On Sun, 10 Aug 2003 11:55:38 +1200 Christopher Sawtell <[EMAIL PROTECTED]> wrote:
> On Sun, 10 Aug 2003 11:12, you wrote: > > I decided to spend a few hours experimenting with compiler options > > in Gentoo last night. I wanted to measure the performance > > difference of"optimal" compiler settings. I built Gentoo from stage > > 1 with the following options: > Expressing an opinion here, I think that you would see a much greater > difference in performance if the test/benchmark programs were 'deep > math' applications. All the proggies following have a fair amount os > disk i/o while the computer is sitting there just waitng for the data > to come off of the disk. This will seriously mask any effect of code > optimisations. In general the faster the cpu, the less overall benefit > you will see from optimisations. Perhaps you might care to repeat the > tests with something like POVray rendering a relatively complex scene > on a less than up to the moment cpu. that would be fine if he wanted to set up his system twice again, with emergeing enough stuff to get povray running. Thats probably a LOT more than just being in a position to compile a kernel! http://www.gentoo.org/dyn/pkgs/media-gfx/povray.xml povray needs xfree (or something else to satisfy virtual/x11) can we think of something to fit the bill which does not require X? some kind of math/physics process? one of those "lets crack 2 million bit encryption together over the net" programs, or setiathome perhaps? (I guess you would want to process the same set of seti data on each run) > > Also it's fairly well documented, in the lore if nowhere else, that > the use of '-O3' can sometimes be rather counter productive on some > chip sets. Unfortunatly I know not what. > > > "Optimal" settings, based on Gentoo documentation: > > CFLAGS="-march=athlon-mp -O3 -pipe -fomit-frame-pointer" > > CHOST="i686-pc-linux-gnu" > > > > "Base" settings: > > CFLAGS="-O2" > > CHOST="i386-pc-linux-gnu" > > > > The configuration of the system was exactly the same other than the > > above settings. I initially built the system with the "optimal" > > settings, performed a benchmark, then rebuilt the entire system with > > the"base" settings. > > [ ... ] > > -- > Sincerely etc., > Christopher Sawtell > >
