>Number:         7678
>Category:       port-sparc
>Synopsis:       Sun 4/110 failing to boot diskless
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    port-sparc-maintainer (NetBSD/sparc Portmaster)
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon May 31 09:35:00 1999
>Last-Modified:
>Originator:     Brad Spencer
>Organization:
	At home
>Release:        1.4C supped week of 5/24/1999
>Environment:
	Sun 4/110, NetBSD 1.4C
System: 


>Description:

My 4/110 failed to boot diskless when trying NetBSD 1.4C.  The
following is a snip of the boot dialog:

NetBSD 1.4C (GENERIC) #0: Sat May 29 21:15:48 EDT 1999
    brad@elrond.eldar.org:/usr/src/sys/arch/sparc/compile/GENERIC
total memory = 8136 KB
avail memory = 5352 KB
using 63 buffers containing 504 KB of memory
bootpath: /obio0/ie0
mainbus0 (root): SUN-4/100 series
cpu0 at mainbus0: MB86900/1A or L64801 @ 14.280 MHz, MB86910 or WTL1164/5 FPU
cpu0: no cache
.
.
.
nfs_boot: trying RARP (and RPC/bootparam)
ie0: TDR detected an open 22 clocks away
ie0: 65 frames 195 bufs
ie0: reserved 56916 bytes
nfs_boot: client_addr=206.21.77.209 (RARP from 206.21.77.209)
ie0: 65 frames 195 bufs
ie0: reserved 56916 bytes
nfs_boot: server_addr=206.21.77.209
nfs_boot: hostname=aragorn
nfs_boot: timeout...
nfs_boot: timeout...
nfs_boot: timeout...
root on mellon.eldar.org:/export2/root/aragorn
root file system type: nfs
data fault: pc=0xf0198e78 addr=0xf0214000 ser=8020<WRITE,TIMEOUT>
panic: kernel fault
syncing disks... done
Frame pointer is at 0xf14bda28
Call traceback:
  pc = 0xf019afd4  args = (0x4000e6, 0x400fe6, 0x0, 0x0, 0xf14bdb40, 0x9000e6, 0xf14bda90) fp = 0xf14bda90
  pc = 0xf0051580  args = (0x100, 0x0, 0x0, 0x0, 0xf14bdbb4, 0x9000e0, 0xf14bdaf8) fp = 0xf14bdaf8
  pc = 0xf01ab18c  args = (0xf01aae38, 0x100, 0xf0214000, 0xf14bdbc0, 0x2d, 0xf01fbc00, 0xf14bdb60) fp = 0xf14bdb60
  pc = 0xf000857c  args = (0x9, 0x8020, 0xf0214000, 0xf0198e78, 0x40, 0xf14bdc68, 0xf14bdc08) fp = 0xf14bdc08
  pc = 0xf01a8a34  args = (0x1fff, 0x0, 0xf0214000, 0xf0002000, 0x0, 0x0, 0xf14bdcb8) fp = 0xf14bdcb8
  pc = 0xf0142fec  args = (0xa0000194, 0x32a000, 0xf0208c00, 0xf01a8928, 0xf01a851c, 0x0, 0xf14bdd20) fp = 0xf14bdd20
  pc = 0xf013c8c8  args = (0xf0227b04, 0xf0227b38, 0xf01fbda8, 0x1, 0x0, 0x0, 0xf14bdd88) fp = 0xf14bdd88
  pc = 0xf01ab054  args = (0xf0227b38, 0x2, 0x7, 0xf14bde00, 0xf14d2050, 0xf01fbc00, 0xf14bdea8) fp = 0xf14bdea8
  pc = 0xf000857c  args = (0xf14c6000, 0x8080, 0x30004, 0x204c, 0x0, 0xf14bdfb0, 0xf14bdf50) fp = 0xf14bdf50
  pc = 0x0  args = (0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xeffffd08) fp = 0xeffffd08
rebooting


It appears that it is crashing around the time that init forks.  It
should be noted that the userland for this machine is 1.4_ALPHA.

I do not suspect the userland code, as 5 other Suns [all Sun 4c
machines] are using it in a disk or diskless configuration.  The
1.4_ALPHA promlib code does not support the PROM on the 4/110
properly, otherwise that would have been used.

>How-To-Repeat:
	Try booting a 4/110 diskless
>Fix:

I am not sure, however, I could test and work with someone who is more
in the know then I am.
>Audit-Trail:
>Unformatted: