Bug#275627: [sparc][netinst][20041007] failure: sun blade 100
Vincent McIntyre
Package: installation-reports
Severity: critical
INSTALL REPORT
Debian-installer-version:
netinst 20041007 from
http://cdimage.debian.org/pub/cdimage-testing/daily/sparc/20041007/sarge-sparc-netinst.iso (md5sum b370379763a9e7efbe3439ab5c16113e )
% file sarge-sparc-netinst.iso
sarge-sparc-netinst.iso: Sun disk label 'CD-ROM Disc with Sun sparc boot created by mkisofs' 768 alts/cyl, 768 data cyls, 127 alt cyls, 8323840 blocks, boot block present
uname -a: 2.4.27
Date: 2004 10 8
Method:
hit L1-A after openboot banner appears
type 'boot cdrom' in OpenBoot Prom 4.0
type <return> to start installer
Machine: Sun SunBlade 100
Processor: Ultrasparc-IIe (ie sun4u, or "sparc64" arch)
Memory: 384Mb
Root Device: ide
Root Size/partition table:
Output of lspci and lspci -n:
Base System Installation Checklist:
[O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it
Initial boot worked: [E]
Configure network HW: [ ]
Config network: [ ]
Detect CD: [ ]
Load installer modules: [ ]
Detect hard drives: [ ]
Partition hard drives: [ ]
Create file systems: [ ]
Mount partitions: [ ]
Install base system: [ ]
Install boot loader: [ ]
Reboot: [ ]
Comments/Problems:
The netinst cdrom is _still_ panicing with this message:
Allocated 8 Megs of memory at 0x40000000 for kernel
Loading initial ram disk of 2890906 bytes at 0x2F800000 phys ...something...
.... lots of messages missing ....
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
RAMDISK: couldn't find valid ram disk image starting at 0
Freeing initrd memory: 2823k freed
cramfs: wrong magic
Kernel panic: VFS: Unable to mount root fs on 00:00
Press L1-A to return to the boot prom
However if I boot _again_, I get to the "Choose Language" prompt.
Unfortunately at this point the USB keyboard does not work, not even L1-A.
If I boot a third time, I get to the same situation.
I tried unplugging the usb mouse, leaving just the keyboard in.
This made no difference.
I could not obtain any boot logs, apart from what I could transcribe
from the screen.
The woody installer fails too, with
Cannot find /boot/sparc64.gz
(this may be my fault, I am not sure if the disc was sparc32-only)
--
To UNSUBSCRIBE, email to debian-bugs-...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Herbert Kaminski
>Debian-installer-version:
> netinst 20041007 from
I have seen a similar boot problem with my Ultra5 for several days.
This was fixed with next day's (2004-10-08) businesscard iso.
Later on, however, the installer is unable to mount the cdrom.
Joshua Kwans's mini-iso:
http://people.debian.org/~joshk/d-i/images/daily/
currently does a better job, though stage 2 still got some bugs.
Regards
Hebert
Colleen Hatfield
Problem 1:
Same as Vincent; keyboard doesn't work from the first installer
screen. I am using a Sun Type 6 US UNIX layout USB keyboard.
Problem 2:
Same as Herbert. Attempting to work around problem 1, I unplugged
keyboard/video/mouse and connected a serial console to Debian x86 box
(minicom). I get to the "Detecting CD-ROM" step, where it tells me
that it cannot mount the installer CD-ROM, and that it's possible that
it's not in the drive. The CD-ROM is definitely in the drive. ;-)
As requested, lspci & lspci -n (courtesy of Gentoo install):
0000:00:00.0 Host bridge: Sun Microsystems Computer Corp. Ultra IIe
0000:00:03.0 Non-VGA unclassified device: ALi Corporation M7101 Power
Management Controller [PMU]
0000:00:05.0 PCI bridge: Digital Equipment Corporation DECchip 21152 (rev 03)
0000:00:07.0 ISA bridge: ALi Corporation M1533 PCI to ISA Bridge [Aladdin IV]
0000:00:08.0 Multimedia audio controller: ALi Corporation M5451 PCI
AC-Link Controller Audio Device (rev 01)
0000:00:0c.0 Bridge: Sun Microsystems Computer Corp. RIO EBUS (rev 01)
0000:00:0c.1 Ethernet controller: Sun Microsystems Computer Corp. RIO
GEM (rev 01)
0000:00:0c.2 FireWire (IEEE 1394): Sun Microsystems Computer Corp. RIO
1394 (rev 01)
0000:00:0c.3 USB Controller: Sun Microsystems Computer Corp. RIO USB (rev 01)
0000:00:0d.0 IDE interface: ALi Corporation M5229 IDE (rev c3)
0000:00:13.0 VGA compatible controller: ATI Technologies Inc Rage XL (rev 65)
0000:00:00.0 Class 0600: 108e:a001
0000:00:03.0 Class 0000: 10b9:7101
0000:00:05.0 Class 0604: 1011:0024 (rev 03)
0000:00:07.0 Class 0601: 10b9:1533
0000:00:08.0 Class 0401: 10b9:5451 (rev 01)
0000:00:0c.0 Class 0680: 108e:1100 (rev 01)
0000:00:0c.1 Class 0200: 108e:1101 (rev 01)
0000:00:0c.2 Class 0c00: 108e:1102 (rev 01)
0000:00:0c.3 Class 0c03: 108e:1103 (rev 01)
0000:00:0d.0 Class 0101: 10b9:5229 (rev c3)
0000:00:13.0 Class 0300: 1002:4752 (rev 65)
I can also provide my dmesg output from the Gentoo boot, or my working
2.4.27 kernel config if that would help anyone.
Vincent, re. your "cramfs: wrong magic" intermittent problem:
I have this problem on my Blade 100 as well, when I don't hit Stop+A
at precisely the instant that the machine wants it. On my machine I
have to send break after initializing memory, and *immediately* after
it sends the sync signal to the monitor (Boot device: disk a or
whatever is the last line above the "ok" I get post-break). If I wait
any longer than this, I get the "cramfs: wrong magic" issue when
trying to boot from the installer CDROM. If I wait too long before
sending the break when trying to boot disk2 (my Gentoo installation),
I get a really ugly "kernel panic: Aiee!" regarding the swapper. This
is reproducible 100% of the time on my Blade.
Thanks,
Colleen
Vincent....@csiro.au
I made some more (unsuccessful) attempts.
> Vincent, re. your "cramfs: wrong magic" intermittent problem:
> I have this problem on my Blade 100 as well, when I don't hit Stop+A
> at precisely the instant that the machine wants it. On my machine I
> have to send break after initializing memory, and *immediately* after
> it sends the sync signal to the monitor (Boot device: disk a or
> whatever is the last line above the "ok" I get post-break). If I wait
> any longer than this, I get the "cramfs: wrong magic" issue when
> trying to boot from the installer CDROM. If I wait too long before
> sending the break when trying to boot disk2 (my Gentoo installation),
> I get a really ugly "kernel panic: Aiee!" regarding the swapper. This
> is reproducible 100% of the time on my Blade.
Attempt 1.
netinst 20041015 (md5 ed4ce23d8135ec1ed287d307230e12d2)
http://cdimage.debian.org/pub/cdimage-testing/daily/sparc/20041015/sarge-sparc-netinst.iso
Power up.
Hit L1-A just after it says
Boot device: disk:a File and args:
ok boot cdrom
boot: <return>
still fails, with cramfs and kernel panic as before
In case it is helpful, this is what I could transcribe of the boot log
....
PCI: found SABRE, main regs at 0000001fe00000000, wsync at
000001fe00001c20
SABRE: Shared PCI config space at 000001fe01000000
SABRE: DVMA at c0000000 [20000000]
PCI-IRQ: Routing bus[ 0] slot[ 8] map[0] to INO[23]
PCI-IRQ: Routing bus[ 0] slot[ c] map[0] to INO[06]
PCI-IRQ: Routing bus[ 0] slot[ c] map[0] to INO[1c]
PCI-IRQ: Routing bus[ 0] slot[ c] map[0] to INO[24]
PCI-IRQ: Routing bus[ 0] slot[ d] map[0] to INO[0c]
PCI-IRQ: Routing bus[ 0] slot[13] map[0] to INO[0f]
PCIO(PBMA): Bus running at 33MHz
isa0: [dma -> (floppy) (parallel)] ]power] [serial] [serial]
ebus0: [flashprom] [idprom]
PCIO serial driver version 1.54
su(serial) at 0x1fe020003f8 (tty 0 irq 12.7eb) is a 16550A
su(serial) at 0x1fe020002e8 (tty 0 irq 12.7eb) is a 16550A
Linux NET4.0 for Linux 2.4
Based on Swansea University Computer Society NET3.039
Initialising RT netlink socket
Starting kswapd
VFS: Disk quotas vdquot_6.5.1
devfs: v1.12c (20020818) Richard Gooch (rgo...@atnf.csiro.au)
devfs: boot_options: 0x1
atyfb: 3D Rage (XL) [0x4752 rev 0x27] 8M SDRAM, 29.498928 MHz XTAL, 235
MHz PLL, 83 MHz MCLK, 63 MHz XCLK
Console: switching to colour frame buffer device 144x56
fb0: ATY Mach64 frame buffer device on PCI
kbd_init: Assuming USB keyboard.
isa(speaker): iobase[000001fe02000042:000001fe02000061]
pty: 256 Unix96 ptys configured
rtc_init: no PC rtc found
RAMDISK driver initialized: 16 RAM disks of 16384K size 1024 blocksize
Uniform Multi-Platform E-IDE Driver Revision: 7.00beta4-2.4
ide: Assuming 33MHz system bus speed for PIO modes: override with
idebus=xx
usb.c: registered new driver usbdevfs
usb.c: registered new driver hub
usb.c: registered new driver usbmouse
usbmouse.c: v1.6:USB HID Boot Protocol mouse driver
usb.c: registered new driver usbkbd
usbkbd.c: USB HID Boot Protocol keyboard driver
Initializing Cryptographic API
NET4: Linux TCP/IP 1.0 for NET4.0
IP: routing cache hash table of 8192 buckets, 64Kbytes
TCP: Hash tables configured (established 65536 bind 65536)
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0
RAMDISK: Couldn't find a valid RAM disk image starting at 0.
Freeing initrd memory: 2027k freed
cramfs: wrong magic
Kernel panic: VFS: Unable to mount root fs on 00:00
Press L1-A to return to the boot prom
At this point, the keyboard does not work.
Power down.
Power up.
Hit L1-A in the middle of "Initializing memory"
ok boot cdrom
boot: <return>
successful boot, but the keyboard is nonfunctional in "Choose Language".
Attempt 2.
JoshK mini-iso from 20041015 (md5 53d25f088f26bee748e85dab4da48b88)
Power up.
L1-A in middle of boot device: disk: a File and args:
ok boot cdrom
Boots successfully, but same problem with kbd at "Choose Language" step.
So Josh seems to have solved the cramfs problem.
Power down.
Power up.
hit L1-A earlier, after "Initializing Memory" but before the
boot device: message appears.
ok boot cdrom
boot: <return>
Fails in same way - boots ok but kbd is nonresponsive at "Choose
Language" step.
I haven't had a go with Gentoo yet, I'd rather not introduce more
variables.
I will have a go with a serial console soon and let you know if
I get to the point where I can get lspci output.
Cheers
Vince
Vincent....@csiro.au
These were all done with the serial port connected, but that's all.
No monitor, no keyboard, no mouse.
1. JoshK mini-iso (20041015)
Power up.
Let solaris boot part way (I had the wrong baud rate set)
send <BRK>
ok boot cdrom
system starts booting, panics unable to mount root.
see joshk.log in attachment.
The system does not respond to <BRK> or whatever, I had to power down.
I tried again but had trouble after sending the BRK, the
serial port was sometimes left in a mangled state; I had an
ok prompt from the PROM but was unable to type anything.
Sending <BRK> during the solaris boot always worked.
2. sarge netinst 20041015
Tried sending <BRK> during the solaris boot, as above.
Fails to boot, cramfs problem.
See daily.log
Could not get around this by finding a magic moment to interrupt
the boot.
Should I be giving debian-installer/probe/usb=false ?
It's not really clear to me usb is the problem in the tests above.
Can I get the lspci data from within solaris?
Regards