This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH i386] fix x86_64 bootstrap failure
On Friday, 19. January 2007 12:58, Paolo Bonzini wrote:
> > A simpler patch would be the following. I think it qualifies as obvious,
> > unless somebody complains in the next few hours :)
> Dirk, did you consider running the pass only in the last VRP invocation?
Yes. Unfortunately, the warning *comes* from the second vrp run. Although the
code is apparently the same, vrp1 correctly detects the BB as dead and skips
it. during vrp2 however it can not detect this anymore and produces the
warning.
Unfortunately, running it *only* in the first vrp run causes both quite some
false negatives. Might still be better of course in case I can't manage to
fix the fortran regression soon (currently trying to deduct a testcase).
Dirk