This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Scheduling for one processor variant, but reporting times for others
- From: Bradley Lucier <lucier at math dot purdue dot edu>
- To: gcc at gcc dot gnu dot org
- Cc: Bradley Lucier <lucier at math dot purdue dot edu>
- Date: Tue, 16 Sep 2003 14:27:49 -0500
- Subject: Scheduling for one processor variant, but reporting times for others
Would anyone be interested in implementing an option that reports, via
-dR, the cycle timing results for a CPU that the instructions weren't
scheduled for? I.e., I'm playing around with powerpc cpus, and I'd
like to know how a specific instruction sequence that was scheduled for
the G5 actually runs on the G4. It's hard to predict now without that
information---extra execution units on the G5 help, but the longer
pipeline hurts, etc. Perhaps -mtune=970 -dR=7400?
Brad