On 24.01.25 15:11, Jason Gunthorpe wrote:
On Fri, Jan 24, 2025 at 11:44:28AM +0100, David Hildenbrand wrote:
There are other concerns I have (what if the page is pinned and access
outside of the user space page tables?). Maybe there was not need to handle
these cases so far.
I think alot of this depends on userspace following some restrictions
so that the pages are always convertible. Presumably if the userspace
breaks things then their atomic using GPU kernels will fault.
So, from a kernel perspective, I'd suggest that creating a reasonable
set of conditions that userspace can follow to have it work reliably
is a reasonable goal.
Yes, that's my assumption as well. If you register a page using io_uring
as a fixed buffer and then trigger "device_exclusive" access, the page
can still be read/written using io_uring and atomics might not work as
expected. "Not supported".
--
Cheers,
David / dhildenb
