On Sun, 10 Nov 2002 20:42:30 +0100 Arpi <[EMAIL PROTECTED]> wrote: > Hi, > [snip] > Not. gcc (3.1+) will use MMX for integer, SSE for float code, even with > normal instructions, if both > - code optimization enabled (so it will recognize normal array operations > (example: for(i=0;i<15;i++) y[i]+=x[i];) and replace them with SIMD) > - SIMD usage enabled (-msse, -mmmx, -msse2, -m3dnow, > -mfpmath=sse (use SSE for float) / -mfpmath=sse,387 (use SSE and FPU in > parallel for float) > > so, by default it won't use SIMD. > > > Many other architectures use a compiler flag like "-msoft-float" or > > "-mno-fpu" to explicitly tell gcc that it cannot use the FP unit. If gcc > > starts using MMX on its own, we'll have to find the switch that tells it > > it has -mno-mmx, -mno-sse etc. > > > not to do that. You might want to ask some gcc people about it.. > > ehh. i'm not gcc people... :) > (but i did RTFM :)
So did I ;-) All I found about -mmmx and friends is that they enable builtin functions. The documentation doesn't say anything about their being used automagically. Also, it's not quite clear from the docs what happens if you specify neither -mmmx nor -mno-mmx. > A'rpi / Astral & ESP-team > > -- > Developer of MPlayer, the Movie Player for Linux - http://www.MPlayerHQ.hu Felix __\|/__ ___ ___ ___ __Tschüß_______\_6 6_/___/__ \___/__ \___/___\___You can do anything,___ _____Felix_______\Ä/\ \_____\ \_____\ \______U___just not everything____ [EMAIL PROTECTED] >o<__/ \___/ \___/ at the same time! ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel