On 2017年03月16日 11:32, Jason Wang wrote: > > > On 2017年03月16日 09:42, Lan Tianyu wrote: >>> Also I presume that this vIOMMU use case is not the only case where a >>> driver would want to be notified of IOMMU faults, in-kernel drivers >>> might want this too. Drivers making use of the DMA API don't really >>> have any visibility to the IOMMU domain in use, so the framework we use >>> to connect drivers with the IOMMU faults probably needs to abstract >>> that. >> Yes, device page request(part of SVM support) on native also requires >> that device driver to receive IOMMU fault event(page request event) from >> IOMMU driver. So it's necessary to add such abstract layer between IOMMU >> driver and device driver(include VFIO-PCI driver). > > Hi: > > Do you have a real case that requires PRS for kernel drivers? > Yes, Intel GPU supports this feature and patches is upstreaming. https://lists.freedesktop.org/archives/intel-gfx/2017-January/116292.html -- Best regards Tianyu Lan
