On 6/19/24 5:10 PM, smitterl(a)redhat.com wrote:
Thank you, Boris.
Is the following test scenario sufficient to describe this change, are you aware of any
other scenarios that might be at risk of seeing a regression, maybe some scenario for
other storage devices like SCSI?
Given a DASD disk with identifier 0.0.4024 on subchannel 0.0.0030
When I deconfigure the device (chzdev -d 0.0.4024)
And I configure the device (chzdev -e 0.0.4204)
And I request the xml details of node devices (virsh nodedev-list --storage)
I am
not aware of a "storage" option on command nodedev-list. My guess
is that your are referring to "nodedev-list --tree".
Then the DASD disk is listed
And its parent is a device with capability ccw and identifier 0.0.4024
And the parent of that device is another device with capability css and identifier
0.0.0030
The first three patches do not actually change existing behavior but
instead fix the problem described in the referenced issue.
When the ccw device is configured (online) the relationships are and
also should have been as you describe above.
Theses patches are not modifying SCSI storage device node behavior.
--
Mit freundlichen Grüßen/Kind regards
Boris Fiuczynski
IBM Deutschland Research & Development GmbH
Vorsitzender des Aufsichtsrats: Wolfgang Wendt
Geschäftsführung: David Faller
Sitz der Gesellschaft: Böblingen
Registergericht: Amtsgericht Stuttgart, HRB 243294