This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] Fix force_single_succ_latches
- From: Andreas Jaeger <aj at suse dot de>
- To: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Sat, 05 Jul 2003 12:39:28 +0200
- Subject: Re: [PATCH] Fix force_single_succ_latches
- References: <20030705102916.GA15252@atrey.karlin.mff.cuni.cz>
Zdenek Dvorak <rakdver@atrey.karlin.mff.cuni.cz> writes:
> Hello,
>
> I have been using the fact that I force loop latches to have exactly one
> successor to conclude that every loop has at least two nodes; this
> argument does not work for loops without exit. The best way to fix it
> seems to enforce that latch differs from header in this case. I am
> commiting the patch as obvious.
>
> It fixes gcc.c-torture/compile/20000728-1.c.
This one?
/usr/src/aj/cvs/gcc-hammer-3_3-branch/gcc/testsuite/gcc.c-torture/compile/20000728-1.c:15: internal compiler error: virtual array dominanc
e info[10]: element 18446744071923247890 out of bounds in set_immediate_dominator, at dominance.c:643
Please submit a full bug report,
then it might fix even more:
gcc.c-torture/compile/20020605-1.c
compile/920729-1.c
compile/930503-1.c
compile/931102-1.c
compile/940611-1.c
compile/940718-1.c
compile/960220-1.c
compile/990617-1.c
compile/990801-1.c
[...]
Please commit also on the hammer-branch.
Thanks,
Andreas
--
Andreas Jaeger, aj@suse.de, http://www.suse.de/~aj
SuSE Linux AG, Deutschherrnstr. 15-19, 90429 Nürnberg, Germany
GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126