Re: [PATCH] floppy: convert to delayed work and single-thread wq

[Jiri Kosina]

On Wed, 16 May 2012, Tejun Heo wrote:

> > > > On Wed, May 16, 2012 at 12:37 PM, Jiri Kosina <jkosina@xxxxxxx> wrote:
> > > > > +       cancel_delayed_work_sync(&fd_timeout);
> > > > > +       cancel_delayed_work_sync(&fd_timer);
> > > > > +       destroy_workqueue(system_nrt_wq);
> > > > 
> > > > Well, *that* doesn't look right.
> > > > 
> > > > I think just a "flush_workqueue()" is in order.
> > > 
> > > System wqs shouldn't be flushed (nothing guarantees that flush will
> > > finish in fixed amount of time).  We probably should make that
> > > explicit by whining when someone tries to flush one of the system wqs.
> > > Here, the two cancel_delayed_work_sync() calls should be enough.
> > 
> > Well, then I think this might be an issue for straightforward conversion 
> > of floppy driver to system_nrt_wq -- see floppy_grab_irq_and_dma().
> 
> Hmmm? flush_work() is fine.  flush_workqueue() isn't.  Am I missing
> something?

In floppy_grab_irq_and_dma() the point is to drain the workqueue 
completely (before the conversion, we were just using 
flush_work_sync(&floppy_work) for particular work item), and for that 
flush_work() is not sufficient any more.

So I am really considering going back to driver-specific singlethreaded 
workqueue.

Thanks,

-- 
Jiri Kosina
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/