[Bug 276002] nfscl: data corruption using both copy_file_range and mmap'd I/O

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276002

--- Comment #6 from Rick Macklem <rmacklem@FreeBSD.org> ---
Yes. The patch looks good to me and, yes, since this
would be a rare occurrence, starting with LK_SHARED
on the invp seems like a good idea.

It also makes sense to invalidate the outvp's buffers.
The change in the NFSv4.2 Change attribute would have
triggered that before a read from the now bogus buffers
occurred, but why wait for that when you know the buffers
are now invalid?

-- 
You are receiving this mail because:
You are the assignee for the bug.