On 04.11.2021 15:37, Keith Busch wrote:

On Thu, Nov 04, 2021 at 06:34:31PM +0100, Christoph Hellwig wrote:

quoted

On Thu, Nov 04, 2021 at 10:32:35AM -0700, Darrick J. Wong wrote:

quoted

I also wonder if it would be useful (since we're already having a
discussion elsewhere about data integrity syscalls for pmem) to be able
to call this sort of thing against files?  In which case we'd want
another preadv2 flag or something, and then plumb all that through the
vfs/iomap as needed?

IFF we do this (can't answer if there is a need) we should not
overload read with it.  It is an operation that does not return
data but just a status, so let's not get into that mess.

If there is a need for this, a new io_uring opcode seems like the
appropirate user facing interface for it.

+1 to this. I was looking at the patchset yesterday and this was one of
the questions I had. Any reasons to not do it this way Chaitanya?