Date: Tue, 11 Nov 2003 04:31:56 -0000 From: Jeff Roberson <jroberson@chesapeake.net> To: Dag-Erling =?iso-8859-1?q?Sm=F8rgrav?= <des@des.no> Cc: Peter Wemm <peter@wemm.org> Subject: Re: cvs commit: src/sys/i386/i386 pmap.c Message-ID: <20031111041442.31931.qmail@exxodus.fedaykin.here> In-Reply-To: <xzp3cdgb6am.fsf@dwp.des.no>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 26 Oct 2003, Dag-Erling [iso-8859-1] Sm=F8rgrav wrote: > Peter Wemm <peter@wemm.org> writes: > > Massively deep pipelines help get the MHz up, and careful optimization = can > > stop it affecting frame rates. But it blows chunks if you mispredict a > > branch in typical gcc generated code. Or take our libc syscall stubs.. > > every single one will be mispredicted because the usual case (no errors= ) > > has an opposite direction branch to what intel's static branch predicti= on > > expects. > > Is there any way to teach (or trick) gcc to generate a branch which > the p4 will predict correctly? I doubt the effect would be measurable. We could simply rearrange the code. I do not know why it is arranged the way it is now but you would expect that there is some good reason. > > DES > -- > Dag-Erling Sm=F8rgrav - des@des.no >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20031111041442.31931.qmail>