Created attachment 70947 [details] gem_ringfill failed dmesg Environment: ------------------------------ The Linux Kernel, the operating system core itself Kernel: (drm-intel-next-queued)42dcedd4f2e715dc0313e359c8288e6397843fff Some additional commit info: Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Thu Nov 15 11:32:30 2012 +0000 Console print: ------------------------------ filling blt ring: 100% verifying filling render ring: 100% verifying render: copy #0 at 0,0 failed: read 0x00d00000 render: copy #1 at 1,0 failed: read 0x00d00001 render: copy #2 at 2,0 failed: read 0x00d00002 render: copy #3 at 3,0 failed: read 0x00d00003 render: copy #4 at 4,0 failed: read 0x00d00004 render: copy #5 at 5,0 failed: read 0x00d00005 render: copy #6 at 6,0 failed: read 0x00d00006 render: copy #7 at 7,0 failed: read 0x00d00007 render: copy #8 at 8,0 failed: read 0x00d00008 Description: ------------------------------ This issue caused by I-G-T, not the kernel. It's like gem_ringfill.c add some function to verify the case, make it failed. Process: ------------------------------ 1. git pull 2. git log gem_ringfill.c 3. ./gem_ringfill.c Result: ------------------------------ latest commit: 64f669f7a94d459437a4dd47142ef2bb9870c5e ---> failed the second commit: ca2851f827d5f60283dc8c202006a2f5095c738 ---> succeed
pro(In reply to comment #0) > Created attachment 70947 [details] > gem_ringfill failed dmesg > > Environment: > ------------------------------ > The Linux Kernel, the operating system core itself > Kernel: (drm-intel-next-queued)42dcedd4f2e715dc0313e359c8288e6397843fff > Some additional commit info: > Author: Chris Wilson <chris@chris-wilson.co.uk> > Date: Thu Nov 15 11:32:30 2012 +0000 > > > Console print: > ------------------------------ > filling blt ring: 100% > verifying > filling render ring: 100% > verifying > render: copy #0 at 0,0 failed: read 0x00d00000 > render: copy #1 at 1,0 failed: read 0x00d00001 > render: copy #2 at 2,0 failed: read 0x00d00002 > render: copy #3 at 3,0 failed: read 0x00d00003 > render: copy #4 at 4,0 failed: read 0x00d00004 > render: copy #5 at 5,0 failed: read 0x00d00005 > render: copy #6 at 6,0 failed: read 0x00d00006 > render: copy #7 at 7,0 failed: read 0x00d00007 > render: copy #8 at 8,0 failed: read 0x00d00008 > > > Description: > ------------------------------ > This issue caused by I-G-T, not the kernel. It's like gem_ringfill.c add > some function to verify the case, make it failed. > > > Process: > ------------------------------ > 1. git pull > > 2. git log gem_ringfill.c > > 3. ./gem_ringfill.c ----> make ----> ./gem_ringfill > > Result: > ------------------------------ > latest commit: 64f669f7a94d459437a4dd47142ef2bb9870c5e ---> failed > > the second commit: ca2851f827d5f60283dc8c202006a2f5095c738 ---> succeed
Does this patch here help? https://patchwork.kernel.org/patch/1812281/ Also, can you please supply the bisect result? And somehow I can't find your sha1 in my kernel (ca2851f827d5f60283dc8c202006a2f5095c738 and 64f669f7a94d459437a4dd47142ef2bb9870c5e)
(In reply to comment #2) > Does this patch here help? > > https://patchwork.kernel.org/patch/1812281/ > > Also, can you please supply the bisect result? And somehow I can't find your > sha1 in my kernel (ca2851f827d5f60283dc8c202006a2f5095c738 and > 64f669f7a94d459437a4dd47142ef2bb9870c5e) It's not a kernel issue, these two commit is in I-G-T git source, just like bug #57826 described.
(In reply to comment #3) > It's not a kernel issue, these two commit is in I-G-T git source, just like > bug #57826 described. Sorry, I've missed that. Can you please confirm that only the gem_ringfill/render subtest fails, but the gem_ringfill/blitter should work. Note that you need the new piglit testrunner to list and run subtests.
Not a regression, the recent change was to start testing the render ring on HSW, so the failure was a previously unknown issue.
Mika is signed up to fix rendercpy.
(In reply to comment #4) > (In reply to comment #3) > > It's not a kernel issue, these two commit is in I-G-T git source, just like > > bug #57826 described. > > Sorry, I've missed that. Can you please confirm that only the > gem_ringfill/render subtest fails, but the gem_ringfill/blitter should work. > Note that you need the new piglit testrunner to list and run subtests. Decription: -------------------- Exactly you are right. Below is the result. [root@x-hsw24 piglit]# /GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests/gem_ringfill --run-subtest render filling render ring: 100% verifying render: copy #0 at 0,0 failed: read 0x00d00000 render: copy #1 at 1,0 failed: read 0x00d00001 render: copy #2 at 2,0 failed: read 0x00d00002 render: copy #3 at 3,0 failed: read 0x00d00003 render: copy #4 at 4,0 failed: read 0x00d00004 render: copy #5 at 5,0 failed: read 0x00d00005 render: copy #6 at 6,0 failed: read 0x00d00006 render: copy #7 at 7,0 failed: read 0x00d00007 render: copy #8 at 8,0 failed: read 0x00d00008 [root@x-hsw24 piglit]# ./piglit-print-commands.py tests/igt.tests | less [root@x-hsw24 piglit]# /GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests/gem_ringfill --run-subtest blitter filling blt ring: 100% verifying
Annotating with the failing subtest.
Fixed in i-g-t commit 1945e2a7f18dc2e4c46010ab8b7774679fc217d3 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Thu Dec 6 17:18:52 2012 +0000 Implement a working gen7 rendercopy Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=57825 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Yeah, it's exactly worked. It's perfect now.
Closing old Verified+Fixed.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.