Created attachment 115111 [details] dmesg ==System Environment== -------------------------- Regression: not sure Non-working platforms: SKL ==kernel== -------------------------- drm-intel-nightly/5ea91de4ff45adb60031853d64314c3405378fbd commit 5ea91de4ff45adb60031853d64314c3405378fbd Author: Daniel Vetter <daniel.vetter@ffwll.ch> Date: Tue Apr 14 20:00:06 2015 +0200 drm-intel-nightly: 2015y-04m-14d-17h-59m-22s UTC integration manifest ==Bug detailed description== ----------------------------- Run ./gem_reloc_vs_gpu --run-subtest forked-interruptible-thrash-inactive-hang 20 cycles on SKL, it happens on the 16th cycle. output error: Failed assertion: gtt_relocs Last errno: 5, Input/output error dmesg： [ 4656.471553] ------------[ cut here ]------------ [ 4656.471606] WARNING: CPU: 0 PID: 8632 at drivers/gpu/drm/i915/i915_gem.c:1999 i915_gem_object_put_pages_gtt+0x47/0x14d [i915]() [ 4656.471610] [drm:i915_reset_and_wakeup] resetting chip [ 4656.471618] WARN_ON(ret != -EIO) [ 4656.471622] Modules linked in: dm_mod snd_hda_codec_realtek ppdev snd_hda_codec_generic pcspkr snd_hda_intel snd_hda_controller snd_hda_codec snd_hda_core i2c_i801 snd_hwdep snd_pcm snd_timer snd soundcore joydev wmi battery parport_pc parport ac acpi_cpufreq i915 button video drm_kms_helper drm [ 4656.471688] CPU: 0 PID: 8632 Comm: gem_reloc_vs_gp Not tainted 4.0.0_drm-intel-nightly_5ea91d_20150415+ #1272 [ 4656.471694] Hardware name: Intel Corporation Skylake Client platform/Skylake Y LPDDR3 RVP3, BIOS SKLSE2R1.86C.B067.R00.1412310711 12/31/2014 [ 4656.471699] 0000000000000000 0000000000000009 ffffffff81795847 ffff880086623d38 [ 4656.471707] ffffffff8103bd5a 00000000000004f4 ffffffffa009e288 0000000000000001 [ 4656.471715] ffff880088384700 0000000000000000 ffffffffa010ce40 ffff880086623e18 [ 4656.471723] Call Trace: [ 4656.471737] [<ffffffff81795847>] ? dump_stack+0x40/0x50 [ 4656.471749] [<ffffffff8103bd5a>] ? warn_slowpath_common+0x98/0xb0 [ 4656.471792] [<ffffffffa009e288>] ? i915_gem_object_put_pages_gtt+0x47/0x14d [i915] [ 4656.471804] [<ffffffff8103bdb7>] ? warn_slowpath_fmt+0x45/0x4a [ 4656.471842] [<ffffffffa009dfbb>] ? i915_gem_object_set_to_cpu_domain+0x27/0x133 [i915] [ 4656.471917] [<ffffffffa009e288>] ? i915_gem_object_put_pages_gtt+0x47/0x14d [i915] [ 4656.471966] [<ffffffffa009fb2e>] ? i915_gem_object_put_pages+0x77/0xcf [i915] [ 4656.472029] [<ffffffffa00a2b60>] ? i915_gem_shrink+0x177/0x1dc [i915] [ 4656.472065] [<ffffffffa008fea3>] ? i915_drop_caches_set+0x87/0xb7 [i915] [ 4656.472085] [<ffffffff8112c632>] ? simple_attr_write+0x81/0xae [ 4656.472098] [<ffffffff8110f29d>] ? vfs_write+0xb0/0x186 [ 4656.472106] [<ffffffff8110fc0c>] ? SyS_write+0x41/0x84 [ 4656.472115] [<ffffffff8179b0f2>] ? system_call_fastpath+0x12/0x17 [ 4656.472121] ---[ end trace fd9a6691bb0a99a2 ]--- [ 4656.472206] ------------[ cut here ]------------ [ 4656.528352] kernel BUG at drivers/gpu/drm/i915/i915_gem.c:2247! ==Reproduce steps== ---------------------------- 1. run ./gem_reloc_vs_gpu --run-subtest forked-interruptible-thrash-inactive-hang 20 cycles.
Created attachment 115112 [details] test log
Would actually be fixed by http://patchwork.freedesktop.org/patch/46607/
(In reply to Chris Wilson from comment #2) > Would actually be fixed by http://patchwork.freedesktop.org/patch/46607/ Which is the best for this patch? Apply it on the latest drm-intel-nightly kernel: patching file drivers/gpu/drm/i915/i915_drv.c patching file drivers/gpu/drm/i915/i915_drv.h Hunk #1 succeeded at 1295 (offset -13 lines). Hunk #2 succeeded at 2049 (offset -20 lines). Hunk #3 FAILED at 2765. Hunk #4 succeeded at 2800 (offset -14 lines). 1 out of 4 hunks FAILED -- saving rejects to file drivers/gpu/drm/i915/i915_drv.h.rej patching file drivers/gpu/drm/i915/i915_gem.c Hunk #1 succeeded at 97 (offset -3 lines). Hunk #2 succeeded at 118 (offset -3 lines). Hunk #3 succeeded at 1132 (offset 1 line). Hunk #4 succeeded at 1203 (offset 1 line). Hunk #5 succeeded at 1217 (offset 1 line). Hunk #6 succeeded at 1264 (offset -1 lines). Hunk #7 FAILED at 1408. Hunk #8 FAILED at 1493. Hunk #9 FAILED at 1502. Hunk #10 FAILED at 1538. Hunk #11 succeeded at 2356 (offset -127 lines). Hunk #12 succeeded at 2541 (offset -96 lines). Hunk #13 FAILED at 3002. Hunk #14 FAILED at 3040. Hunk #15 FAILED at 3053. Hunk #16 FAILED at 3091. Hunk #17 succeeded at 4107 (offset -194 lines). Hunk #18 succeeded at 4126 (offset -194 lines). Hunk #19 succeeded at 4133 (offset -194 lines). 8 out of 19 hunks FAILED -- saving rejects to file drivers/gpu/drm/i915/i915_gem.c.rej patching file drivers/gpu/drm/i915/i915_irq.c Hunk #1 succeeded at 2353 (offset 54 lines). Hunk #2 succeeded at 2369 (offset 54 lines). Hunk #3 succeeded at 2386 (offset 54 lines). Hunk #4 succeeded at 2414 (offset 54 lines). patching file drivers/gpu/drm/i915/intel_display.c Hunk #1 succeeded at 3265 (offset 69 lines). Hunk #2 succeeded at 10220 (offset 532 lines). Hunk #3 FAILED at 10107. 1 out of 3 hunks FAILED -- saving rejects to file drivers/gpu/drm/i915/intel_display.c.rej patching file drivers/gpu/drm/i915/intel_lrc.c Hunk #1 succeeded at 774 (offset 98 lines). patching file drivers/gpu/drm/i915/intel_ringbuffer.c Hunk #1 succeeded at 2183 (offset -6 lines).
I am not sure which commit is the best for this patch. Fail to apply it on the latest drm-intel-nightly branch. patching file drivers/gpu/drm/i915/i915_drv.c Hunk #1 succeeded at 858 (offset 31 lines). Hunk #2 succeeded at 869 (offset 31 lines). Hunk #3 succeeded at 898 (offset 31 lines). Hunk #4 succeeded at 917 (offset 31 lines). Hunk #5 succeeded at 935 (offset 31 lines). patching file drivers/gpu/drm/i915/i915_drv.h Hunk #1 succeeded at 1311 (offset 3 lines). Hunk #2 succeeded at 2075 (offset 6 lines). Hunk #3 FAILED at 2765. Hunk #4 succeeded at 2839 (offset 25 lines). 1 out of 4 hunks FAILED -- saving rejects to file drivers/gpu/drm/i915/i915_drv.h.rej patching file drivers/gpu/drm/i915/i915_gem.c Hunk #1 succeeded at 94 (offset -6 lines). Hunk #2 succeeded at 115 (offset -6 lines). Hunk #3 succeeded at 1129 (offset -2 lines). Hunk #4 succeeded at 1200 (offset -2 lines). Hunk #5 succeeded at 1214 (offset -2 lines). Hunk #6 succeeded at 1261 (offset -4 lines). Hunk #7 FAILED at 1408. Hunk #8 FAILED at 1493. Hunk #9 FAILED at 1502. Hunk #10 FAILED at 1538. Hunk #11 succeeded at 2383 (offset -100 lines). Hunk #12 succeeded at 2569 (offset -68 lines). Hunk #13 FAILED at 3002. Hunk #14 FAILED at 3040. Hunk #15 FAILED at 3053. Hunk #16 FAILED at 3091. Hunk #17 succeeded at 4116 (offset -185 lines). Hunk #18 succeeded at 4135 (offset -185 lines). Hunk #19 succeeded at 4142 (offset -185 lines). 8 out of 19 hunks FAILED -- saving rejects to file drivers/gpu/drm/i915/i915_gem.c.rej patching file drivers/gpu/drm/i915/i915_irq.c Hunk #1 succeeded at 2353 (offset 54 lines). Hunk #2 succeeded at 2369 (offset 54 lines). Hunk #3 succeeded at 2386 (offset 54 lines). Hunk #4 succeeded at 2414 (offset 54 lines). patching file drivers/gpu/drm/i915/intel_display.c Hunk #1 succeeded at 3307 (offset 111 lines). Hunk #2 succeeded at 10350 (offset 662 lines). Hunk #3 FAILED at 10107. 1 out of 3 hunks FAILED -- saving rejects to file drivers/gpu/drm/i915/intel_display.c.rej patching file drivers/gpu/drm/i915/intel_lrc.c Hunk #1 succeeded at 800 (offset 124 lines). patching file drivers/gpu/drm/i915/intel_ringbuffer.c Hunk #1 succeeded at 2201 (offset 12 lines).
I cherry-picked that patch on top of recent -nightly. Please test. https://github.com/anderco/linux/tree/fdo/90050
Test on the latest drm-intel-nightly kernel 5 cycles, I don't meet this issue. I will give it more try(one cycle takes about 4 minutes).
run 20 cycles, I don't see this issue.
Verified.Fixed.
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.