Date: Wed, 26 May 1999 11:38:54 +0300 From: Maxim Sobolev <sobomax@altavista.net> To: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> Cc: current@FreeBSD.ORG Subject: Re: More compiler option comparisons Message-ID: <374BB31E.96E21EE5@altavista.net> References: <199905251635.MAA11353@khavrinen.lcs.mit.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
Garrett Wollman wrote: > Just for completeness, I did one final run of HINT with just `-O' > specified (our usual default). `-O' results in significantly better > integer performance than `-O4'. (Floating-point performance is just > the opposite.) > > This suggests that compiling the world with `-O' levels higher than > one is probably a bad idea. (The generated assembly is identical from > `-O2' to `-O4'.) The `-O2' code appears to be less efficient at > register allocation; about twice as much stack temporary space is > required. > > For the graph, see <http://bostonradio.org/wollman/O4-is-bad.ps>. As my own test shown hint reporded benchmarks depend highly on options used to compile libm, so to get more accurate results your must consider to make libm with the same options used to compile hint (probably better to make several static versions of hint executable linked with the corresponding libm). Also please report your processor model when reporting benchmarks result, because non-Intel processors (AMD for the example) can pepform better with 486 optimisation that with pentium optimisation) Sincerely, Maxim 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?374BB31E.96E21EE5>