Emmanuel Dreyfus <manu@xxxxxxxxxx> wrote: > > That definitely seems like something that should never happen, so I'd be > > a bit concerned. Perhaps more pertinently, ISTRC that some of the code > > I looked at in our last round of investigation had to do with inodelk, > > so there's a decent chance that they're related. > > Keep in mind this is on NetBSD, with a FUSE emulation that is not yet > finished. I first tend to suspect bugs there. In fact this is a funny semantics difference: on process exit, NetBSD will clear locks on all descriptors, calling ADVLOCK. It does it too for files open read-only, and glusterfs is unhappy with ADVLOCK called on a read-only file. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz manu@xxxxxxxxxx