Good news! I'm running upstream 3.2.32-rc1 + c07496fa61f4c5cb2addd1c57f6b. - The tiling corruption is fixed; - 8 tests are still failing (our friend gem_cs_tlb is the only "regression" if run after other tests when compared to the "Debian 3.2.23" kernel --- I'm assuming this kernel is similar to the upstream 3.2.23) -- so I'm counting that as 7; - 6 tests that failed on 3.2.23 are now passing; - gem_set_tiling_vs_pwrite passes.
So, all in all, we have a very positive outcome here. :-) $ diff fails-debian-unpatched fails-3.2.32+c07496fa61f4 0a1 > FAIL: gem_cs_tlb 4,8d4 < FAIL: gem_set_tiling_vs_blt < FAIL: gem_tiled_pread < FAIL: gem_tiled_pread_pwrite < FAIL: gem_tiled_partial_pwrite_pread < FAIL: gem_tiled_swapping 10d5 < FAIL: gem_tiled_fence_blits $ cat fails-3.2.32+c07496fa61f4 FAIL: gem_cs_tlb FAIL: gem_exec_bad_domains FAIL: gem_mmap_gtt FAIL: gem_mmap_offset_exhaustion FAIL: gem_threaded_access_tiled FAIL: flip_test FAIL: prime_self_import FAIL: ZZ_check_dmesg The test logs: http://dl.dropbox.com/u/15755099/intel_gpu_tests_3.tar.bz2 Best regards, Luís Picciochi On 16/10/2012, Luís Picciochi Oliveira <[email protected]> wrote: > On 15/10/2012, Ben Hutchings <[email protected]> wrote: >> >> This is the version that's under review for 3.2.y: >> http://git.kernel.org/?p=linux/kernel/git/bwh/linux-3.2.y-queue.git;a=blob_plain;f=queue-3.2/drm-i915-mark-untiled-blt-commands-as-fenced-on-gen2-3.patch;h=4ec2d388af7168d7bd659453e62dafd0f6b5e02b;hb=b6e32828805b35efce711f75f95f360770157f32 >> >> You can test what is due to become 3.2.32 by taking 3.2.31 and applying >> the patch in <[email protected]>, >> archived at http://article.gmane.org/gmane.linux.kernel/1375450/raw Gah! Darn you, =\n, =3D and =20s! -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
