Bug 61059

Summary:	Hampton Beta Install Hangs on "Loading aic7xxx"
Product:	[Retired] Red Hat Public Beta	Reporter:	Richard Schaal <richard.schaal>
Component:	kernel	Assignee:	Arjan van de Ven <arjanv>
Status:	CLOSED WONTFIX	QA Contact:	Brian Brock <bbrock>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	fisher
Target Milestone:	---
Target Release:	---
Hardware:	i386
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2002-03-21 07:38:40 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Richard Schaal 2002-03-12 21:22:42 UTC

Description of Problem:
Installing Hampton Beta via network install hangs on message "loading aic7xxx"
for an extended period of time.  Finally, the installation says that it can't
find a disk to install on.  Had installed RedHat 7.2 two weeks earlier without
incident.  This also occurs after booting the CDROM directly.

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


How Reproducible:
Every time.


Steps to Reproduce:
1. Boot CDROM 1
2. 
3. 

Actual Results:
Long delay after loading scsi driver and ultimately a claim that no installable
disk exists.


Expected Results:
Should have found SCSI disk 0, and proceeded with partitioning.


Additional Information:
System is a Intel ISP2150G - 2X Pentium!!! 850MHz

Comment 1 Arjan van de Ven 2002-03-12 21:31:30 UTC

The intel ISP2150's have the 440GX chipset if I remember it right.
There is a known bios bug in this 440GX/ISP21xx series machine which gets the
IRQ routing ($PIR table) wrong. Instead of fixing this trivial bug, Intel has
decided to declare this machine unsupported by Linux.

In the past we had a (manually activatable) workaround for this bug; I will port
this to the hampton kernel in the next few weeks; I'm surprised someone from
Intel cares....it has been 10 months or so since we reported this bios bug.

Comment 2 Richard Schaal 2002-03-21 01:28:46 UTC

Just for grins, I built the "off the shelf" linux-2.4.18 kernel and 
booted it without difficulty.  - Richard

Comment 3 Arjan van de Ven 2002-03-21 07:38:35 UTC

Sure if you build a kernel with UP APIC enabled it'll work; same for a SMP kernel

Comment 4 Michael K. Johnson 2002-05-20 17:02:34 UTC

We've put the workaround in, but this is a bug in the Intel-provided BIOS
which we won't fix because we can't fix...

Comment 5 Richard Schaal 2002-05-20 22:14:27 UTC

I'm probably just obtuse, but just what _IS_ the work-around?  
Thanks.

Comment 6 Arjan van de Ven 2002-05-21 07:53:35 UTC

The workaround is to ignore the buggy $PIR table for this exact bios by adding
the DMI identifiers to the hall-of-shame.