I cannot approve the collect2.c, gcc.c, opts.c, doc/invoke.texi. I'll
include some comments anyway, since the patch needs some more work.
> 2011-06-27 Doug Kwan
>
> Google ref 41164-p2
> Backport upstream patch under review.
>
> 2011-01-19 Nick Clifton
>
This should fix PR55327.
Committed as obvious.
Richard.
2012-11-28 Richard Biener
PR testsuite/55327
* gcc.dg/vect/slp-perm-8.c: Mark worker as noinline.
* gcc.dg/vect/slp-perm-9.c: Likewise.
Index: gcc/testsuite/gcc.dg/vect/slp-perm-8.c
Il 16/11/2012 01:53, Andrew Pinski ha scritto:
> Hi,
> If the PATH contains the current working directory (yes a bad idea
> but it could happen with our users), the build fails because it finds
> the newly created g++ which might not find the correct cc1plus. See
> the thread starting at
> http:
On Tue, Nov 27, 2012 at 8:29 PM, Uros Bizjak wrote:
>> As long as volatile asms and UNSPEC_VOLATILE insns (aka.
>> barriers) are handled the same way and consistently throughout
>> gcc, I'm fine. It seems your patch does that, so thanks!
>>
>> > But the question is also what effects your patch c
On 28/11/12 04:49, Andrew Pinski wrote:
On Tue, Nov 27, 2012 at 8:47 PM, Hurugalawadi, Naveen
wrote:
Hi,
Please find attached the patch “asm-subreg.patch” for aarch64 which
fixes ICE for /gcc.dg/torture/asm-subreg-1.c testcase.
The error was the result of reload problems in
“aarch64_load_symr
> We then end up with
> Redirecting fallthru edge 3->4 to 6
> JUMP-BYPASS: Proved reg 59 in jump_insn 15 equals constant (const_int 1
> [0x1]) Bypass edge from 3->4 to 6
> Redirecting fallthru edge 9->4 to 5
> JUMP-BYPASS: Proved reg 59 in jump_insn 15 equals constant (const_int 3
> [0x3]) Bypass e
I'd like to understand our long-term strategy wrt the asan/tsan tests in gcc.
Most of the tests we have today are not specific to the compiler and
so can potentially be used with any compiler.
The problem is the testing harness (FileCheck/gtest vs dejagnu).
I understand that using alien testing har
Hi!
On Wed, Nov 28, 2012 at 01:15:20AM -0800, Wei Mi wrote:
> I try to migrate the left asan lit-tests from llvm (class3). This is a
> preliminary version patch. Please forgive it has many mistakes.
Thanks for working on it.
> A known problems: I hardcoded -m32 in (set link_flags
> "[asan_link_f
On Tue, Nov 27, 2012 at 10:19 PM, H.J. Lu wrote:
> On Tue, Nov 27, 2012 at 2:07 AM, Richard Biener
> wrote:
>> On Tue, Nov 27, 2012 at 9:32 AM, Markus Trippelsdorf
>> wrote:
>>> On 2012.11.26 at 13:58 -0800, H.J. Lu wrote:
Hi,
OPT_SPECIAL_unknown, OPT_SPECIAL_ignore, OPT_SPECIAL_p
On Wed, Nov 28, 2012 at 02:10:05PM +0400, Konstantin Serebryany wrote:
> I'd like to understand our long-term strategy wrt the asan/tsan tests in gcc.
> Most of the tests we have today are not specific to the compiler and
> so can potentially be used with any compiler.
> The problem is the testing
On Tue, Nov 27, 2012 at 11:45 PM, Ramana Radhakrishnan
wrote:
> On Tue, Nov 27, 2012 at 8:22 PM, Richard Sandiford
> wrote:
>> Ramana Radhakrishnan writes:
Tested on x86_64-linux-gnu, powerpc64-linux-gnu and mipsisa64-elf.
Also tested by making sure that there were no changes in assemb
On 11/28/2012 04:12 AM, Ed Smith-Rowland wrote:
Here is a final patch for PR52654 hopefully.
Ultimately we should probably revisit this and 54413 by parsing the
literal
numbers in the C++ front end. That won't happen for 4.8.
This builds and passes on x86_64-linux
Thanks. Let's make sure Ja
On Wed, Nov 28, 2012 at 2:24 PM, Jakub Jelinek wrote:
> On Wed, Nov 28, 2012 at 02:10:05PM +0400, Konstantin Serebryany wrote:
>> I'd like to understand our long-term strategy wrt the asan/tsan tests in gcc.
>> Most of the tests we have today are not specific to the compiler and
>> so can potentia
On Tue, Nov 27, 2012 at 5:39 PM, Eric Botcazou wrote:
> Hi,
>
> this fixes the couple of gnat.dg regressions on all platforms introduced by
> the sizetype changes, as well as finishes the conversion of the Ada compiler
> to the new scheme by invoking valid_constant_size_p and size_binop in the
> a
Hi,
committed to mainline.
Thanks,
Paolo.
/
2012-11-28 Paolo Carlini
PR c++/55497
* g++.dg/init/pr55497.C: New.
Index: g++.dg/init/pr55497.C
===
--- g++.dg/init/pr55497.C (revision 0)
+
On Wed, Nov 28, 2012 at 02:40:55PM +0400, Konstantin Serebryany wrote:
> I fully agree about "minimal testsuite".
> But, for example, porting the asan's gtest test (2+ KLOC) to another
> harness is probably too much.
Depends on how significant changes to the test body are actually needed,
and if w
Hi,
The backport for this patch was not entirely clean,
__builtin_thread_pointer was made a front end builtin on the
4.8 branch, but remains a back-end builtin on aarch64-4.7.
Otherwise, there are no differences between this patch and the
patch which was committed to mainline last week.
OK to c
That's a bit scary (and will be slower than with gtest).
But if we can limit the changes by replacing
asan/tests/asan_test_config.h (and maybe some minimal other changes)
that may work.
--kcc
On Wed, Nov 28, 2012 at 3:03 PM, Jakub Jelinek wrote:
> On Wed, Nov 28, 2012 at 02:40:55PM +0400, Konsta
On Wed, Nov 28, 2012 at 12:03:27PM +0100, Jakub Jelinek wrote:
> Then the test would be run once without ASAN_DIE_IF in environment (or =0),
> that would produce output full of
> EXPECT_DEATH1 AddressSanitizer:.*heap-use-after-free EXPECT_DEATHEND1
> ...
> which tcl could parse, and figure from it
On Tue, 2012-11-27 at 19:49 +0100, Ulrich Weigand wrote:
> Mark Wielaard wrote:
>
> > Which other unwinders are out there, that might rely on the current
> > numbering?
>
> Well, runtime unwinders using .eh_frame should be fine, since this
> uses (and has always used) consistently the GCC numberi
On Wed, Nov 28, 2012 at 3:24 PM, Jakub Jelinek wrote:
> On Wed, Nov 28, 2012 at 12:03:27PM +0100, Jakub Jelinek wrote:
>> Then the test would be run once without ASAN_DIE_IF in environment (or =0),
>> that would produce output full of
>> EXPECT_DEATH1 AddressSanitizer:.*heap-use-after-free EXPECT_
On 11/28/12 10:25, Richard Biener wrote:
On Tue, Nov 27, 2012 at 11:45 PM, Ramana Radhakrishnan
wrote:
On Tue, Nov 27, 2012 at 8:22 PM, Richard Sandiford
wrote:
Ramana Radhakrishnan writes:
Tested on x86_64-linux-gnu, powerpc64-linux-gnu and mipsisa64-elf.
Also tested by making sure that th
This fixes the FAIL of gcc.dg/tree-ssa/pr37508.c on arm. We fail
to canonicalize anti-ranges of 1-bit precision types correctly.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
Richard.
2012-11-28 Richard Biener
PR tree-optimization/54547
* tree-vrp.c
On Wed, Nov 28, 2012 at 1:06 PM, Ramana Radhakrishnan wrote:
> On 11/28/12 10:25, Richard Biener wrote:
>>
>> On Tue, Nov 27, 2012 at 11:45 PM, Ramana Radhakrishnan
>> wrote:
>>>
>>> On Tue, Nov 27, 2012 at 8:22 PM, Richard Sandiford
>>> wrote:
Ramana Radhakrishnan writes:
>>
On Tue, Nov 27, 2012 at 9:37 PM, Martin Jambor wrote:
> Hi,
>
> On Tue, Nov 27, 2012 at 02:02:42PM +0100, Richard Biener wrote:
>> On Wed, Nov 21, 2012 at 5:58 PM, Martin Jambor wrote:
>> > Hi,
>> >
>> > On Tue, Nov 20, 2012 at 09:24:20AM -0800, Richard Henderson wrote:
>> >> The get_pointer_alig
On 11/23/2012 03:24 PM, Jason Merrill wrote:
On 11/23/2012 04:58 AM, Florian Weimer wrote:
Okay, this might work in the sense that it flags the relevant cases. I'm
still not convinced that this actually helps programmers that much
because it pretty much separates the two worlds. If this is the
Hi,
>> Not to mention the ChangeLog entry.
Sorry about the missed patch and ChangeLog Entry.
Please review the patch and let me know if its OK.
Regression Tested on aarch64-elf. No new Regressions.
2012-11-28 Naveen H.S
* aarch64.c (aarch64_load_symref_appropriately): Handle
On Tue, 27 Nov 2012, Jakub Jelinek wrote:
> 2012-11-26 Jakub Jelinek
>
> PR debug/36728
> PR debug/55467
> * cselib.c (cselib_process_insn): If cselib_preserve_constants,
> don't reset table and exit early on volatile insns and setjmp.
> Reset table afterwards on se
On Wed, Nov 28, 2012 at 08:44:18AM -0500, Hans-Peter Nilsson wrote:
> On Tue, 27 Nov 2012, Jakub Jelinek wrote:
> > 2012-11-26 Jakub Jelinek
> >
> > PR debug/36728
> > PR debug/55467
> > * cselib.c (cselib_process_insn): If cselib_preserve_constants,
> > don't reset table and exi
Ramana Radhakrishnan writes:
> On Tue, Nov 27, 2012 at 8:22 PM, Richard Sandiford
> wrote:
>> Ramana Radhakrishnan writes:
Tested on x86_64-linux-gnu, powerpc64-linux-gnu and mipsisa64-elf.
Also tested by making sure that there were no changes in assembly
output for a set of gcc .
On 27/11/12 13:41, Richard Biener wrote:
> On Mon, Nov 19, 2012 at 3:33 AM, Tom de Vries wrote:
>> Richard,
>>
>> Consider the PR55124 example test.c:
>> ...
>> int a, b;
>> long c;
>>
>> static void inline
>> f2 (void)
>> {
>> unsigned long k = 1;
>>
>> foo (b ? k = 0 : 0);
>>
>> b = (((c =
On Wed, Nov 28, 2012 at 3:05 PM, Tom de Vries wrote:
> On 27/11/12 13:41, Richard Biener wrote:
>> On Mon, Nov 19, 2012 at 3:33 AM, Tom de Vries wrote:
>>> Richard,
>>>
>>> Consider the PR55124 example test.c:
>>> ...
>>> int a, b;
>>> long c;
>>>
>>> static void inline
>>> f2 (void)
>>> {
>>>
On Wed, Nov 28, 2012 at 3:29 PM, Richard Biener
wrote:
> On Wed, Nov 28, 2012 at 3:05 PM, Tom de Vries wrote:
>> On 27/11/12 13:41, Richard Biener wrote:
>>> On Mon, Nov 19, 2012 at 3:33 AM, Tom de Vries
>>> wrote:
Richard,
Consider the PR55124 example test.c:
...
int a
On Wed, Nov 28, 2012 at 3:35 PM, Richard Biener
wrote:
> On Wed, Nov 28, 2012 at 3:29 PM, Richard Biener
> wrote:
>> On Wed, Nov 28, 2012 at 3:05 PM, Tom de Vries wrote:
>>> On 27/11/12 13:41, Richard Biener wrote:
On Mon, Nov 19, 2012 at 3:33 AM, Tom de Vries
wrote:
> Richard,
>
On 22/11/12 07:52, Jakub Jelinek wrote:
On Thu, Nov 22, 2012 at 12:26:01AM +0100, Eric Botcazou wrote:
2012-11-21 Jakub Jelinek
* ree.c (struct ext_modified): Add ATTRIBUTE_PACKED.
Are you sure that this will compile with non-GCC compilers?
I'm not sure, but I hope it will. ansi
On Wed, Nov 28, 2012 at 03:02:12PM +, Richard Earnshaw wrote:
> >#if (GCC_VERSION < 2007)
> ># define __attribute__(x)
> >#endif
> >...
> >#ifndef ATTRIBUTE_PACKED
> ># define ATTRIBUTE_PACKED __attribute__ ((packed))
> >#endif
> >
> >So in theory it should expand to nothing for non-GCC compile
On 28/11/12 15:05, Jakub Jelinek wrote:
On Wed, Nov 28, 2012 at 03:02:12PM +, Richard Earnshaw wrote:
#if (GCC_VERSION < 2007)
# define __attribute__(x)
#endif
...
#ifndef ATTRIBUTE_PACKED
# define ATTRIBUTE_PACKED __attribute__ ((packed))
#endif
So in theory it should expand to nothing for
Hi,
another issue pointed out by valgrind:
==12724== Invalid write of size 8
==12724==at 0xD03071: rest_of_handle_dse() (dse.c:2873)
==12724==by 0x82824A: execute_one_pass(opt_pass*) (passes.c:2328)
==12724==by 0x8286B4: execute_pass_list(opt_pass*) (passes.c:2386)
==12724==by 0x8
On 06/11/12 22:05, Andrew Pinski wrote:
Hi,
The problem here is with section anchors turned on, we generate a
BSS rather than a local common symbol and we no longer match the
pattern: "_ZGVZ3foovE1x,8,8". This fixes this testcase by just adding
-fno-section-anchors.
Why is -fsection-anchor
The following proposed change adds a new mode of operation where
the discriminator field in DWARF debugging information is set from
information provided by a language front-end to discriminate
between distinct code instances coming from instances of a given
template. (Currently the discriminator is
Hi Richard,
> This looks like it's working around a bug elsewhere.
Originally, that fail is caused by function join_c_conditions(const
char *cond1, const char *cond2) - if cond1 is "" and cond2 is NULL,
then the function returns cond2, i.e. NULL. Attached patch fixes it -
specifically, with it the
On 11/25/2012 07:26 AM, Kai Tietz wrote:
> - if ((ccvt & (IX86_CALLCVT_FASTCALL | IX86_CALLCVT_THISCALL)) != 0)
> + if ((ccvt & IX86_CALLCVT_FASTCALL) != 0)
> {
> /* Fastcall functions use ecx/edx for arguments, which leaves
>us with EAX for the static chain.
> @
On Wed, Nov 28, 2012 at 8:44 AM, Eric Botcazou wrote:
>> Again: Is this really the direction we should be going in? (I haven't
>> any better ideas...)
>
> Do you mean for self-referencing notes? If so, definitely, these notes are
> either wrong or almost always redundant so we should eliminate th
Latest results for 4.7.x
The replaces the previous update that was never applied.
-tgc
Testresults for 4.7.2
alphaev68-dec-osf5.1a (2)
hppa2.0w-hp-hpux11.00
hppa2.0w-hp-hpux11.11
hppa64-hp-hpux11.00
hppa64-hp-hpux11.11
i386-apple-darwin10.8.0
i386-pc-solaris2.8
i386-pc-solaris2.1
On 11/28/2012 07:55 AM, Michael Zolotukhin wrote:
> Originally, that fail is caused by function join_c_conditions(const
> char *cond1, const char *cond2) - if cond1 is "" and cond2 is NULL
Joining a null condition suggests that's not the original fail.
Where was the null condition created in the f
On Wed, Nov 28, 2012 at 4:46 PM, Thomas Quinot wrote:
> The following proposed change adds a new mode of operation where
> the discriminator field in DWARF debugging information is set from
> information provided by a language front-end to discriminate
> between distinct code instances coming from
On 2012-11-16 13:24 , Caroline Tice wrote:
Index: gcc/cp/decl2.c
===
--- gcc/cp/decl2.c (revision 193571)
+++ gcc/cp/decl2.c (working copy)
@@ -69,8 +69,8 @@ typedef struct priority_info_s {
static void mark_vtable_entrie
Hi!
As discussed on IRC, this testcase clobbers memory before malloc returned
chunk, fixed thusly, tested on x86_64-linux, committed to trunk.
2012-11-28 Jakub Jelinek
PR testsuite/55504
* gcc.c-torture/execute/20071018-1.c (foo): Add noinline/noclone
attributes. Avoi
Hi!
Another broken testcase, I really wonder whether it is worth to keep it,
given that in many places uses uninitialized return values, but for now
I just increased the buffers, so that asan doesn't complain on it.
2012-11-28 Jakub Jelinek
PR testsuite/55505
* gcc.c-torture/e
Hi,
The attached patch implements vector alignment hooks
TARGET_VECTOR_ALIGNMENT
TARGET_VECTORIZE_VECTOR_ALIGNMENT_REACHABLE
TARGET_VECTORIZE_SUPPORT_VECTOR_MISALIGNMENT
Regression tested on aarch64-none-elf. OK for aarch64-4.7 and trunk?
Thanks,
Tejas Belagod.
ARM.
Changelog:
2012-11-28 T
On 28/11/12 13:18, Hurugalawadi, Naveen wrote:
Hi,
Not to mention the ChangeLog entry.
Sorry about the missed patch and ChangeLog Entry.
Please review the patch and let me know if its OK.
Regression Tested on aarch64-elf. No new Regressions.
2012-11-28 Naveen H.S
* aarch64.c (aa
On Wed, Nov 28, 2012 at 04:18:44PM +0100, Markus Trippelsdorf wrote:
> Bootstrapped and tested on x86_64-pc-linux-gnu. Please apply.
> Thanks.
>
> 2012-11-28 Markus Trippelsdorf
>
> PR other/55358
> * dse.c (rest_of_handle_dse): Remove superfluous clearing.
Committed. Thanks.
> Where was the null condition created in the first place?
The reason it's happening is following. Before introduction of
define_subst, function apply_iterators had the following loop:
condition = NULL;
FOR_EACH_VEC_ELT (iterator_uses, i, iuse)
{
v = iuse->iterator->cu
The following patch fixes
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55512
The patch was successfully tested and bootstrapped on x86/x86-64.
Committed as rev. 193901.
2012-11-28 Vladimir Makarov
PR rtl-optimization/55512
* lra-assigns.c (assign_by_spills): Assigned a
2012/11/28 Richard Henderson :
> On 11/25/2012 07:26 AM, Kai Tietz wrote:
>> - if ((ccvt & (IX86_CALLCVT_FASTCALL | IX86_CALLCVT_THISCALL)) != 0)
>> + if ((ccvt & IX86_CALLCVT_FASTCALL) != 0)
>> {
>> /* Fastcall functions use ecx/edx for arguments, which leaves
>>
Ping
2012/11/25 Kai Tietz :
> Hi,
>
> the issue here is that the predefined macro __SEH__ does just indicate
> that SEH-infrastructure is present. It doesn't mean that SEH is used
> as exception-mechansim. Therefore the checks in libgcc's (and as
> followup in libstdc++'s) eh exception-mechansim
Ping
2012/11/25 Kai Tietz :
> Hi,
>
> this patch fixes used exception-mechanism for SEH-enabled targets,
> which are requesting for SjLj-exception-mechanism.
> See also patch for libgcc.
>
> ChangeLog
>
> 2012-11-25 Kai Tietz
>
> PR target/55445
> * libsupc++/eh_personaltity.cc (_
Hello!
> The following patch fixes
>
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55458
>
> The patch was successfully bootstrapped and tested on x86/x86-64.
>
> Committed as rev. 193871.
>
> 2012-11-27 Vladimir Makarov
>
> PR rtl-optimization/55458
> * lra-assigns.c: Include rtl
On 12-11-28 1:10 PM, Uros Bizjak wrote:
Hello!
The following patch fixes
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55458
The patch was successfully bootstrapped and tested on x86/x86-64.
Committed as rev. 193871.
2012-11-27 Vladimir Makarov
PR rtl-optimization/55458
*
On 11/28/2012 10:01 AM, Kai Tietz wrote:
>> 2012-11-25 Kai Tietz
>>
>> PR target/55445
>> * unwind-c (__SEH__): Make sure SjLj isn't
>> active.
>> * unwind-generic.h: Likewise.
>> * unwind-seh.c: Likewise.
>>
Ok.
r~
On 11/28/2012 09:53 AM, Kai Tietz wrote:
> 2012/11/28 Richard Henderson :
>> On 11/25/2012 07:26 AM, Kai Tietz wrote:
>>> - if ((ccvt & (IX86_CALLCVT_FASTCALL | IX86_CALLCVT_THISCALL)) != 0)
>>> + if ((ccvt & IX86_CALLCVT_FASTCALL) != 0)
>>> {
>>> /* Fastcall functions use e
On 11/28/2012 09:20 AM, Michael Zolotukhin wrote:
>> Where was the null condition created in the first place?
> The reason it's happening is following. Before introduction of
> define_subst, function apply_iterators had the following loop:
> condition = NULL;
> FOR_EACH_VEC_ELT (iterato
On Wed, Nov 28, 2012 at 10:52:17AM +0100, Eric Botcazou wrote:
> No, I don't think that's the problem. The above messages are admittedly a
> bit
> terse, they should say:
>
> JUMP-BYPASS: Proved reg 59 in jump_insn 15 equals constant (const_int 3 [0x3])
> when BB 4 is entered from
Fine. Here is a adjusted version which keeps old abi behavior for
chain-registers.
2012-11-28 Kai Tietz
PR target/55171
* i386.c (get_scratch_register_on_entry): Handle
thiscall-convention.
(split_stack_prologue_scratch_regno): Likewise.
(ix86_static_chai
On 11/28/2012 07:02 PM, Kai Tietz wrote:
Ping
2012/11/25 Kai Tietz :
Hi,
this patch fixes used exception-mechanism for SEH-enabled targets,
which are requesting for SjLj-exception-mechanism.
See also patch for libgcc.
ChangeLog
2012-11-25 Kai Tietz
PR target/55445
* libsu
On Sun, Nov 25, 2012 at 7:40 AM, Diego Novillo wrote:
> On Sat, Nov 24, 2012 at 5:47 PM, Doug Evans wrote:
>>
>> Hi.
>> This third patch passes options.results to GetSumFiles when fetching the
>> results
>> for the clean build.
>> It is useful in my use cases, but I'm not sure it's useful for ups
On Wed, Nov 28, 2012 at 1:09 AM, Paolo Bonzini wrote:
> I cannot approve the collect2.c, gcc.c, opts.c, doc/invoke.texi. I'll
> include some comments anyway, since the patch needs some more work.
>
>> 2011-06-27 Doug Kwan
>>
>> Google ref 41164-p2
>> Backport upstream patch under
As per Eric's suggestiong, this patch makes one debug message more
clear. Eric, I know you you pre-approved this, but still I'd
appreciate if you could take a look. Thanks,
2012-11-28 Marek Polacek
* cprop.c (bypass_block): Improve debug message.
--- gcc/cprop.c.mp 2012-11-28 1
> Well, there does seem to be a mistake -- the use of NULL in the first
> place. It seems to me that the easiest fix is
>
> condition = "";
>
> right at the beginning.
Yep, that'll work too, you're right.
On 28 November 2012 22:36, Richard Henderson wrote:
> On 11/28/2012 09:20 AM, Michael Zol
On 11/28/2012 10:52 AM, Kai Tietz wrote:
> 2012-11-28 Kai Tietz
>
> PR target/55171
> * i386.c (get_scratch_register_on_entry): Handle
> thiscall-convention.
> (split_stack_prologue_scratch_regno): Likewise.
> (ix86_static_chain): Likewise.
> (x86_o
Eric Botcazou writes:
>> I strongly disagree with this. Outputs and clobbers have significant
>> meaning even on volatile asms, asm volatile doesn't mean all registers and
>> all memory are supposed to be considered clobbered. For memory you have
>> "memory" clobber for that, for registers it is
On Wed, Nov 28, 2012 at 1:55 PM, Doug Evans wrote:
> In gdb-land, parallel make check runs collect all the subdirectory
> .sum files and reconstruct testsuite/${tool}.sum.
>
> There's more than one solution of course: alternatively one could have
> gdb stop doing this. But I'm not sure which is
On 11/27/2012 09:35 AM, Eric Botcazou wrote:
> It's well established in the RTL middle-end that UNSPEC_VOLATILE can do
> pretty
> much everything behind the back of the compiler.
This is false. U_V is a scheduling barrier, and is never deleted as
dead (as opposed to unreachable) code. Certainl
On 11/27/2012 04:03 AM, Jakub Jelinek wrote:
> 2012-11-26 Jakub Jelinek
>
> PR debug/36728
> PR debug/55467
> * cselib.c (cselib_process_insn): If cselib_preserve_constants,
> don't reset table and exit early on volatile insns and setjmp.
> Reset table afterwards o
Hi,
When we are generating debug info, we still scan all global
declarations and call debug_nothing_tree on each declaration.
It is a waste of time and can add up when there are many
global declarations. This patch skips the scan if
debug_hooks->global_decl == &debug_nothing_tree. OK to install?
Richard Earnshaw writes:
> On 06/11/12 22:05, Andrew Pinski wrote:
>> Hi,
>>The problem here is with section anchors turned on, we generate a
>> BSS rather than a local common symbol and we no longer match the
>> pattern: "_ZGVZ3foovE1x,8,8". This fixes this testcase by just adding
>> -fno-se
Thanks to Torvald for the many sanity checks analyzing this problem.
Tested on x86_64 and alpha-linux.
r~
PR libgcc/48076
* emutls.c (__emutls_get_address): Avoid race condition between
obj->loc.offset read and emutls_key initialization.
diff --git a/libgcc/emutls.c b/l
Il 28/11/2012 20:11, H.J. Lu ha scritto:
> Here is the updated patch. OK for trunk?
Looks good to me, though of course there are many parts I cannot
approve. Thanks!
Paolo
On Wed, Nov 28, 2012 at 1:13 PM, Paolo Bonzini wrote:
> Il 28/11/2012 20:11, H.J. Lu ha scritto:
>> Here is the updated patch. OK for trunk?
>
> Looks good to me, though of course there are many parts I cannot
> approve. Thanks!
>
Joseph, can you take a look at driver change?
Thanks.
--
H.J
Ramana Radhakrishnan writes:
> On Sun, Oct 7, 2012 at 8:56 AM, Richard Sandiford
> wrote:
>> Eric Botcazou writes:
I think modelling it as a TRUNCATE operation is correct for
!TRULY_NOOP_TRUNCATION (it's the bug that Andrew pointed out).
And we shouldn't generate an actual TRUNCAT
Stupid bug of mine: When I cleaned up the code at some point, I forgot
to change se.expr to "image" (se.expr is alway NULL_TREE at that point.)
Committed as Rev. 193908
Tobias
Index: gcc/fortran/ChangeLog
===
--- gcc/fortran/Change
On Wed, 28 Nov 2012, H.J. Lu wrote:
> On Wed, Nov 28, 2012 at 1:13 PM, Paolo Bonzini wrote:
> > Il 28/11/2012 20:11, H.J. Lu ha scritto:
> >> Here is the updated patch. OK for trunk?
> >
> > Looks good to me, though of course there are many parts I cannot
> > approve. Thanks!
> >
>
> Joseph, c
> Well, I'm not sure I agree that they are wrong. Consider:
>
> S0: r1 = ...
> S1: r2 = r1 + 10
> S2: r1 = r2 + 10 { REG_EQUAL r1 + 20 }
> S3: r3 = r1 + 10
>
> Clearly it would be wrong to use the REG_EQUAL note on S2 to optimize S3 to:
>
> S0: r1 = ...
> S1: r2 = r1 + 10
> S2: r1 = r1 + 20
> S3
> As per Eric's suggestiong, this patch makes one debug message more
> clear. Eric, I know you you pre-approved this, but still I'd
> appreciate if you could take a look. Thanks,
>
> 2012-11-28 Marek Polacek
>
> * cprop.c (bypass_block): Improve debug message.
That's fine, thanks.
--
On Wed, Nov 28, 2012 at 12:35 PM, Diego Novillo wrote:
> On Wed, Nov 28, 2012 at 1:55 PM, Doug Evans wrote:
>
>> In gdb-land, parallel make check runs collect all the subdirectory
>> .sum files and reconstruct testsuite/${tool}.sum.
>>
>> There's more than one solution of course: alternatively on
> This is because the structure we are given is:
>
> (mem/v/j:SI (reg/v/f:SI 110 [ t ]) [3 t_2(D)->a+0 S1 A32])
>
> i.e. a 1-byte SImode reference. The strange size comes from the
> set_mem_attributes_minus_bitpos code that was mentioned earlier
> in the series, where we set the size based o
On Wed, Nov 28, 2012 at 6:02 PM, Doug Evans wrote:
> On Wed, Nov 28, 2012 at 12:35 PM, Diego Novillo wrote:
>>
>> Am I right in understanding that you are intending to call
>>
>> $ validate_failures.py --clean_build=/path/to/gdb/bld
>> --manifest=/path/to/gdb.sum ?
>>
>> If that's not what you ar
On Wed, Nov 28, 2012 at 11:10 PM, Eric Botcazou wrote:
>> Well, I'm not sure I agree that they are wrong. Consider:
>>
>> S0: r1 = ...
>> S1: r2 = r1 + 10
>> S2: r1 = r2 + 10 { REG_EQUAL r1 + 20 }
>> S3: r3 = r1 + 10
>>
>> Clearly it would be wrong to use the REG_EQUAL note on S2 to optimize S3 to:
Ping #1 for:
http://gcc.gnu.org/ml/gcc-patches/2012-11/msg01914.html
http://gcc.gnu.org/ml/gcc-patches/2012-11/msg01915.html
Johann
PR ada/55243
* Makefile.in (TOOLS_FLAGS_TO_PASS_CROSS): Set STAMP.
On 11/28/2012 06:13 PM, Ed Smith-Rowland wrote:
Ultimately we should probably revisit this and 54413 by parsing the literal
numbers in the C++ front end. That won't happen for 4.8.
That sounds like you're proposing to duplicate code?
- /* Both C and C++ require a diagnostic for a floating c
In this testcase:
__attribute__((transaction_callable))
void q1()
{
ringo=666;
__transaction_atomic {
george=999;
}
}
...the clone for q1() ends up with instrumented code on both pathways:
:
_12 = tm_state.5_11 & 2;
if (_12 != 0)
goto ;
else
goto ;
:
_13 = 999;
Below is a first attempt at native TLS support on AIX. It produces
the correct syntax and works for small testcases. All of GCC can
build with it enabled, but libstdc++ and libgomp do not run correctly,
so I am not enabling it by default.
The implementation treats local-dynamic as general-dynami
This patch ensures that the histograms from the profile summary are streamed
through the LTO files so that the working set can be computed for use in
downstream optimizations.
Bootstrapped and tested on x86_64-unknown-linux-gnu. Ok for trunk?
Thanks,
Teresa
2012-11-28 Teresa Johnson
From: "H.J. Lu"
To: gcc-patches@gcc.gnu.org
Cc: Joseph Myers , Paolo Bonzini
Bcc:
Subject: [PATCH] Support -fuse-ld=bfd and -fuse-ld=gold
Reply-To:
Hi,
Binutils supports 2 linkers, ld.gold and ld.bfd. One of them is
configured as the default linker, ld, which is used by GCC. Sometimes,
we
Thanks for the input, the patch seems to be much more cute now.
Do we still need to play with release/non-release builds, or is it ok
to commit this patch to the trunk as it is?
Changelog:
2012-11-29 Michael Zolotukhin
* Makefile.in: Add target mddump, build/genmddump.o. Extend
95 matches
Mail list logo
