imx-drm hang issue with etnaviv (GC3000)

imx-drm hang issue with etnaviv (GC3000)

[Wladimir J. van der Laan]

Hello,

After running kmscube (or another KMS executable) on a i.MX6 QuadPlus (etnaviv,
GC3000) a few times on I get the below crash in the drm kernel driver.
This is on a device with LVDS panel. It is always reproducible, although the
number of invocations needed differs.

The only way to get rendering to work again after the crash is to reboot.
Repeated tries only get the "flip_done timed out".

This always happens while the program is exiting.

Versions:

- mesa: https://github.com/etnaviv/mesa 9a09984

- libdrm: https://cgit.freedesktop.org/mesa/drm/ fe4579e

- Kernel: 4.8.0 or 4.8.4 + Pengutronix patches (20161007).

Does anyone have an idea what could be the problem?

Regards,
Wladimir van der Laan

[  130.026973] ------------[ cut here ]------------
[  130.031630] WARNING: CPU: 1 PID: 222 at drivers/gpu/drm/drm_atomic_helper.c:1127 drm_atomic_helper_wait_for_vblanks+0x1e4/0x200
[  130.043149] [CRTC:24] vblank wait timed out
[  130.047367] Modules linked in: hid_generic usbhid hid ci_hdrc_imx ci_hdrc extcon_core ehci_hcd usbcore usb_common usbmisc_imx coda videobuf2_vmalloc
[  130.060915] CPU: 1 PID: 222 Comm: kmscube Not tainted 4.8.4+ #1
[  130.066844] Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree)
[  130.073378] Backtrace: 
[  130.075863] [<8010b6c0>] (dump_backtrace) from [<8010b908>] (show_stack+0x20/0x24)
[  130.083439]  r7:80b441d8 r6:600c0013 r5:00000000 r4:80b441d8
[  130.089187] [<8010b8e8>] (show_stack) from [<803a33fc>] (dump_stack+0x78/0x94)
[  130.096422] [<803a3384>] (dump_stack) from [<8011a9b8>] (__warn+0xdc/0x110)
[  130.103389]  r7:00000009 r6:8042fda4 r5:00000000 r4:ed47fd08
[  130.109132] [<8011a8dc>] (__warn) from [<8011aa34>] (warn_slowpath_fmt+0x48/0x50)
[  130.116620]  r9:00000000 r8:ee1b9418 r7:edc76600 r6:00000000 r5:edffc500 r4:00000000
[  130.124470] [<8011a9f0>] (warn_slowpath_fmt) from [<8042fda4>] (drm_atomic_helper_wait_for_vblanks+0x1e4/0x200)
[  130.134563]  r3:00000018 r2:80898166
[  130.138194] [<8042fbc0>] (drm_atomic_helper_wait_for_vblanks) from [<80457fc8>] (imx_drm_atomic_commit_tail+0x58/0x68)
[  130.148895]  r10:8086686b r9:ee1b923c r8:0000003f r7:80b6bf22 r6:00000000 r5:ee1b9000
[  130.156823]  r4:edffc500
[  130.159390] [<80457f70>] (imx_drm_atomic_commit_tail) from [<804323e8>] (commit_tail+0x4c/0x68)
[  130.168094]  r5:80b4a014 r4:edffc500
[  130.171719] [<8043239c>] (commit_tail) from [<8043249c>] (drm_atomic_helper_commit+0x98/0xb0)
[  130.180249]  r5:00000000 r4:edffc500
[  130.183875] [<80432404>] (drm_atomic_helper_commit) from [<8045810c>] (imx_drm_atomic_commit+0x134/0x144)
[  130.193447]  r7:80b6bf22 r6:edffc800 r5:edffc500 r4:00000006
[  130.199191] [<80457fd8>] (imx_drm_atomic_commit) from [<804556d0>] (drm_atomic_commit+0x60/0x70)
[  130.207981]  r10:00000004 r9:ee1b923c r8:0000003f r7:ee1b9000 r6:edffc500 r5:ee1b9000
[  130.215908]  r4:edffc500
[  130.218475] [<80455670>] (drm_atomic_commit) from [<80435660>] (drm_fb_helper_restore_fbdev_mode_unlocked+0x130/0x29c)
[  130.229175]  r5:eeba9f00 r4:00000000
[  130.232799] [<80435530>] (drm_fb_helper_restore_fbdev_mode_unlocked) from [<80436904>] (drm_fbdev_cma_restore_mode+0x20/0x24)
[  130.244107]  r10:400c0013 r9:ede68f88 r8:ee2053c0 r7:ee1b911c r6:ede68f7c r5:80bc5648
[  130.252033]  r4:ee1b9000
[  130.254598] [<804368e4>] (drm_fbdev_cma_restore_mode) from [<804581e8>] (imx_drm_driver_lastclose+0x20/0x24)
[  130.264439] [<804581c8>] (imx_drm_driver_lastclose) from [<8043a5b0>] (drm_lastclose+0x4c/0xfc)
[  130.273149] [<8043a564>] (drm_lastclose) from [<8043a938>] (drm_release+0x2d8/0x324)
[  130.280898]  r7:ee1b911c r6:ede68f7c r5:ede68f00 r4:ee1b9000
[  130.286642] [<8043a660>] (drm_release) from [<801fa8a8>] (__fput+0xe8/0x1bc)
[  130.293696]  r10:ede069c8 r9:00000008 r8:ee21b190 r7:ee428ee0 r6:00000000 r5:ee197bc0
[  130.301622]  r4:ede069c0
[  130.304183] [<801fa7c0>] (__fput) from [<801fa9ec>] (____fput+0x18/0x1c)
[  130.310890]  r10:00000000 r9:00000000 r8:80107ac4 r7:ed47ff58 r6:edf36a80 r5:80b70eb8
[  130.318817]  r4:ee0bf800
[  130.321381] [<801fa9d4>] (____fput) from [<80134e6c>] (task_work_run+0xc8/0xdc)
[  130.328704] [<80134da4>] (task_work_run) from [<8011caec>] (do_exit+0x438/0x960)
[  130.336105]  r7:ed47ff58 r6:ee0bfc08 r5:eeb88a80 r4:ee0bf800
[  130.341847] [<8011c6b4>] (do_exit) from [<8011e188>] (do_group_exit+0x5c/0xcc)
[  130.349075]  r7:ffffe000
[  130.351638] [<8011e12c>] (do_group_exit) from [<8011e218>] (__wake_up_parent+0x0/0x30)
[  130.359560]  r7:000000f8 r6:76ec6750 r5:00000001 r4:00000001
[  130.365305] [<8011e1f8>] (SyS_exit_group) from [<80107900>] (ret_fast_syscall+0x0/0x3c)
[  130.373357] ---[ end trace 5678aedcbb762e5c ]---
[  140.667136] [drm:drm_atomic_helper_commit_cleanup_done] *ERROR* [CRTC:24:crtc-0] flip_done timed out
[  150.907117] [drm:drm_atomic_helper_commit_cleanup_done] *ERROR* [CRTC:24:crtc-0] flip_done timed out
[  161.147143] [drm:drm_atomic_helper_commit_cleanup_done] *ERROR* [CRTC:24:crtc-0] flip_done timed out

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: imx-drm hang issue with etnaviv (GC3000)

[Lucas Stach]

Am Donnerstag, den 27.10.2016, 19:26 +0200 schrieb Wladimir J. van der
Laan:
> Hello,
> 
> After running kmscube (or another KMS executable) on a i.MX6 QuadPlus (etnaviv,
> GC3000) a few times on I get the below crash in the drm kernel driver.
> This is on a device with LVDS panel. It is always reproducible, although the
> number of invocations needed differs.
> 
> The only way to get rendering to work again after the crash is to reboot.
> Repeated tries only get the "flip_done timed out".
> 
> This always happens while the program is exiting.
> 
> Versions:
> 
> - mesa: https://github.com/etnaviv/mesa 9a09984
> 
> - libdrm: https://cgit.freedesktop.org/mesa/drm/ fe4579e
> 
> - Kernel: 4.8.0 or 4.8.4 + Pengutronix patches (20161007).
> 
> Does anyone have an idea what could be the problem?
> 
I think I've seen this problem a few times already. I'll have a look at
this today.

Regards,
Lucas

> Regards,
> Wladimir van der Laan
> 
> [  130.026973] ------------[ cut here ]------------
> [  130.031630] WARNING: CPU: 1 PID: 222 at drivers/gpu/drm/drm_atomic_helper.c:1127 drm_atomic_helper_wait_for_vblanks+0x1e4/0x200
> [  130.043149] [CRTC:24] vblank wait timed out
> [  130.047367] Modules linked in: hid_generic usbhid hid ci_hdrc_imx ci_hdrc extcon_core ehci_hcd usbcore usb_common usbmisc_imx coda videobuf2_vmalloc
> [  130.060915] CPU: 1 PID: 222 Comm: kmscube Not tainted 4.8.4+ #1
> [  130.066844] Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree)
> [  130.073378] Backtrace: 
> [  130.075863] [<8010b6c0>] (dump_backtrace) from [<8010b908>] (show_stack+0x20/0x24)
> [  130.083439]  r7:80b441d8 r6:600c0013 r5:00000000 r4:80b441d8
> [  130.089187] [<8010b8e8>] (show_stack) from [<803a33fc>] (dump_stack+0x78/0x94)
> [  130.096422] [<803a3384>] (dump_stack) from [<8011a9b8>] (__warn+0xdc/0x110)
> [  130.103389]  r7:00000009 r6:8042fda4 r5:00000000 r4:ed47fd08
> [  130.109132] [<8011a8dc>] (__warn) from [<8011aa34>] (warn_slowpath_fmt+0x48/0x50)
> [  130.116620]  r9:00000000 r8:ee1b9418 r7:edc76600 r6:00000000 r5:edffc500 r4:00000000
> [  130.124470] [<8011a9f0>] (warn_slowpath_fmt) from [<8042fda4>] (drm_atomic_helper_wait_for_vblanks+0x1e4/0x200)
> [  130.134563]  r3:00000018 r2:80898166
> [  130.138194] [<8042fbc0>] (drm_atomic_helper_wait_for_vblanks) from [<80457fc8>] (imx_drm_atomic_commit_tail+0x58/0x68)
> [  130.148895]  r10:8086686b r9:ee1b923c r8:0000003f r7:80b6bf22 r6:00000000 r5:ee1b9000
> [  130.156823]  r4:edffc500
> [  130.159390] [<80457f70>] (imx_drm_atomic_commit_tail) from [<804323e8>] (commit_tail+0x4c/0x68)
> [  130.168094]  r5:80b4a014 r4:edffc500
> [  130.171719] [<8043239c>] (commit_tail) from [<8043249c>] (drm_atomic_helper_commit+0x98/0xb0)
> [  130.180249]  r5:00000000 r4:edffc500
> [  130.183875] [<80432404>] (drm_atomic_helper_commit) from [<8045810c>] (imx_drm_atomic_commit+0x134/0x144)
> [  130.193447]  r7:80b6bf22 r6:edffc800 r5:edffc500 r4:00000006
> [  130.199191] [<80457fd8>] (imx_drm_atomic_commit) from [<804556d0>] (drm_atomic_commit+0x60/0x70)
> [  130.207981]  r10:00000004 r9:ee1b923c r8:0000003f r7:ee1b9000 r6:edffc500 r5:ee1b9000
> [  130.215908]  r4:edffc500
> [  130.218475] [<80455670>] (drm_atomic_commit) from [<80435660>] (drm_fb_helper_restore_fbdev_mode_unlocked+0x130/0x29c)
> [  130.229175]  r5:eeba9f00 r4:00000000
> [  130.232799] [<80435530>] (drm_fb_helper_restore_fbdev_mode_unlocked) from [<80436904>] (drm_fbdev_cma_restore_mode+0x20/0x24)
> [  130.244107]  r10:400c0013 r9:ede68f88 r8:ee2053c0 r7:ee1b911c r6:ede68f7c r5:80bc5648
> [  130.252033]  r4:ee1b9000
> [  130.254598] [<804368e4>] (drm_fbdev_cma_restore_mode) from [<804581e8>] (imx_drm_driver_lastclose+0x20/0x24)
> [  130.264439] [<804581c8>] (imx_drm_driver_lastclose) from [<8043a5b0>] (drm_lastclose+0x4c/0xfc)
> [  130.273149] [<8043a564>] (drm_lastclose) from [<8043a938>] (drm_release+0x2d8/0x324)
> [  130.280898]  r7:ee1b911c r6:ede68f7c r5:ede68f00 r4:ee1b9000
> [  130.286642] [<8043a660>] (drm_release) from [<801fa8a8>] (__fput+0xe8/0x1bc)
> [  130.293696]  r10:ede069c8 r9:00000008 r8:ee21b190 r7:ee428ee0 r6:00000000 r5:ee197bc0
> [  130.301622]  r4:ede069c0
> [  130.304183] [<801fa7c0>] (__fput) from [<801fa9ec>] (____fput+0x18/0x1c)
> [  130.310890]  r10:00000000 r9:00000000 r8:80107ac4 r7:ed47ff58 r6:edf36a80 r5:80b70eb8
> [  130.318817]  r4:ee0bf800
> [  130.321381] [<801fa9d4>] (____fput) from [<80134e6c>] (task_work_run+0xc8/0xdc)
> [  130.328704] [<80134da4>] (task_work_run) from [<8011caec>] (do_exit+0x438/0x960)
> [  130.336105]  r7:ed47ff58 r6:ee0bfc08 r5:eeb88a80 r4:ee0bf800
> [  130.341847] [<8011c6b4>] (do_exit) from [<8011e188>] (do_group_exit+0x5c/0xcc)
> [  130.349075]  r7:ffffe000
> [  130.351638] [<8011e12c>] (do_group_exit) from [<8011e218>] (__wake_up_parent+0x0/0x30)
> [  130.359560]  r7:000000f8 r6:76ec6750 r5:00000001 r4:00000001
> [  130.365305] [<8011e1f8>] (SyS_exit_group) from [<80107900>] (ret_fast_syscall+0x0/0x3c)
> [  130.373357] ---[ end trace 5678aedcbb762e5c ]---
> [  140.667136] [drm:drm_atomic_helper_commit_cleanup_done] *ERROR* [CRTC:24:crtc-0] flip_done timed out
> [  150.907117] [drm:drm_atomic_helper_commit_cleanup_done] *ERROR* [CRTC:24:crtc-0] flip_done timed out
> [  161.147143] [drm:drm_atomic_helper_commit_cleanup_done] *ERROR* [CRTC:24:crtc-0] flip_done timed out
> 


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: imx-drm hang issue with etnaviv (GC3000)

[Lucas Stach]

Am Freitag, den 28.10.2016, 10:18 +0200 schrieb Lucas Stach:
> Am Donnerstag, den 27.10.2016, 19:26 +0200 schrieb Wladimir J. van der
> Laan:
> > Hello,
> > 
> > After running kmscube (or another KMS executable) on a i.MX6 QuadPlus (etnaviv,
> > GC3000) a few times on I get the below crash in the drm kernel driver.
> > This is on a device with LVDS panel. It is always reproducible, although the
> > number of invocations needed differs.
> > 
> > The only way to get rendering to work again after the crash is to reboot.
> > Repeated tries only get the "flip_done timed out".
> > 
> > This always happens while the program is exiting.
> > 
> > Versions:
> > 
> > - mesa: https://github.com/etnaviv/mesa 9a09984
> > 
> > - libdrm: https://cgit.freedesktop.org/mesa/drm/ fe4579e
> > 
> > - Kernel: 4.8.0 or 4.8.4 + Pengutronix patches (20161007).
> > 
> > Does anyone have an idea what could be the problem?
> > 
> I think I've seen this problem a few times already. I'll have a look at
> this today.

And it had to wait a bit, but I was finally able to track down what's
wrong here.

You want patch "[PATCH 1/2] drm/imx: disable planes before DC" to get
rid of this issue.

Regards,
Lucas

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel