On Tue, Nov 18, 2014 at 11:19 AM, Paolo Carlini
wrote:
> Jonathan lately is following your work much better than me, but naively
> seems weird that _M_begin is non-const and _M_end is const, a different type
> anyway.
Hmm. The current regex_search algorithm is implemented as try match
starting fr
On November 18, 2014 10:59:28 PM CET, Jakub Jelinek wrote:
>Hi!
>
>The case ASHIFTRT: which is meant for this optimization is preceeded by
>/* FALLTHRU */ from ROTATE cases, which can't be optimized this way.
>Thus, this patch limits this optimization to ASHIFTRT.
>
>Bootstrapped/regtested on x86_
> On 17 November 2014 06:58, Yangfei (Felix) wrote:
> > PING?
> > BTW: It seems that Alan's way of improving vld1(q?)_dup intrinsic is more
> elegant.
> > So is the improvement of vcls(q?) vcnt(q?) OK for trunk? Thanks.
>
> Please rebase over Alan's patch and repost, thank you /Marcus
I rebase
On 11/19/14 11:03, Chen Gang wrote:
> On 11/19/14 0:44, Joseph Myers wrote:
>> On Sun, 16 Nov 2014, Chen Gang wrote:
>>
>>> The maximize 64-bits integer decimal string length excluding NUL is 20 (
>>> '-9223372036854775808'), so need use 20 instead of 18 for HOST_WIDE_INT.
>>>
>>> 2014-11-16 Chen
The first patch fixes 63924; I'm not sure whether or not it is really
valid, but it's easy enough to allow "copying" a non-constant empty
class object.
The second patch fixes GNU statement-expressions in
constant-expressions. This is mostly interesting statement-expressions
are used internal
On 11/19/14 0:44, Joseph Myers wrote:
> On Sun, 16 Nov 2014, Chen Gang wrote:
>
>> The maximize 64-bits integer decimal string length excluding NUL is 20 (
>> '-9223372036854775808'), so need use 20 instead of 18 for HOST_WIDE_INT.
>>
>> 2014-11-16 Chen Gang
>>
>> * c-family/c-cppbuiltin.c
Current thumb2 -Os generates suboptimal code for following tail call case:
int f4(int b, int a, int c, int d);
int g(int a, int b, int c, int d)
{ return f4(b, a, c, d); }
arm-none-eabi-gcc -Os -mthumb -mcpu=cortex-m3 test.c
push
{r4, lr}
mov r4, r1
mov r1, r0
mov r0, r4
pop {r4, lr}
b f4
Ther
On Tue, 18 Nov 2014, Matthew Fortune wrote:
> > > With a quick look at `mips_process_sync_loop' it looks to me the
> > > other NOP is produced from here:
> > >
> > > else if (!(required_oldval && cmp))
> > > mips_multi_add_insn ("nop", NULL);
> > >
> > > -- correct? If so, then can't you j
On Nov 18, 2014, at 3:42 PM, Jakub Jelinek wrote:
> No, I'm not touching tmp array at all in that case, only look at vp
> individual bytes. Either they are all 0, or all 0xff, or I return NULL.
Oh, sorry, I misread where the break; in your code was going. I might have
been misled by:
> -
On Tue, Nov 18, 2014 at 07:59:26PM +0100, Jan Hubicka wrote:
> > Hi,
> >
> > On Tue, Nov 18, 2014 at 04:39:00PM +0100, Jan Hubicka wrote:
> > > > On Fri, Nov 14, 2014 at 08:59:10PM +0100, Jan Hubicka wrote:
> > > > > > >
> > > > > > > > b) with GTY, we cannot call destructor
> > > > > > >
> > >
On Tue, Nov 18, 2014 at 03:30:56PM -0800, Mike Stump wrote:
> > Before wide-int got merged, the testcase worked, though the code didn't
> > bother checking anything, just created 0 or constm1_rtx for the two cases
> > that could happen and if something else appeared, could just return what
> > matc
Hi,
On Mon, Nov 17, 2014 at 01:05:23PM +0100, Richard Biener wrote:
> On Sat, Nov 15, 2014 at 2:04 AM, Martin Jambor wrote:
> > Hi,
> >
> > this patch adds very simple propagation of alignment of pointers to
> > IPA-CP. Because I have not attempted to estimate profitability of
> > such propagati
On Nov 18, 2014, at 1:52 PM, Jakub Jelinek wrote:
> OImode/XImode on i?86/x86_64 are not <= MAX_BITSIZE_MODE_ANY_INT, because
> they are never used for integer arithmetics (and there is no way
> to represent all their values in RTL if not using CONST_WIDE_INT).
> As the following testcase shows, s
On 11/18/14 09:57, Tom de Vries wrote:
Richard,
this (trunk) patch fixes PR62167.
The patch fixes a problem that triggers with the test-case on the 4.8
branch, when tail-merge makes a dead type-unsafe load alive.
I'm not able to reproduce this bug on 4.9 and trunk with the same
test-case. On t
On Tue, Nov 18, 2014 at 09:48:52PM +0100, Jan Hubicka wrote:
> Hi,
> this patch goes through most of ipa passes: ipa-devirt, ipa-cp,
> ipa-pure-const,
> ipa-profile and ipa-inline and audits them for opt_for_fn.
> I did not converted yet ipa-reference because the code is organized in a way
> makin
This patch updates ssa and inline summary in the correct location for AutoFDO.
Bootstrapped and passed regression test. OK for trunk?
Thanks,
Dehao
gcc/ChangeLog:
2014-11-18 Dehao Chen
* auto-profile.c (afdo_annotate_cfg): Invoke update_ssa in the right
place.
(auto_p
On Fri, Oct 17, 2014 at 6:08 AM, Yuri Rumyantsev wrote:
> Jeff,
>
> I prepared another patch that includes test-case as you requested.
>
> Below are answers on your questions.
>
>> First, for the benefit of anyone trying to understand what you're doing,
>> defining what "cd equivalent" means woul
On Tue, Nov 18, 2014 at 07:59:26PM +0100, Jan Hubicka wrote:
> > Hi,
> >
> > On Tue, Nov 18, 2014 at 04:39:00PM +0100, Jan Hubicka wrote:
> > > > On Fri, Nov 14, 2014 at 08:59:10PM +0100, Jan Hubicka wrote:
> > > > > > >
> > > > > > > > b) with GTY, we cannot call destructor
> > > > > > >
> > >
Richard Biener wrote:
> On Tue, Nov 11, 2014 at 2:14 AM, Sebastian Pop wrote:
> > Hi Jeff,
> >
> > I have adapted the code generation part from James' patch to current trunk,
> > and
> > the resulting patch gets the 30% speedup on coremark and passes bootstrap
> > of GCC.
> >
> > Ok for trunk?
>
On Tue, Nov 18, 2014 at 09:57:37PM +, James Greenhalgh wrote:
> 2014-11-18 James Greenhalgh
>
> PR target/63937
> * target.def (use_by_pieces_infrastructure_p): Take unsigned
> HOST_WIDE_INT as the size parameter.
> * targhooks.c (default_use_by_pieces_infrastructure
Hi!
This PR got fixed by a fix for another PR, but I've committed
the testcase to the testsuite for better coverage.
2014-11-18 Jakub Jelinek
PR tree-optimization/61042
* gcc.c-torture/compile/pr61042.c: New test.
--- gcc/testsuite/gcc.c-torture/compile/pr61042.c.jj2014-1
On 11/17/14 03:16, Bob Dunlop wrote:
Hi,
This patch prevents optimizing libraries for space from overriding all
previous CFLAGS_FOR_TARGET settings. Add optimization flags to any
pre-existing values.
I hit this problem whilst building a Gcc cross compiler and newlib
library for ARM with crosst
Hi!
Simd clone vectorization uses vect_finish_stmt_generation which
handles adding the new calls properly to EH tables, but afterwards
we replace the original call with a normal assignment, and if the original
call can throw, we need to remove it from the EH tables.
Fixed thusly, bootstrapped/reg
Hi!
Apparently, expand_expr with EXPR_WRITE can return
a SUBREG with SUBREG_PROMOTED_VAR_P set on it. For
UBSAN_CHECK_{ADD,SUB,MUL} expansion, I've been doing just
emit_move_insn into it, which apparently is wrong in that case,
store_expr instead uses convert_move for it. The
{ADD,SUB,MUL}_OVERF
OK.
Jason
Hi!
The case ASHIFTRT: which is meant for this optimization is preceeded by
/* FALLTHRU */ from ROTATE cases, which can't be optimized this way.
Thus, this patch limits this optimization to ASHIFTRT.
Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
2014-11-18 Jakub Jelinek
Hi,
The TARGET_USE_BY_PIECES_INFRASTRUCTURE_P hook takes an unsigned
int for the size of memory operation. This is dangerous, as all the
callers of this hook pass an unsigned HOST_WIDE_INT.
This causes pr63937, where the compiler ends up trying to emit
an unfeasible number of instructions for a
Hi!
As the testcase shows, if there is a reinterpret_cast, ubsan
reference instrumentation can ICE, because for NOP_EXPR
to REFERENCE_TYPE we pass argument of that NOP_EXPR, and if it is not
a POINTER_TYPE or POINTER_TYPE to a different type, we can get the
type to use in the diagnostics wrong or
Hi!
OImode/XImode on i?86/x86_64 are not <= MAX_BITSIZE_MODE_ANY_INT, because
they are never used for integer arithmetics (and there is no way
to represent all their values in RTL if not using CONST_WIDE_INT).
As the following testcase shows, simplify_immed_subreg can be called
with such modes tho
Hi!
This patch fixes instrumentation of bool/enum loads if they could
throw. We want to keep the EH on the load, and push
further statements on the fallthru edge.
Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
2014-11-18 Jakub Jelinek
PR sanitizer/63913
I'm not sure why this wasn't showing up on x86_64-linux, but here's a
fix. When we lower SIZEOF_EXPR during genericization, we don't then
fold the containing expression, so the constexpr evaluator shouldn't
assume that everything has been folded.
Tested x86_64-pc-linux-gnu, applying to trunk.
On 11/18/14 09:48, Ilya Enkovich wrote:
On 15 Nov 00:10, Jeff Law wrote:
On 11/14/14 10:26, Ilya Enkovich wrote:
Hi,
This patch introduces a simple library with several wrappers to be used with
MPX and Pointer Bounds Checker. Wrappers allow to obtain, copy and just keep
alive bounds whrough
On Tue, 2014-11-18 at 10:53 +0100, Richard Biener wrote:
> On Tue, Nov 18, 2014 at 2:59 AM, David Malcolm wrote:
> > On Mon, 2014-11-17 at 11:06 +0100, Richard Biener wrote:
> >> On Sat, Nov 15, 2014 at 12:00 PM, David Malcolm
> >> wrote:
> >> > On Thu, 2014-11-13 at 11:45 +0100, Richard Biener
On 11/12/2014 06:53 AM, Markus Trippelsdorf wrote:
Anyway, I will defer working on this until next stage1, because I don't
have time to implement this before stage1 closes on Saturday.
I think this should be OK for stage 3, especially given that you first
sent the patch before the end of stage
On 11/18/2014 02:32 PM, Andrew MacLeod wrote:
So it effectively does nothing. Unless Jason can think of a good
reason for it, we probably ought to turf it. Its effectively a NOP.
Yeah, go ahead.
Jason
Hi,
this patch goes through most of ipa passes: ipa-devirt, ipa-cp, ipa-pure-const,
ipa-profile and ipa-inline and audits them for opt_for_fn.
I did not converted yet ipa-reference because the code is organized in a way
making it bit difficult to hook the tests in and ipa-icf that I want to do
sepa
On Tue, 18 Nov 2014, James Greenhalgh wrote:
On Wed, Nov 05, 2014 at 11:31:24PM +, James Greenhalgh wrote:
On Wed, Nov 05, 2014 at 09:50:52PM +, Marc Glisse wrote:
Thanks. Do you know if anyone is planning to "port" this patch to the arm
target (which IIRC has the same issue)? No press
> > From: Maciej W. Rozycki [mailto:ma...@codesourcery.com]
> >
> > On Tue, 18 Nov 2014, Andrew Bennett wrote:
> >
> > > Produces (for the atomic operation):
> > >
> > >.setnoat
> > > sync
> > > 1:
> > > ll $3,0($5)
> > > and $1,$3,$4
> > > bne
On 11/18/2014 05:56 AM, Renlin Li wrote:
> Hi all,
>
> This patch should fix PR63762.
>
> Hard_reg allocated during IRA pass conflicts with hard_reg allocated
> in LRA pass because of inconsistent information between allocno and
> reg_pref.
>
> The staled reg_class information in reg_pref is used b
On Mon, Nov 17, 2014 at 11:53 PM, Jakub Jelinek wrote:
> On Mon, Nov 17, 2014 at 11:39:47PM -0800, Alexey Samsonov wrote:
>> > That is not what I think we've agreed on and what is implemented in GCC.
>> > -fsanitize-recover only enables recovery of the undefined plus
>> > undefined-like sanitizers
On 11/18/2014 01:36 PM, Jeff Law wrote:
On 11/18/14 09:30, Andrew MacLeod wrote:
I tried doing the if before chaning to TREE_TYPE... absolutely no
effect on the testsuite or anything else :-)
What do you think, should I check this in? What is there is clearly
incorrect.we could also rev
On 11/18/14 12:25, Steve Ellcey wrote:
On Mon, 2014-11-17 at 09:24 +, James Greenhalgh wrote:
For what it is worth, I've bootstrapped and tested this patch on
aarch64-none-linux-gnu and arm-none-linux-gnueabi with no issues, and
both targets get the expected speedup in the interesting bench
On Mon, 2014-11-17 at 09:24 +, James Greenhalgh wrote:
> For what it is worth, I've bootstrapped and tested this patch on
> aarch64-none-linux-gnu and arm-none-linux-gnueabi with no issues, and
> both targets get the expected speedup in the interesting benchmark.
> I've also thrown some of the
> > -Original Message-
> > From: Matthew Fortune [mailto:matthew.fort...@imgtec.com]
> > Sent: Tuesday, November 18, 2014 12:22 PM
> > To: Rozycki, Maciej
> > Cc: gcc-patches@gcc.gnu.org; Moore, Catherine; Eric Christopher
> > Subject: RE: [PATCH] MIPS/GCC: Unconditional jump generation bug
Hi,
On 11/18/2014 08:12 PM, Tim Shen wrote:
Bootstrapped and tested.
Jonathan lately is following your work much better than me, but naively
seems weird that _M_begin is non-const and _M_end is const, a different
type anyway.
Paolo.
Bootstrapped and tested.
Thanks!
--
Regards,
Tim Shen
commit f17155183b9ae1283d04f3bbdb61d05d9279ebe4
Author: timshen
Date: Tue Nov 18 00:07:28 2014 -0800
PR libstdc++/63920
* include/bits/regex_executor.h: Make _M_begin non const.
* include/bits/regex_executor.tcc (
The middle-end doesn't like when I give it a PLUS_EXPR with pointer type.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit 98e8254b2e89e1e34521a7cf9f027bd78eddd3ab
Author: Jason Merrill
Date: Tue Nov 18 13:36:46 2014 -0500
PR c++/63925
* constexpr.c (cxx_eval_increment_expressi
> Hi,
>
> On Tue, Nov 18, 2014 at 04:39:00PM +0100, Jan Hubicka wrote:
> > > On Fri, Nov 14, 2014 at 08:59:10PM +0100, Jan Hubicka wrote:
> > > > > >
> > > > > > > b) with GTY, we cannot call destructor
> > > > > >
> > > > > > Everything in symbol table is expecitely memory managed (i.e. enver
On Tue, Nov 18, 2014 at 1:44 PM, Jeff Law wrote:
> On 11/18/14 05:18, Richard Biener wrote:
>
>>>
>>> As I understand the main problem is that fixing dbxout in trunk won't
>>> help to build stage1 compiler.
>>
>>
>> Simply build stage1 without debug info on AIX then.
>>
>> Btw, you have to start f
On 11/18/14 05:18, Richard Biener wrote:
As I understand the main problem is that fixing dbxout in trunk won't
help to build stage1 compiler.
Simply build stage1 without debug info on AIX then.
Btw, you have to start fixing the bug at some point ... (we can
backport to 4.8 and 4.9). Of cour
On 11/18/14 06:42, Segher Boessenkool wrote:
On Tue, Nov 18, 2014 at 06:38:49AM -0600, Segher Boessenkool wrote:
Does it help pr52714 where we'd like to rip apart a PARALLEL with two
sets, one of which is dead. If so, it might allow us to optimize that
code better.
It does not seem to fix th
On 11/18/14 09:30, Andrew MacLeod wrote:
I tried doing the if before chaning to TREE_TYPE... absolutely no
effect on the testsuite or anything else :-)
What do you think, should I check this in? What is there is clearly
incorrect.we could also revert the original patch since that is what
On 18/11/14 10:45 +0100, Richard Biener wrote:
Looking at all these issues that just pop up inside libstdc++ I wonder
if this whole business will not blow up in our face once out in the wild...
I'm trying to ensure that most of the the pain is dealt with inside
libstdc++ and so users won't have
Hi,
On Tue, Nov 18, 2014 at 04:39:00PM +0100, Jan Hubicka wrote:
> > On Fri, Nov 14, 2014 at 08:59:10PM +0100, Jan Hubicka wrote:
> > > > >
> > > > > > b) with GTY, we cannot call destructor
> > > > >
> > > > > Everything in symbol table is expecitely memory managed (i.e. enver
> > > > > left
>
The Go frontend was slightly incorrect in the way that it handled the
order of initialization (http://golang.org/issue/8052). Fortunately
it rarely made a difference in real code. This patch by Chris
Manghane fixes it. Bootstrapped and ran Go testsuite on
x86_64-unknown-linux-gnu. Committed to
> On Tue, Nov 18, 2014 at 9:27 AM, Jan Hubicka wrote:
> >> https://gcc.gnu.org/ml/gcc-regression/2014-11/msg00473.html
> >>
> >> /export/gnu/import/git/gcc-test-profiled/bld/./prev-gcc/xg++
> >> -B/export/gnu/import/git/gcc-test-profiled/bld/./prev-gcc/
> >> -B/usr/5.0.0/x86_64-unknown-linux-gnu/b
On Tue, Nov 18, 2014 at 9:27 AM, Jan Hubicka wrote:
>> https://gcc.gnu.org/ml/gcc-regression/2014-11/msg00473.html
>>
>> /export/gnu/import/git/gcc-test-profiled/bld/./prev-gcc/xg++
>> -B/export/gnu/import/git/gcc-test-profiled/bld/./prev-gcc/
>> -B/usr/5.0.0/x86_64-unknown-linux-gnu/bin/ -nostdin
> https://gcc.gnu.org/ml/gcc-regression/2014-11/msg00473.html
>
> /export/gnu/import/git/gcc-test-profiled/bld/./prev-gcc/xg++
> -B/export/gnu/import/git/gcc-test-profiled/bld/./prev-gcc/
> -B/usr/5.0.0/x86_64-unknown-linux-gnu/bin/ -nostdinc++
> -B/export/gnu/import/git/gcc-test-profiled/bld/prev
> -Original Message-
> From: Matthew Fortune [mailto:matthew.fort...@imgtec.com]
> Sent: Tuesday, November 18, 2014 12:22 PM
> To: Rozycki, Maciej
> Cc: gcc-patches@gcc.gnu.org; Moore, Catherine; Eric Christopher
> Subject: RE: [PATCH] MIPS/GCC: Unconditional jump generation bug fix
>
>
Maciej W. Rozycki writes:
> On Mon, 17 Nov 2014, Matthew Fortune wrote:
>
> > > gcc/
> > > * gcc/config/mips/mips.md (*jump_absolute): Use a branch when in
> > > range, a jump otherwise.
> >
> > OK.
> >
> > I only got my head around this code last week otherwise I wouldn't
> > have known wh
On 15 November 2014 10:23, Tobias Burnus wrote:
> Especially since color diagnostic is now the default [1], it makes sense to
> convert more gfortran diagnostics to use the common diagnostics.
>
> For an example, see [1]. That also brings all the nice features like placing
> the warning option in
On ARM constructors return a pointer, so the test for void return
breaks. We also don't need to mess with the CONSTRUCTOR we built up in
the new scheme.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit ddfea2b5af06860433d32ad440673c2df37dc8f1
Author: Jason Merrill
Date: Tue Nov 18 11:1
Richard,
this (trunk) patch fixes PR62167.
The patch fixes a problem that triggers with the test-case on the 4.8 branch,
when tail-merge makes a dead type-unsafe load alive.
I'm not able to reproduce this bug on 4.9 and trunk with the same test-case. On
those branches, the tail-merge already
Jason,
After your recent C++ patches from yesterday, I am encountering a huge
number of libstdc++ failures on AIX preventing creation of
executables. They all are of the form:
/tmp/20141118/powerpc-ibm-aix7.1.0.0/libstdc++-v3/include/bits/regex_compiler.h:
424:31: in constexpr expansion of
Hi,
On Sun, Nov 16, 2014 at 12:56:45AM +0100, Jan Hubicka wrote:
> Hi,
> this patch enables propagation of speculative contextes I promised to fix
> after Martin's
> merge. There were few bugs that ended up disturbing testsuite:
Wonderful, thanks a lot.
>
> 1) ipa_polymorphic_call_context::co
On Mon, 17 Nov 2014, Moore, Catherine wrote:
> > * config/mips/mips.md (compression): Add `micromips32' setting.
> > (enabled, length): Handle it.
> > (shift_compression): Replace `micromips' with `micromips32' in
> > the `compression' attribute.
> > (*add3, sub3): Likewise.
>
On Mon, 17 Nov 2014, Matthew Fortune wrote:
> > gcc/
> > * gcc/config/mips/mips.md (*jump_absolute): Use a branch when in
> > range, a jump otherwise.
>
> OK.
>
> I only got my head around this code last week otherwise I wouldn't have
> known whether this was correct!
Committed now
On 15 Nov 00:10, Jeff Law wrote:
> On 11/14/14 10:26, Ilya Enkovich wrote:
> >Hi,
> >
> >This patch introduces a simple library with several wrappers to be used with
> >MPX and Pointer Bounds Checker. Wrappers allow to obtain, copy and just
> >keep alive bounds whrough widely use library calls.
On Sun, 16 Nov 2014, Chen Gang wrote:
> The maximize 64-bits integer decimal string length excluding NUL is 20 (
> '-9223372036854775808'), so need use 20 instead of 18 for HOST_WIDE_INT.
>
> 2014-11-16 Chen Gang
>
> * c-family/c-cppbuiltin.c (builtin_define_with_int_value): Use
>
On 18 Nov 16:23, Joseph Myers wrote:
> On Tue, 18 Nov 2014, Ilya Enkovich wrote:
>
> > +@item -fcheck-pointer-bounds
> > +@opindex fcheck-pointer-bounds
> > +@opindex fno-check-pointer-bounds
> > +Enable Pointer Bounds Checker instrumentation. Each memory reference
> > +is instrumented with check
On 11/18/2014 09:52 AM, Andrew MacLeod wrote:
On 11/18/2014 09:40 AM, Jason Merrill wrote:
On 11/18/2014 09:26 AM, Andrew MacLeod wrote:
I was poking around attribs.c while trial running my tree-type-safety
stuff, and it triggered something in decl_attributes() that seems fishy
to me. It looks
On Mon, 17 Nov 2014, Mike Stump wrote:
> > http://gcc.gnu.org/ml/gcc-patches/2014-09/msg00242.html
> >
> > is still waiting, please review.
>
> Wait no more.
>
> Ok.
Applied now, thanks for your review.
Maciej
> From: Richard Biener [mailto:richard.guent...@gmail.com]
> Sent: Monday, November 17, 2014 12:47 PM
>
> Hmm. I am a little bit concerned about the malloc traffic generated here.
> So why not use a vec, get rid of the ->next pointer and
> use a hash_map to associate the stmt with
> an index int
On Tue, 18 Nov 2014, Ilya Enkovich wrote:
> +@item -fcheck-pointer-bounds
> +@opindex fcheck-pointer-bounds
> +@opindex fno-check-pointer-bounds
> +Enable Pointer Bounds Checker instrumentation. Each memory reference
> +is instrumented with checks of pointer used for memory access against
> +boun
On 11/18/14 02:55, Richard Biener wrote:
On Tue, Nov 18, 2014 at 3:59 AM, Jeff Law wrote:
On 11/17/14 13:05, Ilya Enkovich wrote:
How comes you emit debug info for functions that do not exist and thus
are never used? Is problem caused by builtins going after
END_CHKP_BUILTINS? Or some info
On 17/11/14 17:14, Marcus Shawcroft wrote:
On 14 November 2014 15:06, Kyrill Tkachov wrote:
Hi all,
Considering that the erratum workaround option was backported to 4.9, I
assume we'll need an item for that
in the changes.html for that branch?
The text is the same as in the trunk version tha
On Tue, Nov 11, 2014 at 11:59 AM, Kyrill Tkachov wrote:
> Hi all,
>
> This patch models the latency of moves between FP and GP registers on the
> A15 and A57 a bit more accurately by splitting the reservations for FP->GP
> and GP->FP moves and adding an appropriate bypass.
>
> Bootstrapped and tes
Hi,
On Mon, 17 Nov 2014, Richard Biener wrote:
> This means I can no longer interrupt a compile that is running too long?
No, that's not what it means, cc1 will also get the SIGINT.
> You should instead debug the actual compiler, not the driver.
-wrapper is specifically also for invoking cc1
> From: Maciej W. Rozycki [mailto:ma...@codesourcery.com]
>
> On Tue, 18 Nov 2014, Andrew Bennett wrote:
>
> > Produces (for the atomic operation):
> >
> >.setnoat
> > sync
> > 1:
> > ll $3,0($5)
> > and $1,$3,$4
> > bne $1,$7,2f
> >
On Tue, 18 Nov 2014, Andrew Bennett wrote:
> Produces (for the atomic operation):
>
>.setnoat
> sync
> 1:
> ll $3,0($5)
> and $1,$3,$4
> bne $1,$7,2f
> and $1,$3,$6
> or $1,$1,$8
> sc $1,0($5)
>
On Fri, Nov 14, 2014 at 2:19 AM, Richard Biener
wrote:
> On Fri, Nov 14, 2014 at 12:15 AM, H.J. Lu wrote:
>> On Tue, Nov 11, 2014 at 8:02 AM, H.J. Lu wrote:
>>> On Mon, Nov 10, 2014 at 11:42 AM, H.J. Lu wrote:
On Mon, Nov 10, 2014 at 5:44 AM, Richard Biener
wrote:
> On Mon, Nov 1
On Mon, Nov 17, 2014 at 10:38 AM, Jan Hubicka wrote:
> Hi,
> this patch makes us to store default optimization node same way as we stream
> target node. This means that command line options given at compile time
> prevail those given at linktime. Previously we sort of combined both.
>
> We still
On Mon, Nov 17, 2014 at 5:13 PM, Marcus Shawcroft
wrote:
> On 14 November 2014 14:35, Wilco Dijkstra wrote:
>
>> 2014-11-14 Wilco Dijkstra
>>
>> * gcc/config/aarch64/aarch64.c (generic_regmove_cost):
>> Increase FP move cost.
>
> OK /Marcus
Changelog should probably indicate P
Hi, this patch adds a new plugin event PLUGIN_START_PARSE_FUNCTION and
PLUGIN_FINISH_PARSE_FUNCTION that are invoked at start_function and
finish_function respectively in the C and C++ frontends.
PLUGIN_START_PARSE_FUNCTION is called before parsing the function body.
PLUGIN_FINISH_PARSE_FUNCTION
> On Fri, Nov 14, 2014 at 08:59:10PM +0100, Jan Hubicka wrote:
> > > >
> > > > > b) with GTY, we cannot call destructor
> > > >
> > > > Everything in symbol table is expecitely memory managed (i.e. enver left
> > > > to be freed by garbage collector). It resists in GTY only to allow
> > > > link
On Tue, Nov 18, 2014 at 06:31:59PM +0300, Ilya Verbin wrote:
> @@ -8287,7 +8289,9 @@ expand_omp_target (struct omp_region *region)
>if (kind == GF_OMP_TARGET_KIND_REGION)
> {
>unsigned srcidx, dstidx, num;
> +#ifdef ENABLE_OFFLOADING
>struct cgraph_node *node;
> +#endif
Pl
Hi,
This patch disables the generation of sections with offload IR and offload
tables by default, when a compiler is configured without
--enable-offload-targets= .
Bootsrap and regtesting in progress, OK after it finished?
-- Ilya
PR regression/63868
gcc/
* cgraph.c (cgraph_no
On 17/11/14 18:15, Jason Merrill wrote:
On 11/17/2014 05:29 AM, Richard Biener wrote:
can you rename it to copy_fn please? It really copies parameter and
result and then the body.
Ok with that change.
Done. Here's what I'm checking in, along with a second patch to enable
the new code for C+
On Nov 18, 2014, at 03:29 , Jason Merrill wrote:
> What happens when the outer loop hits a register that we've already seen as
> part of a span?
Hmm, I was under the impression that this was supposed never to happen. I can
see that this is not so clear though.
What would happen with the cur
> OK, this does look to me like the correct way to address the issue, but
> where is the second NOP that you previously mentioned? I fail to see it
> here and this code can't be made any better, there isn't anything you
> could possibly schedule into the delay slot as there is nothing else to
> d
On 15 Nov 00:16, Jeff Law wrote:
> On 11/14/14 10:29, Ilya Enkovich wrote:
> >Hi,
> >
> >This patch adds wrapper calls. It's simply achieved by using proper name
> >for builtin clones. Patches apply over builtins instrumentation series.
> >
> >Thanks,
> >Ilya
> >--
> >2014-11-14 Ilya Enkovich
On 18/11/14 02:48, Terry Guo wrote:
+
+ The Thumb-1 assembly code are now generated in unified syntax. The
new option
+-masm-syntax-unified can be used to specify whether
inline assembly
+code are using unified syntax. By default the option is off which means
+
On Tue, 18 Nov 2014, Andrew Bennett wrote:
> My fix places a nop in the delay slot of the branch likely instruction
> by using the %~ output operation. This then causes the sync code for the
> previous example to be correct:
>
> .setnoat
> sync # 15 sync_new_addsi/2
> The atomic-compare-exchange-3.c and atomic-op-3.c tests are failing when
> using the -mfix-r1 option. This is due to the fact that the delay
> slot of the branch instruction that checks if the atomic operation was
> not successful can be filled with an operation that returns the output
> res
While looking at 55992 I noticed that we weren't pushing into dependent
scope properly; we need to pass the "entering scope" flag to
tsubst_aggr_type so that we get the version of the type with the
members. This isn't enough to fix 55992, but may well fix other bugs.
Tested x86_64-pc-linux-gn
> -Original Message-
> From: Maciej W. Rozycki [mailto:ma...@codesourcery.com]
> Sent: 18 November 2014 13:48
> To: Andrew Bennett
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH] If using branch likelies in MIPS sync code fill the delay
> slot with a nop
>
> On Tue, 18 Nov 2014, Andre
On Tue, Nov 18, 2014 at 9:07 AM, Richard Biener
wrote:
> Can we emit this particular STABS piece in "raw" form then? Thus
> does it work to have sth like
>
>.stabs ...
>.string "..."
>.stabs ...
>
> or whatever way to emit assembled data there?
The stabstring debug information is sa
Jason Merrill writes:
> commit e52e82e56507d1de1932abcafd80683c4dc00d1e
> Author: Jason Merrill
> Date: Sun Nov 16 17:14:12 2014 -0500
>
> * constexpr.c (use_new_call): Always use new call handling.
>
> diff --git a/gcc/cp/constexpr.c b/gcc/cp/constexpr.c
> index 57d0c46..8881271 100644
On 11/18/14 0:38, Joseph Myers wrote:
> On Sat, 15 Nov 2014, Chen Gang wrote:
>
>> Also in c_common_read_pch(), when failure occurs, also need be sure the
>> 'fd' is not '-1' for the next close operation.
>
> Please clarify how c_common_read_pch gets called with fd == -1.
> Certainly checking a
On Tue, Nov 18, 2014 at 2:44 PM, David Edelsohn wrote:
> On Tue, Nov 18, 2014 at 8:33 AM, Ilya Enkovich wrote:
>> 2014-11-18 15:18 GMT+03:00 Richard Biener :
>>> On Tue, Nov 18, 2014 at 1:13 PM, Ilya Enkovich
>>> wrote:
2014-11-18 15:04 GMT+03:00 Richard Biener :
> On Tue, Nov 18, 2014
1 - 100 of 180 matches
Mail list logo
