installing on rx2620

[Anton Shterenlikht]

Marcel

Here are some observations regarding installing on rx2620,
more precisely, the difference between installing on a local
disk and on a fibre disk. In both cases mpt(4) is used.

When installing on a local 17G disk, I get stopped at:

	Extracting base into / directory... 

	4170752 bytes read from base dist, chunk  3 of 59 @ 1390.2 KBytes/sec.


no panic, no reboot, just sits there indefinitely

At the same time VFP looks fine:

	LEDs    | LOCATOR      | SYSTEM       | POWER
	-----------------------------------------------------------------------------
	        | OFF          | ON GREEN     | ON GREEN
	-----------------------------------------------------------------------------
	Status  |  System running normally.
	-----------------------------------------------------------------------------


Now, what is really interesting is that if I try to install
over fibre, the installation passes "chunk 3" stage (where
the install onto a local disk hangs), but
on the console I see at that moment this LOR:

lock order reversal:
 1st 0xa00000007d58cc18 bufwait (bufwait) @ /usr/src/sys/kern/vfs_bio.c:2559
 2nd 0xe00000001197ae00 dirhash (dirhash) @ /usr/src/sys/ufs/ufs/ufs_dirhash.c:2
85
KDB: stack backtrace:
db_trace_self(0xe000000004130af0) at db_trace_self+0x20
db_trace_self_wrapper(0xe0000000046b6bd0) at db_trace_self_wrapper+0x70
kdb_backtrace(0xe000000004d6a470, 0xe0000000046e0280) at kdb_backtrace+0xc0
_witness_debugger(0x1, 0xe000000004bba030, 0xe0000000046e1b20, 0x999, 0xe0000000
04be2fc8) at _witness_debugger+0x60
witness_checkorder(0xe00000001197ae00, 0x9, 0xffffffffffffffff, 0x11d, 0x0) at w
itness_checkorder+0x12c0
_sx_xlock(0xe00000001197ae00, 0x0, 0xe000000004be2fc8, 0x11d) at _sx_xlock+0xc0
ufsdirhash_acquire(0xe000000011a565e8, 0xe00000001197ae00, 0xe000000004a7ecc0) a
t ufsdirhash_acquire+0x50
ufsdirhash_add(0xe000000011a565e8, 0xa0000000c5e73020, 0x810, 0xa0000000c5e72ff0
, 0xe000000004a86a20, 0xda1) at ufsdirhash_add+0x20
ufs_direnter(0xe000000011a5a5a0, 0xe000000011a66780, 0xa0000000c5e73020, 0xa0000
000c5e73378, 0x0) at ufs_direnter+0x10e0
ufs_makeinode(0xa0000000c5e73128, 0xe000000011a5a5a0, 0xa0000000c5e73350, 0xa000
0000c5

the best match I could find is this:
http://sources.zabbadoz.net/freebsd/lor/261.html

but the line numbers are quite different.


Finally, when I install on a disk via fibre, and
all was fine, no problems, I cannot find a bootable
partition on the disk. Below is an extract from
EFI map:

Device mapping table
  fs0  : Acpi(HWP0002,100)/Pci(1|0)/Scsi(Pun1,Lun0)/HD(Part1,Sig79A0351D-DB3C-11
DE-8F43-00110A31E60A)
  fs1  : Acpi(HWP0002,600)/Pci(1|0)/Fibre(WWN500805F3000EC221,Lun2000000000000)/
HD(Part1,Sig0044C4A1-3543-11DC-81DF-AA000400FEFF)
  fs2  : Acpi(HWP0002,600)/Pci(1|0)/Fibre(WWN500805F3000EC221,Lun2000000000000)/
HD(Part4,Sig0044C4A0-3543-11DC-81E0-AA000400FEFF)

On fs1 I've VMS bootable, fs2 must be the freebsd disk, but it's empty.
If I manually copy loader.efi from anothe ia64 box to fs2, I can't
then find a kernel.

will update

many thanks
anton

-- 
Anton Shterenlikht
Room 2.6, Queen's Building
Mech Eng Dept
Bristol University
University Walk, Bristol BS8 1TR, UK
Tel: +44 (0)117 331 5944
Fax: +44 (0)117 929 4423