Re: [PATCH v2 0/5] guest_memfd fixes for bind and populate

[Sean Christopherson <seanjc@google.com>]

On Fri, May 22, 2026, Ackerley Tng wrote:
> This series is a group of fixes for the bind and populate flows for
> guest_memfd, and fixes some issues reported by Sashiko after reviewing the
> guest_memfd in-place conversions series [1] and another fixup series Sean
> posted [3].

In the future, please don't bundle unrelated changes.  The SNP specific changes
are related and should be a series, but the signed integer thing and the lack of
error handling on xa_store_range() are completely unrelated, because the fact
that Sashiko kept complaining about pre-existing issues.

I totally understand why you bundled these together, but that obviously didn't
stop Sashiko from complaining about pre-existing issues, over and over.

Unnecessarily bundling can lead to exactly what's happening here: the three SNP
changes are ready to go, but the two unrelated guest_memfd changes need new
versions.  Which isn't hard to deal with, but it's extra friction that is easily
avoided.

I'll apply the SNP changes, and send a new version of the signed vs. unsigned
issue.  Please send a new version of the xa_store_range() error handling (or
prove that I'm wrong).

Thanks!