All of lore.kernel.org
 help / color / mirror / Atom feed
* USB devices should not oops after disconnect now
@ 2002-11-25 18:15 Jaroslav Kysela
  2002-11-27  4:34 ` Patrick Shirkey
  2002-11-28 11:50 ` Takashi Iwai
  0 siblings, 2 replies; 6+ messages in thread
From: Jaroslav Kysela @ 2002-11-25 18:15 UTC (permalink / raw)
  To: ALSA development

Hello all,

	the subject says everything. CVS contains first code to avoid 
oops / systems hangs when a "running" hotplug device is removed from the 
system. Please, test and report problems.

						Jaroslav

-----
Jaroslav Kysela <perex@suse.cz>
Linux Kernel Sound Maintainer
ALSA Project, SuSE Labs



-------------------------------------------------------
This SF.net email is sponsored by: Get the new Palm Tungsten T 
handheld. Power & Color in a compact size! 
http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: USB devices should not oops after disconnect now
  2002-11-25 18:15 USB devices should not oops after disconnect now Jaroslav Kysela
@ 2002-11-27  4:34 ` Patrick Shirkey
  2002-11-27  9:33   ` Jaroslav Kysela
  2002-11-28 11:50 ` Takashi Iwai
  1 sibling, 1 reply; 6+ messages in thread
From: Patrick Shirkey @ 2002-11-27  4:34 UTC (permalink / raw)
  To: Jaroslav Kysela; +Cc: ALSA development

Jaroslav Kysela wrote:
> Hello all,
> 
> 	the subject says everything. CVS contains first code to avoid 
> oops / systems hangs when a "running" hotplug device is removed from the 
> system. Please, test and report problems.
> 

I have updated and now when I disconnect the device it will not register 
on reconnect. No "connection found" messages are written to dmesg an it 
doesn't appear in /proc/asound/cards. However it doesn't hang or cause 
an oops either :) This is with the maudio quattro.



-- 
Patrick Shirkey - Boost Hardware Ltd.
For the discerning hardware connoisseur
Http://www.boosthardware.com
Http://www.djcj.org - The Linux Audio Users guide
========================================

Being on stage with the band in front of crowds shouting, "Get off! No! 
We want normal music!", I think that was more like acting than anything 
I've ever done.

Goldie, 8 Nov, 2002
The Scotsman



-------------------------------------------------------
This SF.net email is sponsored by: Get the new Palm Tungsten T 
handheld. Power & Color in a compact size! 
http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: USB devices should not oops after disconnect now
  2002-11-27  4:34 ` Patrick Shirkey
@ 2002-11-27  9:33   ` Jaroslav Kysela
  2002-11-27 10:42     ` Patrick Shirkey
  0 siblings, 1 reply; 6+ messages in thread
From: Jaroslav Kysela @ 2002-11-27  9:33 UTC (permalink / raw)
  To: Patrick Shirkey; +Cc: ALSA development

On Wed, 27 Nov 2002, Patrick Shirkey wrote:

> Jaroslav Kysela wrote:
> > Hello all,
> > 
> > 	the subject says everything. CVS contains first code to avoid 
> > oops / systems hangs when a "running" hotplug device is removed from the 
> > system. Please, test and report problems.
> > 
> 
> I have updated and now when I disconnect the device it will not register 
> on reconnect. No "connection found" messages are written to dmesg an it 
> doesn't appear in /proc/asound/cards. However it doesn't hang or cause 
> an oops either :) This is with the maudio quattro.

Note that you'll have to terminate all applications using that soundcard
before you'll try to reconnect the device again. In next code, hopefully
the applications will be somehow terminated to make the cleanup process
automatic.

						Jaroslav

-----
Jaroslav Kysela <perex@suse.cz>
Linux Kernel Sound Maintainer
ALSA Project, SuSE Labs



-------------------------------------------------------
This SF.net email is sponsored by: Get the new Palm Tungsten T 
handheld. Power & Color in a compact size! 
http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: USB devices should not oops after disconnect now
  2002-11-27  9:33   ` Jaroslav Kysela
@ 2002-11-27 10:42     ` Patrick Shirkey
  2002-11-27 11:06       ` Jaroslav Kysela
  0 siblings, 1 reply; 6+ messages in thread
From: Patrick Shirkey @ 2002-11-27 10:42 UTC (permalink / raw)
  To: Jaroslav Kysela; +Cc: ALSA development

Jaroslav Kysela wrote:
> 
> Note that you'll have to terminate all applications using that soundcard
> before you'll try to reconnect the device again. In next code, hopefully
> the applications will be somehow terminated to make the cleanup process
> automatic.
> 


Ok. That's the probable cause as I am having problems with jack not 
releasing the device.

It seems that hotpluging out is a tricky issue because we have to decide 
whether the application/s that are using the device should also be 
terminated.

Is it possible to transparently swap snd-usb-audio with snd-dummy if it 
has been installed? That could be a secondary default if people would 
prefer their apps do not automatically segv.




-- 
Patrick Shirkey - Boost Hardware Ltd.
For the discerning hardware connoisseur
Http://www.boosthardware.com
Http://www.djcj.org - The Linux Audio Users guide
========================================

Being on stage with the band in front of crowds shouting, "Get off! No! 
We want normal music!", I think that was more like acting than anything 
I've ever done.

Goldie, 8 Nov, 2002
The Scotsman



-------------------------------------------------------
This SF.net email is sponsored by: Get the new Palm Tungsten T 
handheld. Power & Color in a compact size! 
http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: USB devices should not oops after disconnect now
  2002-11-27 10:42     ` Patrick Shirkey
@ 2002-11-27 11:06       ` Jaroslav Kysela
  0 siblings, 0 replies; 6+ messages in thread
From: Jaroslav Kysela @ 2002-11-27 11:06 UTC (permalink / raw)
  To: Patrick Shirkey; +Cc: ALSA development

On Wed, 27 Nov 2002, Patrick Shirkey wrote:

> Jaroslav Kysela wrote:
> > 
> > Note that you'll have to terminate all applications using that soundcard
> > before you'll try to reconnect the device again. In next code, hopefully
> > the applications will be somehow terminated to make the cleanup process
> > automatic.
> > 
> 
> 
> Ok. That's the probable cause as I am having problems with jack not 
> releasing the device.
> 
> It seems that hotpluging out is a tricky issue because we have to decide 
> whether the application/s that are using the device should also be 
> terminated.
> 
> Is it possible to transparently swap snd-usb-audio with snd-dummy if it 
> has been installed? That could be a secondary default if people would 

It's too complicated and probably an overdesign. Also, what to do, if
another hardware is plugged to the system than previous one? The full
handshake between driver and application have to be triggered all times
(old "disconnected" file descriptors have to be closed).

> prefer their apps do not automatically segv.

We have two choices:

1) terminate all applications using a signal (SIG_KILL or SIG_TERM)
2) disable all operations with a file descriptor and give an error code
   to applications when disconnect occured so they can decide what to 
   do - terminate - or wait until hardware is not available again - (this
   is my goal)

						Jaroslav

-----
Jaroslav Kysela <perex@suse.cz>
Linux Kernel Sound Maintainer
ALSA Project, SuSE Labs



-------------------------------------------------------
This SF.net email is sponsored by: Get the new Palm Tungsten T 
handheld. Power & Color in a compact size! 
http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: USB devices should not oops after disconnect now
  2002-11-25 18:15 USB devices should not oops after disconnect now Jaroslav Kysela
  2002-11-27  4:34 ` Patrick Shirkey
@ 2002-11-28 11:50 ` Takashi Iwai
  1 sibling, 0 replies; 6+ messages in thread
From: Takashi Iwai @ 2002-11-28 11:50 UTC (permalink / raw)
  To: Jaroslav Kysela; +Cc: ALSA development

Hi Jaroslav,

i'm back from vacation now.

At Mon, 25 Nov 2002 19:15:41 +0100 (CET),
Jaroslav wrote:
> 
> Hello all,
> 
> 	the subject says everything. CVS contains first code to avoid 
> oops / systems hangs when a "running" hotplug device is removed from the 
> system. Please, test and report problems.

tried now and still got oops when OSS mixer is being used and
unplugged.

we need a globak lock and "stale" flag in the card, so that every 
device can check the activity of the card before actually accessing to
it.

later on, i'll change the vx driver for your new disconnect stuff.

so far, i'm too busy to read a pile of mails :)


ciao,

Takashi		(... still 12,000 mails left)


-------------------------------------------------------
This SF.net email is sponsored by: Get the new Palm Tungsten T 
handheld. Power & Color in a compact size! 
http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2002-11-29 20:50 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-11-25 18:15 USB devices should not oops after disconnect now Jaroslav Kysela
2002-11-27  4:34 ` Patrick Shirkey
2002-11-27  9:33   ` Jaroslav Kysela
2002-11-27 10:42     ` Patrick Shirkey
2002-11-27 11:06       ` Jaroslav Kysela
2002-11-28 11:50 ` Takashi Iwai

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.