Roman Drahtmueller wrote:
The latest versions of rpm appear to have what is currently believed to be either a race condition or stale lock problem that causes it to lock up in a select call:
select(0, NULL, NULL, NULL, {1,0}) = 0 (Timeout) select(0, NULL, NULL, NULL, {1,0}) = 0 (Timeout) etc.
These select calls on NULL sockets are used as an equivalent to sleep(3) or usleep(2), nothing more. The suspicion about a deadlock holds, though, because it seems to sit on a resource and wait for the weather to become warmer. But a deadlock requires two parties to fight for a resource. Schizophrenic, isn't it?
I'll try and see what else is going on next time it happens, but it's not related to any particular package.
But it doesn't bring us any further. See if --rebuilddb brings about a change.
Unfortunately not, I've already been using --rebuilddb as part of recovery after terminating rpm. Of course, package operations go through after that, at least for a little while. It must be said that this isn't so much of a problem under SuSE, I've had far more problems with rpm-4.1. Dave Alfar Networks