[ofa-general] Re: [GIT PULL] please pull infiniband.git for-linus

[Dave Dillow]

On Thu, 2008-01-03 at 18:11 -0500, Rik van Riel wrote:
> On Thu, 03 Jan 2008 15:20:09 -0500
> David Dillow <dillowda at ornl.gov> wrote:
> 
> > diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
> > index 950228f..6e7e3c8 100644
> > --- a/drivers/infiniband/ulp/srp/ib_srp.c
> > +++ b/drivers/infiniband/ulp/srp/ib_srp.c
> > @@ -423,8 +423,8 @@ static void srp_remove_work(struct work_struct *work)
> >  	list_del(&target->list);
> >  	spin_unlock(&target->srp_host->target_lock);
> >  
> > -	srp_remove_host(target->scsi_host);
> >  	scsi_remove_host(target->scsi_host);
> > +	srp_remove_host(target->scsi_host);
> >  	ib_destroy_cm_id(target->cm_id);
> >  	srp_free_target_ib(target);
> >  	scsi_host_put(target->scsi_host);
> 
> These last two look suspicious.  Are you freeing target before
> freeing target->scsi_host or does the code simply not do what
> it looks like it's doing? :)
> 
> (no, I haven't looked at the IB code - I'm probably wrong)

srp_free_target_ib() just frees the buffers for the target, and
scsi_host_put() does the actual cleanup once the refcount drops to zero.

Dave