Re: HP NetServer boot trouble

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

>> [ACPI breakpoint opcode in NetServer LP 1000r]
>> Looking at the (NetBSD) code gives me the feeling this is some kind
>> of debugging facility and I could "fix" it by just having
>> AcpiOsSignal handle ACPI_SIGNAL_BREAKPOINT by silently returning
>> without doing anything.
> There were some ACPI tables in productive BIOSes that contained ACPI
> break points.  Those are skipped now (since 5.0 or so?), so if that
> really is the problem, try a 5.x kernel first.

No joy.  (It also appears they're not as skipped as you think.)

I found and burnt the 5.1 i386 ISO to a CD. Booting this the default
way, ie, with ACPI enabled, gives me (ten-finger copy)

Loading /miniroot.kmod
Copyright ...blah blah...
NetBSD 5.1 (GENERIC) #0: Sun Nov  7 14:39:56 UTC 2010
        
builds%b6.netbsd.org@localhost:/home/builds/ab/netbsd-5-1-RELEASE/i386/20101106194
32-obj/home/builds/ab/netbsd-5-1-RELEASE/src/sys/arch/i386/compile/GENERIC
total memory = 1024 MB
avail memory = 989 MB
Hewlett-Packard HP NetServer (LP 1000r)
mainbus0 (root)
cpu0 at mainbus0 apid 3: Intel 686-class, 1266MHz, id 0x6b1
cpu1 at mainbus0 apid 0: Intel 686-class, 1266MHz, id 0x6b1
ioapic0 at mainbus0 apid 1
ioapic1 at mainbus0 apid 2
acpi0 at mainbus0: Intel ACPICA 20080321
LNKU: BIOS IRQ 10 for 0.15.INTA is invalid
Executed AML Breakpoint opcode
fatal breakpoint trap in supervisor mode
trap type 1 code 0 eip c057a54c cs 8 eflags 282 cr2 0 ilevel 8
Stopped in pid 0.1 (system) at  netbsd:breakpoint+0x4:  popl    %ebp
db{0}> 

If I "c" here, it carries on with autoconf but wedges later (more
ten-finger copying):

scsibus0: waiting 2 seconds for devices to settle...
scsibus1: waiting 2 seconds for devices to settle...
cpu1: failed to start
fd0 at fdc0 drive 0: 1.44MB, 80 cyl, 2 head, 18 sec
uhub0 at usb0: vendor 0x1166 OHCI root hub, class 9/0, rev 1.00/1.00, addr 1
sd0 at scsibus0 target 0 lun 0: <DEC, RZ1CF-AF (C) DEC, 1614> disk fixed
sd0: 4091 MB, 3708 cyl, 20 head, 113 sec, 512 bytes/sect x 8380080 sectors
sd0: sync (100.00ns offset 31), 8-bit (10.000MB/s) transfers
ses0 at scsibus0 target 11 lun 0: <SDR, GEM318, 0> processor fixed
ses0: SAF-TE Compliant Device
ses0: async, 8-bit transfers
atapibus0 at atabus0: 2 targets
cd0 at atapibus0 drive 0: <CD-224E, , 1.5A> cdrom removable

If I instead pick the "no ACPI" option (ie, without also suppressing
SMP), it boots apparently normally - it comes up far enough for me to
NFS-mount another machine and dump dmesg output to it.

I also have reason to suspect that just disabling
ACPI_SIGNAL_BREAKPOINT won't fix 4.0.1; I tried just "c"ing from the
ddb prompt, and it appears to come up but gets errors trying to probe
the SCSI bus, then more errors trying to talk to the CD.  If I boot -c
and "disable acpi" these errors go away.

I guess I need to figure out whether MP support on this hardware is
worth trying to rebase all my changes atop 5.1.  Ugh.  :(  I think I'll
try disabling ACPI_SIGNAL_BREAKPOINT in 4.0.1 in the (admittedly faint)
hope that that is relevantly different from continuing from ddb....

/~\ 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