On Mon, Oct 07, 2024 at 09:47:53AM +0100, David Woodhouse wrote: > On Mon, 2024-10-07 at 08:39 +0000, Gowans, James wrote: > > > > I think we have two other possible approaches here: > > > > 1. What this RFC is sketching out, serialising fields from the structs > > and setting those fields again on deserialise. As you point out this > > will be complicated. > > > > 2. Get userspace to do the work: userspace needs to re-do the ioctls > > after kexec to reconstruct the objects. My main issue with this approach > > is that the kernel needs to do some sort of trust but verify approach to > > ensure that userspace constructs everything the same way after kexec as > > it was before kexec. We don't want to end up in a state where the > > iommufd objects don't match the persisted page tables. > > To what extent does the kernel really need to trust or verify? If iommufd is going to adopt an existing iommu_domain then that iommu_domain must have exactly the IOPTEs it expects it to have otherwise there will be functional problems in iommufd. So, IMHO, some kind of validation would be needed to ensure that userspace has created the same structure as the old kernel had. >At LPC we seemed to speak of a model where userspace builds a "new" > address space for each device and then atomically switches to the > new page tables instead of the original ones inherited from the > previous kernel. The hitless replace model would leave the old translation in place while userspace builds up a replacement translation that is equivalent. Then hitless replace would adopt the new translation and we discard the old ones memory. IMHO this is easiest to make correct and least maintenance burden because the only kernel thing you are asking for in iommufd is hitless iommu_domain replace, which we already want to add to the drivers anyhow. (ARM already has it) Jason
