Date: Sun, 12 May 1996 00:50:48 -0400 (EDT) From: "matthew c. mead" <mmead@Glock.COM> To: phk@critter.tfs.com (Poul-Henning Kamp) Cc: hackers@freebsd.org Subject: Re: 6x86 120+ Message-ID: <199605120450.AAA00407@neon.Glock.COM> In-Reply-To: <5891.831836840@critter.tfs.com> from "Poul-Henning Kamp" at May 11, 96 05:47:20 pm
next in thread | previous in thread | raw e-mail | index | archive | help
Poul-Henning Kamp writes: > > I just got one of these the other day and installed it. I've > > read before that they accept the full pentium instruction set. > > For all but a few things, it's a fair amount faster than my > > P90(@100). I did notice that these applications that seem slower > > are built with pgcc pentium optimization. Does anyone know if I > > need to build with -m486 to get better performance out of this > > chip, or should I stick with -mpentium? Thanks! > >From what I have heard so far you should only use -mpentium for the P5. > Most of the newer chips run slower with the pentium opts. because the > attempted dual parallel sequences of instructions defeat the lookahead > of the logic in charge of the "RISCification". Oops. I totally missed this in my skim earlier - the way you positioned it, I ended up looking only at the following. :-) I'll try building things without pgcc. > Can we have a number for the FreeBSD-stones ? > cd /usr/src > rm -rf /usr/obj/* > time make world Just about to sup things up to tonight's current and start. However, I realized that I need a little more information to give you a fair number. Mount filesystem with -o async? Use CC="cc -pipe"? I have all of /usr/src and /usr/obj going to a JAZ drive disk, which is slower than my system disks. Any instructions you wanna give me to make the number fairer will help. -matt -- Matthew C. Mead mmead@Glock.COM http://www.Glock.COM/~mmead/
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199605120450.AAA00407>