On 06-06-22, 20:54, Adrian Larumbe wrote:
> >On 06.06.2022 23:23, Vinod Koul wrote:
> >On 06-06-22, 09:47, Christoph Hellwig wrote:
> >> This was removed before due to the complete lack of users, but
> >> 3218910fd585 ("dmaengine: Add core function and capability check for
> >> DMA_MEMCPY_SG") and 29cf37fa6dd9 ("dmaengine: Add consumer for the new
> >> DMA_MEMCPY_SG API function.") added it back despite still not having
> >> any users whatsoever.
> >>
> >> Fixes: 3218910fd585 ("dmaengine: Add core function and capability check for DMA_MEMCPY_SG")
> >> Fixes: 29cf37fa6dd9 ("dmaengine: Add consumer for the new DMA_MEMCPY_SG API function.")
> >
> >This is consumer of the driver API and it was bought back with the
> >premise that user will also come...
>
> It's commit 29cf37fa6dd9 ("dmaengine: Add consumer for the new DMA_MEMCPY_SG API function.")
>
> The two previous commits add the new driver API callback and document it.
>
> >Adrianm, Michal any reason why user is not mainline yet..?
>
> Just double checked the mainline, and all three commits are there.
There are no clients in mainline which call this API. There is a driver
which implements this API, but no users...
$ git grep dmaengine_prep_dma_memcpy_sg
include/linux/dmaengine.h:static inline struct dma_async_tx_descriptor *dmaengine_prep_dma_memcpy_sg(
--
~Vinod
