Re: [PATCH] dma: ti: fix a missing check in omap_dma_prep_dma_cyclic

[Peter Ujfalusi]

On 24/03/2019 0.39, Kangjie Lu wrote:
> It is invalid when "buf_len" is not aligned with "period_len".
> 
> The fix adds a check for the alignment.
> 
> Signed-off-by: Kangjie Lu <kjlu@xxxxxxx>
> ---
>  drivers/dma/ti/omap-dma.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/dma/ti/omap-dma.c b/drivers/dma/ti/omap-dma.c
> index a4a931ddf6f6..5f0ce1975e52 100644
> --- a/drivers/dma/ti/omap-dma.c
> +++ b/drivers/dma/ti/omap-dma.c
> @@ -1065,6 +1065,9 @@ static struct dma_async_tx_descriptor *omap_dma_prep_dma_cyclic(
>  	unsigned es;
>  	u32 burst;
>  
> +	if (buf_len % period_len)
> +		return NULL;
> +

if it really happens, then it might be better to do the check in
dmaengine_prep_dma_cyclic() rather than fixing _all_ drivers?

>  	if (dir == DMA_DEV_TO_MEM) {
>  		dev_addr = c->cfg.src_addr;
>  		dev_width = c->cfg.src_addr_width;
> 

- PÃter

Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki.
Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki