This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: The new scheduler and x86 CPUs
- To: bernds at redhat dot com, dewar at gnat dot com
- Subject: Re: The new scheduler and x86 CPUs
- From: dewar at gnat dot com
- Date: Wed, 29 Aug 2001 06:31:24 -0400 (EDT)
- Cc: gcc at gcc dot gnu dot org, jh at suse dot cz, vmakarov at toke dot toronto dot redhat dot com
<<There are several cases where less extensive analysis is needed at run-time
(e.g. to determine memory aliases), or where certain aspects of the program
are only available at run-time, since they are variable (e.g. load
latencies).
>>
Well of course!
But "several cases" is a far cry from your original unqualified statement
that an OOO core can always find the parallelism just as well! Your argument
in favor of superscalar OOO architectures is strong without needing obviously
excessive exaggeration :-)
<<Don't forget that if your compiler can do something very clever, it can do
so for both EPIC and OOO>
>>
Again, you are making an absolute statement without qualification that is
not by any means always true.