Bug 1402951
Summary: | libvirt nodedev doesn't rescan the pci_devices | ||
---|---|---|---|
Product: | Red Hat Enterprise Linux 7 | Reporter: | Robin Cernin <rcernin> |
Component: | libvirt | Assignee: | Laine Stump <laine> |
Status: | CLOSED DUPLICATE | QA Contact: | Jingjing Shao <jishao> |
Severity: | high | Docs Contact: | |
Priority: | unspecified | ||
Version: | 7.3 | CC: | alex.williamson, berrange, dyuan, jsuchane, laine, rbalakri, rcernin, vcojot, vromanso, xuzhang |
Target Milestone: | rc | ||
Target Release: | --- | ||
Hardware: | Unspecified | ||
OS: | Unspecified | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | If docs needed, set a value | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2017-12-04 18:11:39 UTC | Type: | Bug |
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: | 1420851 |
Description
Robin Cernin
2016-12-08 17:37:40 UTC
(In reply to Robin Cernin from comment #0) > > Isn't that any nodedev command should rescan the PCI devices? No, libvirt is supposed to be notified of any changes by udev, so its view ought to be fully up2date without need for any manual refresh command. The reproduction steps in the Description don't match what I understood on IRC. In IRC it was explained that the host had a PF with 5 VFs when libvirtd started, then the PF was changed to have 63 VFs but libvirt still only showed 5 of them in the dumpxml for the PF. What I see here is that the dumpxml of the PF *does* show the new VFs, but that the VFs themselves can't be dumped. If that's the case, and if the host journal is showing errors like the following right after the VF count is changed: libvirtd[1605]: Device 0000:03:00.0 not found: could not access /sys/bus/pci/devices/0000:03:00.0/config: No such file or directory then I think this actually *is* the problem reported in Bug 1376907 (in spite of our best attempts to convince you otherwise :-) Can you please confirm/deny Laine's suspicions in comment 2? Thanks. Lacking other feedback, I'm assuming this is the same problem as Bug 1409957, and mark it as a duplicate. If you learn otherwise, feel free to reopen. *** This bug has been marked as a duplicate of bug 1409957 *** |