[rescue] mysterious hard hangs, two different sun4m's

[Skeezics Boondoggle]

On Tue, 3 Sep 2002, James Lockwood wrote:

[snip]

> Unless your desk has a Tesla coil directly underneath it I wouldn't
> worry about the problem migrating with position.  :)

damn!  i completely forgot about that pesky Tesla coil. :-)

reminds me of the story about the utterly exasperated tech support person
who could not diagnose the trouble, and finally asked the caller, "has
ANYTHING changed in the last few days that might explain what's gone
wrong?"  to which she replied, "well, it did catch fire yesterday..."

no, my machines haven't caught fire.  i checked that. :-)

thanks to all for the suggestions; i'll have to try a combination of 
things, obviously - latest patches, more methodical approach to debugging 
the hardware end of things...

but i realized that the other common feature is that i've been building
all my solaris code (as much as possible, anyway) under the sparcworks 5.0
compilers, not gcc, for both the 2.6 and 7 environments.  since these hard
hangs have happened on both the ss4/tcx and the ss20/sx i'm suspecting
that it's something i've built - or maybe a compiler bug, or a solaris bug
that's being tickled by olvwm or openssl/openssh or xconq/tcl/tk or who
knows what..."doctor, it hurts when i run xconq!"  "so don't do that..."

it used to be easy:  blame netscape.  netscape has crashed every X server
i've ever used on every platform at one time or another, so that was an
obvious suspect.  when starting up java under netscape used to cause
freezes, patches alleviated that somewhat - now i get core dumps, or the X
server crashes instead... at least i avoid having to power cycle the
machine. :-)  but netscape hasn't been giving me grief, lately.

thanks for the tips.  i'd like to solve this mystery.  i mean, if i wanted
to have to reboot my machine every day, i'd run w... ew, i can't even say
it. :-)

cheers,

-- skeez