]> git.baikalelectronics.ru Git - kernel.git/commit
Revert "ALSA: usb-audio: Reduce latency at playback start"
authorLinus Torvalds <torvalds@linux-foundation.org>
Sat, 3 Jul 2021 05:05:52 +0000 (22:05 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Sat, 3 Jul 2021 05:05:52 +0000 (22:05 -0700)
commit0706b2cbe0df0c9f04b921bae84e5ab4d3e90b98
treee2bcfe89b46608d89972d539d27958705350085c
parenta95bcf6fce8686bd59d560ac0cf41c20bd62b9a9
Revert "ALSA: usb-audio: Reduce latency at playback start"

This reverts commit 9d9ea7e1fdaf030f20f95331fa99a0cd3e3017ba.

This commit causes watchdog lockups on my machine, and while I have no
idea what the cause is, it bisected right to this commit, and reverting
the change promptly fixes it.

At least occasionally one of the watchdog call traces was

  Call Trace:
    _raw_spin_lock_irqsave+0x35/0x40
    snd_pcm_period_elapsed+0x1b/0xa0 [snd_pcm]
    snd_usb_endpoint_start+0x1a0/0x3c0 [snd_usb_audio]
    start_endpoints+0x23/0x90 [snd_usb_audio]
    snd_usb_substream_playback_trigger+0x7b/0x1a0 [snd_usb_audio]
    snd_pcm_common_ioctl+0x1c44/0x2360 [snd_pcm]
    snd_pcm_ioctl+0x2e/0x40 [snd_pcm]
    __se_sys_ioctl+0x72/0xc0
    do_syscall_64+0x4c/0xa0
    entry_SYSCALL_64_after_hwframe+0x44/0xae

so presumably it's a locking error on that substream spinlock that
snd_pcm_period_elapsed() takes.  But at this point I just want to have a
working system so that I can continue the merge window work tomorrow.

Cc: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
sound/usb/pcm.c