Just in-case anyone was ever curious about how well Wine performs with different C/XXFLAGs I did a test today with RC5 to see how much of a difference it makes with 3DMark 2001 SE, nothing major but if someone can think of a better benchmark to try, please let me know (I had hoped to try Oblivion or some such but it has no benchmark feature).
Core2Quad Q9450 @ 3.4Ghz || 4GB PC8500 5-5-5-15 || GCC 4.3.1 || Linux 2.6.25 3DMark 2001 SE B300. Wine 1.0-rc5.
28576 -march=native -O3 -pipe -fomit-frame-pointer 28522 -march=native -O3 -pipe -fomit-frame-pointer -mfpmath=sse,387 28511 -march=native -O2 -pipe -fomit-frame-pointer 28427 -march=native -O3 -fomit-frame-pointer -ffast-math -funroll-loops -Wall -pipe 28426 -march=native -O2 -pipe -mfpmath=sse,387 28311 -march=native -O2 -pipe 28270 -march=native -Os -pipe -fomit-frame-pointer 28126 -O3 -pipe -fomit-frame-pointer 28072 Wine Default (-g -O2) 27984 -march=native -O3 -pipe 27646 -march=native -Os -pipe
For reference I tried the same with Lame just to show how it would affect other programs:
Core2Quad Q9450 @ 3.4Ghz || 4GB PC8500 5-5-5-15 || GCC 4.3.1 || Linux 2.6.25 ~30 minute stereo wave file encoded in Lame Lame-3.97 64bit.
real 1m28.739s -O3 -ffast-math -funroll-loops -Wall -pipe (Lame defaults) real 1m29.168s -march=native -O3 -ffast-math -funroll-loops -Wall -pipe real 1m35.633s -march=native -O3 -pipe -ffast-math real 1m35.923s -march=native -O3 -pipe -ftracer -ffast-math real 1m36.346s -march=native -O3 -pipe -ffast-math -falign-functions=64 real 1m36.768s -march=native -O2 -pipe -ftracer -ffast-math real 1m37.242s -march=native -O3 -pipe -ftracer real 1m38.127s -march=native -O3 -pipe -fomit-frame-pointer real 1m38.235s -march=native -O3 -pipe -fvect-cost-model real 1m38.243s -march=native -O2 -pipe -mfpmath=sse,387 real 1m38.278s -march=native -O2 -pipe real 1m38.423s -march=native -O2 -pipe -falign-functions=64 real 1m38.527s -march=native -O2 -pipe -ffast-math real 1m38.573s -march=native -O2 -pipe -falign-functions=128 real 1m38.575s -march=native -O2 -pipe -mfpmath=sse real 1m39.241s -march=native -O2 -pipe -ftracer real 1m39.261s -march=native -O2 -pipe -fomit-frame-pointer real 1m40.946s -march=native -Os -pipe -ffast-math real 1m41.385s -march=native -Os -pipe real 1m41.513s -march=native -Os -pipe -fomit-frame-pointer real 1m41.602s -march=native -Os -pipe -fomit-frame-pointer -mfpmath=sse,387 real 1m46.577s -march=native -O2 -pipe -mfpmath=387
Slightly more influential there. Ben H.
Ben Hodgetts (Enverex) wrote:
Just in-case anyone was ever curious about how well Wine performs with different C/XXFLAGs I did a test today with RC5 to see how much of a difference it makes with 3DMark 2001 SE, nothing major but if someone can think of a better benchmark to try, please let me know (I had hoped to try Oblivion or some such but it has no benchmark feature).
Core2Quad Q9450 @ 3.4Ghz || 4GB PC8500 5-5-5-15 || GCC 4.3.1 || Linux 2.6.25 3DMark 2001 SE B300. Wine 1.0-rc5.
28576 -march=native -O3 -pipe -fomit-frame-pointer 28522 -march=native -O3 -pipe -fomit-frame-pointer -mfpmath=sse,387 28511 -march=native -O2 -pipe -fomit-frame-pointer 28427 -march=native -O3 -fomit-frame-pointer -ffast-math -funroll-loops -Wall -pipe 28426 -march=native -O2 -pipe -mfpmath=sse,387 28311 -march=native -O2 -pipe 28270 -march=native -Os -pipe -fomit-frame-pointer 28126 -O3 -pipe -fomit-frame-pointer 28072 Wine Default (-g -O2) 27984 -march=native -O3 -pipe 27646 -march=native -Os -pipe
-pipe only speeds up compilation, not the resulting machine code. -Wall makes gcc show more warnings about the code and again, it does not speed up the code.
tom
Tomas Carnecky wrote:
Ben Hodgetts (Enverex) wrote:
Just in-case anyone was ever curious about how well Wine performs with different C/XXFLAGs I did a test today with RC5 to see how much of a difference it makes with 3DMark 2001 SE, nothing major but if someone can think of a better benchmark to try, please let me know (I had hoped to try Oblivion or some such but it has no benchmark feature).
Core2Quad Q9450 @ 3.4Ghz || 4GB PC8500 5-5-5-15 || GCC 4.3.1 || Linux 2.6.25 3DMark 2001 SE B300. Wine 1.0-rc5.
28576 -march=native -O3 -pipe -fomit-frame-pointer 28522 -march=native -O3 -pipe -fomit-frame-pointer -mfpmath=sse,387 28511 -march=native -O2 -pipe -fomit-frame-pointer 28427 -march=native -O3 -fomit-frame-pointer -ffast-math -funroll-loops -Wall -pipe 28426 -march=native -O2 -pipe -mfpmath=sse,387 28311 -march=native -O2 -pipe 28270 -march=native -Os -pipe -fomit-frame-pointer 28126 -O3 -pipe -fomit-frame-pointer 28072 Wine Default (-g -O2) 27984 -march=native -O3 -pipe 27646 -march=native -Os -pipe
-pipe only speeds up compilation, not the resulting machine code. -Wall makes gcc show more warnings about the code and again, it does not speed up the code.
tom
I know, I was just listing "all" the flags I was using. Wall was only in there because I copied the CFLAGs from Lame which got a massive boost using a certain set which I copied over entirely.
Ben H.
-
Ben Hodgetts -
Ben Hodgetts (Enverex)
-
Tomas Carnecky