linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* devtmpfs [censored] oddities
@ 2009-12-16  8:04 Al Viro
  2009-12-17  0:39 ` Greg KH
  0 siblings, 1 reply; 4+ messages in thread
From: Al Viro @ 2009-12-16  8:04 UTC (permalink / raw)
  To: linux-kernel; +Cc: Kay Sievers, Greg Kroah-Hartman

	May I respectfully suggest that a blocking operation (such as
kstrdup with GFP_KERNEL, or grabbing a mutex, or, say it, pathname resolution)
is not quite the thing to do while holding an rwlock?

	As it is, any device_add() is an embarrassingly obvious deadlock
waiting to happen...

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: devtmpfs [censored] oddities
  2009-12-16  8:04 devtmpfs [censored] oddities Al Viro
@ 2009-12-17  0:39 ` Greg KH
  2009-12-17  0:46   ` Al Viro
  0 siblings, 1 reply; 4+ messages in thread
From: Greg KH @ 2009-12-17  0:39 UTC (permalink / raw)
  To: Al Viro; +Cc: linux-kernel, Kay Sievers

On Wed, Dec 16, 2009 at 08:04:31AM +0000, Al Viro wrote:
> 	May I respectfully suggest that a blocking operation (such as
> kstrdup with GFP_KERNEL, or grabbing a mutex, or, say it, pathname resolution)
> is not quite the thing to do while holding an rwlock?
> 
> 	As it is, any device_add() is an embarrassingly obvious deadlock
> waiting to happen...

Thomas has posted a patch to fix this now.

Sorry for not catching it sooner, we should just delete the rwlocks so
no one tries to ever use it again.

thanks,

greg k-h

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: devtmpfs [censored] oddities
  2009-12-17  0:39 ` Greg KH
@ 2009-12-17  0:46   ` Al Viro
  2009-12-17  1:00     ` Greg KH
  0 siblings, 1 reply; 4+ messages in thread
From: Al Viro @ 2009-12-17  0:46 UTC (permalink / raw)
  To: Greg KH; +Cc: linux-kernel, Kay Sievers

On Wed, Dec 16, 2009 at 04:39:22PM -0800, Greg KH wrote:
> On Wed, Dec 16, 2009 at 08:04:31AM +0000, Al Viro wrote:
> > 	May I respectfully suggest that a blocking operation (such as
> > kstrdup with GFP_KERNEL, or grabbing a mutex, or, say it, pathname resolution)
> > is not quite the thing to do while holding an rwlock?
> > 
> > 	As it is, any device_add() is an embarrassingly obvious deadlock
> > waiting to happen...
> 
> Thomas has posted a patch to fix this now.
> 
> Sorry for not catching it sooner, we should just delete the rwlocks so
> no one tries to ever use it again.

Say again?  Why would we delete rwlocks?

-- 
If you stare into drivers/staging long enough, drivers/staging stares back at
you...

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: devtmpfs [censored] oddities
  2009-12-17  0:46   ` Al Viro
@ 2009-12-17  1:00     ` Greg KH
  0 siblings, 0 replies; 4+ messages in thread
From: Greg KH @ 2009-12-17  1:00 UTC (permalink / raw)
  To: Al Viro; +Cc: linux-kernel, Kay Sievers

On Thu, Dec 17, 2009 at 12:46:43AM +0000, Al Viro wrote:
> On Wed, Dec 16, 2009 at 04:39:22PM -0800, Greg KH wrote:
> > On Wed, Dec 16, 2009 at 08:04:31AM +0000, Al Viro wrote:
> > > 	May I respectfully suggest that a blocking operation (such as
> > > kstrdup with GFP_KERNEL, or grabbing a mutex, or, say it, pathname resolution)
> > > is not quite the thing to do while holding an rwlock?
> > > 
> > > 	As it is, any device_add() is an embarrassingly obvious deadlock
> > > waiting to happen...
> > 
> > Thomas has posted a patch to fix this now.
> > 
> > Sorry for not catching it sooner, we should just delete the rwlocks so
> > no one tries to ever use it again.
> 
> Say again?  Why would we delete rwlocks?

Sorry, no _new_ use of rwlocks should be added, generally they are used
incorrectly, like this one was.

thanks,

greg k-h

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2009-12-17  1:01 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-12-16  8:04 devtmpfs [censored] oddities Al Viro
2009-12-17  0:39 ` Greg KH
2009-12-17  0:46   ` Al Viro
2009-12-17  1:00     ` Greg KH

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).