Simon 'corecode' Schubert wrote:
Ok, I tried to play a bit Sherlock Holmes (on the live system).
Conclusion:
tcsh is terminating, calling exit(2). The kernel closes all fdesc,
including the controlling terminal ttyp5. Trying to flush ttyp5, the
vnode gets locked, but ttylclose/ttywflush blocks while the output queue
is draining.
At the same time sshd, the process controlling ptsp5, does a chown of
ttyp5. For some reason it holds the namecache entry locked [*].
kern_chown tries to lock the vnode of ttyp5 and blocks on this attempt.
sshd can't read from ptsp5 and thus drain the output queue tcsh is
waiting for.
Deadlock.
But I have no clue where do remove that deadlock. Probably somewhere in
the tty code and not in chown. Any hints?
[*] The fact that the namecache entry is locked, doesn't lead to the
deadlock, but it is the reason that dozens of find(1)s lock on /dev/ttyp5.
cheers
simon
--
Serve - BSD +++ RENT this banner advert +++ ASCII Ribbon /"\
Work - Mac +++ space for low $$$ NOW!1 +++ Campaign \ /
Party Enjoy Relax | http://dragonflybsd.org Against HTML \
Dude 2c 2 the max ! http://golden-apple.biz Mail + News / \
