[ Petter Nilsen writes: ]
>
> In article <199705021341.IAA14412@miaow.risc.sps.mot.com>,
> dwolfe@risc.sps.mot.com (Dave Wolfe) wrote:
>
> > If the MTA doesn't check locks, then it doesn't matter what MJ does
> > about locking because it'll still be accessing the file even when Mj
> > has it locked.
>
> My point was that Mj shouldn't start the MTA if the lock was set, but
> wait until it was cleared again.
And my point is that is doesn't matter whether or not the lock is
set when the MTA is started, only if the lock is checked by the MTA
when it accesses the file. We're talking about preemptive scheduling
systems here, so the sequence of events among independent processes is
non-deterministic. So what if Mj waits to start the MTA until the lock
is cleared? By the time the MTA accesses the file, the lock could be set
again.
--
Dave Wolfe
Follow-Ups:
References:
|
|
