Backend dies suddenly

I have posted the following on the ADMIN list, and having received no
answer, I'm re-posting here.

This is the second time this month that the backend dies on me, and I hope
someone can tell me what the reason is.

I look at the log (debug level 2), and the last message is:

/www/server/pgsql/bin/postmaster: reaping dead processes...
/www/server/pgsql/bin/postmaster: CleanupProc: pid 22650 exited with status 0
/www/server/pgsql/bin/postmaster: ServerLoop: select failed
proc_exit(1) [#0]
shmem_exit(1) [#0]
exit(1)

It occured to me that this is not in Postgres's SELECT, but rather in
select(3C). Which suggests that it has something to do with file
descriptors. The question is what I have to change to prevent this from
happening.

My platform: Solaris 2.6 SPARC, Postgres v. 6.4.2

Herouth

--
Herouth Maoz, Internet developer.
Open University of Israel - Telem project
http://telem.openu.ac.il/~herutma