()) already in the
> header?
> Otherwise looks good to me.
>
Fixed befo
On 01/15/2014 08:37 AM, H.J. Lu wrote:
> We should add a testcase to verify this.
>
I included the following testcase with the commit. I couldn't find a way
to test this properly generically, so I just went with the obvious i386 test.
r~
diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index c93bf23
On Jan 14, 2014, at 7:25 AM, Richard Sandiford
wrote:
> Mike Stump writes:
>> diff --git a/gcc/expmed.c b/gcc/expmed.c
>> index ce063eb..720d8c1 100644
>> --- a/gcc/expmed.c
>> +++ b/gcc/expmed.c
>> @@ -4963,6 +4963,7 @@ make_tree (tree type, rtx x)
>> return t;
>>
>> case CONST_DOUBL
Hi!
As discussed in the PR, when combine combines something across a setter
of a likely spilled non-fixed hard register, we may get RA ICE, as the
combined insn might need for reload a hard reg that is live at the point
where combiner has combined the insn.
The following patch attempts to fix tha
On Wed, Jan 15, 2014 at 01:43:18PM -0800, Richard Henderson wrote:
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/i386/pr54694.c
> @@ -0,0 +1,11 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O" } */
> +
> +register void *hfp __asm__("%ebp"); /* { dg-message "note: for" } */
Shouldn't that
On Wed, Jan 15, 2014 at 01:43:18PM -0800, Richard Henderson wrote:
> On 01/15/2014 08:37 AM, H.J. Lu wrote:
> > We should add a testcase to verify this.
> >
>
> I included the following testcase with the commit. I couldn't find a way
> to test this properly generically, so I just went with the o
-ENOPATCH
Hello Everyone,
Attached, please find a patch that will fix PR 59825. The main issue
was array notations occurring in COMPOUND_EXPR. This patch should fix that and
fix the rank_mismatch2.c test-case ICE.
Ok for trunk?
Thanks,
Balaji V. Iyer.
diff --git a/gcc/c/ChangeLog
On Wed, Jan 15, 2014 at 9:34 AM, Vladimir Makarov wrote:
> The following patch fixes:
>
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59511
>
> Register move cost calculations were too conservative for some cases
> and it resulted into overflows. The patch fixes also some wrong
> frequencies usag
On Wed, Jan 15, 2014 at 10:37:04PM +, Iyer, Balaji V wrote:
> Hello Everyone,
> Attached, please find a patch that will fix PR 59825. The main issue was
> array notations occurring in COMPOUND_EXPR. This patch should fix that
> and fix the rank_mismatch2.c test-case ICE.
> --- a/gcc/c/c-arr
On Wed, Jan 15, 2014 at 10:37:04PM +, Iyer, Balaji V wrote:
> +2014-01-15 Balaji V. Iyer
> +
> + PR c/59825
> + * c-array-notation.c (expand_array_notation_exprs): Added COMPOUND_EXPR
> + case.
> +
"Add"
> --- a/gcc/c/c-array-notation.c
> +++ b/gcc/c/c-array-notation.c
> @@ -12
ok for google branch.
David
On Wed, Jan 15, 2014 at 2:58 PM, Rong Xu wrote:
> attached is the updated patch.
>
> On Wed, Jan 15, 2014 at 10:51 AM, Rong Xu wrote:
>> On Wed, Jan 15, 2014 at 10:40 AM, Xinliang David Li
>> wrote:
>>> In libgcov-driver.c,
>>>
>>> 1) there are couple of places wit
On 15-01-14 19:40, Richard Sandiford wrote:
Tom de Vries writes:
This patch adds scanning of clobbers in CALL_INSN_FUNCTION_USAGE to
find_all_hard_reg_sets.
For MIPS, calls are split at some point. After the split, one of the
resulting insns may clobber $6. But before the split, that's not
exp
1 - 100 of 107 matches
Mail list logo
