Steve Revilak wrote:
> A few days Ago I reported some oddities with X windows under NetBSD 1.4,
> adn a few crashes using the MRG-6 and GENERIC-6 kernels.
> 
> Unfortunately, I haven't been able to duplicate the "startx and crash
> immediately" problem, then again, since my libraries are too old for
> this kernel I'm not sure how worthwhile this would be.
> 
> However, I am seeing a "startx and hang forever" problem.... and it's
> not confined to the July 1 kernels --> it also happens with the 1.4
> GENERIC one as well.  
> 
> "startx" never makes it to the "screen 640 x 480...." line.  While it's
> hung, I can jump into the debugger.  The stack trace is pretty
> consistent:
> 
> 	_debugger(13940e, 37c2, 83ae04,358a,0 ) +6
> 	_nmi_hand(0,fffffffe,0,0,1) + 26
> 	_lev7intr(?)
>         _mi_switch(0,2004,4,8,167) + 12
> 	_tsleep(16f8dc,118,35548,167) + 162
> 	_sys_select(7e918,83af88,83af80) +27e
> 	_syscall(5d) +14c
> 	_trap0() +e
> 
> (you can see where I hit the interrupt switch).  

well, whatever made the select() call is waiting for something to occur
and never gets it
 
> A typical process list.  Nothing that strikes me as terribly unusual:
> 
> 	Xmac68k
> 	xinit
> 	sh

one of these three is probably the one that's hung...does the ps listing
give a wait channel (wchan)?  if so, what are the channels these are
waiting on?  this will give you some clue as to where exactly in the code
the process is waiting...
 
later.

colin