Re: [PATCH] dmaengine: imx-sdma: fix deadlock in interrupt handler

[Sascha Hauer]

On Mon, Sep 25, 2023 at 11:20:04AM +0200, Uwe Kleine-König wrote:
> Hello,
> 
> On Thu, Sep 21, 2023 at 09:57:11AM +0000, Tim van der Staaij | Zign wrote:
> > dev_warn internally acquires the lock that is already held when
> > sdma_update_channel_loop is called. Therefore it is acquired twice and
> > this is detected as a deadlock. Temporarily release the lock while
> > logging to avoid this.
> > 
> > Signed-off-by: Tim van der Staaij <tim.vanderstaaij@xxxxxxxxxxxxx>
> > Link: https://lore.kernel.org/all/AM0PR08MB308979EC3A8A53AE6E2D3408802CA@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/
> > ---
> >  drivers/dma/imx-sdma.c | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
> > index 51012bd39900..3a7cd783a567 100644
> > --- a/drivers/dma/imx-sdma.c
> > +++ b/drivers/dma/imx-sdma.c
> > @@ -904,7 +904,10 @@ static void sdma_update_channel_loop(struct sdma_channel *sdmac)
> >  	 * owned buffer is available (i.e. BD_DONE was set too late).
> >  	 */
> >  	if (sdmac->desc && !is_sdma_channel_enabled(sdmac->sdma, sdmac->channel)) {
> > +		spin_unlock(&sdmac->vc.lock);
> >  		dev_warn(sdmac->sdma->dev, "restart cyclic channel %d\n", sdmac->channel);
> > +		spin_lock(&sdmac->vc.lock);
> > +
> 
> I don't know if Sascha's patch helps

I do ;)

I inserted a dev_info() into the imx-sdma driver and got said circular
locking warning. With my patch applied it's gone. Nevertheless I would
wait for Tims feedback and resend it with some more people on Cc. I
never used lockdep_set_subclass() and I am not sure if it's legal to
put it into the UART startup function like I did.

Sascha

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |