--
Bug 21884 depends on bug 21155, which changed state.
Bug 21155 Summary: [4.1 Regression] ICE in ssa tree check compiling crafty with
-ftree-vectorize -maltivec
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21155
What|Old Value |New Value
-
--- Additional Comments From reichelt at gcc dot gnu dot org 2005-06-11
14:19 ---
All three testcases (from comment #1, #4, and #7)
got fixed by Keith's patch.
--
What|Removed |Added
---
--- Additional Comments From dorit at il dot ibm dot com 2005-06-09 08:47
---
this patch: http://gcc.gnu.org/ml/gcc-patches/2005-06/msg00850.html seems to
fix this ICE.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21884
--- Additional Comments From dorit at il dot ibm dot com 2005-06-07 09:58
---
> I tried on powerpc-apple-darwin and on i686-pc-linux-gnu - couldn't reproduce
> the failure
correction - can't reproduce the failure on powerpc,
but I do see the ICE on i686-pc-linux-gnu with both testcases
--- Additional Comments From stefaan dot deroeck at gmail dot com
2005-06-07 07:54 ---
Another testcase possibly, that triggers the fault on 20050605 with parameters
-march=pentium4 -O3 -ftree-vectorize, but not with -O2 or -O1. Bug 21933 seems
related but doesn't trigger with -O3 or -
--- Additional Comments From dorit at il dot ibm dot com 2005-06-07 07:35
---
> I tried compiling x.ii on power-pc and couldn't reproduce the problem.
> I don't know if this means that the problem goes away with recent
> patches or that it only occurs when vectorizing on pentium. I curr
--- Additional Comments From kbesaw at us dot ibm dot com 2005-06-06 23:41
---
I tried compiling x.ii on power-pc and couldn't reproduce the problem.
I don't know if this means that the problem goes away with recent
patches or that it only occurs when vectorizing on pentium. I currently
--
What|Removed |Added
CC||stefaan dot deroeck at gmail
||dot com
http://gcc.gnu.org/bugzi
--- Additional Comments From bangerth at dealii dot org 2005-06-02 20:17
---
Created an attachment (id=9018)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=9018&action=view)
Much shorter testcase
I attach another testcase that I managed to get down to approximately 10kloc.
It is on
--- Additional Comments From bangerth at dealii dot org 2005-06-02 19:36
---
I don't know -- the failure appears in a completely different place. I see
this particular ICE on at least half a dozen of my files, but I'll be happy
holding off finding a reduced testcase until the other PR
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-06-02
19:29 ---
This is most likely related to PR 21155 which is another bug with the same ICE.
--
What|Removed |Added
---
--- Additional Comments From bangerth at dealii dot org 2005-06-02 19:24
---
Created an attachment (id=9017)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=9017&action=view)
Preprocessed source
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21884
12 matches
Mail list logo
