Re: [PATCH 1/1] dmaengine: change dmaengine_get_unmap_data() first arg to dma_chan

[Frank Li <Frank.li@oss.nxp.com>]

On Fri, Oct 09, 2026 at 09:19:43AM -0700, Dave Jiang wrote:
>
>
> On 10/9/26 2:43 AM, Vinod Koul wrote:
> > On 22-09-26, 16:09, Dave Jiang wrote:
> >>
> >>
> >> On 9/22/26 11:45 AM, Frank.Li@oss.nxp.com wrote:
> >>> From: Frank Li <Frank.Li@nxp.com>
> >>>
> >>> Replace the struct device *dev argument with struct dma_chan *chan so
> >>> callers do not need to reach into the dma_device internals to obtain
> >>> the device pointer.  The implementation now derives the device via
> >>> dmaengine_get_dma_device(chan), which correctly handles the per-channel
> >>> IOMMU mapping case.
> >>>
> >>> Prepare to make dma_device private and make dmaengine API consistent.
> >>>
> >>> Signed-off-by: Frank Li <Frank.Li@nxp.com>
> >>
> >> For the NTB bits:
> >> Acked-by: Dave Jiang <dave.jiang@intel.com>
> >>
> >>
> >>> ---
> >>>  crypto/async_tx/async_memcpy.c      | 4 ++--
> >>>  crypto/async_tx/async_pq.c          | 4 ++--
> >>>  crypto/async_tx/async_raid6_recov.c | 4 ++--
> >>>  crypto/async_tx/async_xor.c         | 8 ++++----
> >>
> >> Maybe one day Dan's wishes will become true and someone will finally deprecate async_tx....
> >
> > Does Intel still have use it. I am looking at fixing the names and we
> > can drop async_tx completely...?
> >
> I don't think so. Not since ioatdma days. Dan has talked about wanting to remove async_tx for quite a few years now. No idea if there are ARM dma devices that still use that or not.

Maybe we can create patch to skip build it firstly, let's see if there are
really user.

Frank