Current-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

locking woes with -current



People,

I updated to -current this week on an Intel Core Duo based laptop and
I have been having some nasty troubles:

1) sometimes file accesses make the requesting process get stuck in the D
state - I enabled LOCKDEBUG and the waitchannel is uvn_fp2.  It looks
like something is not unlocking the uvm_object (v_uobj) that is
associated with the vnode.  I have tried looking for a cause but
cannot spot one, I have not found an easy way of getting the open
files for a process and looking at which is locked (I may write
something for ddb to do this..)

2) When LOCKDEBUG was enabled I would get some random panics about
things trying to be locked at an interrupt level.  These seemed to be
coming from softclock() processing callouts - the ones I saw were
sched_pstats() and pfslowtimo().  They seemed pretty random, I am not
sure if this is just a problem with the LOCKDEBUG code or the callouts
really should not be doing locking

3) Loading a lot of tabs all at once in firefox3 wedged my machine -
this was quite repeatable.  What I would do was click the button to
recover a crashed firefox session - the tabs would all load for a few
seconds and then the machine would just wedge.  I tried typing some
ddb commands blind but the machine appeared unresponsive and had to be
power-cycled

4) Building without MULTIPROCESSOR appears to be broken - I would get
a compile error in kern_stub.c saying you could not have
__HAVE_PREEMPTION without MULTIPROCESSOR.  I may have messed up my
kernel config to do this but I am not sure how.

I am currently running:

NetBSD rover 4.99.69 NetBSD 4.99.69

last updated a day or so ago.

-- 
Brett Lymn
"Warning:
The information contained in this email and any attached files is
confidential to BAE Systems Australia. If you are not the intended
recipient, any use, disclosure or copying of this email or any
attachments is expressly prohibited.  If you have received this email
in error, please notify us immediately. VIRUS: Every care has been
taken to ensure this email and its attachments are virus free,
however, any loss or damage incurred in using this email is not the
sender's responsibility.  It is your responsibility to ensure virus
checks are completed before installing any data sent in this email to
your computer."




Home | Main Index | Thread Index | Old Index