Race condition in shlock.c (time window between ValidLock and unlink)

[greg andruk]

Katsuhiro Kondou wrote:
> I don't think shlock is used so heavily that above case
> may happen.  So, I think it should be noted in man pages
> rather than fixing at the moment.  Any comments?

I'm not sure that it _could_ be fixed, short of losing some 
functionality (leading to hangs) or giving up on the idea of having 
uucp-style locks.

I have a vague memory of Majordomo having a race problem with its shlock 
several years back.  If someone has a copy handy, maybe it would be 
helpful to peek at how that one currently works and see if it contains 
some useful solution to the problem (unless, of course, their race 
problem was in some other detail).