RE: [PATCH v4 6/7] ALSA: aloop: Support selection of snd_timer instead of jiffies

[Andrew Gabbasov <andrew_gabbasov@mentor.com>]

Hello Takashi,

> -----Original Message-----
> From: Takashi Iwai [mailto:tiwai@suse.de]
> Sent: Wednesday, November 20, 2019 6:32 PM
> To: Gabbasov, Andrew
> Cc: alsa-devel@alsa-project.org; linux-kernel@vger.kernel.org; Jaroslav
> Kysela; Takashi Iwai; Timo Wischer
> Subject: Re: [PATCH v4 6/7] ALSA: aloop: Support selection of snd_timer
> instead of jiffies
> 
> On Wed, 20 Nov 2019 16:21:36 +0100,
> Andrew Gabbasov wrote:
> >
> > Hello Takashi,
> >
> > > -----Original Message-----
> > > From: Takashi Iwai [mailto:tiwai@suse.de]
> > > Sent: Wednesday, November 20, 2019 5:34 PM
> > > To: Gabbasov, Andrew
> > > Cc: alsa-devel@alsa-project.org; linux-kernel@vger.kernel.org;
Jaroslav
> > > Kysela; Takashi Iwai; Timo Wischer
> > > Subject: Re: [PATCH v4 6/7] ALSA: aloop: Support selection of
snd_timer
> > > instead of jiffies
> > >
> > > On Wed, 20 Nov 2019 12:58:55 +0100,
> > > Andrew Gabbasov wrote:
> > > > +/* call in loopback->cable_lock */
> > > > +static int loopback_snd_timer_open(struct loopback_pcm *dpcm)
> > > > +{
> > > > +	int err = 0;
> > > > +	struct snd_timer_id tid = {
> > > > +		.dev_class = SNDRV_TIMER_CLASS_PCM,
> > > > +		.dev_sclass = SNDRV_TIMER_SCLASS_APPLICATION,
> > > > +	};
> > > > +	struct snd_timer_instance *timeri;
> > > > +	struct loopback_cable *cable = dpcm->cable;
> > > > +
> > > > +	spin_lock_irq(&cable->lock);
> > > > +
> > > > +	/* check if timer was already opened. It is only opened once
> > > > +	 * per playback and capture subdevice (aka cable).
> > > > +	 */
> > > > +	if (cable->snd_timer.instance)
> > > > +		goto unlock;
> > > > +
> > > > +	err = loopback_parse_timer_id(dpcm->loopback->timer_source,
> &tid);
> > > > +	if (err < 0) {
> > > > +		pcm_err(dpcm->substream->pcm,
> > > > +			"Parsing timer source \'%s\' failed with
%d",
> > > > +			dpcm->loopback->timer_source, err);
> > > > +		goto unlock;
> > > > +	}
> > > > +
> > > > +	cable->snd_timer.stream = dpcm->substream->stream;
> > > > +	cable->snd_timer.id = tid;
> > > > +
> > > > +	timeri = snd_timer_instance_new(dpcm->loopback->card->id);
> > > > +	if (!timeri) {
> > > > +		err = -ENOMEM;
> > > > +		goto unlock;
> > > > +	}
> > > > +	/* The callback has to be called from another tasklet. If
> > > > +	 * SNDRV_TIMER_IFLG_FAST is specified it will be called from
> the
> > > > +	 * snd_pcm_period_elapsed() call of the selected sound card.
> > > > +	 * snd_pcm_period_elapsed() helds
> snd_pcm_stream_lock_irqsave().
> > > > +	 * Due to our callback loopback_snd_timer_function() also
> calls
> > > > +	 * snd_pcm_period_elapsed() which calls
> > > snd_pcm_stream_lock_irqsave().
> > > > +	 * This would end up in a dead lock.
> > > > +	 */
> > > > +	timeri->flags |= SNDRV_TIMER_IFLG_AUTO;
> > > > +	timeri->callback = loopback_snd_timer_function;
> > > > +	timeri->callback_data = (void *)cable;
> > > > +	timeri->ccallback = loopback_snd_timer_event;
> > > > +
> > > > +	/* snd_timer_close() and snd_timer_open() should not be
called
> with
> > > > +	 * locked spinlock because both functions can block on a
> mutex. The
> > > > +	 * mutex loopback->cable_lock is kept locked. Therefore
> > > snd_timer_open()
> > > > +	 * cannot be called a second time by the other device of the
> same
> > > cable.
> > > > +	 * Therefore the following issue cannot happen:
> > > > +	 * [proc1] Call loopback_timer_open() ->
> > > > +	 *	   Unlock cable->lock for snd_timer_close/open()
call
> > > > +	 * [proc2] Call loopback_timer_open() -> snd_timer_open(),
> > > > +	 *	   snd_timer_start()
> > > > +	 * [proc1] Call snd_timer_open() and overwrite running timer
> > > > +	 *	   instance
> > > > +	 */
> > > > +	spin_unlock_irq(&cable->lock);
> > > > +	err = snd_timer_open(timeri, &cable->snd_timer.id, current-
> >pid);
> > > > +	if (err < 0) {
> > > > +		pcm_err(dpcm->substream->pcm,
> > > > +			"snd_timer_open (%d,%d,%d) failed with %d",
> > > > +			cable->snd_timer.id.card,
> > > > +			cable->snd_timer.id.device,
> > > > +			cable->snd_timer.id.subdevice,
> > > > +			err);
> > > > +		snd_timer_instance_free(timeri);
> > > > +		return err;
> > > > +	}
> > > > +	spin_lock_irq(&cable->lock);
> > > > +
> > > > +	cable->snd_timer.instance = timeri;
> > > > +
> > > > +	/* initialise a tasklet used for draining */
> > > > +	tasklet_init(&cable->snd_timer.event_tasklet,
> > > > +		     loopback_snd_timer_tasklet, (unsigned
> long)timeri);
> > >
> > > This has to be set before snd_timer_open().  The callback might be
> > > called immediately after snd_timer_open().
> >
> > This tasklet is used/scheduled only in ccallback (not regular tick
> > callback),
> > and only for SNDRV_TIMER_EVENT_MSTOP event. Can this event really happen
> > immediately after snd_timer_open()?
> 
> Why not?  The master timer can be stopped at any time, even between
> these two lines.
> 
> Beware that there are fuzzer programs that can trigger such racy
> things, and you're adding the code to the target that is actively
> slapped by them :)

OK, got it.
I'll move this initialization to before snd_timer_open() in the next
update together with the fixes for the other issues you will find
in this version.

Thanks!

Best regards,
Andrew