diff options
| author | Jeremy Allison <jra@samba.org> | 2001-09-10 22:54:57 +0000 |
|---|---|---|
| committer | Jeremy Allison <jra@samba.org> | 2001-09-10 22:54:57 +0000 |
| commit | f8a4e20e70cb0f92f8c35b33e0aa5a0293b95638 (patch) | |
| tree | c48c2ff4c2eb93bf348f0056d8ccdcb3b018f50c /source/pam_smbpass/support.c | |
| parent | 9d53fa9ede9abf1f9cf33a63dc36be5d37419c15 (diff) | |
| download | samba-f8a4e20e70cb0f92f8c35b33e0aa5a0293b95638.tar.gz samba-f8a4e20e70cb0f92f8c35b33e0aa5a0293b95638.tar.xz samba-f8a4e20e70cb0f92f8c35b33e0aa5a0293b95638.zip | |
Nailed the "oplock hang" problem. Found the logic bug. It was an
optimisation someone... :-) added to process an incoming smbd if both the oplock
fd and server fd because active simultaneously. This of course is
wrong as processing the oplock break message can involve I/O on
the server fd which makes the stored FD_SET state completely bogus.
Also added debug 10 messages I will remove once Gerry has confirmed
the fix.
Jeremy.
Diffstat (limited to 'source/pam_smbpass/support.c')
0 files changed, 0 insertions, 0 deletions