On Wed, Dec 31, 2014 at 10:48:28AM +0800, Zhigang Gong wrote: > On Mon, Dec 29, 2014 at 11:17:43AM +0800, Zhenyu Wang wrote: > > On 2014.12.29 10:06:58 +0800, Zhigang Gong wrote: > > > When does the kernel do instruction cache invalidate for us? > > > Could you give a pointer to the related code. > > > > > > > You may check i915_gem_execbuffer_move_to_gpu() function. > > I just checked and confirmed that it does flush and invalidate all type of > caches. > > > > > > I also tested it and found many utests failures. My kernel version is > > > as below: > > > > > > Linux ivb-gt2-rev4 3.17.1 #1 SMP Fri Oct 24 14:45:46 CST 2014 x86_64 > > > x86_64 x86_64 GNU/Linux > > > > > > > I tested with drm-intel-nightly tree which seems fine on HSW and BDW. > > I haven't run it on IVB, might try that later. > > I also tested your patch on HSW, and it works fine. But it doesn't work on > IVB. > Don't know why, but need further investigation.
further investigation indicate on IVB we need another pipe control after the media state flush command. I will modify the patch accordingly and push it latter. Thanks. > > > > > -- > > Open Source Technology Center, Intel ltd. > > > > $gpg --keyserver wwwkeys.pgp.net --recv-keys 4D781827 > > > > > _______________________________________________ > > Beignet mailing list > > [email protected] > > http://lists.freedesktop.org/mailman/listinfo/beignet > > _______________________________________________ > Beignet mailing list > [email protected] > http://lists.freedesktop.org/mailman/listinfo/beignet _______________________________________________ Beignet mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/beignet
