Re: [patch 0/4] DEBUG_RCU_HEAD: Debug and fix racy call_rcu() users

[Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>]

* Paul E. McKenney (paulmck@linux.vnet.ibm.com) wrote:
> On Tue, Oct 06, 2009 at 10:37:28AM -0400, Mathieu Desnoyers wrote:
> > Here is a patchset, done on 2.6.30.9, which permits to detect and fix racy
> > call_rcu() users.
> > 
> > Fix for vunmap, which happens to be one of them, is included.
> > 
> > Note that some false-positives might come up when call_rcu() is called on
> > uninitialized struct rcu_head. The solution is, surprisingly enough, to
> > initialize them.
> 
> Very cool, Mathieu!!!
> 
> I do like adding the debug-only field to struct rcu_head.  Lots of
> additional places need initialization, or am I missing a trick here?
> 

In my case, only 2 sites in the process management and 1 site in the now
deprecated marker.c complained about not being initialized.

I can guess we will find others though. Please see the v2 of my patch,
which adds the supplementary "debug" field rather than depending on
word-aligned pointers. I volountarily complain about the .debug pointer
being non-NULL when we find an uninitialized site. I could be more
specific and complain "differently" if the pointer happens to be random
vs if it's LIST_POISON1.

Thanks,

Mathieu

> 							Thanx, Paul

-- 
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68