Date: Sat, 25 Sep 1999 22:24:29 +0200 (CEST) From: Oliver Fromme <olli@dorifer.heim3.tu-clausthal.de> To: freebsd-current@FreeBSD.ORG Subject: Re: gcc optimizer in -current system ... Message-ID: <199909252024.WAA22407@dorifer.heim3.tu-clausthal.de>
next in thread | raw e-mail | index | archive | help
Just some numbers that I got with a small testsuite. This
is the setiathome client running on an AMD Athlon-500 (K7),
FreeBSD 4.0-current (as of 1999-09-18). Of course, I've
used the same work unit for all tests. I also changed the
code to stop after a certain amount of data has processed,
because I don't want to wait some hours for each test to
complete. ;-)
-O1 302.72 real 301.96 user 0.28 sys
-O2 315.20 real 314.42 user 0.18 sys
-O3 320.01 real 319.23 user 0.35 sys
-Os 315.99 real 315.19 user 0.35 sys
As you can see, -O1 is fastest, the higher optimization levels
are slower. I'm currently running a testsuite with a larger
number of optimizer flags combinations, but it'll take a while.
On a side note, I downloaded AMD's document about optimizing
code for the Athlon, and I found out that gcc does a pretty
poor job. :-( Unfortunately, I'm not an expert in writing
compilers, and I don't dare to touch the gcc source code.
After all, the compiler is one of the most critical parts of
the system.
Regards
Oliver
--
Oliver Fromme, Leibnizstr. 18/61, 38678 Clausthal, Germany
(Info: finger userinfo:olli@dorifer.heim3.tu-clausthal.de)
"In jedem Stück Kohle wartet ein Diamant auf seine Geburt"
(Terry Pratchett)
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199909252024.WAA22407>