On 07-09-26, 10:55, Amelie Delaunay wrote: > Hi Frank, > > On 9/4/26 23:49, [email protected] wrote: > [...] > > I don't think removing chan2dev() function from the STM32 DMA drivers is an > improvement, since the parameter is a `stm32` chan, not a `dma_chan`. > chan2dev() was meant to expose the channel device behind the `stm32` wrapper > and keep logging concise. > > > diff --git a/drivers/dma/stm32/stm32-dma.c b/drivers/dma/stm32/stm32-dma.c > > index 3e8151e2ad75a..4688cebaf7f7a 100644 > > --- a/drivers/dma/stm32/stm32-dma.c > > +++ b/drivers/dma/stm32/stm32-dma.c > > @@ -248,11 +248,6 @@ static struct stm32_dma_desc *to_stm32_dma_desc(struct > > virt_dma_desc *vdesc) > > return container_of(vdesc, struct stm32_dma_desc, vdesc); > > } > > -static struct device *chan2dev(struct stm32_dma_chan *chan) > > -{ > > - return &chan->vchan.chan.dev->device; > > -} > > - > > I would rather keep it as: > > static struct device *chan2dev(struct stm32_dma_chan *chan) > { > return vchan_chan_dev(&chan->vchan); > } > > and then drop all further updates.
I would say better to use a common macro than keep one for yours -- ~Vinod
