Note: This is a public test instance of Red Hat Bugzilla. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback at bugzilla.redhat.com.
Bug 1768551 - qemu exits immediately on power8 box (with older firmware?) on F31, with "Requested safe cache capability level not supported by kvm, try cap-cfpc=broken"
Summary: qemu exits immediately on power8 box (with older firmware?) on F31, with "Req...
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: Fedora
Classification: Fedora
Component: qemu
Version: 31
Hardware: ppc64le
OS: Linux
unspecified
urgent
Target Milestone: ---
Assignee: Fedora Virtualization Maintainers
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
: 1769445 (view as bug list)
Depends On:
Blocks: PPCTracker
TreeView+ depends on / blocked
 
Reported: 2019-11-04 17:15 UTC by Adam Williamson
Modified: 2020-11-03 22:27 UTC (History)
12 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2020-11-03 22:27:53 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)

Description Adam Williamson 2019-11-04 17:15:14 UTC
Since I upgraded it to Fedora 31, I cannot run virt-install on one of the ppc64le worker host boxes for openQA:

ERROR    internal error: qemu unexpectedly closed the monitor: 2019-11-04T17:10:51.239226Z qemu-system-ppc64: Requested safe cache capability level not supported by kvm, try cap-cfpc=broken
Removing disk 'disk_f31_minimal_3_ppc64le.img.tmp'                                             |    0 B  00:00:00     
Domain installation does not appear to have been successful.
If it was, you can restart your domain by running:
  virsh --connect qemu:///system start createhdds
otherwise, please restart your installation.
virt-install command virt-install --disk size=20,path=disk_f31_minimal_3_ppc64le.img.tmp --os-variant fedora-unknown -x inst.ks=file:/minimal.ks --initrd-inject /root/createhdds/minimal.ks --location https://download.fedoraproject.org/pub/fedora-secondary/releases/31/Everything/ppc64le/os/ --name createhdds --memory 4096 --noreboot --wait -1 --graphics vnc --noautoconsole --network user failed!

this worked fine in F30. Googling around for the error message suggests we may be missing a patch for qemu:

https://github.com/qemu/qemu/commit/b2540203bdf4a390c3489146eae82ce237303653#diff-b0796c6a577709f249c7c5d96b9ea049

via: https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1765364

Comment 1 Adam Williamson 2019-11-06 17:40:36 UTC
Hmm. I note in this relevant-looking qemu commit:

https://github.com/qemu/qemu/commit/2782ad4c4102d57f7f8e135dce0c1adb0149de77

"By now machine firmware should have been upgraded to allow these settings."

I wonder if perhaps the firmware on this machine...hasn't?

Comment 2 Adam Williamson 2019-11-06 19:04:40 UTC
Indeed, running with `-M pseries-3.1` works, which supports my theory.

Comment 3 Adam Williamson 2019-11-07 15:42:56 UTC
*** Bug 1769445 has been marked as a duplicate of this bug. ***

Comment 4 Adam Williamson 2019-11-07 15:46:33 UTC
Note this is somewhat similar to, but I think not the *same* as, https://bugzilla.redhat.com/show_bug.cgi?id=1769600 . This case is on power8, and qemu *exits* almost immediately on launch with the error message. That case is on power9, and qemu runs, but cannot successfully boot much past the bootloader. Also, this bug can only be worked around with the pseries-3.1 or lower machine type: trying with `-M pseries-4.0` still produces the error. On the other hand, bug#1769600 can be worked around with pseries-4.0.

Comment 5 Laurent Vivier 2019-11-07 17:56:33 UTC
(In reply to Adam Williamson from comment #1)
> Hmm. I note in this relevant-looking qemu commit:
> 
> https://github.com/qemu/qemu/commit/2782ad4c4102d57f7f8e135dce0c1adb0149de77
> 
> "By now machine firmware should have been upgraded to allow these settings."
> 
> I wonder if perhaps the firmware on this machine...hasn't?

So you could restore previous behavior with:

  ... -M pseries,cap-cfpc=broken,cap-sbbc=broken,cap-ibs=broken ...

What is you firmware version: sudo update_flash -d ?

Comment 6 Laurent Vivier 2019-11-07 18:09:40 UTC
Sam,

do you know which firmware version we need to support cap-cfpc=workaround,cap-sbbc=workaround,cap-ibs=workaround?

Comment 7 Adam Williamson 2019-11-07 19:30:53 UTC
Oh, sorry, thought that info was here already but it wasn't (I mentioned it in one of the dupes). It is old indeed:

[root@openqa-ppc64le-01 adamwill][PROD]# lsmcode
Version of System Firmware : 
 Product Name          : OpenPOWER Firmware
 Product Version       : IBM-habanero-ibm-OP8_v1.7_1.62
 Product Extra         : 	hostboot-bc98d0b-1a29dff
 Product Extra         : 	occ-0362706-16fdfa7
 Product Extra         : 	skiboot-5.1.13
 Product Extra         : 	hostboot-binaries-43d5a59
 Product Extra         : 	habanero-xml-a71550e-cdd3b31
 Product Extra         : 	capp-ucode-105cb8f

I filed an infra issue to get the firmware updated, but doing that seems to require a PhD in ppc64-ology :) https://pagure.io/fedora-infrastructure/issue/8365

Comment 8 Michel Normand 2019-11-12 15:44:43 UTC
I verified on my local Habanero machine that after firmware upgrade to 8348_820.1923.20190613n, I do not have anymore qemu-4.1 failure with default pseries model.

I searched FW OP820 for 8348-21C from 
https://www-945.ibm.com/support/fixcentral/
After installation new lsmcode reports
===
$sudo lsmcode
Version of System Firmware :
 Product Name          : OpenPOWER Firmware
 Product Version       : IBM-habanero-OP8_v1.12_2.96
 Product Extra         :        op-build-v2.3-5-g98aa884
 Product Extra         :        skiboot-v6.3.1
 Product Extra         :        bmc-firmware-version-2.16
 Product Extra         :        occ-p8-a2856b7
 Product Extra         :        hostboot-p8-c893515-pd6f049d
 Product Extra         :        buildroot-2019.02.1-16-ge01dcd0
 Product Extra         :        capp-ucode-p9-dd2-v4
 Product Extra         :        machine-xml-c3998a5
 Product Extra         :        hostboot-binaries-hw041519a.opv23
 Product Extra         :        petitboot-v1.10.3
 Product Extra         :        linux-5.0.7-openpower1-p8e31f00
===

Comment 9 Adam Williamson 2019-11-12 15:52:23 UTC
Thanks! Any chance you can help us upgrade ours, now?

Comment 10 Michel Normand 2019-11-12 16:34:45 UTC
(In reply to Adam Williamson from comment #9)
> Thanks! Any chance you can help us upgrade ours, now?

I will add comments in the related infra issue.

Comment 11 Sam Bobroff (IBM) 2019-11-14 04:49:22 UTC
(In reply to Laurent Vivier from comment #6)
> Sam,
> 
> do you know which firmware version we need to support
> cap-cfpc=workaround,cap-sbbc=workaround,cap-ibs=workaround?

I'm sorry but I don't.

Do you still need to know where support begins, or can you continue using the version that Michel Normand posted above?
(I can probably find out, but it's not a trivial task for me.)

Comment 12 Ben Cotton 2020-11-03 16:47:44 UTC
This message is a reminder that Fedora 31 is nearing its end of life.
Fedora will stop maintaining and issuing updates for Fedora 31 on 2020-11-24.
It is Fedora's policy to close all bug reports from releases that are no longer
maintained. At that time this bug will be closed as EOL if it remains open with a
Fedora 'version' of '31'.

Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version' 
to a later Fedora version.

Thank you for reporting this issue and we are sorry that we were not 
able to fix it before Fedora 31 is end of life. If you would still like 
to see this bug fixed and are able to reproduce it against a later version 
of Fedora, you are encouraged  change the 'version' to a later Fedora 
version prior this bug is closed as described in the policy above.

Although we aim to fix as many bugs as possible during every release's 
lifetime, sometimes those efforts are overtaken by events. Often a 
more recent Fedora release includes newer upstream software that fixes 
bugs or makes them obsolete.

Comment 13 Adam Williamson 2020-11-03 22:27:53 UTC
We got the firmwares on the affected boxes updated, I believe, and now we don't really use them any more. Don't think there's anything else to be done here.


Note You need to log in before you can comment on or make changes to this bug.