Re: [PATCH 1/2] drm/i2c/tda998x: Switch to atomic operations

[Fabio Estevam <festevam@gmail.com>]

Hi Laurent,

On Mon, Jan 3, 2022 at 8:48 AM Laurent Pinchart
<laurent.pinchart@ideasonboard.com> wrote:

> With the comment from 2/2 taken into account,
>
> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Thanks for the review. I addressed your feedback and sent v2.

I noticed a problem when removing/inserting the HDMI cable.

If I boot the board with the HDMI cable connected, then after
removal/insertion of the HDMI cable, the following
kernel warning is observed:

# [   23.201080] ------------[ cut here ]------------
[   23.207275] WARNING: CPU: 0 PID: 56 at
drivers/gpu/drm/drm_atomic_helper.c:1514
drm_atomic_helper_wait_for_vblanks.part.0+0x27c/0x294
[   23.221469] [CRTC:35:crtc-0] vblank wait timed out
[   23.226448] Modules linked in:
[   23.230255] CPU: 0 PID: 56 Comm: kworker/0:3 Not tainted
5.15.12-00003-g27f29fb60028 #94
[   23.238508] Hardware name: Freescale i.MX6 SoloX (Device Tree)
[   23.244457] Workqueue: events output_poll_execute
[   23.249377] [<c01118f8>] (unwind_backtrace) from [<c010bde0>]
(show_stack+0x10/0x14)
[   23.257316] [<c010bde0>] (show_stack) from [<c0e8d38c>]
(dump_stack_lvl+0x58/0x70)
[   23.265059] [<c0e8d38c>] (dump_stack_lvl) from [<c0125e2c>]
(__warn+0xd8/0x114)
[   23.272533] [<c0125e2c>] (__warn) from [<c0e87f94>]
(warn_slowpath_fmt+0x90/0xc4)
[   23.280166] [<c0e87f94>] (warn_slowpath_fmt) from [<c0707c98>]
(drm_atomic_helper_wait_for_vblanks.part.0+0x27c/0x294)
[   23.291054] [<c0707c98>]
(drm_atomic_helper_wait_for_vblanks.part.0) from [<c07092cc>]
(drm_atomic_helper_commit_tail_rpm+0x5c/0x6c)
[   23.303150] [<c07092cc>] (drm_atomic_helper_commit_tail_rpm) from
[<c0709e44>] (commit_tail+0x9c/0x190)
[   23.312717] [<c0709e44>] (commit_tail) from [<c070a098>]
(drm_atomic_helper_commit+0x158/0x18c)
[   23.321588] [<c070a098>] (drm_atomic_helper_commit) from
[<c073f4e4>] (drm_client_modeset_commit_atomic+0x238/0x284)
[   23.332314] [<c073f4e4>] (drm_client_modeset_commit_atomic) from
[<c073f600>] (drm_client_modeset_commit_locked+0x60/0x1cc)
[   23.343615] [<c073f600>] (drm_client_modeset_commit_locked) from
[<c073f790>] (drm_client_modeset_commit+0x24/0x40)
[   23.354218] [<c073f790>] (drm_client_modeset_commit) from
[<c071050c>] (__drm_fb_helper_restore_fbdev_mode_unlocked+0x9c/0xc8)
[   23.365803] [<c071050c>]
(__drm_fb_helper_restore_fbdev_mode_unlocked) from [<c07105a0>]
(drm_fb_helper_set_par+0x38/0x68)
[   23.377015] [<c07105a0>] (drm_fb_helper_set_par) from [<c0710674>]
(drm_fb_helper_hotplug_event.part.0+0xa4/0xc0)
[   23.387443] [<c0710674>] (drm_fb_helper_hotplug_event.part.0) from
[<c073ea80>] (drm_client_dev_hotplug+0x6c/0xb4)
[   23.397959] [<c073ea80>] (drm_client_dev_hotplug) from [<c06fc5e0>]
(output_poll_execute+0x200/0x21c)
[   23.407346] [<c06fc5e0>] (output_poll_execute) from [<c0147f54>]
(process_one_work+0x298/0x7cc)
[   23.416224] [<c0147f54>] (process_one_work) from [<c01484b8>]
(worker_thread+0x30/0x50c)
[   23.424479] [<c01484b8>] (worker_thread) from [<c015138c>]
(kthread+0x154/0x17c)
[   23.432039] [<c015138c>] (kthread) from [<c010011c>]
(ret_from_fork+0x14/0x38)
[   23.439413] Exception stack(0xc42a1fb0 to 0xc42a1ff8)
[   23.444588] 1fa0:                                     00000000
00000000 00000000 00000000
[   23.452888] 1fc0: 00000000 00000000 00000000 00000000 00000000
00000000 00000000 00000000
[   23.461182] 1fe0: 00000000 00000000 00000000 00000000 00000013 00000000
[   23.468734] irq event stamp: 43775
[   23.472305] hardirqs last  enabled at (43783): [<c0197408>]
__up_console_sem+0x50/0x60
[   23.480785] hardirqs last disabled at (43792): [<c01973f4>]
__up_console_sem+0x3c/0x60
[   23.489224] softirqs last  enabled at (43774): [<c010150c>]
__do_softirq+0x2ec/0x5a4
[   23.497163] softirqs last disabled at (43747): [<c012efdc>]
irq_exit+0x18c/0x210
[   23.505106] ---[ end trace 86572327287ca501 ]---

I haven't managed to fix this yet, but if you have any suggestions,
please let me know.

Thanks,

Fabio Estevam