Re: [PATCH 1/5] iio: buffer-dmaengine: add dev-managed calls for buffer alloc/free

[Jonathan Cameron]

On Tue, 25 Feb 2020 13:33:59 +0000
"Ardelean, Alexandru" <alexandru.Ardelean@xxxxxxxxxx> wrote:

> On Fri, 2020-02-21 at 12:21 +0000, Jonathan Cameron wrote:
> > [External]
> > 
> > On Thu, 20 Feb 2020 17:03:13 +0200
> > Alexandru Ardelean <alexandru.ardelean@xxxxxxxxxx> wrote:
> >   
> > > Currently, when using a 'iio_dmaengine_buffer_alloc()', an matching call to
> > > 'iio_dmaengine_buffer_free()' must be made.
> > > 
> > > With this change, this can be avoided by using
> > > 'devm_iio_dmaengine_buffer_alloc()'. The buffer will get free'd via the
> > > device's devres handling.
> > > 
> > > Signed-off-by: Alexandru Ardelean <alexandru.ardelean@xxxxxxxxxx>
> > > ---
> > >  .../buffer/industrialio-buffer-dmaengine.c    | 70 +++++++++++++++++++
> > >  include/linux/iio/buffer-dmaengine.h          |  5 ++
> > >  2 files changed, 75 insertions(+)
> > > 
> > > diff --git a/drivers/iio/buffer/industrialio-buffer-dmaengine.c
> > > b/drivers/iio/buffer/industrialio-buffer-dmaengine.c
> > > index b129693af0fd..eff89037e3f5 100644
> > > --- a/drivers/iio/buffer/industrialio-buffer-dmaengine.c
> > > +++ b/drivers/iio/buffer/industrialio-buffer-dmaengine.c
> > > @@ -229,6 +229,76 @@ void iio_dmaengine_buffer_free(struct iio_buffer
> > > *buffer)
> > >  }
> > >  EXPORT_SYMBOL_GPL(iio_dmaengine_buffer_free);
> > >  
> > > +static void __devm_iio_dmaengine_buffer_free(struct device *dev, void *res)
> > > +{
> > > +	iio_dmaengine_buffer_free(*(struct iio_buffer **)res);
> > > +}
> > > +
> > > +/**
> > > + * devm_iio_dmaengine_buffer_alloc() - Resource-managed
> > > iio_dmaengine_buffer_alloc()
> > > + * @dev: Parent device for the buffer
> > > + * @channel: DMA channel name, typically "rx".
> > > + *
> > > + * This allocates a new IIO buffer which internally uses the DMAengine
> > > framework
> > > + * to perform its transfers. The parent device will be used to request the
> > > DMA
> > > + * channel.
> > > + *
> > > + * Once done using the buffer iio_dmaengine_buffer_free() should be used to
> > > + * release it.  
> > 
> > Umm.  It really shouldn't!  
> 
> Yeah.
> Copy+paste.
> My bad.
> 
> 
> > > + */
> > > +struct iio_buffer *devm_iio_dmaengine_buffer_alloc(struct device *dev,
> > > +	const char *channel)
> > > +{
> > > +	struct iio_buffer **bufferp, *buffer;
> > > +
> > > +	bufferp = devres_alloc(__devm_iio_dmaengine_buffer_free,
> > > +			       sizeof(*bufferp), GFP_KERNEL);
> > > +	if (!bufferp)
> > > +		return ERR_PTR(-ENOMEM);
> > > +
> > > +	buffer = iio_dmaengine_buffer_alloc(dev, channel);
> > > +	if (!IS_ERR(buffer)) {
> > > +		*bufferp = buffer;
> > > +		devres_add(dev, bufferp);  
> > 
> > From a flow point of view I'd prefer.
> > 
> > 	if (IS_ERR(buffer) {
> > 		devres_free(buferp)
> > 		return buffer;
> > 	}
> > 
> > 	*bufferp = buffer;
> > 	devres_add(dev, bufferp);
> > 
> > 	return buffer;
> > 
> >   
> > > +	} else {
> > > +		devres_free(bufferp);
> > > +	}
> > > +
> > > +	return buffer;
> > > +}
> > > +EXPORT_SYMBOL_GPL(devm_iio_dmaengine_buffer_alloc);
> > > +
> > > +static int devm_iio_dmaengine_buffer_match(struct device *dev, void *res,
> > > +	void *data)
> > > +{
> > > +	struct iio_buffer **r = res;
> > > +
> > > +	if (!r || !*r) {
> > > +		WARN_ON(!r || !*r);
> > > +		return 0;
> > > +	}
> > > +
> > > +	return *r == data;
> > > +}
> > > +
> > > +/**
> > > + * devm_iio_dmaengine_buffer_free - iio_dmaengine_buffer_free
> > > + * @dev: Device this iio_buffer belongs to
> > > + * @buffer: The iio_buffer associated with the device
> > > + *
> > > + * Free buffer allocated with devm_iio_dmaengine_buffer_alloc().
> > > + */
> > > +void devm_iio_dmaengine_buffer_free(struct device *dev,
> > > +	struct iio_buffer *buffer)
> > > +{
> > > +	int rc;
> > > +
> > > +	rc = devres_release(dev, __devm_iio_dmaengine_buffer_free,
> > > +			    devm_iio_dmaengine_buffer_match, buffer);
> > > +	WARN_ON(rc);
> > > +}
> > > +EXPORT_SYMBOL_GPL(devm_iio_dmaengine_buffer_free);  
> 
> Should I remove devm_iio_dmaengine_buffer_free() ?
> There was a comment on the AXI ADC that may apply here as well, about maybe
> removing it.

yup. Drop it.

> 
> > > +
> > >  MODULE_AUTHOR("Lars-Peter Clausen <lars@xxxxxxxxxx>");
> > >  MODULE_DESCRIPTION("DMA buffer for the IIO framework");
> > >  MODULE_LICENSE("GPL");
> > > diff --git a/include/linux/iio/buffer-dmaengine.h
> > > b/include/linux/iio/buffer-dmaengine.h
> > > index b3a57444a886..8dcd973d76c1 100644
> > > --- a/include/linux/iio/buffer-dmaengine.h
> > > +++ b/include/linux/iio/buffer-dmaengine.h
> > > @@ -14,4 +14,9 @@ struct iio_buffer *iio_dmaengine_buffer_alloc(struct
> > > device *dev,
> > >  	const char *channel);
> > >  void iio_dmaengine_buffer_free(struct iio_buffer *buffer);
> > >  
> > > +struct iio_buffer *devm_iio_dmaengine_buffer_alloc(struct device *dev,
> > > +	const char *channel);
> > > +void devm_iio_dmaengine_buffer_free(struct device *dev,
> > > +	struct iio_buffer *buffer);  
> > Please align parameters with opening bracket where possible.
> >   
> 
> ack
> 
> > Thanks,
> > 
> > Jonathan
> >   
> > > +
> > >  #endif