This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: 2.95, x86: severe performance problems with short arithmetic
- To: Zack Weinberg <zack@bitmover.com>
- Subject: Re: 2.95, x86: severe performance problems with short arithmetic
- From: Richard Henderson <rth@cygnus.com>
- Date: Thu, 12 Aug 1999 14:16:11 -0700
- Cc: law@cygnus.com, John Wehle <john@feith.com>, gcc@gcc.gnu.org
- References: <rth@cygnus.com> <199908122020.NAA24484@zack.bitmover.com>
On Thu, Aug 12, 1999 at 01:20:03PM -0700, Zack Weinberg wrote:
> What we actually want is to do everything in SImode until the results
> become visible outside a function - returning, or write to memory.
Or comparison, or anything else that could detect overflow,
like another addition.
> I thought that inc MEM was discouraged on post-486.
No, it's still very useful.
E.g. on P2, inc mem is a 4 uop insn. Your three insn sequence
also generates 4 uops, but takes up more decoders so we can only
issue 4 uops in that cycle instead of 6.
Plus, inc mem is 3 bytes min whereas your sequence is 7 bytes min.
r~