On Mon, Feb 09, 2015 at 01:54:16PM -0800, Ben Widawsky wrote: > If we're moving a bunch of buffers from the CPU domain to the GPU domain, and > we've already blown out the entire cache via a wbinvd, there is nothing more > to > do. > > With this and the previous patches, I am seeing a 3x FPS increase on a certain > benchmark which uses a giant 2d array texture. Unless I missed something in > the > code, it should only effect non-LLC i915 platforms.
Out of curiosity, have you compared with how this performs with an improved userspace? There are several techniques which userspace can do that are much higher performance than either clflush or wbinvd. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/intel-gfx
