David Turner <David.Turner@xxxxxxxxxxxx> writes: > If the writer has the smaller HOST_NAME_MAX, this will work fine. If the reader > has the smaller HOST_NAME_MAX, and the writer's actual value is too long, > then there's no way the strcmp would succeed anyway. So I don't think we need > to worry about it. Hmph, I have to agree with that reasoning, only because the value we read into locking_host[] is not used for error reporting at all. I would have insisted to read what is on the filesystem anyway if that were not the case. Thanks.