Bug 688604

Summary: qemu-kvm: Add virtio disk identification support
Product: Red Hat Enterprise Linux 6 Reporter: Vivek Goyal <vgoyal>
Component: qemu-kvmAssignee: Virtualization Maintenance <virt-maint>
Status: CLOSED DUPLICATE QA Contact: Virtualization Bugs <virt-bugs>
Severity: medium Docs Contact:
Priority: medium    
Version: 6.2CC: amwang, ddutile, ehabkost, john.cooper, mkenneth, tburke, virt-maint
Target Milestone: rc   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-06-09 20:10:32 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Bug Depends On:    
Bug Blocks: 687042    

Description Vivek Goyal 2011-03-17 13:33:07 UTC
Description of problem:

Add virtio disk identification support in rhel6 qemu-kvm. I noticed same is present in upstream qemu-kvm.

Commit 2930b313dd602d67a568815b0b031b824916cec9
Author: john cooper <john.cooper>
Date:   Fri Jul 2 13:44:25 2010 -0400

     Add virtio disk identification support

In kdump we need to make sure that we are dumping to right device and due to device renaming issues we don't end up dumping to wrong device. Hence we are planning to read /sys/block/<virtio-device>/serial and store in kdump-initrd
and re-read it again in kdump kernel and match serial numbers before we proceed with saving vmcore.

Current plan is to read /sys directly. Is there a use space tool which can
do the job of getting the serial number?


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


How reproducible:


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


Expected results:


Additional info:

Comment 1 Vivek Goyal 2011-03-17 13:36:02 UTC
I have opened a kernel size bz 687042 for tracking the kernel virtio-blk changes to export device serial through /sys.

Comment 4 Vivek Goyal 2011-03-18 13:30:31 UTC
- i think you can try running lsscsi and see if device appears there or not.
- Or parse the device name. I am assuming that all virtio block devices are
  going to start with "v".
- Or if scsi_id does not report the device identity, then try to read serial
  number.

Comment 5 Vivek Goyal 2011-03-18 13:44:32 UTC
for virtio devices, udevadm is reporting following path info.

[root@rhel6-vm1 ~]# udevadm info --query=all --name=/dev/vdb
P: /devices/pci0000:00/0000:00:07.0/virtio3/block/vdb


I think you can grep for virtio keyword for recognzing virtio devices.

Comment 6 Vivek Goyal 2011-03-18 13:46:07 UTC
I think it is boiling down to trial and error thing. We can also probably look into udev and try to figure out what they are doing to identify different kind
of devices.

In long run, we really need udev in second kernel so that all this device identification logic is handled by udev and we don't end up duplicating it in
kexec-tools.

Comment 7 Cong Wang 2011-03-21 03:49:07 UTC
IIRC, the reason why we don't include udev is that it can cause OOM easily. Don't know how much mdev can do to replace udev...

Comment 9 Don Dutile (Red Hat) 2011-06-09 20:10:32 UTC
Closing as dupe of 710349, patch posted to rhel6.2 by Amit Shah

*** This bug has been marked as a duplicate of bug 710349 ***