[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 #19 from Konstantin Belousov <kib@FreeBSD.org> ---
(In reply to Alan Somers from comment #18)
> Whether or not copy_file_range is enabled, if nomsyncafterwrite is not set, then
> there is no crash.
But if nomsyncafterwrite is set, is there an inconsistency even without
copy_file_range?

For the second part, I mean that most likely this knobs works because msyncs
are interlined with writes, and really msync affects the next write result,
not the previous.

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