| Summary: | [PNV/BDW Bisected]igt/kms_flip subcases rcs-flip-vs-panning and rcs-flip-vs-panning-interruptible fail | ||||||
|---|---|---|---|---|---|---|---|
| Product: | DRI | Reporter: | Guo Jinxian <jinxianx.guo> | ||||
| Component: | DRM/Intel | Assignee: | Intel GFX Bugs mailing list <intel-gfx-bugs> | ||||
| Status: | CLOSED FIXED | QA Contact: | Intel GFX Bugs mailing list <intel-gfx-bugs> | ||||
| Severity: | normal | ||||||
| Priority: | highest | CC: | intel-gfx-bugs, yi.sun | ||||
| Version: | unspecified | ||||||
| Hardware: | Other | ||||||
| OS: | All | ||||||
| Whiteboard: | |||||||
| i915 platform: | i915 features: | ||||||
| Attachments: |
|
||||||
|
Description
Guo Jinxian
2014-05-22 05:30:57 UTC
Bisected result form prts shows that was igt issue: c75dcbdac51dcc2d631a4a45049c9421e0229e16 is the first bad commit commit c75dcbdac51dcc2d631a4a45049c9421e0229e16 Author: Daniel Vetter <daniel.vetter@ffwll.ch> AuthorDate: Tue May 20 15:45:31 2014 +0200 Commit: Daniel Vetter <daniel.vetter@ffwll.ch> CommitDate: Tue May 20 15:45:31 2014 +0200 tests/kms_flip: various improvements - Some debug logging for the ts continuity checks. - Add a plain vblank-vs-suspend test where kms_flip doesn't switch off the displays first. - Ensure that we do at least 2 rounds, the suspend tests bailed out after 1 round and so didn't test anything. - Frob the testnames a bit. Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> . This bug is able to reproduce on BDW too. ./kms_flip --run-subtest bcs-flip-vs-panning-interruptible IGT-Version: 1.6-gc75dcbd (x86_64) (Linux: 3.15.0-rc6_drm-intel-nightly_de544d_20140523+ x86_64) Using monotonic timestamps Beginning bcs-flip-vs-panning-interruptible on crtc 5, connector 14 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 ..Test assertion failure function run_test_step, file kms_flip.c:929: Last errno: 16, Device or resource busy Failed assertion: set_mode(o, o->fb_ids[o->current_fb_id], x_ofs, 0) == 0 failed to pan (1920x1080@60Hz)+10: Device or resource busy Subtest bcs-flip-vs-panning-interruptible: FAIL (In reply to comment #1) > Bisected result form prts shows that was igt issue: > > c75dcbdac51dcc2d631a4a45049c9421e0229e16 is the first bad commit > commit c75dcbdac51dcc2d631a4a45049c9421e0229e16 > Author: Daniel Vetter <daniel.vetter@ffwll.ch> > AuthorDate: Tue May 20 15:45:31 2014 +0200 > Commit: Daniel Vetter <daniel.vetter@ffwll.ch> > CommitDate: Tue May 20 15:45:31 2014 +0200 > > tests/kms_flip: various improvements > > - Some debug logging for the ts continuity checks. > - Add a plain vblank-vs-suspend test where kms_flip doesn't switch off > the displays first. > - Ensure that we do at least 2 rounds, the suspend tests bailed out > after 1 round and so didn't test anything. > - Frob the testnames a bit. > > Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Bisect is a bit suspicious. Can you please test whether the revert makes the regression go away? It reverts cleanly here. If this indicates a wrong bisect result, please restart the bisect. Test still failed on latest -nightly(af1aaba219fdd90ca1b30f9b8d8d19352224f170) on BDW root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./kms_flip --run-subtest busy-flip IGT-Version: 1.7-gbffef22 (x86_64) (Linux: 3.16.0-rc6_drm-intel-nightly_af1aab_20140724_debug+ x86_64) Using monotonic timestamps Beginning busy-flip on crtc 7, connector 18 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 ............................................................................................................................................................................................................................................................................................................. busy-flip on crtc 7, connector 18: PASSED Beginning busy-flip on crtc 11, connector 18 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 .Test assertion failure function run_test_step, file kms_flip.c:900: Failed assertion: __wait_for_vblank(TEST_VBLANK_BLOCK, o->pipe, 1, 0, &reply) == 0 Last errno: 16, Device or resource busy Subtest busy-flip: FAIL I have a similar traceback on Sandy Bridge (i5-2520M) with fullscreen videos on 3.16 Is it the same bug or should I fill a new one ? No issue with 3.15.x so I downgraded. [49879.496132] WARNING: CPU: 1 PID: 1562 at drivers/gpu/drm/i915/intel_display.c:3313 intel_crtc_wait_for_pending_flips+0x157/0x160() [49879.496204] Modules linked in: ctr ccm hidp uvcvideo rfcomm videobuf2_vmalloc videobuf2_memops vboxpci(O) bnep videobuf2_core v4l2_common vboxnetadp(O) videodev vboxnetflt(O) vboxdrv(O) binfmt_misc nfsd auth_rpcgss oid_registry nfs_acl ecb nfs lockd btusb bluetooth sunrpc crc16 bridge stp llc tun xt_REDIRECT xt_nat ipt_MASQUERADE iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack ipt_REJECT xt_tcpudp iptable_filter ip_tables x_tables arc4 iwldvm mac80211 snd_hda_codec_hdmi iwlwifi snd_hda_codec_conexant snd_hda_codec_generic x86_pkg_temp_thermal intel_powerclamp crc32c_intel psmouse serio_raw sdhci_pci sdhci sr_mod sg cdrom mmc_core i2c_i801 xhci_hcd cfg80211 thinkpad_acpi snd_hda_intel snd_hda_controller snd_hda_codec wmi snd_pcm nvram rfkill snd_timer snd soundcore ehci_pci ehci_hcd zram tp_smapi(O) thinkpad_ec(O) kvm_intel kvm usblp fuse loop autofs4 nbd e1000e ptp pps_core [49879.496223] CPU: 1 PID: 1562 Comm: Xorg Tainted: G W O 3.16.0+ #1 [49879.496225] Hardware name: LENOVO 4171CTO/4171CTO, BIOS 8CET58WW (1.38 ) 07/18/2013 [49879.496231] 0000000000000009 ffffffff8168b7bd 0000000000000000 ffffffff81066552 [49879.496233] 0000000000000000 ffff8800d8119000 ffff8800d8058180 ffff8800d81ad800 [49879.496236] ffff8800d81ad800 ffffffff814b0787 0000000000000000 ffff8800cdbed000 [49879.496238] Call Trace: [49879.496248] [<ffffffff8168b7bd>] ? dump_stack+0x41/0x51 [49879.496258] [<ffffffff81066552>] ? warn_slowpath_common+0x72/0x90 [49879.496262] [<ffffffff814b0787>] ? intel_crtc_wait_for_pending_flips+0x157/0x160 [49879.496268] [<ffffffff810a0090>] ? __wake_up_sync+0x10/0x10 [49879.496274] [<ffffffff814ba05c>] ? intel_crtc_set_config+0xb1c/0xd90 [49879.496280] [<ffffffff81452cca>] ? drm_mode_set_config_internal+0x5a/0xd0 [49879.496284] [<ffffffff81441fec>] ? restore_fbdev_mode+0xac/0xd0 [49879.496288] [<ffffffff81442029>] ? drm_fb_helper_restore_fbdev_mode_unlocked+0x19/0x30 [49879.496291] [<ffffffff8144408d>] ? drm_fb_helper_set_par+0x1d/0x60 [49879.496296] [<ffffffff813b3d5a>] ? fb_set_var+0x16a/0x3a0 [49879.496303] [<ffffffff81266f96>] ? btrfs_delalloc_release_metadata+0x146/0x150 [49879.496307] [<ffffffff812661a6>] ? btrfs_free_reserved_data_space+0x36/0xd0 [49879.496313] [<ffffffff81095de0>] ? update_curr+0x80/0x110 [49879.496321] [<ffffffff813aaf55>] ? fbcon_blank+0x225/0x2e0 [49879.496330] [<ffffffff81423dde>] ? do_unblank_screen+0xae/0x1d0 [49879.496334] [<ffffffff8141a65f>] ? complete_change_console+0x4f/0xd0 [49879.496338] [<ffffffff8141b7b2>] ? vt_ioctl+0x10d2/0x1350 [49879.496342] [<ffffffff814477ee>] ? drm_ioctl+0x6e/0x570 [49879.496347] [<ffffffff81410854>] ? tty_ioctl+0x264/0xb70 [49879.496352] [<ffffffff8117250f>] ? new_sync_write+0x6f/0xa0 [49879.496356] [<ffffffff811aa4f6>] ? fsnotify+0x2b6/0x390 [49879.496363] [<ffffffff81183800>] ? do_vfs_ioctl+0x2d0/0x4a0 [49879.496367] [<ffffffff81174dd8>] ? __sb_end_write+0x28/0x60 [49879.496371] [<ffffffff81172c62>] ? vfs_write+0x1b2/0x1f0 [49879.496376] [<ffffffff81183a06>] ? SyS_ioctl+0x36/0x80 [49879.496381] [<ffffffff816916d6>] ? system_call_fastpath+0x1a/0x1f [49879.496384] ---[ end trace aeb7f0ab716be7f4 ]--- [49879.496388] [drm:intel_pipe_set_base] *ERROR* pipe is still busy with an old pageflip (In reply to comment #3) > (In reply to comment #1) > > Bisected result form prts shows that was igt issue: > > > > c75dcbdac51dcc2d631a4a45049c9421e0229e16 is the first bad commit > > commit c75dcbdac51dcc2d631a4a45049c9421e0229e16 > > Author: Daniel Vetter <daniel.vetter@ffwll.ch> > > AuthorDate: Tue May 20 15:45:31 2014 +0200 > > Commit: Daniel Vetter <daniel.vetter@ffwll.ch> > > CommitDate: Tue May 20 15:45:31 2014 +0200 > > > > tests/kms_flip: various improvements > > > > - Some debug logging for the ts continuity checks. > > - Add a plain vblank-vs-suspend test where kms_flip doesn't switch off > > the displays first. > > - Ensure that we do at least 2 rounds, the suspend tests bailed out > > after 1 round and so didn't test anything. > > - Frob the testnames a bit. > > > > Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> > > Bisect is a bit suspicious. Can you please test whether the revert makes the > regression go away? It reverts cleanly here. > > If this indicates a wrong bisect result, please restart the bisect. Please try this. Test passes on latest -nightly(ab7f1ab9c4d1c8af46517d60e54eeda9d8673f6a) on both PNV and BDW. [root@x-pnv2 tests]# ./kms_flip --run-subtest rcs-flip-vs-panning IGT-Version: 1.8-g137877f (i686) (Linux: 3.17.0-rc5_drm-intel-nightly_ab7f1a_20140917+ i686) Using monotonic timestamps Beginning rcs-flip-vs-panning on crtc 10, connector 11 1024x600 60 1024 1072 1104 1200 600 603 609 625 0xa 0x48 45000 Beginning rcs-flip-vs-panning on crtc 7, connector 17 1024x768 60 1024 1048 1184 1344 768 771 777 806 0xa 0x40 65000 Beginning rcs-flip-vs-panning on crtc 10, connector 17 1024x768 60 1024 1048 1184 1344 768 771 777 806 0xa 0x40 65000 Subtest rcs-flip-vs-panning: SUCCESS (1.584s) [root@x-pnv2 tests]# ./kms_flip --run-subtest rcs-flip-vs-panning-interruptible IGT-Version: 1.8-g137877f (i686) (Linux: 3.17.0-rc5_drm-intel-nightly_ab7f1a_20140917+ i686) Using monotonic timestamps Beginning rcs-flip-vs-panning-interruptible on crtc 10, connector 11 1024x600 60 1024 1072 1104 1200 600 603 609 625 0xa 0x48 45000 Beginning rcs-flip-vs-panning-interruptible on crtc 7, connector 17 1024x768 60 1024 1048 1184 1344 768 771 777 806 0xa 0x40 65000 Beginning rcs-flip-vs-panning-interruptible on crtc 10, connector 17 1024x768 60 1024 1048 1184 1344 768 771 777 806 0xa 0x40 65000 Subtest rcs-flip-vs-panning-interruptible: SUCCESS (0.830s) root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./kms_flip --run-subtest rcs-flip-vs-panning IGT-Version: 1.8-g137877f (x86_64) (Linux: 3.17.0-rc4_drm-intel-nightly_43df30_20140916+ x86_64) Using monotonic timestamps Beginning rcs-flip-vs-panning on crtc 8, connector 19 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 .. rcs-flip-vs-panning on crtc 8, connector 19: PASSED Beginning rcs-flip-vs-panning on crtc 12, connector 19 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 .. rcs-flip-vs-panning on crtc 12, connector 19: PASSED Beginning rcs-flip-vs-panning on crtc 16, connector 19 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 .. rcs-flip-vs-panning on crtc 16, connector 19: PASSED Subtest rcs-flip-vs-panning: SUCCESS (133.456s) root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./kms_flip --run-subtest rcs-flip-vs-panning-interruptible IGT-Version: 1.8-g137877f (x86_64) (Linux: 3.17.0-rc4_drm-intel-nightly_43df30_20140916+ x86_64) Using monotonic timestamps Beginning rcs-flip-vs-panning-interruptible on crtc 8, connector 19 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 .. rcs-flip-vs-panning-interruptible on crtc 8, connector 19: PASSED Beginning rcs-flip-vs-panning-interruptible on crtc 12, connector 19 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 .. rcs-flip-vs-panning-interruptible on crtc 12, connector 19: PASSED Beginning rcs-flip-vs-panning-interruptible on crtc 16, connector 19 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 .. rcs-flip-vs-panning-interruptible on crtc 16, connector 19: PASSED Subtest rcs-flip-vs-panning-interruptible: SUCCESS (133.453s) Closing (>2 years) 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.