Re: [long] panic on a recent current

At 20:05 18/05/2004, Thierry Herbelot wrote:
>on a recent -current, running KDE and simultaneously building the world.
>[snip]

Short answer: Blame jhb. :-)

Long answer: When he split sleepq_wakeup_thread() in revision 1.6 of
subr_sleepqueue.c, he left a window open between sleepq_remove_thread
and sleepq_resume_thread where td_wchan has been set to NULL but
the TDI_SLEEPING bit has not yet been removed from td_inhibitors.
This causes a panic when tdsigwakeup() calls sleepq_abort() on a
thread which has been removed from the sleepqueue but not yet resumed.

Colin Percival