Date: Sun, 6 Sep 1998 12:20:36 +0000 From: Niall Smart <rotel@indigo.ie> To: Luigi Rizzo <luigi@labinfo.iet.unipi.it>, mark@grondar.za (Mark Murray) Cc: dag-erli@ifi.uio.no, ormonde@aker.com.br, hackers@FreeBSD.ORG Subject: Re: Assembler with FreeBSD Message-ID: <199809061120.MAA00740@indigo.ie> In-Reply-To: <199809031917.VAA12329@labinfo.iet.unipi.it>; Luigi Rizzo <luigi@labinfo.iet.unipi.it>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sep 3, 9:17pm, Luigi Rizzo wrote: } Subject: Re: Assembler with FreeBSD > > It is very CPU and algorithm-dependent, and not always true. > > I did something similar with an FEC code (see my web page for > more) and while i could almost double performance by recoding the core > loop in assembler _on an old Pentium_, things were completely different > on a PentiumII/PentiumPro, were the C code were actually much faster > than my Pentium-optimized assembler, and the obvious asm optimizations > did not gain anything over the C code. As you probably already know the main optimisation technique in the Pentium is to pair instructions so they simultaneously execute in the U and V pipelines, as for optimising on the PII I haven't yet read Intels docs but PII-optimised code will be significantly different to the P-optimised. Do you know if the same potential for performance gain due to hand-coding exists? For anyone interested: the "multimedia" tutorials available at ftp://download.intel.com/design/perftool/cbts/exes/ are good, but require windows. Niall -- Niall Smart, rotel@indigo.ie. Amaze your friends and annoy your enemies: echo '#define if(x) if (!(x))' >> /usr/include/stdio.h To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199809061120.MAA00740>