> 2023年2月21日 18:14，Conor Dooley <conor@xxxxxxxxxx> 写道： > > On Tue, Feb 21, 2023 at 12:46:12PM +0000, Jiaxun Yang wrote: >> For RISCV we always assume devices are DMA coherent. > > "Always assume", I'm not keen on that wording as it is unclear as to > whether you are suggesting that a) we always take devices to be DMA > coherent, or b) unless a device states it is non-coherent, we take it to > be DMA coherent. > I think you mean b, but being exact would be appreciated, thanks. Yep I meant option b, thanks for the point. > >> Select ARCH_DMA_DEFAULT_COHERENT to ensure dev->dma_conherent >> is always initialized to true. >> >> Signed-off-by: Jiaxun Yang <jiaxun.yang@xxxxxxxxxxx> > > Why was this not sent to the riscv list? > When (or if) you send v2, can you please make sure that you do cc it? Will do. Thanks - Jiaxun > > Thanks, > Conor.
