At work, a colleague has been researching some problems we encountered in
8.2-R; he found that some commits made to stable/8 since 8.2 was
released has changed the behavior, but that there are still some issues.

He has filed 3 PRs (yesterday) for different aspects of what he's seen:

kern/165923: Writing to NFS-backed mmapped files fails if flushed automatically 
kern/165927: msync reports success after a failed pager flush 
docs/165929: msync(2) man page indicates obsolescence, but it's useful

Here are the links (to make things a bit easier:

<http://www.freebsd.org/cgi/query-pr.cgi?pr=165923&cat=>
<http://www.freebsd.org/cgi/query-pr.cgi?pr=165927&cat=>
<http://www.freebsd.org/cgi/query-pr.cgi?pr=165929&cat=>

The basic result is that attempts to write to an mmap()ed file may fail
silently.

If it turns out that his analysis is incorrect, we would sure like to
know about it.  On the other hand, if the analysis is (largely) correct,
folks may want to be thinkng about whether 8.3 should be released with
the behavior in question, whether it should be documented via Errata, or
whether we should try to fix the problem before 8.3 is released.

Peace,
david
-- 
David H. Wolfskill                              da...@catwhisker.org
Depriving a girl or boy of an opportunity for education is evil.

See http://www.catwhisker.org/~david/publickey.gpg for my public key.

Attachment: pgpO68k5P2Xo2.pgp
Description: PGP signature

Reply via email to