And there's that first "truncate" close releasing the lock gained on the
first open.
Alright, stepping back from the gigantic mess that is fcntl() and looking
at the core problem .... can anyone think of a reason that creat() was used
in preference to ftruncate() way back when? I see it was that way in
mh-6.8.5, but that also used ftruncate() in another file so it's not like
that interface hasn't existed forever.
I think just using ftruncate() would solve this problem, right? Ralph,
could you give this a try?
--Ken
_______________________________________________
Nmh-workers mailing list
Nmh-workers(_at_)nongnu(_dot_)org
https://lists.nongnu.org/mailman/listinfo/nmh-workers