Bug 21190

Summary: AMI MegaRAID 471 card doesn't work in Intel Lion slots 2-10
Product: [Retired] Red Hat Linux Reporter: Matt Domsch <matt_domsch>
Component: kernelAssignee: Michael K. Johnson <johnsonm>
Status: CLOSED RAWHIDE QA Contact: Brock Organ <borgan>
Severity: high Docs Contact:
Priority: high    
Version: 7.3CC: peterj
Target Milestone: ---   
Target Release: ---   
Hardware: ia64   
OS: Linux   
Whiteboard: Florence Gold
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2001-01-12 17:48:14 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 Matt Domsch 2000-11-21 17:02:58 UTC
The megaraid 471 card doesn't work in Intel Lion slots 2-10.
Using kernel 2.4.0-test9-ia64-001004 + megaraid v1.10c driver + IOSAPIC 
patch from Asit Mallic of Intel.  Card works in slots 1 & 2.

Failure mode is that after ~1400 interrupts, the card stops responding.   
This is easily demonstrated on a 4x9GB RAID5 array, one partition 
at /dev/sdb1, mke2fs /dev/sdb1.  SCSI timeouts will occur before the 
command completes.

PeterJ, Asit Mallic @ Intel, and Mike Stephens @ Intel are working 
to address this issue.

Comment 1 Glen Foster 2001-01-11 21:03:17 UTC
This defect is considered MUST-FIX for Florence Gold release

Comment 2 Matt Domsch 2001-01-11 21:40:48 UTC
megaraid driver v1.13s seems to address the problem.  MKJ says it'll be 
included in the public beta kernel.


Comment 3 Michael K. Johnson 2001-01-12 15:05:36 UTC
The latest megaraid driver is in and should fix this.  It breaks
Azusa, so a better fix needs to be provided, which means that we
need to continue to pay attention for now.

Comment 4 Matt Domsch 2001-01-12 17:48:10 UTC
The Azuza issue was with the qla1280 driver, not the megaraid driver.
In any case, BJ released a qla1280 driver today (v3.23) that addresses it.


Comment 5 Michael K. Johnson 2001-01-12 19:39:56 UTC
Oops, cache coherency issue in my mind.  Thanks, Matt!

Anyway, 1.13s is included.  My thinko.