On 05/19/14 21:39, Maxim Kuvyrkov wrote:
This patch improves debug printouts for scheduling passes.
Tested on arm-linux-gnueabihf and x86_64-linux-gnu.
OK to apply?
Yes. This is fine.
jeff
On 05/19/14 22:00, Maxim Kuvyrkov wrote:
Hi,
This patch removes msched-prefer-non-data-spec-insns and
msched-prefer-non-control-spec-insns IA64 options and handling
thereof. I have implemented these option during my youthful
indiscrimination and, as far as I can tell, no one uses them. Yet,
On 05/19/14 18:47, Maxim Kuvyrkov wrote:
Hi,
This patch fixes sched_insn debug counter that was broken by
modulo/backtracking changes.
Tested on arm-linux-gnueabihf and x86_64-linux-gnu, though testsuite
does not exercise this execution path. Real testing was done by me
debugging an unrelated
Hi
The last operand of instruction sqdmulh can only be low fp registers,
so we should use constraint x. But the intrinsic functions use w.
This patch fixed the constrains in these intrinsics.
Passed dejagnu test on aarch64 qemu. OK for trunk, 4.9 and 4.8?
thanks
Guozhi Wei
2014-05-19 Guozhi
On Tue, May 20, 2014 at 07:18:40AM +0100, Carrot Wei wrote:
Hi
Hi,
The last operand of instruction sqdmulh can only be low fp registers,
so we should use constraint x. But the intrinsic functions use w.
This patch fixed the constrains in these intrinsics.
This restriction is only on the
Why duplicating the merger functions in dyn-ipa.c? Should those in
libgcov-merge.c be reused?
The refactoring of gcov_exit_write_gcda should probably be done in a
separate patch -- preferably submitted to trunk too.
David
On Mon, May 19, 2014 at 10:08 PM, Teresa Johnson tejohn...@google.com
On Mon, May 19, 2014 at 04:50:35PM -0400, Jason Merrill wrote:
On 05/13/2014 03:21 AM, Mark Wielaard wrote:
So the debugger doesn't have to guess the properties of the enum's
underlying base type, like size, encoding and signedness.
Well, the enum already has DW_AT_byte_size. It seems to me
Hi,
The patch fix ICE issue triggered by shrink-wrapping enhancement.
Bootstrap and no make check regression on X86-64.
OK for trunk?
Thanks!
-Zhenqiang
2014-05-20 Zhenqiang Chen zhenqiang.c...@linaro.org
PR rtl-optimization/61220
Part of PR rtl-optimization/61225
On Tue, May 20, 2014 at 1:30 AM, Jeff Law l...@redhat.com wrote:
On 05/19/14 00:38, Bin.Cheng wrote:
On Sat, May 17, 2014 at 12:32 AM, Jeff Law l...@redhat.com wrote:
On 05/16/14 04:07, Bin.Cheng wrote:
But can't you go through movXX to generate either the simple insn on the
ARM
or the
On Tue, May 20, 2014 at 5:02 AM, Mike Stump mikest...@comcast.net wrote:
On May 19, 2014, at 10:30 AM, Jeff Law l...@redhat.com wrote:
Yes, I think it's more than upsizing the mode. There is another
example from one of x86's candidate peephole patch at
Richard Biener writes:
The following is my current idea on progressing on the HOST_WIDE_INT
removal
...
And HOST_WIDEST_FAST_INT for which
I don't have a very good suggestion other than either keeping
it, unconditionally using 'long' (thus simply remove
On Tue, 20 May 2014, Maciej W. Rozycki wrote:
Ian,
On Sat, 17 May 2014, Richard Biener wrote:
On May 17, 2014 12:22:23 AM CEST, Maciej W. Rozycki
ma...@codesourcery.com wrote:
On Fri, 16 May 2014, Joseph S. Myers wrote:
2014-05-16 Maciej W. Rozycki ma...@codesourcery.com
On Mon, 19 May 2014, Richard Biener wrote:
On May 19, 2014 6:57:52 PM CEST, Jeff Law l...@redhat.com wrote:
On 05/19/14 06:54, Richard Biener wrote:
In this PR we run into the issue that releasing SSA names from
FRE/PRE elimination corrupts the VN lattice and thus the VN lookup
we
get_attr_enabled was showing up high in a -O0 compile of fold-const.ii.
At the moment, extract_insn calls this function for every alternative
on each extraction, which can be expensive for instructions like
moves that have many alternatives.
The attribute is only supposed to depend on the insn
Honza,
Consider this program:
...
int
main(void)
{
#pragma omp parallel
{
extern void foo(void);
foo ();
}
return 0;
}
...
When compiling this program with -fopenmp, the ompexp pass splits off a new
function called main._omp_fn.0 containing the call to foo. The new function is
Richard Sandiford rdsandif...@googlemail.com writes:
get_attr_enabled was showing up high in a -O0 compile of fold-const.ii.
At the moment, extract_insn calls this function for every alternative
on each extraction, which can be expensive for instructions like
moves that have many alternatives.
Following on from (and depending on) the last patch, process_constraints
also shows up high in the profile. This patch caches the recog_op_alt
information by insn code too. It also shrinks the size of the structure
from 1 pointer + 5 ints to 1 pointer + 2 ints:
- no target should have more than
2014-05-09 16:36 GMT+08:00 Sebastian Huber sebastian.hu...@embedded-brains.de:
On 2014-04-28 18:24, Mike Stump wrote:
On Apr 28, 2014, at 12:32 AM, Sebastian
Hubersebastian.hu...@embedded-brains.de wrote:
I would like to add tests for C/C++ compatibility of atomic operations.
Hi,
On 05/20/2014 04:50 AM, Jason Merrill wrote:
On 05/19/2014 06:43 PM, Paolo Carlini wrote:
if (unqualified_id)
- error (field %qD has incomplete type %qT,
- unqualified_id, type);
+ cxx_incomplete_type_error (unqualified_id, type);
else
Backportet the fix for PR60991 also to 4.7 so that the issue can be closed:
http://gcc.gnu.org/r210635
Johann
gcc/
2014-05-20 Senthil Kumar Selvaraj senthil_kumar.selva...@atmel.com
Backport from mainline r210325
2014-05-12 Senthil Kumar Selvaraj
On Tue, May 20, 2014 at 4:46 AM, Thomas Preud'homme
thomas.preudho...@arm.com wrote:
From: Richard Biener [mailto:richard.guent...@gmail.com]
Agreed, but I am happy with doing that as a followup. Btw,
a very simple one would be to reject unaligned
SLOW_UNALIGNED_ACCESS (TYPE_MODE
On 19/05/14 14:57 -0600, Sandra Loosemore wrote:
On 05/17/2014 04:07 AM, Jonathan Wakely wrote:
On 17 May 2014 10:50, Jonathan Wakely wrote:
On 17 May 2014 01:16, Sandra Loosemore wrote:
It appears that this patch from last fall never got reviewed.
Hi Guys
I am checking in the patch below to fix a failure in the GCC testsuite
for the MSP430. The problem was that PSI mode values only occupy 1
hard register whereas SI mode values occupy 2 registers. This meant
that when gcc generated:
(set (reg:SI r7) (zero_extend:SI (reg:PSI
Charles Baylis charles.bay...@linaro.org writes:
On 19 May 2014 19:07, Richard Sandiford rdsandif...@googlemail.com wrote:
Sorry for the breakage. I wanted to make the script as picky as I could
get away with though, so that results aren't lost accidentally.
Could you try the attached?
From: Richard Biener [mailto:richard.guent...@gmail.com]
It may do three aligned loads, char, short, char and combine them
while doing an unaligned int load may end up being slower. Though
very probable the RTL expansion machinery for unaligned loads
is way more clever to emit an optimal
From: Marcus Shawcroft [mailto:marcus.shawcr...@gmail.com]
Sent: 19 May 2014 11:45
To: Ian Bolton
Cc: gcc-patches
Subject: Re: [PATCH, AArch64] Fix macro in vdup_lane_2 test case
On 8 May 2014 18:41, Ian Bolton ian.bol...@arm.com wrote:
gcc/testsuite
*
Can't I have sometihng like
loop (i as iterator)
begin
I := F(0);
exception
when E = N := N + 1;
end;
here I would expect PRE to discover I as loop invariant that will cause
exception to be called just once, not at every iteration of loop?
Yet the following
On Tue, May 20, 2014 at 6:29 AM, Jan Hubicka hubi...@ucw.cz wrote:
On 05/18/2014 08:45 PM, Sandra Loosemore wrote:
On 05/18/2014 02:59 PM, Jan Hubicka wrote:
For cases like local-statics-7 your approach can be saved by adding
simple IPA analysis
to look for static vars that are used only by
The following is my current idea on progressing on the HOST_WIDE_INT
removal
1) https://gcc.gnu.org/ml/gcc-patches/2014-05/msg00381.html (ping)
2) make sure [u]int64_t is available and use that to define HOST_WIDE_INT
3) s/HOST_WIDE_INT/int64_t/ (same for unsigned HOST_WIDE_INT)
Does
DOM bits are good. Richi wants the redundant test to be kept around,
so OK with that change.
Thanks. The test is more duplicated than redundant so I'll remove it anyway.
--
Eric Botcazou
On Tue, 20 May 2014, Eric Botcazou wrote:
The following is my current idea on progressing on the HOST_WIDE_INT
removal
1) https://gcc.gnu.org/ml/gcc-patches/2014-05/msg00381.html (ping)
2) make sure [u]int64_t is available and use that to define HOST_WIDE_INT
3)
Same as for going C++.
Not to the same extent, this will be worse because done en masse throughout
the code instead of gradually.
Make the code base easier to understand for newcomers. It's also a
documentation improvement (you see what a HOST_WIDE_INT really is),
alongside with [u]int64_t
On Tue, 20 May 2014, Eric Botcazou wrote:
Same as for going C++.
Not to the same extent, this will be worse because done en masse throughout
the code instead of gradually.
Like the gimple - gimple * change pending or the various
gimple - gswitch,glabel,etc. stuff? It's on a similar scale
On Mon, May 19, 2014 at 5:24 PM, Xinliang David Li davi...@google.com wrote:
Sorry about it. Here is the patch. There is one remaining case where
cgraph_dump_file and dump_enable_p are checked separately --
cgraph_dump_file is set up differently from 'dump_file'.
But there you check with an
Hello,
On 19 May 09:58, H.J. Lu wrote:
On Mon, May 19, 2014 at 9:45 AM, Uros Bizjak ubiz...@gmail.com wrote:
On Mon, May 19, 2014 at 6:42 PM, H.J. Lu hjl.to...@gmail.com wrote:
Uros,
I am looking into libreoffice size and the data alignment seems to make
huge
difference. Data section
We can't really code-generate those (and thus such patterns are
simply skipped during code-gen ...). Reject them early.
Committed to the branch.
Richard.
2014-05-20 Richard Biener rguent...@suse.de
* genmatch.c (parse_match_and_simplify): Reject outermost
expressions that
On 16/05/14 14:56, Alexey Merzlyakov wrote:
On 07.05.2014 13:28, Ramana Radhakrishnan wrote:
On 05/07/14 09:19, Yury Gribov wrote:
Original Message
Subject: [PING] [PATCH] Fix for PR libstdc++/60758
Date: Thu, 17 Apr 2014 17:48:12 +0400
From: Alexey Merzlyakov
On 20/05/14 08:57, Richard Biener wrote:
On Tue, 20 May 2014, Maciej W. Rozycki wrote:
Ian,
On Sat, 17 May 2014, Richard Biener wrote:
On May 17, 2014 12:22:23 AM CEST, Maciej W. Rozycki
ma...@codesourcery.com wrote:
On Fri, 16 May 2014, Joseph S. Myers wrote:
2014-05-16 Maciej W.
OK.
Jason
Rainer Orth r...@cebitec.uni-bielefeld.de writes:
Prompted by the recent failures of c-c++-common/gomp/pr60823-2.c on
Solaris/x86 with Sun as
http://gcc.gnu.org/ml/gcc-patches/2014-05/msg00943.html
I've reworked the clearing of hardware capabilities via linker maps for
Sun ld, which
This adds $subject.
Committed to branch.
Richard.
2014-05-20 Richard Biener rguent...@suse.de
* genmatch.c (error_cb, fatal_at): New functions.
(expect, parse_expr, parse_op, parse_match_and_simplify,
main): Provide error locations.
Index: gcc/genmatch.c
On 15/05/14 09:47, Kyrill Tkachov wrote:
Hi all,
Shifted arithmetic operations can never be encoded in 16-bits in and
therefore
can not appear in Thumb2 IT blocks under ARMv8-A rules (and the -mrestrict-it
rules). This patch adjusts the relevant pattern for that purpose.
Tested and
On 20.05.2014 16:25, Richard Earnshaw wrote:
On 16/05/14 14:56, Alexey Merzlyakov wrote:
On 07.05.2014 13:28, Ramana Radhakrishnan wrote:
On 05/07/14 09:19, Yury Gribov wrote:
Original Message
Subject: [PING] [PATCH] Fix for PR libstdc++/60758
Date: Thu, 17 Apr 2014
The following PR is opened for this problem:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61223
Thumb1 failure was also detected and reported in pr60758.
I've proposed a thumb1 bugfix there. Regtest for the fix currently is in
progress.
Patches must be proposed on gcc-patches and / or in
PR 61210 is about a case where a templated wi:: function had x | y
and where both x and y needed temporary variables. The recursive
instantation of x and y were done as two arguments to the same function
call, which meant that the order of instantiation depended on the host
compiler. The uids of
On 20/05/14 14:12, Ramana Radhakrishnan wrote:
The following PR is opened for this problem:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61223
Thumb1 failure was also detected and reported in pr60758.
I've proposed a thumb1 bugfix there. Regtest for the fix currently is in
progress.
Hi,
On 05/13/2014 02:19 AM, Paolo Carlini wrote:
Hi,
in this issue Marc noticed that the warning for ignored attribute
visibility is truncated when the previous declaration belongs to a
system header. Luckily the problem can be neatly fixed because it's
just a consequence of not using
On Mon, May 19, 2014 at 11:51 PM, Xinliang David Li davi...@google.com wrote:
Why duplicating the merger functions in dyn-ipa.c? Should those in
libgcov-merge.c be reused?
The merger functions in libgcov-merge.c use a macro to either read the
counter to merge from a buffer in memory (when
This makes the dumps easier to follow.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied.
Richard.
2014-05-20 Richard Biener rguent...@suse.de
* tree-ssa-sccvn.c (process_scc): Dump SCC here, when
iterating,
(extract_and_process_scc_for_name): not here.
OK.
Jason
OK.
Jason
Please look then on the following patch.
Regtested successfully on x86_64.
Is it ok for trunk and 4.9?
gcc/cp/ChangeLog:
2014-05-20 Igor Zamyatin igor.zamya...@intel.com
PR c/60189
* parser.c (cp_parser_postfix_expression): Move handling of cilk_sync
from here to...
(cp_parser_statement):
Hi,all
There are some small problems that block the ira to make the best
choice, as PR61241 shows.
This patch fix PR61241. I have done make bootstrap and make check
on x86, nothing changed after the patch.
Is the patch OK for trunk?
Thanks
Ma Jiang
2014-05-20 Ma Jiang ma.ji...@zte.com.cn
Hi all!
The following patch fixes the ICE for the cilk code with syntax error.
Regtested on x86_64.
Ok for trunk and 4.9?
Thanks,
Igor
gcc/c/ChangeLog:
2014-05-20 Igor Zamyatin igor.zamya...@intel.com
* c-array-notation.c (fix_builtin_array_notation_fn): Check invalid
function
On 05/19/2014 05:37 PM, James Greenhalgh wrote:
On Fri, May 16, 2014 at 06:49:45PM +0100, Vladimir Makarov wrote:
The following patch fixes
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60969
The patch was bootstrapped and tested on x86/x86-64.
Committed as rev. 210519 to gcc 4.9 branch
On May 20, 2014, at 1:17 AM, Richard Sandiford rdsandif...@googlemail.com
wrote:
The patch gives a consistent compile-time improvement of about ~3.5%
on the -O0 fold-const.ii test case.
3.5 alone is really nice and 3.5 on top of 3.5 is amazing.
On 05/20/2014 02:55 AM, Mark Wielaard wrote:
On Mon, May 19, 2014 at 04:50:35PM -0400, Jason Merrill wrote:
On 05/13/2014 03:21 AM, Mark Wielaard wrote:
So the debugger doesn't have to guess the properties of the enum's
underlying base type, like size, encoding and signedness.
Well, the enum
Hi,
On 05/19/2014 08:28 PM, Jason Merrill wrote:
How about doing digest_init in get_nsdmi, so that the conversion is
also exposed to walk_field_subobs?
Thus, good news: something as simple as the below passes testing, works
for the 2 bugs and for c++/58704 too. Thus, what else? Personally, I'm
The first part of this patch fixes the PR, we should be adding
target-libstdc++-v3 to noconfigdirs, not libstdc++-v3.
The second part disables the target libs that can't be bootstrapped
without libstdc++.
libcilkrts and libsanitizer use lots of std:: stuff.
libitm only uses std::atomic, but in
On Tue, May 20, 2014 at 03:53:14PM +0100, Jonathan Wakely wrote:
The first part of this patch fixes the PR, we should be adding
target-libstdc++-v3 to noconfigdirs, not libstdc++-v3.
The second part disables the target libs that can't be bootstrapped
without libstdc++.
libcilkrts and
On 20/05/14 15:53 +0100, Jonathan Wakely wrote:
The first part of this patch fixes the PR, we should be adding
target-libstdc++-v3 to noconfigdirs, not libstdc++-v3.
The second part disables the target libs that can't be bootstrapped
without libstdc++.
libcilkrts and libsanitizer use lots of
On 05/20/2014 10:50 AM, Paolo Carlini wrote:
+ if (TREE_TYPE (init) != type)
+ {
+ if (BRACE_ENCLOSED_INITIALIZER_P (init)
+ CP_AGGREGATE_TYPE_P (type))
+ init = reshape_init (type, init, tf_warning_or_error);
+ init = digest_init (type, init,
On Tue, 20 May 2014, Eric Botcazou wrote:
Make the code base easier to understand for newcomers. It's also a
documentation improvement (you see what a HOST_WIDE_INT really is),
alongside with [u]int64_t being less to type ...
I personally find the abstraction and the separation with the
On Tue, May 20, 2014 at 5:00 AM, Kirill Yukhin kirill.yuk...@gmail.com wrote:
Hello,
On 19 May 09:58, H.J. Lu wrote:
On Mon, May 19, 2014 at 9:45 AM, Uros Bizjak ubiz...@gmail.com wrote:
On Mon, May 19, 2014 at 6:42 PM, H.J. Lu hjl.to...@gmail.com wrote:
Uros,
I am looking into
On May 20, 2014, at 6:03 AM, Rainer Orth r...@cebitec.uni-bielefeld.de wrote:
The following patch implements what I've outlined there
Ok for mainline and 4.9 branch if those pass?
For the test suite bits, Ok.
On Tue, May 20, 2014 at 7:31 AM, Zamyatin, Igor igor.zamya...@intel.com wrote:
Hi all!
The following patch fixes the ICE for the cilk code with syntax error.
Regtested on x86_64.
Ok for trunk and 4.9?
Thanks,
Igor
gcc/c/ChangeLog:
2014-05-20 Igor Zamyatin igor.zamya...@intel.com
PR 61208 is a wrong code bug in Thumb2 where we can generate out of
range branches due to incorrect instruction size calculations. It's
mostly gone latent on 4.9 and trunk (though could still happen at -O0
where splitting is done during final instruction generation).
Complicating things slightly
On Tue, May 20, 2014 at 6:32 AM, Teresa Johnson tejohn...@google.com wrote:
On Mon, May 19, 2014 at 11:51 PM, Xinliang David Li davi...@google.com
wrote:
Why duplicating the merger functions in dyn-ipa.c? Should those in
libgcov-merge.c be reused?
The merger functions in libgcov-merge.c use
Hi all,
This is a fix for thumb1 build fail on trunk appeared since
rev.210515(pr60758).
On thumb1 targets the LR can not be used as argument of POP instruction.
To keep a support of __cxa_end_cleanup backtracing on thumb1
we can make additional register operations before push/pop.
But I
On 20.05.2014 17:16, Richard Earnshaw wrote:
On 20/05/14 14:12, Ramana Radhakrishnan wrote:
The following PR is opened for this problem:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61223
Thumb1 failure was also detected and reported in pr60758.
I've proposed a thumb1 bugfix there. Regtest
Hi,
On 05/20/2014 05:08 PM, Jason Merrill wrote:
On 05/20/2014 10:50 AM, Paolo Carlini wrote:
+ if (TREE_TYPE (init) != type)
+{
+ if (BRACE_ENCLOSED_INITIALIZER_P (init)
+ CP_AGGREGATE_TYPE_P (type))
+init = reshape_init (type, init, tf_warning_or_error);
+
On 05/20/14 17:05, Alexey Merzlyakov wrote:
Hi all,
This is a fix for thumb1 build fail on trunk appeared since
rev.210515(pr60758).
On thumb1 targets the LR can not be used as argument of POP instruction.
To keep a support of __cxa_end_cleanup backtracing on thumb1
we can make additional
Hi,
libgo currently does not build because of duplicated symbols. This is because of
confusion in the output machinery, where comdat local is output as a global
symbol. This is because of confusing use of DECL_ONE_ONLY. This bug seems to be
there since introduction of comdat locals in gcc 4.9,
On 05/20/14 09:27, H.J. Lu wrote:
On Tue, May 20, 2014 at 7:31 AM, Zamyatin, Igor igor.zamya...@intel.com wrote:
Hi all!
The following patch fixes the ICE for the cilk code with syntax error.
Regtested on x86_64.
Ok for trunk and 4.9?
Thanks,
Igor
gcc/c/ChangeLog:
2014-05-20 Igor
The test in get_call_reg_set_usage for flag_use_caller_save and the hook is
strictly speaking not necessary. But it's the interface function to retrieve
the collected register usage information, so it seems a good location to do
an early-out. I've left it in for now.
But the test for
On 05/20/14 08:10, Zamyatin, Igor wrote:
Please look then on the following patch.
Regtested successfully on x86_64.
Is it ok for trunk and 4.9?
gcc/cp/ChangeLog:
2014-05-20 Igor Zamyatin igor.zamya...@intel.com
PR c/60189
* parser.c (cp_parser_postfix_expression): Move handling of
On 05/20/14 02:06, Richard Biener wrote:
We're still going to have problems if SSA names are re-used
(as even released SSA names have to remain valid lattice
entries after this). But currently nothing will create new
SSA names (fingers crossing) during eliminate. A followup
patch of mine will
This is an updated version of the patch to replace void_zero_node with
void_node. The main controversial point/hack in the original was that
it handled void_node in the gimplifier as a zero constant. Jason sent
me a patch that traps the dummy object in the front end, meaning that
it no longer
On 05/20/14 01:13, Bin.Cheng wrote:
The idea being that common cases where a pair moves can be turned into a
single wider move without having to write target code to make that happen
much of the time. ie 2xQI-HI, 2xHI-SI, 2xSI-DI 2xSF-DF. For things
outside those simple cases, fall back to a
Hi Guys,
I am checking in the patch below (to both the mainline and the 4.9
branch) in order to fix some unexpected failures in the gcc testsuite
with the msp430-elf toolchain. With this patch in place the following
tests now pass:
gcc.c-torture/execute/builtins/memcpy-chk.c
On 05/20/14 01:11, Zhenqiang Chen wrote:
Hi,
The patch fix ICE issue triggered by shrink-wrapping enhancement.
Bootstrap and no make check regression on X86-64.
OK for trunk?
Thanks!
-Zhenqiang
2014-05-20 Zhenqiang Chen zhenqiang.c...@linaro.org
PR rtl-optimization/61220
On Tue, May 20, 2014 at 12:13 AM, Bin.Cheng amker.ch...@gmail.com wrote:
On Tue, May 20, 2014 at 1:30 AM, Jeff Law l...@redhat.com wrote:
On 05/19/14 00:38, Bin.Cheng wrote:
On Sat, May 17, 2014 at 12:32 AM, Jeff Law l...@redhat.com wrote:
On 05/16/14 04:07, Bin.Cheng wrote:
But can't
On 05/20/14 11:14, Wei Mi wrote:
On Tue, May 20, 2014 at 12:13 AM, Bin.Cheng amker.ch...@gmail.com wrote:
On Tue, May 20, 2014 at 1:30 AM, Jeff Law l...@redhat.com wrote:
On 05/19/14 00:38, Bin.Cheng wrote:
On Sat, May 17, 2014 at 12:32 AM, Jeff Law l...@redhat.com wrote:
On 05/16/14
For now, please revert your original patch
Alex asked me to revert patch for him. Done in r210650.
-Y
On 05/19/14 22:27, Maxim Kuvyrkov wrote:
Hi,
This patch cleans up haifa-sched.c:choose_ready() function while
allow more powerful customization by backends at the same time.
The primary change is that
targetm.sched.first_cycle_multipass_dfa_lookahead_guard hook is
converted from returning a
OK. I've committed my patch.
Jason
Hi James
Thank you for pointing this out. In the new patch I removed the
modification of vqdmulh_n_s32 and vqdmulhq_n_s32.
Passed dejagnu testing on aarch64 qemu again. OK for trunk, 4.9 and 4.8?
2014-05-20 Guozhi Wei car...@google.com
* config/aarch64/arm_neon.h (vqdmulh_n_s16):
OK.
Jason
On 05/20/14 02:16, Richard Sandiford wrote:
get_attr_enabled was showing up high in a -O0 compile of fold-const.ii.
At the moment, extract_insn calls this function for every alternative
on each extraction, which can be expensive for instructions like
moves that have many alternatives.
The
On Tue, May 20, 2014 at 10:43:22AM -0400, Jason Merrill wrote:
On 05/20/2014 02:55 AM, Mark Wielaard wrote:
On Mon, May 19, 2014 at 04:50:35PM -0400, Jason Merrill wrote:
On 05/13/2014 03:21 AM, Mark Wielaard wrote:
So the debugger doesn't have to guess the properties of the enum's
underlying
On Tue, May 20, 2014 at 8:39 AM, Xinliang David Li davi...@google.com wrote:
On Tue, May 20, 2014 at 6:32 AM, Teresa Johnson tejohn...@google.com wrote:
On Mon, May 19, 2014 at 11:51 PM, Xinliang David Li davi...@google.com
wrote:
Why duplicating the merger functions in dyn-ipa.c? Should
The current code converts every tree to signed hwi; this ICEs with
values not representable as shwi, like 999ULL in
pr34154.c (and if it didn't ICE, it would print the wrong value).
This fixes it. Bootstrapped and tested on powerpc64-linux. Okay
to apply?
Segher
2014-05-20
On 05/20/14 02:19, Richard Sandiford wrote:
Following on from (and depending on) the last patch, process_constraints
also shows up high in the profile. This patch caches the recog_op_alt
information by insn code too. It also shrinks the size of the structure
from 1 pointer + 5 ints to 1
On May 20, 2014 6:47:44 PM CEST, Jeff Law l...@redhat.com wrote:
On 05/20/14 02:06, Richard Biener wrote:
We're still going to have problems if SSA names are re-used
(as even released SSA names have to remain valid lattice
entries after this). But currently nothing will create new
SSA names
On 05/20/14 12:13, Richard Biener wrote:
I think it's still important for things like keeping SSA names
densely allocated and for reuse on SSA rewrite of virtual operands
for example. So I'm thinking of an explicit stop-/restart-reuse
interface.
Fair enough. WRT dense allocation, we could
The following patch fixes a semi-latent bug for the HTM pattern used with
the __builtin_ttest() builtin. This is supposed to expand to a tabortwci.
instruction which sets cr0 and then some code that copies the cr0 value
into a gpr and then shifts and masks it into the lowest 2 bits in the gpr.
On 19/05/14 22:27 +0200, François Dumont wrote:
On 15/05/2014 22:52, Jonathan Wakely wrote:
Does this get initialized in the constructors?
Would it make sense to give it an initializer?
__bucket_type_M_single_bucket = nullptr;
This bucket is replacing those normally allocated
On 05/20/2014 02:11 AM, Jonathan Wakely wrote:
On 19/05/14 14:57 -0600, Sandra Loosemore wrote:
On 05/17/2014 04:07 AM, Jonathan Wakely wrote:
On 17 May 2014 10:50, Jonathan Wakely wrote:
On 17 May 2014 01:16, Sandra Loosemore wrote:
It appears that this patch from last fall never got
On 05/20/2014 01:51 PM, Mark Wielaard wrote:
The DWARF part isn't what this patch is blocked on. That has already
been discussed on the DWARF standard list, coordinated with the gdb
hackers and approved some months ago.
Fair enough.
The part that hasn't been reviewed
and approved yet is the
Hi,
as disucssed some time ago, our assumption that every symbol of shared library
can
be interposed at runtime is expensive and prevents a lot of useful
optimizations,
including inlining or IPA propagation.
While this is useful feature, it is rather incommon to use it for bigger C++
projects,
1 - 100 of 128 matches
Mail list logo
