linux-kernel - Re: [PATCH] ACPIPHP: fix device destroying order issue in handling dock notification

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 12 Jun 2013 23:05:59 +0600
From:	"Alexander E. Patrakov" <patrakov@...il.com>
To:	Jiang Liu <liuj97@...il.com>
Cc:	Jiang Liu <jiang.liu@...wei.com>,
	Bjorn Helgaas <bhelgaas@...gle.com>,
	Yinghai Lu <yinghai@...nel.org>,
	"Rafael J. Wysocki" <rafael.j.wysocki@...el.com>,
	linux-pci@...r.kernel.org,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] ACPIPHP: fix device destroying order issue in handling
 dock notification

2013/6/12 Alexander E. Patrakov <patrakov@...il.com>:
> 2013/6/12 Jiang Liu <liuj97@...il.com>:
>> On Wed 12 Jun 2013 12:51:59 AM CST, Alexander E. Patrakov wrote:
>>> In the initially-docked case, it exhibits the following problem: when
>>> I press the undock button, only one PCI device disappears, and the
>>> "docked" LED does not turn off. Additionally, there is a hung task.
>> 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'm not familiar with Intel HDA driver,  so please help to fire another
>> bug for it.
>
> Thanks for pointing the finger to snd-hda-intel. With that driver
> blacklisted, the lspci output matches the expectations even after
> undocking the initially-docked laptop. Redocking re-adds the devices,
> too.

> As for snd-hda-intel bug, I will file it later today and let you know.

Actually, I debugged it further and want some input from you before
filing. Here is the new information: the incomplete undocking and hung
task does not appear if I test from a virtual console, not from an
XFCE session. The obvious difference is that in the XFCE session some
processes (the mixer applet and pulseaudio) keep the mixer device
always open.

I am not qualified enough to say what should happen if the user
presses the undock button while a program has a to-be-undocked device
open. Or, for that matter and for an analogy, if the user unplugs a
USB sound card while it is playing.

And in my case, this is going to be an issue not only because of the
snd-hda-intel driver. Xorg would sometimes keep the DRM node of the
Radeon card (that is in the dock station) open in addition to the
onboard Intel card.

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