Here is new version of sse2 clamping, one patch grew into small set. Now sse2 stuff is separated into its own object which will get -msse2 compile flag. I did sse'ize also rest of _mesa_map_rgba function. As previously there are ifdefs which I don't think look nice but I tried to keep looks of the code (new versus old) similar.
What is now the biggest 'rfc' part here is where should all this sse2 stuff really exist in. This patch set show my suggestion, I made 'x86' folder under src/mesa/main. The idea here being if there is optimization targeting architecture it'd exist directly under the place where it was used, in its own subdirectly indicating targeted architecture. I don't think majority of such code would be generic code thus this approach. /Juha-Pekka Juha-Pekka Heikkila (3): configure.ac: Add detection for sse2 compilation support mesa/main/x86: Add sse2 streaming clamping mesa/main: Clamp rgba with streamed sse configure.ac | 7 +++ src/mesa/Makefile.am | 8 +++ src/mesa/main/pixeltransfer.c | 62 ++++++++++++++++------- src/mesa/main/x86/sse2_clamping.c | 103 ++++++++++++++++++++++++++++++++++++++ src/mesa/main/x86/sse2_clamping.h | 49 ++++++++++++++++++ 5 files changed, 210 insertions(+), 19 deletions(-) create mode 100644 src/mesa/main/x86/sse2_clamping.c create mode 100644 src/mesa/main/x86/sse2_clamping.h -- 1.8.5.1 _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
