cvs commit: src/sys/fs/devfs devfs_int.h devfs_vnops.c src/sys/kern kern_conf.c src/sys/sys conf.h

[Kostik Belousov]

On Tue, Jul 03, 2007 at 11:24:41AM -0700, Nate Lawson wrote:
> Konstantin Belousov wrote:
> > kib         2007-07-03 17:42:37 UTC
> > 
> >   FreeBSD src repository
> > 
> >   Modified files:
> >     sys/fs/devfs         devfs_int.h devfs_vnops.c 
> >     sys/kern             kern_conf.c 
> >     sys/sys              conf.h 
> >   Log:
> >   Since rev. 1.199 of sys/kern/kern_conf.c, the thread that calls
> >   destroy_dev() from d_close() cdev method would self-deadlock.
> >   devfs_close() bump device thread reference counter, and destroy_dev()
> >   sleeps, waiting for si_threadcount to reach zero for cdev without
> >   d_purge method.
> >   
> >   destroy_dev_sched() could be used instead from d_close(), to
> >   schedule execution of destroy_dev() in another context. The
> >   destroy_dev_sched_drain() function can be used to drain the scheduled
> >   calls to destroy_dev_sched(). Similarly, drain_dev_clone_events() drains
> >   the events clone to make sure no lingering devices are left after
> >   dev_clone event handler deregistered.
> >   
> >   make_dev_credf(MAKEDEV_REF) function should be used from dev_clone
> >   event handlers instead of make_dev()/make_dev_cred() to ensure that created
> >   device has reference counter bumped before cdev mutex is dropped inside
> >   make_dev().
> >   
> >   Reviewed by:    tegge (early versions), njl (programming interface)
> >   Debugging help and testing by:  Peter Holm
> >   Approved by:    re (kensmith)
> >   
> >   Revision  Changes    Path
> >   1.4       +6 -0      src/sys/fs/devfs/devfs_int.h
> >   1.148     +13 -0     src/sys/fs/devfs/devfs_vnops.c
> >   1.204     +147 -18   src/sys/kern/kern_conf.c
> >   1.233     +10 -0     src/sys/sys/conf.h
> 
> Thanks for all your work on this.  Now to review all callers of
> destroy_dev() to see if they should use the simpler interface.

I committed the version of destroy_dev() that shall support
autodetection of deadlock (hopefully, got it right second time).
It is done in separate commit to ease the reversal.

See my HEADS UP on current at .
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20070703/c140bd18/attachment.pgp