>> Anyone have enough idea what the differences between the ELC and IPX
>> are to take even a wild stab at where to look for the problem?

> ELC == monochrome;

I rather doubt this is relevant, since the kernel in use has neither
bwtwo nor cgsix configured into it (I get a "not configured" message at
boot time).  I stripped out everything not needed for running the mud
when I set up this kernel.

Here's dmesg from the live system (the IPX).  Note how many things are
"not configured".

NetBSD 1.4J (VGER) #1: Sun Jan 16 02:51:12 EST 2000
    mouse@Sparkle.Rodents.Montreal.QC.CA:/big/sources/working-usr-src/sys/arch/sparc/compile/VGER
total memory = 40644 KB
avail memory = 37928 KB
using 128 buffers containing 512 KB of memory
bootpath: /sbus@1,f8000000/esp@0,800000/sd@0,0
mainbus0 (root): SUNW,Sun 4/50
cpu0 at mainbus0: cache chip bug; trap page uncached: W8601/8701 or MB86903 @ 40 MHz, on-chip FPU
cpu0: 64K byte write-through, 32 bytes/line, hw flush: cache enabled
memreg0 at mainbus0 ioaddr 0xf4000000
clock0 at mainbus0 ioaddr 0xf2000000: mk48t02 (eeprom)
timer0 at mainbus0 ioaddr 0xf3000000 ipl 10 delay constant 17
auxreg0 at mainbus0 ioaddr 0xf7400003
zs0 at mainbus0 ioaddr 0xf1000000 ipl 12 softpri 6
zstty0 at zs0 channel 0 (console)
zstty1 at zs0 channel 1
zs at mainbus0 ioaddr 0xf0000000 ipl 12 not configured
audio at mainbus0 ioaddr 0xf7201000 ipl 13 not configured
sbus0 at mainbus0 ioaddr 0xf8000000: clock = 20 MHz
dma0 at sbus0 slot 0 offset 0x400000: rev 1+
esp0 at sbus0 slot 0 offset 0x800000 level 3: ESP100A, 25MHz, SCSI ID 7
scsibus0 at esp0: 8 targets, 8 luns per target
probe(esp0:0:0): max sync rate 5.00Mb/s
sd0 at scsibus0 targ 0 lun 0: <SGI, SEAGATE ST51080N, 0950> SCSI2 0/direct fixed
sd0: 1010 MB, 4826 cyl, 4 head, 107 sec, 512 bytes/sect x 2070235 sectors
le0 at sbus0 slot 0 offset 0xc00000 level 4 (ipl 5): address 08:00:20:0d:88:a9
le0: 8 receive buffers, 2 transmit buffers
cgsix at sbus0 slot 3 offset 0x0 level 7 not configured
fd at mainbus0 ioaddr 0xf7200000 ipl 11 not configured
root on sd0a dumps on sd0b
root file system type: ffs

> ELC caching =? different;

This, if true, could very likely be relevant.

> What's max mem on an ELC?

64M.  The ELC has four SIMM sockets, which show up 16M apart in
physical space.

> Different MMU?

Maybe.  Does it?  I don't know.

> No sbus; possibly taking an sbus access fault?

It looks to software as though it has an sbus, certainly, though there
are no connectors where stuff can be plugged in.  Even the ROMs think
it has an sbus; setenv diag-switch? true and they print messages about
probing the various sbus slots....

> No floopy controller; possibly taking an fdc fault?

I'm inclined to discount this, and the sbus fault theory, largely
because whatever is going wrong doesn't crash the machine, but only the
userland process.

> A/B unified serial interface...

Well, yeah, but if we start in on all the mechanical differences we'll
be here all week. :-)  (I don't *think* the ELC serial interface is any
different from two ordinary serial interfaces that happen to be wired
to the same connector instead of two connectors....)

> Do you have an IPC or other SPARCstation you could test on?

An IPC, no.  I have an SS1+, but (a) I'm not sure it doesn't have
hardware problems already and (b) I'm not sure I can scare up enough 4M
30pin SIMMs to try it (with less than about 30M of RAM the mud will
drive the machine into the ground paging - ps alx reports a VSIZE of
27136 for the mud process).  Good thought though; I'll see if I can
scare up the memory to try that, even if it means taking some other
machine out of service for a little....

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B