--- Comment #20 from hubicka at gcc dot gnu dot org 2010-09-20 15:48
---
Subject: Bug 45605
Author: hubicka
Date: Mon Sep 20 15:48:42 2010
New Revision: 164438
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=164438
Log:
PR tree-optimize/45605
* cgraph.h
--- Comment #21 from hubicka at gcc dot gnu dot org 2010-09-20 15:53
---
OK, we now fold the testcase using obj_type_ref folding. We still should do it
via vtable lookup and we don't but that is for other PR I guess.
--
hubicka at gcc dot gnu dot org changed:
What
--- Comment #19 from hubicka at gcc dot gnu dot org 2010-09-18 21:25
---
Subject: Bug 45605
Author: hubicka
Date: Sat Sep 18 21:25:29 2010
New Revision: 164402
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=164402
Log:
PR tree-optimization/45605
* cgraphunit.c
--- Comment #18 from jamborm at gcc dot gnu dot org 2010-09-17 14:46
---
Honza submitted a patch
(http://gcc.gnu.org/ml/gcc-patches/2010-09/msg01369.html) so I guess it is his
PR now :-)
--
jamborm at gcc dot gnu dot org changed:
What|Removed
--- Comment #10 from rguenther at suse dot de 2010-09-16 08:50 ---
Subject: Re: Missed devirtualization
On Wed, 15 Sep 2010, hubicka at ucw dot cz wrote:
--- Comment #9 from hubicka at ucw dot cz 2010-09-15 22:39 ---
Subject: Re: Missed devirtualization
We fold a stmt
--- Comment #11 from hubicka at gcc dot gnu dot org 2010-09-16 12:25
---
Hmm, so do you have any idea where folding should be added for this particular
case?
It always seemed to me that it would make sense to add verifier that all
statements are folded (locally, not by looking at SSA
--- Comment #12 from rguenther at suse dot de 2010-09-16 12:31 ---
Subject: Re: Missed devirtualization
On Thu, 16 Sep 2010, hubicka at gcc dot gnu dot org wrote:
--- Comment #11 from hubicka at gcc dot gnu dot org 2010-09-16 12:25
---
Hmm, so do you have any idea where
--- Comment #13 from hubicka at ucw dot cz 2010-09-16 12:48 ---
Subject: Re: Missed devirtualization
I'm lost in this PR - for what testcase what statement needs folding
(and what pending patches do I need to apply to see that)?
PR is tracking missed optimization in the testcase in
--- Comment #14 from rguenther at suse dot de 2010-09-16 12:51 ---
Subject: Re: Missed devirtualization
On Thu, 16 Sep 2010, hubicka at ucw dot cz wrote:
--- Comment #13 from hubicka at ucw dot cz 2010-09-16 12:48 ---
Subject: Re: Missed devirtualization
I'm lost in
--- Comment #15 from rguenth at gcc dot gnu dot org 2010-09-16 14:55
---
Like
Index: gimplify.c
===
--- gimplify.c (revision 164333)
+++ gimplify.c (working copy)
@@ -2477,10 +2477,13 @@ gimplify_call_expr (tree
--- Comment #16 from jamborm at gcc dot gnu dot org 2010-09-16 16:00
---
(In reply to comment #15)
Like
Index: gimplify.c
===
--- gimplify.c (revision 164333)
+++ gimplify.c (working copy)
@@ -2477,10 +2477,13 @@
--- Comment #17 from rguenther at suse dot de 2010-09-16 16:06 ---
Subject: Re: Missed devirtualization
On Thu, 16 Sep 2010, jamborm at gcc dot gnu dot org wrote:
--- Comment #16 from jamborm at gcc dot gnu dot org 2010-09-16 16:00
---
(In reply to comment #15)
--- Comment #7 from jamborm at gcc dot gnu dot org 2010-09-15 18:42 ---
Well, it turns out that fold_stmt_1 is never called on that statement
(neither with -O -finline or -O2 or -O3). Where is it supposed to be
called from?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45605
--- Comment #8 from rguenther at suse dot de 2010-09-15 19:09 ---
Subject: Re: Missed devirtualization
On Wed, 15 Sep 2010, jamborm at gcc dot gnu dot org wrote:
--- Comment #7 from jamborm at gcc dot gnu dot org 2010-09-15 18:42
---
Well, it turns out that fold_stmt_1
--- Comment #9 from hubicka at ucw dot cz 2010-09-15 22:39 ---
Subject: Re: Missed devirtualization
We fold a stmt only if it is propagated to (by ccp, copyprop, forwprop,
dom or by inlining).
Well, since fold_stmt is stornger than what fe does, I guess we should fold
each stmt at
--- Comment #6 from hubicka at gcc dot gnu dot org 2010-09-10 09:42 ---
Subject: Bug 45605
Author: hubicka
Date: Fri Sep 10 09:42:20 2010
New Revision: 164148
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=164148
Log:
PR tree-optimization/45605
* cp/class.c
--- Comment #2 from rguenther at suse dot de 2010-09-09 08:43 ---
Subject: Re: New: Missed devirtualization
On Thu, 9 Sep 2010, hubicka at gcc dot gnu dot org wrote:
The following testcase taken from testsuite (and many others):
// PR 14535
// { dg-do run }
// { dg-options -O
--- Comment #3 from hubicka at gcc dot gnu dot org 2010-09-09 11:02 ---
Hmm, is it?
C equivalent IMO is:
int a(void);
typedef void (*ptr) (void);
static const ptr array[1]={(ptr)a};
ptr t;
main()
{
t=array[0];
}
Here we have ctor represented as follows:
nop_expr 0x76b53f30
--- Comment #4 from hubicka at gcc dot gnu dot org 2010-09-09 11:33 ---
testing the following fix:
Index: class.c
===
--- class.c (revision 163947)
+++ class.c (working copy)
@@ -7797,7 +7797,7 @@
--- Comment #5 from hubicka at gcc dot gnu dot org 2010-09-09 11:36 ---
Created an attachment (id=21750)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=21750action=view)
WIP patch. Still misses some of Richi's earlier comments.
--
--- Comment #1 from pinskia at gcc dot gnu dot org 2010-09-09 01:17 ---
I think this is the same issue as PR 19816.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45605
21 matches
Mail list logo
