http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
Jakub Jelinek jakub at gcc dot gnu.org changed:
What|Removed |Added
Status|NEW |RESOLVED
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #16 from Pat Haugen pthaugen at gcc dot gnu.org ---
I tried the patch from Comment 15 and was able to build/run the benchmark
successfully.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #17 from Jan Hubicka hubicka at gcc dot gnu.org ---
Author: hubicka
Date: Thu Feb 6 07:39:24 2014
New Revision: 207529
URL: http://gcc.gnu.org/viewcvs?rev=207529root=gccview=rev
Log:
PR middle-end/60013
*
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #12 from Jakub Jelinek jakub at gcc dot gnu.org ---
Slightly more reduced testcase for -O2:
typedef long int jmp_buf[64];
extern int _setjmp (jmp_buf) __attribute__ ((__nothrow__));
struct S { int a, b, c; };
extern struct S *baz
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #13 from Jakub Jelinek jakub at gcc dot gnu.org ---
Ok, here is what happens on the reduced testcase. We have 9 dynamic conditions
and 14 basic blocks. Let's use letters for the dynamic conditions:
A (op1[ref offset: 0] == 0)
B
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #14 from Jakub Jelinek jakub at gcc dot gnu.org ---
Important thing I forgot to mention is the order or predecessors, bb10 has
first bb9 and then bb11, bb11 has first bb4, then bb12, then bb5, bb6, bb8.
The oscillation is there really
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #15 from Jakub Jelinek jakub at gcc dot gnu.org ---
Untested patch to turn things into true predicate when running out of space is:
--- gcc/ipa-inline-analysis.c.jj2014-02-03 08:53:12.0 +0100
+++ gcc/ipa-inline-analysis.c
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
Richard Biener rguenth at gcc dot gnu.org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #9 from H.J. Lu hjl.tools at gmail dot com ---
Created attachment 32014
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=32014action=edit
A patch
I am testing this patch.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
H.J. Lu hjl.tools at gmail dot com changed:
What|Removed |Added
Attachment #32014|0 |1
is
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #11 from H.J. Lu hjl.tools at gmail dot com ---
A patch is posted at
http://gcc.gnu.org/ml/gcc-patches/2014-02/msg00035.html
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #2 from H.J. Lu hjl.tools at gmail dot com ---
Still unfixed with r207387.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
H.J. Lu hjl.tools at gmail dot com changed:
What|Removed |Added
Attachment #32006|0 |1
is
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
Jakub Jelinek jakub at gcc dot gnu.org changed:
What|Removed |Added
CC||hubicka at gcc
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #5 from H.J. Lu hjl.tools at gmail dot com ---
This in compute_bb_predicates
while (!done)
{
done = true;
FOR_EACH_BB_FN (bb, my_function)
{
struct predicate p = false_predicate ();
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #6 from H.J. Lu hjl.tools at gmail dot com ---
Like this:
diff --git a/gcc/ipa-inline-analysis.c b/gcc/ipa-inline-analysis.c
index 9a4c6df..991a10b 100644
--- a/gcc/ipa-inline-analysis.c
+++ b/gcc/ipa-inline-analysis.c
@@ -1881,9
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
H.J. Lu hjl.tools at gmail dot com changed:
What|Removed |Added
Attachment #32011|0 |1
is
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60013
--- Comment #8 from Jakub Jelinek jakub at gcc dot gnu.org ---
That doesn't look correct, that will not clear done on any basic block that has
abnormal successor with ABNORMAL_DISPATCHER, which is a lot of basic blocks.
If the algorithm isn't
18 matches
Mail list logo