Re: [PATCH] EDAC: core EDAC support code

[Greg KH <greg@kroah.com>]

On Mon, Mar 06, 2006 at 08:45:59PM -0500, Dmitry Torokhov wrote:
> On Monday 06 March 2006 16:53, Greg KH wrote:
> > On Mon, Mar 06, 2006 at 09:32:03PM +0000, Al Viro wrote:
> > > On Mon, Mar 06, 2006 at 01:01:37PM -0800, Dave Peterson wrote:
> > > > Regarding the above problem with the kobject reference count, this
> > > > was recently fixed in the -mm tree (see edac-kobject-sysfs-fixes.patch
> > > > in 2.6.16-rc5-mm2).  The fix I implemented was to add a call to
> > > > complete() in edac_memctrl_master_release() and then have the module
> > > > cleanup code wait for the completion.  I think there were a few other
> > > > instances of this type of problem that I also fixed in the
> > > > above-mentioned patch.
> > > 
> > > This is not a fix, this is a goddamn deadlock.
> > > 	rmmod your_turd </sys/spew/from/your_turd
> > > and there you go.  rmmod can _NOT_ wait for sysfs references to go away.
> > 
> > To be fair, the only part of the kernel that supports the above process,
> > is the network stack.  And they implemented a special kind of lock to
> > handle just this kind of thing.
> > 
> 
> Not so:
> 
> [root@core ~]# rmmod psmouse < /sys/bus/serio/devices/serio0/rate
> ERROR: Module psmouse is in use
> [root@core ~]# rmmod psmouse
> [root@core ~]# modprobe psmouse
> [root@core ~]# 
> 
> It would be nice if more subsystem could handle this, preferably without
> "Waiting for blah to become free" messages (as in W1).

See my previous apology about this :)

Anyway, the network stack does have a special lock for unloading modules
while they are still "in use", but as long as Al was referring to your
above sequence, I have no disagreement.

thanks,

greg k-h