Bug 13322 - KMS - Problem with detaching VGA-output
Summary: KMS - Problem with detaching VGA-output
Status: RESOLVED CODE_FIX
Alias: None
Product: Drivers
Classification: Unclassified
Component: Video(DRI - Intel) (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: drivers_video-dri-intel@kernel-bugs.osdl.org
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2009-05-16 21:06 UTC by Jens Weibler
Modified: 2009-09-21 01:08 UTC (History)
2 users (show)

See Also:
Kernel Version: 2.6.30-rc5
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description Jens Weibler 2009-05-16 21:06:18 UTC
After using a external TFT as additional monitor I detached the VGA-cable but I can't get the desktop return to normal resolution.

71373.889237] [drm:i915_gem_object_unbind] *ERROR* Attempting to unbind pinned buffer
[71373.889247] ------------[ cut here ]------------
[71373.889249] WARNING: at drivers/gpu/drm/i915/i915_gem_tiling.c:320 i915_gem_set_tiling+0x3cd/0x458()
[71373.889251] Hardware name: Latitude E6500
[71373.889253] failed to unbind object for tiling switchModules linked in: iwlagn iwlcore mac80211 cfg80211 uvcvideo e1000e uhci_hcd ehci_hcd snd_hda_codec_intelhdmi snd_hda_codec_idt snd_hda_intel snd_hda_codec snd_pcm snd_page_alloc btusb [last unloaded: cfg80211]
[71373.889271] Pid: 4575, comm: X Tainted: G        W  2.6.30-rc5 #2
[71373.889273] Call Trace:
[71373.889280]  [<ffffffff8105f053>] warn_slowpath_fmt+0xe6/0x131
[71373.889286]  [<ffffffff8146b1a3>] ? printk+0x79/0x92
[71373.889290]  [<ffffffff812956cb>] ? drm_mode_object_put+0x4c/0x6b
[71373.889294]  [<ffffffff812a5ffc>] ? i915_gem_object_unbind+0x5b/0x233
[71373.889298]  [<ffffffff812a9e15>] i915_gem_set_tiling+0x3cd/0x458
[71373.889301]  [<ffffffff812a522d>] ? i915_gem_unpin_ioctl+0xe7/0x109
[71373.889305]  [<ffffffff8128c1e0>] drm_ioctl+0x22e/0x2e5
[71373.889308]  [<ffffffff8107a693>] ? hrtimer_start_range_ns+0x22/0x38
[71373.889312]  [<ffffffff812a9a48>] ? i915_gem_set_tiling+0x0/0x458
[71373.889316]  [<ffffffff8106e77e>] ? get_signal_to_deliver+0xf0/0x356
[71373.889321]  [<ffffffff810345fd>] ? save_i387_xstate+0x135/0x1cf
[71373.889325]  [<ffffffff810dcbb5>] vfs_ioctl+0x6e/0x9e
[71373.889328]  [<ffffffff810dd041>] do_vfs_ioctl+0x45c/0x4b2
[71373.889331]  [<ffffffff810dd0fb>] sys_ioctl+0x64/0x9a
[71373.889335]  [<ffffffff8102aeab>] system_call_fastpath+0x16/0x1b
[71373.889338] ---[ end trace dbb291fa1fdd3c2a ]---
Comment 1 johan_vdp 2009-06-14 18:24:55 UTC
Setting the display resolution (using xfce4-display-settings) causes a kerneloops.
The display resolution is changed.
(But the setting is not persistent. I assume the setting was not written to file. Application behavior is not a kernel issue however.)

Linux blauw 2.6.29.4-167.fc11.x86_64 #1 SMP Wed May 27 17:27:08 EDT 2009 x86_64 x86_64 x86_64 GNU/Linux

WARNING: at drivers/gpu/drm/i915/i915_gem_tiling.c:313 i915_gem_set_tiling+0x491/0x4ed [i915]() (Not tainted)
Hardware name: P5E-V HDMI
failed to unbind object for tiling switchModules linked in: sit tunnel4 ip6table_filter ip6_tables fuse sco bridge stp llc bnep l2cap bluetooth sunrpc nf_conntrack_netbios_ns ip6t_REJECT nf_conntrack_ipv6 ipv6 cpufreq_ondemand acpi_cpufreq freq_table dm_multipath raid456 async_xor async_memcpy async_tx xor uinput snd_hda_codec_intelhdmi snd_hda_codec_realtek ata_generic pata_acpi snd_hda_intel firewire_ohci snd_hda_codec firewire_core pcspkr crc_itu_t snd_hwdep serio_raw iTCO_wdt snd_pcm iTCO_vendor_support i2c_i801 snd_timer sata_promise snd soundcore snd_page_alloc asus_atk0110 atl1 hwmon mii pata_jmicron i915 drm i2c_algo_bit i2c_core video output [last unloaded: ip6_tables]
Pid: 6207, comm: Xorg Not tainted 2.6.29.4-167.fc11.x86_64 #1
Call Trace:
 [<ffffffff810489bf>] warn_slowpath+0xbc/0xf0
 [<ffffffff813a95ff>] ? printk+0x41/0x4a
 [<ffffffffa0063118>] i915_gem_set_tiling+0x491/0x4ed [i915]
 [<ffffffffa0062c87>] ? i915_gem_set_tiling+0x0/0x4ed [i915]
 [<ffffffffa0023b7e>] drm_ioctl+0x20e/0x2c1 [drm]
 [<ffffffff810e0f58>] vfs_ioctl+0x6f/0x87
 [<ffffffff810e13f3>] do_vfs_ioctl+0x462/0x4a3
 [<ffffffff810e148a>] sys_ioctl+0x56/0x79
 [<ffffffff8101133a>] system_call_fastpath+0x16/0x1b
---[ end trace b65a47ec69ee5cc8 ]---
Comment 2 Gordon Jin 2009-09-17 03:26:19 UTC
Does this still exist in 2.6.31?
Comment 3 johan_vdp 2009-09-20 20:42:35 UTC
Using Fedora 11 kernel 2.6.30.5-43.fc11.x86_64
without KMS (boot i915.modeset=0)
KMS causes display hangs.
Comment 4 johan_vdp 2009-09-20 20:50:30 UTC
Just rebooted with KMS enabled
xfce4-display-settings still causes an error in /var/log/messages:
 
Sep 20 22:47:42 blauw kernel: [drm:i915_gem_object_unbind] *ERROR* Attempting to unbind pinned buffer
Sep 20 22:47:42 blauw kernel: ------------[ cut here ]------------
Sep 20 22:47:42 blauw kernel: WARNING: at drivers/gpu/drm/i915/i915_gem_tiling.c:473 i915_gem_set_tiling+0x4ba/0x52c [i915]() (Not tainted)
Sep 20 22:47:42 blauw kernel: Hardware name: P5E-V HDMI
Sep 20 22:47:42 blauw kernel: failed to unbind object for tiling switchModules linked in: fuse w83627ehf hwmon_vid coretemp sunrpc nf_conntrack_netbios_ns ip6t_REJECT nf_conntrack_ipv6 ip6table_filter ip6_tables ipv6 cpufreq_ondemand acpi_cpufreq freq_table dm_multipath raid456 raid6_pq async_xor async_memcpy async_tx xor snd_hda_codec_intelhdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_pcm firewire_ohci snd_timer ata_generic firewire_core snd pata_acpi soundcore iTCO_wdt serio_raw crc_itu_t sata_promise pcspkr iTCO_vendor_support i2c_i801 snd_page_alloc atl1 floppy pata_jmicron asus_atk0110 mii hwmon i915 drm i2c_algo_bit video output i2c_core [last unloaded: microcode]
Sep 20 22:47:42 blauw kernel: Pid: 2420, comm: Xorg Not tainted 2.6.30.5-43.fc11.x86_64 #1
Sep 20 22:47:42 blauw kernel: Call Trace:
Sep 20 22:47:42 blauw kernel: [<ffffffff81057104>] warn_slowpath_common+0x95/0xc3
Sep 20 22:47:42 blauw kernel: [<ffffffff810571bf>] warn_slowpath_fmt+0x50/0x66
Sep 20 22:47:42 blauw kernel: [<ffffffffa006ddcc>] i915_gem_set_tiling+0x4ba/0x52c [i915]
Sep 20 22:47:42 blauw kernel: [<ffffffffa006d912>] ? i915_gem_set_tiling+0x0/0x52c [i915]
Sep 20 22:47:42 blauw kernel: [<ffffffffa00289cc>] drm_ioctl+0x21d/0x2e9 [drm]
Sep 20 22:47:42 blauw kernel: [<ffffffff81123744>] vfs_ioctl+0x7e/0xaa
Sep 20 22:47:42 blauw kernel: [<ffffffff81123bdc>] do_vfs_ioctl+0x46c/0x4c3
Sep 20 22:47:42 blauw kernel: [<ffffffff8111e5b2>] ? path_put+0x31/0x4c
Sep 20 22:47:42 blauw kernel: [<ffffffff81123c98>] sys_ioctl+0x65/0x9c
Sep 20 22:47:42 blauw kernel: [<ffffffff81012082>] system_call_fastpath+0x16/0x1b
Sep 20 22:47:42 blauw kernel: ---[ end trace f8b8546ffa581087 ]---
Sep 20 22:47:42 blauw kernel: [drm] DAC-6: set mode  18
Comment 5 johan_vdp 2009-09-20 22:29:36 UTC
just downloaded and built 2.6.31 (using Fedora provided config-2.6.30.5-43.fc11.x86_64 as .config)

using xfce4-display-config now shows (/var/log/message):
Sep 21 00:20:11 blauw kernel: [drm] DAC-6: set mode  19
Sep 21 00:20:27 blauw kernel: [drm] DAC-6: set mode  1b
Sep 21 00:21:45 blauw kernel: [drm] DAC-6: set mode  1c
Sep 21 00:21:57 blauw kernel: [drm] DAC-6: set mode  1d

So yes, 2.6.31 does solve the kerneloops I reported.
Comment 6 Gordon Jin 2009-09-21 01:08:13 UTC
Then closing. Thanks.

Note You need to log in before you can comment on or make changes to this bug.