Re: [PATCH] ACPIPHP: fix device destroying order issue in handlingdock notification

From: Alexander E. Patrakov
Date: Wed Jun 12 2013 - 14:37:33 EST


2013/6/12 Jiang Liu <liuj97@xxxxxxxxx>:
> Hi Alexander,
> In the initially-docked case, the failure is caused by an issue in
> the intel sound card driver. Seems something is wrong with reference
> count management and it never returns to zero on driver detach.
> Could you please help to disable the Intel sound card driver and try
> again?

I have discussed the problem with pulseaudio developers. According to
my understanding (that may be wrong), we have a classic ABBA deadlock.
The device will go away when its reference count goes to zero. The
reference count will go to zero when there are no open fds. The mixer
fd will be closed when it gets a -EIO, i.e. after the device goes
away.

If the above understanding is correct, I think that waiting for zero
references should be at least questioned.

--
Alexander E. Patrakov
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/