http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
--- Comment #10 from Mikael Pettersson 2012-09-23
22:12:07 UTC ---
Although -fno-tree-fre works for the test case in #c3, adjusting it to use
__builtin_memcpy() for the assignment in f() results in wrong code even with
-fno-tree-fre.
Th
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
--- Comment #9 from Frzderzk Dziarmagowski
2012-09-23 19:52:29 UTC ---
Confirmed, jackd compiled with -O1 -fno-tree-fre works as expected
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
--- Comment #8 from Mikael Pettersson 2012-09-23
18:44:53 UTC ---
(In reply to comment #7)
> It may be yet another SRA problem.
Perhaps not, -fno-tree-sra makes no difference. Looking through the tree dumps
I see the wrong code first a
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
Mikael Pettersson changed:
What|Removed |Added
CC||rguenth at gcc dot gnu.org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
Mikael Pettersson changed:
What|Removed |Added
CC||mikpe at it dot uu.se
--- C
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
Frzderzk Dziarmagowski changed:
What|Removed |Added
CC||fdziarmagowski at gmail d
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
--- Comment #4 from brendan.jones.it at gmail dot com 2012-06-24 07:56:59 UTC
---
Hi,
have you been able to replicate this issue with the second example I have sent?
If so is there a known workaround that we can use?
many thanks
Brendan
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
--- Comment #3 from brendan.jones.it at gmail dot com 2012-06-14 09:00:37 UTC
---
Sorry - wrong file
/* -*- Mode: C ; c-basic-offset: 4 -*- */
/* gcc test.c -Wall -Wextra -o test -O1 && ./test ; echo $?
should print 0, prints 10
gcc known t
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
Richard Guenther changed:
What|Removed |Added
Status|UNCONFIRMED |WAITING
Last reconfirmed|
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53663
Andrew Pinski changed:
What|Removed |Added
Component|c |tree-optimization
--- Comment #1 from And
10 matches
Mail list logo
