At 02:43 PM 10/25/00 -0400, Bill Zumach wrote:
>Unless something has been broken in the AFS client recently, the blocking
>behavior you observe is correct. In fact, an AFS client is supposed to flush
>any dirty data which is under the lock back to the server before releasing the
>lock.

Surely this contradicts the FAQ which says that an ordinary blocking 
flock() issued from another client from one that has a lock will 
immediately return EWOULDBLOCK?

--
Peter Scott
[EMAIL PROTECTED]

Reply via email to