This is the mail archive of the
mailing list for the GCC project.
Re: little epilogue optimization
- To: law at cygnus dot com
- Subject: Re: little epilogue optimization
- From: Robert Wilhelm <robert at physiol dot med dot tu-muenchen dot de>
- Date: Thu, 2 Jul 1998 11:02:10 +0200
- Cc: Ulrich Drepper <drepper at cygnus dot com>, egcs-patches at cygnus dot com
- References: <firstname.lastname@example.org> <email@example.com>
On Thu, Jul 02, 1998 at 01:59:26AM -0600, Jeffrey A Law wrote:
> In message <firstname.lastname@example.org>you write:
> > The little patch appended adds an optimization suggested in Intel's
> > docs. Frames of size 4 or 8 should be removed using pop and not add.
> Why? Does it schedule better or something like that? Just curious.
As a special exception, on Pentium two pushes or pops can be paired even they both
modify the stack pointer.
> Is this transformation going to pessimize other x86 processors? If so
> we'll need to conditionalize it.
I don't think so, but it would be better to test with benchmarks.