This is the mail archive of the
mailing list for the GCC project.
Re: ifcvt/crossjump patch: Fix PR 42496, 21803
On 07/28/2010 08:25 PM, Jeff Law wrote:
I'm saying that optimizations run in cfgcleanup.c must play by the
If your "rules" lead to an absurd result, the rules are bogus. Who
decided those "rules" anyway?
I'm not aware of an such rule. I can see the value in placing such rules
on cfgcleanup.c's worker bees which is part of the reason why I
originally suggested this optimization (if implemented as insn movement)
be placed somewhere other than cfgcleanup.
Personally I don't care about how the pass is implemented, I think it
fits more in cfgcleanup.c anyway than in if-conversion (because it
doesn't remove the conditional execution).
There may be another advantage in putting it in cfgcleanup; using flags
to control head-merging may be more suitable than the relatively rigid