libkse and SMP (was Re: USB bulk read & pthreads)

[Julian Elischer]

On Thu, 22 May 2003, Dan Nelson wrote:

> In the last episode (May 22), Julian Elischer said:
> > On Thu, 22 May 2003, Dan Nelson wrote:
> > > I do read -current, and I'm definitely not going anywhere near
> > > SCHED_ULE.  I just tested both libraries with today's kernel and
> > > libraries, and was able to get a hard lockup with both libthr and
> > > libkse.  Mysql seems to run okay.  Starting a threaded pike process
> > > seems to be the killer.  Unfortunately, pike's a pretty large app so
> > > it's not easy to get a stripped-down testcase.
> > 
> > Ok so we need to get a description of this 'lockup'.
> 
> Yes, the entire system locks.  2-CPU system.
> 
> No ping responses.  CTRL-ALT-ESC doesn't do anything, and neither does
> <cr> ~ ^B from the serial console.  No X.
> 
> I don't have BREAK_TO_DEBUGGER enabled because my console is connected
> to a Windows machine and I don't want my Unix box to hang every time I
> reboot it :)  Do you think a real BREAK might work where <cr> ~ ^B doesn't?

no, I'd say that the lack of ping response is a very good indicator.


what's 'pike'?

I assume it's one of /usr/ports/lang/pike72 or /usr/ports/lang/pike74?
if so what do I need to do to du0plicate the hang.. you said it is 
predictable..

(what you could do is run it from a serial cable under truss..)