http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #32 from Jakub Jelinek jakub at gcc dot gnu.org ---
Author: jakub
Date: Thu Mar 13 09:38:28 2014
New Revision: 208535
URL: http://gcc.gnu.org/viewcvs?rev=208535root=gccview=rev
Log:
PR tree-optimization/59025
PR
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
Jakub Jelinek jakub at gcc dot gnu.org changed:
What|Removed |Added
Status|UNCONFIRMED |RESOLVED
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #27 from Jakub Jelinek jakub at gcc dot gnu.org ---
basic_block bbb = gimple_bb (SSA_NAME_DEF_STMT (oea-op));
in the patch contains a pasto, which I guess can explain most if not all the
-fcompare-debug failures. That said, thinking
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #28 from rguenther at suse dot de rguenther at suse dot de ---
On Tue, 11 Mar 2014, hjl.tools at gmail dot com wrote:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #26 from H.J. Lu hjl.tools at gmail dot com ---
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #29 from Jakub Jelinek jakub at gcc dot gnu.org ---
Created attachment 32338
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=32338action=edit
gcc49-pr60418.patch
New version of the patch, going to bootstrap/regtest it momentarily.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #30 from Jakub Jelinek jakub at gcc dot gnu.org ---
The #c29 patch now passed bootstrap/regtest on x86_64-linux and i686-linux.
So, if it also fixes 435.gromacs on -mx32, I'll submit it to gcc-patches.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #31 from H.J. Lu hjl.tools at gmail dot com ---
(In reply to Jakub Jelinek from comment #29)
Created attachment 32338 [details]
gcc49-pr60418.patch
New version of the patch, going to bootstrap/regtest it momentarily. H.J.,
could
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #23 from rguenther at suse dot de rguenther at suse dot de ---
On Mon, 10 Mar 2014, hjl.tools at gmail dot com wrote:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #20 from H.J. Lu hjl.tools at gmail dot com ---
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #24 from H.J. Lu hjl.tools at gmail dot com ---
(In reply to rguent...@suse.de from comment #23)
On Mon, 10 Mar 2014, hjl.tools at gmail dot com wrote:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #20 from
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #25 from Jakub Jelinek jakub at gcc dot gnu.org ---
Created attachment 32334
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=32334action=edit
gcc49-pr60418.patch
Unfortunately the patch fails to bootstrap due to -fcompare-debug
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #26 from H.J. Lu hjl.tools at gmail dot com ---
This patch avoids removing preheader when optimizing since
preheader will be added back later even if it is removed:
diff --git a/gcc/tree-cfgcleanup.c b/gcc/tree-cfgcleanup.c
index
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #18 from Jakub Jelinek jakub at gcc dot gnu.org ---
Shouldn't we just prefer the original IL if possible? That is not
SSA_NAME_VERSION, but not gimple_uid of the stmt definition either.
If you have:
_4 = something;
_5 =
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #19 from rguenther at suse dot de rguenther at suse dot de ---
On Mon, 10 Mar 2014, jakub at gcc dot gnu.org wrote:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #18 from Jakub Jelinek jakub at gcc dot gnu.org ---
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #20 from H.J. Lu hjl.tools at gmail dot com ---
(In reply to Richard Biener from comment #13)
Huh, adding a pre-header should _never_ do sth like that. Can you produce
a small testcase that exhibits these kind of changes with
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #21 from Jakub Jelinek jakub at gcc dot gnu.org ---
Can you try if sorting on gimple_uid would help this or not? I think it would
be something like:
--- gcc/tree-ssa-reassoc.c.jj2014-02-19 06:59:35.0 +0100
+++
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #22 from H.J. Lu hjl.tools at gmail dot com ---
(In reply to Jakub Jelinek from comment #21)
Can you try if sorting on gimple_uid would help this or not? I think it
would be something like:
Yes, it works.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #17 from rguenther at suse dot de rguenther at suse dot de ---
On 3/7/14 10:57 PM, hjl.tools at gmail dot com wrote:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #14 from H.J. Lu hjl.tools at gmail dot com ---
(In
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #12 from Richard Biener rguenth at gcc dot gnu.org ---
(In reply to H.J. Lu from comment #8)
With LTO, current_loops-state may not be set and
loops_state_satisfies_p () may always be false.
It's zero. It probably should be
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
Richard Biener rguenth at gcc dot gnu.org changed:
What|Removed |Added
CC||rguenth at gcc
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #14 from H.J. Lu hjl.tools at gmail dot com ---
(In reply to Richard Biener from comment #13)
Huh, adding a pre-header should _never_ do sth like that. Can you produce
a small testcase that exhibits these kind of changes with
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
Jakub Jelinek jakub at gcc dot gnu.org changed:
What|Removed |Added
CC||jakub at gcc dot
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #16 from H.J. Lu hjl.tools at gmail dot com ---
This patch:
diff --git a/gcc/tree-ssanames.c b/gcc/tree-ssanames.c
index 2fc8220..56160bd 100644
--- a/gcc/tree-ssanames.c
+++ b/gcc/tree-ssanames.c
@@ -136,7 +136,7 @@ make_ssa_name_fn
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #9 from H.J. Lu hjl.tools at gmail dot com ---
One pass removes a preheader in
if (_47 0)
goto bb 3;
else
goto bb 8;
bb 3:
bb 4:
# n_20 = PHI 1(3), n_269(7)
and becomes:
if (_47 0)
goto bb 3;
else
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #10 from H.J. Lu hjl.tools at gmail dot com ---
Sources have many FP loops contains codes like:
rsq11 = dx11*dx11+dy11*dy11+dz11*dz11;
When they are compiled with
-O3 -funroll-loops -ffast-math -fwhole-program
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #11 from H.J. Lu hjl.tools at gmail dot com ---
sort_by_operand_rank in reassoc.cin has
/* Lastly, make sure the versions that are the same go next to each
other. We use SSA_NAME_VERSION because it's stable. */
if
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
Richard Biener rguenth at gcc dot gnu.org changed:
What|Removed |Added
Target||x32
Target
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #3 from Richard Biener rguenth at gcc dot gnu.org ---
(In reply to H.J. Lu from comment #1)
Do we need to verify that basic block DEST has a single predecessor?
No, that's not necessary. In fact that disables all pre-header
removal
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #4 from Richard Biener rguenth at gcc dot gnu.org ---
(In reply to H.J. Lu from comment #0)
Since r208165, 435.gromacs in SPEC CPU 2006 is miscompiled on x32 with
-mx32 -O3 -funroll-loops -ffast-math -fwhole-program -flto=jobserver
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #5 from H.J. Lu hjl.tools at gmail dot com ---
Created attachment 32280
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=32280action=edit
The assembly difference of good and bad executables
Here is the assembly difference of good
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #6 from H.J. Lu hjl.tools at gmail dot com ---
(In reply to Richard Biener from comment #4)
That's odd. What's the code generation difference in the end
for the minimal range {to, from} that still shows a difference
of the
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #7 from H.J. Lu hjl.tools at gmail dot com ---
current_loops is setup by
(gdb) bt
#0 set_loops_for_fn (fn=0x715901f8, loops=0x7fffef079e20)
at ../../../../gcc/gcc/function.h:747
#1 0x008fe35e in input_cfg
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #8 from H.J. Lu hjl.tools at gmail dot com ---
With LTO, current_loops-state may not be set and
loops_state_satisfies_p () may always be false.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60418
--- Comment #1 from H.J. Lu hjl.tools at gmail dot com ---
Do we need to verify that basic block DEST has a single predecessor?
This patch works for me:
diff --git a/gcc/tree-cfgcleanup.c b/gcc/tree-cfgcleanup.c
index 926d300..0ed15df 100644
---
33 matches
Mail list logo