Quoting r. Roland Dreier <[EMAIL PROTECTED]>: > Subject: Re: Fwd: issues in ipoib > > > + if (atomic_dec_and_test(&client->users)) > > + wake_up(&client->wait); > > > + wait_event(client->wait, atomic_read(&client->users) == 0); > > I think this is vulnerable to the race we fixed up all over the place > a few months ago, where wait_event runs between the > atomic_dec_and_test() and the wake_up().
No, I think there's no problem - user callback has finished running and that's all we care about. But I'd like to point out that using rwsem gets all these corners right automatically. Reconsider that approach? -- MST _______________________________________________ openib-general mailing list [email protected] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
