Description of Problem:
Total lockup (no keyboard led, nothing) during installation of packages
if target disk doesn't sits on hda (if it does lockup happens during
partition read at boot).


Version-Release number of selected component (if applicable):


How Reproducible:
Disk on hda: 100% reproducible. 
Disk on hdg: happens sooner or later.

Steps to Reproduce:
1. boot with "linux ide0=0xd800" -> unable to read partitions
2.
3. 

Actual Results:
Freeze. No visible disk activity (led).


Expected Results:


Additional Information:
Highly suspicious hardware: bp6+hpt366+ibm HD+G400.
IBM HD health software had no problem rewriting the whole
disk.

Caught this one:
	Oops: 0000
CPU:    0
EIP:    0010:[<360ce4e4>]
EFLAGS: 00010017
eax: e147b812   ebx: 0000005b   ecx: 000023ae   edx: 00000b6a
esi: c022e000   edi: 00000000   ebp: c022ff98   esp: c022fedc
ds: 0018   es: 0018   ss: 0018
Process swapper (pid: 0, stackpage=c022f000)
Stack: c011bb43 00000009 c022e000 00000001 c191d850 00000020 00000000
c191d850
       c022e000 00000001 00000000 c011bbaa c022e000 00000000 00000001
00000000
       00002e93 c022ff98 00000000 c022ff98 c011bdd3 00000000 c010b821
c022ff98
Call Trace: [<c011bb43>] [<c011bbaa>] [<c011bdd3>] [<c010b821>]
[<c018960c>]
   [<c0107fed>] [<c010813c>] [<c0105344>] [<c010a1e8>] [<c0105344>]
[<c0105367>]
   [<c01053ce>] [<c0105000>] [<c01f2500>]

Code:  Bad EIP value.
 <0>Kernel panic: Aiee, killing interrupt handler!
In interrupt handler - not syncing