Problems with i915

cataphract · 2013-08-06 20:15:34

I keep losing my session more or less randomly after a few time (hours). xscreensaver seems to be particularly good at triggering this. dmesg shows:

[46714.209068] ------------[ cut here ]------------
[46714.209071] WARNING: at drivers/gpu/drm/i915/intel_display.c:7892 intel_modeset_check_state+0x5fa/0x720 [i915]()
[46714.209072] encoder's computed active state doesn't match tracked active state (expected 1, found 0)
[46714.209072] Modules linked in: fuse md5 ecb ecryptfs cbc sha256_ssse3 sha256_generic encrypted_keys sha1_ssse3 sha1_generic hmac trusted tpm tpm_bios joydev hid_generic snd_hda_codec_realtek usbhid snd_usb_audio snd_usbmidi_lib snd_rawmidi hid snd_seq_device it87 snd_hda_codec_hdmi hwmon_vid nls_cp437 vfat fat ppdev mxm_wmi iTCO_wdt iTCO_vendor_support coretemp kvm_intel kvm crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd microcode psmouse serio_raw pcspkr i2c_i801 fan thermal snd_hda_intel acpi_cpufreq snd_hda_codec mperf e1000e snd_hwdep snd_pcm snd_page_alloc snd_timer parport_pc ptp parport mei_me pps_core wmi snd mei soundcore processor lpc_ich evdev ext4 crc16 mbcache jbd2 sd_mod ahci libahci libata xhci_hcd ehci_pci ehci_hcd
[46714.209089]  scsi_mod usbcore usb_common i915 video button i2c_algo_bit intel_agp intel_gtt drm_kms_helper drm i2c_core
[46714.209092] CPU: 2 PID: 7386 Comm: X Tainted: G        W    3.10.3-1-ARCH #1
[46714.209093] Hardware name: Gigabyte Technology Co., Ltd. Z87MX-D3H/Z87MX-D3H-CF, BIOS F2 05/03/2013
[46714.209093]  0000000000000009 ffff8804109e5a58 ffffffff814bd2df ffff8804109e5a90
[46714.209094]  ffffffff81055881 ffff880428106ca0 0000000000000001 ffff880428106800
[46714.209096]  ffff880428106c88 ffff8804281d8800 ffff8804109e5af0 ffffffff810558ec
[46714.209097] Call Trace:
[46714.209099]  [<ffffffff814bd2df>] dump_stack+0x19/0x1b
[46714.209100]  [<ffffffff81055881>] warn_slowpath_common+0x61/0x80
[46714.209101]  [<ffffffff810558ec>] warn_slowpath_fmt+0x4c/0x50
[46714.209105]  [<ffffffffa00cb5b8>] ? intel_ddi_get_hw_state+0x38/0x160 [i915]
[46714.209110]  [<ffffffffa00c425a>] intel_modeset_check_state+0x5fa/0x720 [i915]
[46714.209112]  [<ffffffff814c0374>] ? __mutex_lock_slowpath+0x284/0x3b0
[46714.209113]  [<ffffffff814c0374>] ? __mutex_lock_slowpath+0x284/0x3b0
[46714.209116]  [<ffffffffa00c43c0>] intel_connector_dpms+0x40/0x70 [i915]
[46714.209120]  [<ffffffffa00223d8>] drm_mode_obj_set_property_ioctl+0x308/0x320 [drm]
[46714.209123]  [<ffffffffa0022420>] drm_mode_connector_property_set_ioctl+0x30/0x40 [drm]
[46714.209126]  [<ffffffffa00111e2>] drm_ioctl+0x532/0x660 [drm]
[46714.209130]  [<ffffffff8119cd55>] do_vfs_ioctl+0x2e5/0x4d0
[46714.209131]  [<ffffffff8118d2f1>] ? __sb_end_write+0x31/0x60
[46714.209133]  [<ffffffff8119cfc1>] SyS_ioctl+0x81/0xa0
[46714.209134]  [<ffffffff814ca75d>] system_call_fastpath+0x1a/0x1f
[46714.209135] ---[ end trace cf9e973eab93dd18 ]---

I also have this xorg configuration file:

==> /etc/X11/xorg.conf.d/20-intel.conf <==
Section "Device"
	Identifier	"Intel Graphics"
	Driver		"Intel"
	Option		"AccelMethod"		"sna"
	Option		"TearFree"		"true"
EndSection

Also:

$ pacman -Qi xf86-video-intel | head -2
Name : xf86-video-intel
Version : 2.21.14-1

Any ideas here?

WonderWoofy · 2013-08-06 20:31:31

Hmmm... you say you have issues with i915 in your title, then post the installed version of the xf86-video-intel package. I think the more relevant bit of info would be the kernel version you are using. There was a kernel update today (3.10.5) that has some work done to the i915 "fences" (don't ask me what those are), which apparently were the purported cause of those suspend issues.

cataphract · 2013-08-06 20:35:59

The kernel version is in the dmesg output and it's 3.10.3-1-ARCH. I'll update and leave the computer on with xscreensaver running tomorrow, so I should have more data in 24h.

Thanks

cataphract · 2013-08-07 18:07:39

Unfortunately, it did not help:

[40111.867386] ------------[ cut here ]------------
[40111.867402] WARNING: at drivers/gpu/drm/i915/intel_display.c:7892 intel_modeset_check_state+0x5fa/0x720 [i915]()
[40111.867404] encoder's computed active state doesn't match tracked active state (expected 1, found 0)
[40111.867406] Modules linked in: fuse md5 ecb ecryptfs cbc sha256_ssse3 sha256_generic encrypted_keys sha1_ssse3 sha1_generic hmac trusted tpm tpm_bios joydev hid_generic snd_hda_codec_realtek it87 hwmon_vid snd_usb_audio usbhid hid snd_usbmidi_lib snd_rawmidi snd_seq_device snd_hda_codec_hdmi nls_cp437 vfat fat ppdev iTCO_wdt mxm_wmi iTCO_vendor_support coretemp kvm_intel kvm crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd microcode psmouse serio_raw pcspkr i2c_i801 snd_hda_intel snd_hda_codec fan e1000e snd_hwdep thermal snd_pcm snd_page_alloc parport_pc acpi_cpufreq ptp snd_timer parport mperf mei_me evdev snd lpc_ich wmi pps_core mei soundcore processor ext4 crc16 mbcache jbd2 sd_mod ahci libahci ehci_pci libata ehci_hcd xhci_hcd
[40111.867488]  scsi_mod usbcore usb_common i915 video button i2c_algo_bit intel_agp intel_gtt drm_kms_helper drm i2c_core
[40111.867502] CPU: 2 PID: 3955 Comm: X Tainted: G        W    3.10.5-1-ARCH #1
[40111.867505] Hardware name: Gigabyte Technology Co., Ltd. Z87MX-D3H/Z87MX-D3H-CF, BIOS F2 05/03/2013
[40111.867507]  0000000000000009 ffff8803d39d9a58 ffffffff814bcf77 ffff8803d39d9a90
[40111.867514]  ffffffff81055881 ffff8804289cf4a0 0000000000000001 ffff8804289cf000
[40111.867520]  ffff8804289cf488 ffff880427f69000 ffff8803d39d9af0 ffffffff810558ec
[40111.867526] Call Trace:
[40111.867534]  [<ffffffff814bcf77>] dump_stack+0x19/0x1b
[40111.867539]  [<ffffffff81055881>] warn_slowpath_common+0x61/0x80
[40111.867545]  [<ffffffff810558ec>] warn_slowpath_fmt+0x4c/0x50
[40111.867568]  [<ffffffffa00cb7b8>] ? intel_ddi_get_hw_state+0x38/0x160 [i915]
[40111.867590]  [<ffffffffa00c445a>] intel_modeset_check_state+0x5fa/0x720 [i915]
[40111.867598]  [<ffffffff814c0014>] ? __mutex_lock_slowpath+0x284/0x3b0
[40111.867603]  [<ffffffff814c0014>] ? __mutex_lock_slowpath+0x284/0x3b0
[40111.867623]  [<ffffffffa00c45c0>] intel_connector_dpms+0x40/0x70 [i915]
[40111.867639]  [<ffffffffa00223d8>] drm_mode_obj_set_property_ioctl+0x308/0x320 [drm]
[40111.867656]  [<ffffffffa0022420>] drm_mode_connector_property_set_ioctl+0x30/0x40 [drm]
[40111.867673]  [<ffffffffa00111e2>] drm_ioctl+0x532/0x660 [drm]
[40111.867688]  [<ffffffff8119cce5>] do_vfs_ioctl+0x2e5/0x4d0
[40111.867695]  [<ffffffff8118d271>] ? __sb_end_write+0x31/0x60
[40111.867702]  [<ffffffff8119cf51>] SyS_ioctl+0x81/0xa0
[40111.867709]  [<ffffffff814ca3dd>] system_call_fastpath+0x1a/0x1f
[40111.867712] ---[ end trace 04f5ea7dcd6bae2c ]---

So what are my options? I'd like to avoid going back to 3.9 since it's been EOLed. I did not have this problem with 3.9 though.

mar04 · 2013-08-07 18:39:56

First of all report that upstream: https://01.org/linuxgraphics/documentat … ort-bugs-0
If you don't want to use EOLed kernel, one option is to build 3.8 maintained by ubuntu: http://kernel.ubuntu.com/git?p=ubuntu/l … ;a=summary

cataphract · 2013-08-07 20:55:32

I'm not convinced the problem is the kernel anymore. I just left the computer for enough time for xscreensaver to start and was greated with the login prompt. But this time dmesg shows nothing. However, /var/log/Xorg.0.log:

[  5351.171] (EE) intel(0): sna_mode_redisplay: page flipping failed, disabling CRTC:3 (pipe=0)
[  5362.320] (EE) 
[  5362.320] (EE) Backtrace:
[  5362.320] (EE) 0: /usr/bin/X (xorg_backtrace+0x3d) [0x57f71d]
[  5362.320] (EE) 1: /usr/bin/X (0x400000+0x183489) [0x583489]
[  5362.320] (EE) 2: /usr/lib/libpthread.so.0 (0x7fc4bb337000+0xf0e0) [0x7fc4bb3460e0]
[  5362.320] (EE) 3: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7fc4b8f01000+0xc4814) [0x7fc4b8fc5814]
[  5362.320] (EE) 4: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7fc4b8f01000+0xc5528) [0x7fc4b8fc6528]
[  5362.320] (EE) 5: /usr/bin/X (DRI2SwapBuffers+0x2ab) [0x552c9b]
[  5362.320] (EE) 6: /usr/bin/X (0x400000+0x154363) [0x554363]
[  5362.321] (EE) 7: /usr/bin/X (0x400000+0x3728e) [0x43728e]
[  5362.321] (EE) 8: /usr/bin/X (0x400000+0x268b5) [0x4268b5]
[  5362.321] (EE) 9: /usr/lib/libc.so.6 (__libc_start_main+0xf5) [0x7fc4ba3bda15]
[  5362.321] (EE) 10: /usr/bin/X (0x400000+0x26bfd) [0x426bfd]
[  5362.321] (EE) 
[  5362.321] (EE) Segmentation fault at address 0x18
[  5362.321] (EE) 
Fatal server error:
[  5362.321] (EE) Caught signal 11 (Segmentation fault). Server aborting

So the problem would be in xf86-video-intel, right?

Thaodan · 2013-08-07 21:10:06

Do you have KMS enabled?

WonderWoofy · 2013-08-07 21:12:57

Without KMS, you can't start X I don't think... at least not using intel_drv.so.

cataphract · 2013-08-07 21:34:47

Thaodan wrote:

Do you have KMS enabled?

Yes, I can't even start X if I add nomodeset to the kernel command line (which is otherwise just "root=UUID=9dac92cf-60d1-4a6d-b9a4-066bb876c9ce ro initrd=\initramfs-linux.img").

cfr · 2013-08-07 21:48:50

Since it is complaining about a feature in sna, have you tried uxa? This is default so you could just move your conf file out of range temporarily, restart X and see if it helps.

mar04 · 2013-08-08 05:11:15

uxa is no longer default in arch package: https://projects.archlinux.org/svntogit … d68def7e48

cfr · 2013-08-08 21:00:26

Hmm... didn't realise that. However, in that case, try setting it explicitly to uxa.

EDIT: I'm confused. Should I have seen an update message from pacman? Because I thought I must have missed it but there is no message in pacman.log either, even though the install script seems to include one.

Last edited by cfr (2013-08-08 21:04:33)

karol · 2013-08-08 21:23:17

Unfortunately it's shown only on install, not on update - no idea why, I'd consider it a packaging bug.
I've learned about the change from a comment on the bugtracker, so I wasn't surprised it's not UXA by default anymore.

cfr · 2013-08-08 22:27:27

https://bugs.archlinux.org/task/36466

Arch Linux

#1 2013-08-06 20:15:34

Problems with i915

#2 2013-08-06 20:31:31

Re: Problems with i915

#3 2013-08-06 20:35:59

Re: Problems with i915

#4 2013-08-07 18:07:39

Re: Problems with i915

#5 2013-08-07 18:39:56

Re: Problems with i915

#6 2013-08-07 20:55:32

Re: Problems with i915

#7 2013-08-07 21:10:06

Re: Problems with i915

#8 2013-08-07 21:12:57

Re: Problems with i915

#9 2013-08-07 21:34:47

Re: Problems with i915

#10 2013-08-07 21:48:50

Re: Problems with i915

#11 2013-08-08 05:11:15

Re: Problems with i915

#12 2013-08-08 21:00:26

Re: Problems with i915

#13 2013-08-08 21:23:17

Re: Problems with i915

#14 2013-08-08 22:27:27

Re: Problems with i915

Board footer