On 29-07-21, 09:18, Juergen Borleis wrote: > Commit dea7a9f you should use 12+ chars for sha1 > dmaengine: imx-dma: remove dma_slave_config direction usage even with this is commit log, you made the patch tile dma: ....! > > changes the method from a "configuration when called" to an "configuration > when used". Due to this, only the cyclic DMA type gets configured > correctly, while the generic DMA type is left non-configured. > > Without this additional call, the struct imxdma_channel::word_size member > is stuck at DMA_SLAVE_BUSWIDTH_UNDEFINED and imxdma_prep_slave_sg() always > returns NULL. I ahve fixed the subsystem name, updated commit in 12 chars and format expected, added it as Fixes line and applied. Please take care of these things! > > Signed-off-by: Juergen Borleis <jbe@xxxxxxxxxxxxxx> > --- > drivers/dma/imx-dma.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c > index 7f116bb..2ddc31e 100644 > --- a/drivers/dma/imx-dma.c > +++ b/drivers/dma/imx-dma.c > @@ -812,6 +812,8 @@ static struct dma_async_tx_descriptor *imxdma_prep_slave_sg( > dma_length += sg_dma_len(sg); > } > > + imxdma_config_write(chan, &imxdmac->config, direction); > + > switch (imxdmac->word_size) { > case DMA_SLAVE_BUSWIDTH_4_BYTES: > if (sg_dma_len(sgl) & 3 || sgl->dma_address & 3) > -- > 2.20.1 -- ~Vinod