Re: 600au: SRM setting doesn't stick?

[Mouse <mouse%Rodents-Montreal.ORG@localhost>]

> The CMOS battery should be on the mainboard, but the mainboard is not
> the mainboard if that makes sense.  [...]
> Just in case, here's the service manual, 5-12 and onward is what
> you're looking for:
> https://theretroweb.com/motherboard/manual/b30wwsma-60841c95f009e303767433.pdf

Thank you!  That helps greatly.  I saw the lifter handles but it looked
as though there were a single board all the way through the machine.
(I found a work machine I could fetch that page from - odd, that
something calling itself "the retro web" would desupport HTTP.  Oh
well, I did find a work machine to fetch it with.)

I've now replaced the battery.  Now there's another issue, leading me
to wonder if I managed to break something in the process.

Before I replaced the battery, when I was using a fast power-cycle as a
reset, I netbooted at least three times and from disk at least once.
Once it loaded the kernel (which failed for at least one netboot), it
was fine.  Now, it panics early in autoconf, regardless of whether I
boot from the net or the disk.  The last working boot, before battery
replacement, looked like

>>>boot dkd0.0.0.1010.0
(boot dkd0.0.0.1010.0 -flags 0)
block 0 of dkd0.0.0.1010.0 is a valid boot block
reading 14 blocks from dkd0.0.0.1010.0
bootstrap code read in
base = 1d6000, image_start = 0, image_bytes = 1c00
initializing HWRPB at 2000
initializing page table at 1c8000
initializing machine state
setting affinity to the primary CPU
jumping to bootstrap code

NetBSD/alpha 1.4T FFS Primary Bootstrap
Jumping to entry point...

NetBSD/alpha 1.4T Secondary Bootstrap, Revision 1.10
(mouse%Nought.Rodents.Montreal.QC.CA@localhost, Feb 12 05:55:30 EST 2002)

VMS PAL rev: 0x1000000010114
OSF PAL rev: 0x1000000020116
Switch to OSF PAL code succeeded.

Boot flags: 0
4064304+319120 [85+262608+151177]=0x4936f8

Entering netbsd at 0xfffffc0000301040...
Unrecognized boot flag '0'.
[ preserving 414776 bytes of netbsd ELF symbol table ]
consinit: not using prom console
Copyright (c) 1996, 1997, 1998, 1999, 2000
    The NetBSD Foundation, Inc.  All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
    The Regents of the University of California.  All rights reserved.

NetBSD 1.4T (GENERIC) #0: Wed Dec  6 00:17:24 EST 2023
    mouse@:/home/mouse/kbuild/GENERIC
Digital Personal WorkStation 600au, 598MHz
8192 byte page size, 1 processor.
total memory = 512 MB
(1888 KB reserved for PROM, 510 MB used by NetBSD)
avail memory = 469 MB
using 3277 buffers containing 26216 KB of memory
mainbus0 (root)
cpu0 at mainbus0: ID 0 (primary), 21164A-0 (unknown minor type 0)
cia0 at mainbus0: DECchip 2117x Core Logic Chipset (Pyxis), pass 1
cia0: extended capabilities: 1<BWEN>
cia0: using BWX for PCI config access
pci0 at cia0 bus 0
pci0: i/o space, memory space enabled, rd/line, rd/mult, wr/inv ok
de0 at pci0 dev 3 function 0
[etc]

In any case, now, I get

>>>boot dkd0
(boot dkd0.0.0.1010.0 -flags 0)
block 0 of dkd0.0.0.1010.0 is a valid boot block
reading 14 blocks from dkd0.0.0.1010.0
bootstrap code read in
base = 1d6000, image_start = 0, image_bytes = 1c00
initializing HWRPB at 2000
initializing page table at 1c8000
initializing machine state
setting affinity to the primary CPU
jumping to bootstrap code

NetBSD/alpha 1.4T FFS Primary Bootstrap
Jumping to entry point...

NetBSD/alpha 1.4T Secondary Bootstrap, Revision 1.10
(mouse%Nought.Rodents.Montreal.QC.CA@localhost, Feb 12 05:55:30 EST 2002)

VMS PAL rev: 0x1000000010114
OSF PAL rev: 0x1000000020116
Switch to OSF PAL code succeeded.

Boot flags: 0
4064304+319120 [85+262608+151177]=0x4936f8

Entering netbsd at 0xfffffc0000301040...
Unrecognized boot flag '0'.
[ preserving 414776 bytes of netbsd ELF symbol table ]
consinit: not using prom console
Copyright (c) 1996, 1997, 1998, 1999, 2000
    The NetBSD Foundation, Inc.  All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
    The Regents of the University of California.  All rights reserved.

NetBSD 1.4T (GENERIC) #0: Wed Dec  6 00:17:24 EST 2023
    mouse@:/home/mouse/kbuild/GENERIC
Digital Personal WorkStation 600au, 598MHz
8192 byte page size, 1 processor.
total memory = 512 MB
(1888 KB reserved for PROM, 510 MB used by NetBSD)
avail memory = 469 MB
using 3277 buffers containing 26216 KB of memory
mainbus0 (root)
cpu0 at mainbus0: ID 0 (primary), 21164A-0 (unknown minor type 0)
cia0 at mainbus0: DECchip 2117x Core Logic Chipset (Pyxis), pass 1
cia0: extended capabilities: 1<BWEN>
cia0: using BWX for PCI config access
sio_intr_setup: 8 LT!
sio_sis: initial: ocw = (fb,ff), elcr = (a0,a9)
         current: ocw = (fb,ff), elcr = (a0,a8)
panic: sio_setirqstat: hosed
Stopped in swapper at   cpu_Debugger+0x4:       ret     zero,(ra)
db> 

I've tried three boots, and there is slight variation.  The last few
lines of the other two boots are

cia0: using BWX for PCI config access
sio_intr_setup: 0 LT!
sio_sis: initial: ocw = (fb,ff), elcr = (a1,a9)
         current: ocw = (fb,ff), elcr = (a0,a9)
panic: sio_setirqstat: hosed
Stopped in swapper at   cpu_Debugger+0x4:       ret     zero,(ra)
db> 

and

cia0: using BWX for PCI config access
sio_intr_setup: 0 LT!
sio_sis: initial: ocw = (fb,ff), elcr = (a1,a9)
         current: ocw = (fb,ff), elcr = (a0,a9)
panic: sio_setirqstat: hosed
Stopped in swapper at   cpu_Debugger+0x4:       ret     zero,(ra)
db> 

I notice BROKEN_PROM_CONSOLE, which controls that test, is gone in
5.2's version of that file; I would tend to just try 5.2 instead, but
when it "worked fine" before the battery change I have to wonder.

I'm going to try reseating the board I removed/reinstalled....

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse%rodents-montreal.org@localhost
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B