Date: Mon, 6 May 1996 09:01:39 -0700 (PDT) From: "Rodney W. Grimes" <rgrimes@GndRsh.aac.dev.com> To: bde@zeta.org.au (Bruce Evans) Cc: asami@cs.berkeley.edu, current@FreeBSD.org, nisha@cs.berkeley.edu Subject: Re: more on fast bcopy Message-ID: <199605061601.JAA00445@GndRsh.aac.dev.com> In-Reply-To: <199605061207.WAA04793@godzilla.zeta.org.au> from Bruce Evans at "May 6, 96 10:07:13 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
> >The FP thing is still by far the fastest for large copies on all the > >Pentiums we've tried over here, but we can't use that in the kernel. > > Why not? :-) It should be possible to use the fpu after saving and > restoring the FP registers reentrantly. ... > > I couldn't get pairs of uncached writes to work at all. Even `pushal' > apparently does 8 separate writes. This is on an ASUS P55TP4XE. Is > this an ASUS or Triton limitation? Pentium limitation would be my best guess, you may be hitting an AGI that is preventing the writes to occur togeather. pushal probably has an AGI on the posted write buffer. It seems other than cache line write backes it is next to impossible to get the Pentium processor to issue 64 bit write operations. :-( -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199605061601.JAA00445>