Knowledge Base

what does it mean if a process is in state waitpage for FFFFFFFF?
Is FFFFFFFF only a hint for a mem manager or is this a bug/feature?
This scenario pops up if we produce a very havy load on TCP/IP.

BTW: io-pkt-v4 hangs forever.

12302 1 mle/sbin/io-pkt-v4 21r SIGWAITINFO
12302 2 mle/sbin/io-pkt-v4 21r WAITPAGE ffffffff
12302 3 mle/sbin/io-pkt-v4 21r REPLY 4103
12302 4 mle/sbin/io-pkt-v4 10r RECEIVE 14
32783 1 e/usr/sbin/fs-nfs3 10r REPLY 12302

System is still alive.

Solution 
You are hitting the dreaded waitpage deadlock, my friend.

The curse of the single threaded resource managers has struck you.

loop:
The problem is caused by io-pkt replying to a client, becoming waitpaged because the client's memory is not faulted in, but the procnto thread that is dispatched to handle it gets blocked waiting on the address space lock of the client process, which is current owned by a procnto thread that was already trying to satisfy a page fault request which involved talking to nfs which can't finish because io-pkt is blocked waiting for the lock to be released.
<deep breath>
goto loop;