Bug 438214

Summary: [QLogic 5.2 bug] qla4xxx: Fix scsi completion, lun reset code
Product: Red Hat Enterprise Linux 5 Reporter: David Somayajulu <david.somayajulu>
Component: kernelAssignee: Marcus Barrow <mbarrow>
Status: CLOSED ERRATA QA Contact: Martin Jenner <mjenner>
Severity: urgent Docs Contact:
Priority: high    
Version: 5.2CC: andriusb, coughlan, david.somayajulu, mbarrow, qlogic-redhat-ext
Target Milestone: rc   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: RHBA-2008-0314 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2008-05-21 15:12:12 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:
Bug Depends On:    
Bug Blocks: 216992, 439316    
Attachments:
Description Flags
qla4xxx: Fix scsi completion, lun reset code none

Description David Somayajulu 2008-03-19 18:18:27 UTC 
Description of problem:
1. Remove extraneous code and fix scsi command completion code so that when the
HBA indicates that the completion status is good, check the iscsi transport
status prior to processing scsi status
2. Firmware requirement is to issue one marker iocb per task management command.

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


How reproducible:


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:
The corresponding patch has already been posted to linux-scsi reflector
Comment 1 David Somayajulu 2008-03-19 18:18:27 UTC 
Created attachment 298561 [details]
qla4xxx: Fix scsi completion, lun reset code
Comment 2 Andrius Benokraitis 2008-03-19 18:37:55 UTC 
Just making sure this is a RHEL 5.3 item... correct?
Comment 3 David Somayajulu 2008-03-19 18:40:10 UTC 
If this can go into 5.2 that would be great!
Comment 4 Andrius Benokraitis 2008-03-19 19:10:07 UTC 
I think this should be assigned to Marcus, since the code is isolated to the
qla4xxx code, not anything Mike would be involved with.

Marcus - I know you're pretty hammered to get everything in - what's this priority?
Comment 5 Tom Coughlan 2008-03-20 01:05:31 UTC 
At this point in the 5.2 schedule, the only fixes we can take are for data
corruption, system crashers, or serious regressions introduced in 5.2. Does this
fit that bill?
Comment 6 Marcus Barrow 2008-03-20 01:19:02 UTC 
I corresponded with David S. earlier today. This is a serious issue where one of several status that 
comprise the success or failure of an I/O was missed. In addition a required handshake " a marker" with
the HBA is not properly performed.

The problem with the status would be that a iSCSI Transport/Completions are not properly identified.
This issue would cause a data corruption, because the error would be missed.

Secondly the problem with the handshake would cause the device queue on the HBA to block, preventing
further I/o.
Comment 9 Don Zickus 2008-03-26 20:32:03 UTC 
in kernel-2.6.18-87.el5
You can download this test kernel from http://people.redhat.com/dzickus/el5
Comment 12 errata-xmlrpc 2008-05-21 15:12:12 UTC 
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2008-0314.html