Re: [PATCH] ALSA: compress: fix the states to check for allowing read

[Richard Fitzgerald <rf@opensource.wolfsonmicro.com>]

On Mon, Apr 29, 2013 at 11:20:42AM +0200, Takashi Iwai wrote:
> At Mon, 29 Apr 2013 14:15:31 +0530,
> Vinod Koul wrote:
> > 
> > On Mon, Apr 29, 2013 at 10:55:04AM +0200, Takashi Iwai wrote:
> > > At Sun, 28 Apr 2013 13:35:22 +0530,
> > > Vinod Koul wrote:
> > > > 
> > > > for reading compressed data, we need to allow when we are paused, draining or
> > > > stopped.
> > > > 
> > > > Signed-off-by: Vinod Koul <vinod.koul@intel.com>
> > > > Cc: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
> > > > Cc: Richard Fitzgerald <rf@opensource.wolfsonmicro.com>
> > > > 
> > > > ---
> > > >  sound/core/compress_offload.c |   10 ++++++++--
> > > >  1 files changed, 8 insertions(+), 2 deletions(-)
> > > > 
> > > > diff --git a/sound/core/compress_offload.c b/sound/core/compress_offload.c
> > > > index a0bc47f..5389b9a 100644
> > > > --- a/sound/core/compress_offload.c
> > > > +++ b/sound/core/compress_offload.c
> > > > @@ -311,8 +311,14 @@ static ssize_t snd_compr_read(struct file *f, char __user *buf,
> > > >  	stream = &data->stream;
> > > >  	mutex_lock(&stream->device->lock);
> > > >  
> > > > -	/* read is allowed when stream is running */
> > > > -	if (stream->runtime->state != SNDRV_PCM_STATE_RUNNING) {
> > > > +	/* read is allowed when stream is running, paused, draining and setup
> > > > +	 * (yes setup is state which we transistion to after stop, so if user
> > > > +	 * wants to read data after stop we allow that
> > > > +	 */
> > > > +	if (stream->runtime->state != SNDRV_PCM_STATE_RUNNING ||
> > > > +			stream->runtime->state != SNDRV_PCM_STATE_DRAINING ||
> > > > +			stream->runtime->state != SNDRV_PCM_STATE_PAUSED ||
> > > > +			stream->runtime->state != SNDRV_PCM_STATE_SETUP) {
> > > >  		retval = -EBADFD;
> > > 
> > > Aren't they "&&"?
> > Yup :(
> > > 
> > > Maybe better to use switch for avoiding such a trivial error...
> > I started with switch and then didnt want to code for all the PCM_STATES... but
> > yes that makes more sense, i will send that right away.
> 
> You can code like
> 	switch (stream->runtime->state) {
> 	case SNDRV_PCM_STATE_RUNNING:
> 	case SNDRV_PCM_STATE_DRAINING:
> 	case SNDRV_PCM_STATE_PAUSED:
> 	case SNDRV_PCM_STATE_SETUP:
> 		break; /* OK */
> 	default:
> 		retval = -EBADFD;
> 		break;
> 	}
> 
> 
> > Is it okay if this get in for 3.10...
> 
> Yes.  It's just a damn simple fix.
> 
> 
> thanks,
> 
> Takashi


Ok with me too.
(You beat me to it with this fix)