On 20/01/05 12:09 -0800, Greg KH wrote:
> On Thu, Jan 20, 2005 at 08:04:13PM +0000, Matthew Wilcox wrote:
> > On Thu, Jan 20, 2005 at 08:50:47AM -0800, Greg KH wrote:
> > > On Thu, Jan 20, 2005 at 07:44:24AM -0800, Randy.Dunlap wrote:
> > > > Matthew Wilcox wrote:
> > > > >I think this is referring to initialisation of spinlocks that are
> > > > >allocated dynamically, not statically.  If a lock validator can't cope
> > > > >with that, it needs to be fixed, IMO.
> > > 
> > > Yes, it is.  See the number of patches that have been flowing into the
> > > kernel lately to fix this issue up.
> > 
> > However, the correct way to fix this is to use DEFINE_SPIN_LOCK, not to
> > make static locks dynamically initialised.
> 
> Where ever possible, yes.

Is "static DEFINE_SPINLOCK(lock);" (which seems right in this case) ok?


	Domen