[ldv-project] [PATCH] input: pcap2: avoid calling mutex_lock() in irq handler

[Dmitry Torokhov]

On Sun, Apr 06, 2014 at 11:37:44PM +0200, Antonio Ospite wrote:
> On Sun, 6 Apr 2014 13:24:36 -0700
> Dmitry Torokhov <dmitry.torokhov at gmail.com> wrote:
> 
> > On Sun, Apr 06, 2014 at 12:54:50AM +0400, Alexey Khoroshilov wrote:
> > > pcap_keys_handler() calls ezx_pcap_read() that calls mutex_lock().
> > > pcap_keys_handler() is registered as nonthreaded irq handler,
> > > that means sleeping function is called in irq handler.
> > > 
> > > The patch makes a switch to threaded irq handling.
> > > Compile tested only.
> > > 
> > > Found by Linux Driver Verification project (linuxtesting.org).
> > > 
> > > Signed-off-by: Alexey Khoroshilov <khoroshilov at ispras.ru>
> > > ---
> > >  drivers/input/misc/pcap_keys.c | 10 ++++++----
> > >  1 file changed, 6 insertions(+), 4 deletions(-)
> > > 
> > > diff --git a/drivers/input/misc/pcap_keys.c b/drivers/input/misc/pcap_keys.c
> > > index cd230365166e..2a10f3a30969 100644
> > > --- a/drivers/input/misc/pcap_keys.c
> > > +++ b/drivers/input/misc/pcap_keys.c
> > > @@ -79,13 +79,15 @@ static int pcap_keys_probe(struct platform_device *pdev)
> > >  	if (err)
> > >  		goto fail_allocate;
> > >  
> > > -	err = request_irq(pcap_to_irq(pcap_keys->pcap, PCAP_IRQ_ONOFF),
> > > -			pcap_keys_handler, 0, "Power key", pcap_keys);
> > > +	err = request_threaded_irq(pcap_to_irq(pcap_keys->pcap, PCAP_IRQ_ONOFF),
> > > +				   NULL, pcap_keys_handler, 0,
> > > +				   "Power key", pcap_keys);
> > >  	if (err)
> > >  		goto fail_register;
> > >  
> > > -	err = request_irq(pcap_to_irq(pcap_keys->pcap, PCAP_IRQ_MIC),
> > > -			pcap_keys_handler, 0, "Headphone button", pcap_keys);
> > > +	err = request_threaded_irq(pcap_to_irq(pcap_keys->pcap, PCAP_IRQ_MIC),
> > > +				   NULL, pcap_keys_handler, 0,
> > > +				   "Headphone button", pcap_keys);
> > 
> > So I guess we do need threaded IRQ here, but I do not see the pcap's
> > irqchip specifying IRQCHIP_ONESHOT_SAFE so I do not think
> > request_threaded_irq() without IRQF_ONESHOT would succeed.
> > 
> > Can someone test the change to be sure?
> > 
> 
> I can try to test it in a couple of weeks, I don't have the hardware
> at hand right now.

Antonio, did you have a chance to try this out yet?

Thanks!

-- 
Dmitry