Bug 510806
| Summary: | KVM: invoke pci_reset_function() when PCI device assigned or deassigned to a guest | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 5 | Reporter: | Don Dutile (Red Hat) <ddutile> | ||||
| Component: | kvm | Assignee: | Don Dutile (Red Hat) <ddutile> | ||||
| Status: | CLOSED ERRATA | QA Contact: | Lawrence Lim <llim> | ||||
| Severity: | medium | Docs Contact: | |||||
| Priority: | medium | ||||||
| Version: | 5.4 | CC: | chrisw, ddutile, lihuang, mjenner, sghosh, tburke, tools-bugs, virt-maint, ykaul | ||||
| Target Milestone: | rc | ||||||
| Target Release: | --- | ||||||
| Hardware: | All | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | kvm-83-94.el5 | Doc Type: | Bug Fix | ||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2009-09-02 09:28:41 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: | 510805 | ||||||
| Bug Blocks: | |||||||
| Attachments: |
|
||||||
|
Description
Don Dutile (Red Hat)
2009-07-10 21:07:24 UTC
Created attachment 354027 [details]
V2 of patch that is compatible w/upstream & RHEL5
Post V2 of patch.
Hi Don Could you tell how to reproduce/verify this bug ? I have tried to assign a NIC to guest. then shutdown the guest. everything seems ok. ( even with the pre-patched kvm-83-90.el5). Is it enough for this bug ? Reply from Don
This is not a bug, it is a feature.
One would have to have a PCIe logic analyzer to prove
the functionality is working 100%, or a broken card that is
only fixed by the reset ... but working good enough not to
crash either a KVM guest or it's host.
The way we checked/verified it's functionality:
(a) ported from upstream, and code source is from Intel
(b) I created a trace version of the code that outputted
what the code was doing as it executed (or didn't execute)
the code (it only is run on a PCIe device that has FLR support,
which is optional). Chris Wright tested it against PCIe devices
that had & didn't have FLR support.
I'm not sure how a qe test could verify or not verify the functionality,
since it is primarily a hw reset of a PCIe device.
- Don
verified the patch is applied in kvm-83-94.el5. rpm -q --changelog kvm | grep 510806 - kvm-Use-pci_reset_function-in-RHEL5.4-so-FLR-done-proper.patch [bz#510806] - Resolves: bz#510806 In the acceptance testing . no regression of this patch was found. setting to verified. 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 therefore 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/RHEA-2009-1272.html |