Returning to PPC, I wonder why issue_rate was set to 1 for the first
scheduling pass. Heck, maybe it was [at least in part] due to this
very
issue. If anyone having a benchmark PPC box is interested, it may be
educating to remove the setting of issue rate to 1 before reload and
run
some tests with this patch applied.
I will give things a try and let you know.
-Pat
David already explained the reasoning behind it, but following up with the
numbers since I said I would gather them. I ran cpu2006 on a Power6 box
using mainline with your patch applied, when I removed the setting of
issue_rate to 1 for the first scheduling pass I saw an overall degradation
of about 1%, with two benchmarks degrading 10% (dealII and calculix).