Re: [PATCH v4 4/5] dma: imx-sdma: add a test for imx8mq multi sdma devices

From: Lucas Stach <l.stach@pengutronix.de>
To: "Angus Ainslie (Purism)" <angus@akkea.ca>
Cc: angus.ainslie@puri.sm, Vinod Koul <vkoul@kernel.org>,
	dmaengine@vger.kernel.org, NXP Linux Team <linux-imx@nxp.com>,
	Pengutronix Kernel Team <kernel@pengutronix.de>,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org,
	Daniel Baluta <daniel.baluta@gmail.com>
Subject: Re: [PATCH v4 4/5] dma: imx-sdma: add a test for imx8mq multi sdma devices
Date: Fri, 25 Jan 2019 10:34:44 +0100	[thread overview]
Message-ID: <1548408884.28802.33.camel@pengutronix.de> (raw)
In-Reply-To: <20190125025528.15645-5-angus@akkea.ca>

Am Donnerstag, den 24.01.2019, 19:55 -0700 schrieb Angus Ainslie (Purism):
> On i.mx8mq, there are two sdma instances, and the common dma framework
> will get a channel dynamically from any available sdma instance whether
> it's the first sdma device or the second sdma device. Some IPs like
> SAI only work with sdma2 not sdma1. To make sure the sdma channel is from
> the correct sdma device, use the phandle to match.
> 
> > Signed-off-by: Angus Ainslie (Purism) <angus@akkea.ca>
> ---
>  drivers/dma/imx-sdma.c                | 8 ++++++++
>  include/linux/platform_data/dma-imx.h | 1 +
>  2 files changed, 9 insertions(+)
> 
> diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
> index ec6fb48f387a..88e112a4aabf 100644
> --- a/drivers/dma/imx-sdma.c
> +++ b/drivers/dma/imx-sdma.c
> @@ -1927,11 +1927,17 @@ static int sdma_init(struct sdma_engine *sdma)
>  static bool sdma_filter_fn(struct dma_chan *chan, void *fn_param)
>  {
> >  	struct sdma_channel *sdmac = to_sdma_chan(chan);
> > +	struct sdma_engine *sdma = sdmac->sdma;
> >  	struct imx_dma_data *data = fn_param;
>  
> >  	if (!imx_dma_is_general_purpose(chan))
> >  		return false;
>  
> > +	/* return false if it's not the right device */
> +	if ((sdmac->sdma->drvdata == &sdma_imx8mq)

Why do this check only on i.MX8M? This is completely generic and can
and should be done for every SDMA engine with a OF node.

Also why use the phandle to match this instead of the of_node pointer
directly?

Regards,
Lucas

> +		&& (sdma->dev->of_node->phandle != data->phandle))
> > +		return false;
> +
> >  	sdmac->data = *data;
> >  	chan->private = &sdmac->data;
>  
> @@ -1942,6 +1948,7 @@ static struct dma_chan *sdma_xlate(struct of_phandle_args *dma_spec,
> >  				   struct of_dma *ofdma)
>  {
> >  	struct sdma_engine *sdma = ofdma->of_dma_data;
> > +	struct device_node *np = ofdma->of_node;
> >  	dma_cap_mask_t mask = sdma->dma_device.cap_mask;
> >  	struct imx_dma_data data;
>  
> @@ -1959,6 +1966,7 @@ static struct dma_chan *sdma_xlate(struct of_phandle_args *dma_spec,
> >  	 * be set to sdmac->event_id1.
> >  	 */
> >  	data.dma_request2 = 0;
> > +	data.phandle = np->phandle;
>  
> >  	return dma_request_channel(mask, sdma_filter_fn, &data);
>  }
> diff --git a/include/linux/platform_data/dma-imx.h b/include/linux/platform_data/dma-imx.h
> index 7d964e787299..eeb5b73ae542 100644
> --- a/include/linux/platform_data/dma-imx.h
> +++ b/include/linux/platform_data/dma-imx.h
> @@ -55,6 +55,7 @@ struct imx_dma_data {
> >  	int dma_request2; /* secondary DMA request line */
> >  	enum sdma_peripheral_type peripheral_type;
> >  	int priority;
> > +	int phandle;
>  };
>  
>  static inline int imx_dma_is_ipu(struct dma_chan *chan)