On Tue, Oct 15, 2024 at 04:13:13PM +0100, Robin Murphy wrote: > On 2024-10-15 1:47 pm, Jason Gunthorpe wrote: > > On Tue, Oct 15, 2024 at 08:13:28AM +0000, Pranjal Shrivastava wrote: > > > > > Umm.. this was specific for rmr not a generic thing. I'd suggest to > > > avoid meddling with the STEs directly for acheiving bypass. Playing > > > with the iommu domain type could be neater. Perhaps, modify the > > > ops->def_domain_type to return an appropriate domain? > > > > Yeah, that is the expected way, to force the def_domain_type to > > IDENTITY and refuse to attach a PAGING/BLOCKED domain. > > There is no domain, this is bypassing an arbitrary StreamID not associated > with any device. Which incidentally is something an IORT RMR can quite > happily achieve already (I think the DT reserved-memory binding does need a > proper device node to relate to, though). +1. I assumed that the use-case was to first attach the streamID to a device and then intall a bypass for that specific streamID. If that's not the case, I'm not sure why are we trying to achieve that. I thought about rmr too, but it looks like the "device" is a DMA and may want to write to more than a fixed region of memory. > > Thanks, > Robin. Thanks, Pranjal
