>Date: Thu, 27 Jan 2000 13:16:50 -0800
>From: Ethan Solomita <ethan@geocast.com>
>
>	Soft Updates is particularly susceptible to eating memory during a
>large series of removes, but that has always led to panics for me, not
>lockups. 

Yes, when I had softdeps enabled, several of my "lockups" were during
large parallel file removals.  Since I'm running X and had DDB
enabled, Jason's explanation would account for why I thought the
system froze, when actually the system was most likely sitting in DDB.

It's worth reiterating here, as someone pointed out earlier in this
tread, that I've probably experienced these "freezes" over the past
few days due to several different causes:

- the kern_proc.c ipl level bug, now fixed by Jason
- softdep livelick bugs
- possible memory shortages caused by softdeps
- in my last freeze, possibly a kernel virtual address space shortage
  caused by all the crap I was running to stress the system.

So I don't think there's any great cause for alarm at this point.
Most likely there were several different things going on, and what I
thought were lockups were most likely just run-of-the-mill panics.  Of
course we don't want the system to panic, but I would think it will be
much easier to hunt and kill any bugs if I can provide a crashdump and
backtrace, instead of just reporting freeze after freeze.

--
entropy