On Wed, 5 Nov 1997, Todd Vierling wrote:

: : Exactly which model of sparc machines have demonstrated the
: : 	panic: crazy interrupts
: : problem with the new zs driver?  I'd like to investigate this,
: : but have been unable to reproduce it.
: 
: SS2 here; I haven't seen it but once, and that was after a ddb entry for a
: different panic (one a bug in my SVR4 compat stuff caused ;) and I typed
: "continue" to let the kernel dump and reboot.  The syncing disks message
: prompted the trail of spurious interrupts followed by the crazy interrupts
: panic.

I take that back.  I got it yesterday, reproduceably, while seeing debug
output from the svr4 compat system (SVR4_DEBUG).  For a little bit of
context, the svr4 debug mechanism, if compiled in, prints to the controlling
tty of the process getting debug dump output.  When both the process and the
kernel wanted to write to the tty at once, I got the stream of spurious
interrupts followed by the panic.  I haven't tried booting a recent install
floppy--the old way I could get the panic was from booting a 1.2.1 install
floppy--because my floppy drive is now kaput. 

Can someone rehash how the old driver fixed this, and we just add a hack to
block the proper interrupt on the sparc (if that's how it was fixed)?  Or do
we lose somehow if that interrupt is blocked?

=====
== Todd Vierling (Personal tv@pobox.com; Business tv@lucent.com)
== I know you like the Internet, Bobby.  Now go eat your Frosted Flakes.