Bug 1899759
| Summary: | kernel install fails when /proc/cmdline is empty | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 8 | Reporter: | Jan Stodola <jstodola> |
| Component: | s390utils | Assignee: | Dan Horák <dhorak> |
| Status: | CLOSED DEFERRED | QA Contact: | Vilém Maršík <vmarsik> |
| Severity: | medium | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | 8.4 | CC: | bugproxy, dhorak, fmartine, rvr, smitterl, tstaudt |
| Target Milestone: | rc | Keywords: | Reopened, Triaged |
| Target Release: | 8.5 | Flags: | pm-rhel:
mirror+
|
| Hardware: | s390x | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2022-12-19 13:34:40 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: | 1815043, 1851111 | ||
| Bug Blocks: | 1609325, 1879091, 1903942, 1916117 | ||
|
Description
Jan Stodola
2020-11-19 22:55:06 UTC
The missing dasd.conf error is tracked in bug 1815043. I can't reproduce the problem when using RHEL-8.4.0-20201118.d.3 compose and using URL as the installation source (no ISO), tested in graphical and text+kickstart installations. I think you need to use ISO mounted in a virtual drive, which will be auto-detected and used as the installation source. The problem is probably caused by the kernel command line, which is empty when using virt-install, see kernel messages in syslog: 16:50:56,642 INFO kernel:Policy zone: DMA 16:50:56,642 NOTICE kernel:Kernel command line: 16:50:56,642 NOTICE kernel:Specific versions of hardware ..... You are right, the prerequisite is using the direct boot from ISO. Which makes me wonder if there is something wrong with the boot image, because it should pass some options to the kernel ... After many tests and discussion with Jan, I think we know what is the actual problem. The code in the kernel installation scripts expects /proc/cmdline to be non-empty (or /etc/kernel/cmdline) which unfortunately happens when virt-install is called with --location <path_to_iso_file>. It extracts kernel and initrd and doesn't set any other kernel option so as a result /proc/cmdline is empty and the kernel installation scriptlet then fails during the actual instalation. I believe this behaviour could be reproduced with a pure qemu command too. Javier, is there really a strict requirement for /proc/cmdline to contain some data in the kernel installation scripts? workarounds for virt-install - use --cdrom instead of --location, or add --extra-args=foo Hello Red Hat / Dan, a quick question ... ... is this Red Hat bugzilla still correctly assigned against 's390utils' ...? Please advise ... Thanks for your support. I think it is correct, because the kernel install scripts causing the issue are part of the s390utils package. ------- Comment From tstaudt.com 2021-11-24 09:03 EDT------- Hello Dan, what do you think should be the next steps here? Thanks. ------- Comment From tstaudt.com 2021-11-25 03:26 EDT------- (In reply to comment #15) > Hello Dan, > > what do you think should be the next steps here? > Thanks. We talked yesterday and Dan agreed that we could lower the severity for this, but leave it open. Doing so. *** Bug 2052695 has been marked as a duplicate of this bug. *** After evaluating this issue, there are no plans to address it further or fix it in an upcoming release. Therefore, it is being closed. If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened. ------- Comment From tstaudt.com 2022-05-19 03:40 EDT------- Hi Dan, this was automatically closed - if you still plan to look at this, please reopen it yourself - I'm fine either way. I am taking a look and based on my analysis so far I think the check for the empty cmdline is superfluous. Also it doesn't exist on non-s390x platforms, if I understand it right. fix implemented in rawhide with https://src.fedoraproject.org/rpms/s390utils/c/cc536faf67c5c1a363a5eb9ceda6161e0e6cc6e1?branch=rawhide, let's give it some time for testing After evaluating this issue, there are no plans to address it further or fix it in an upcoming release. Therefore, it is being closed. If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened. (In reply to Dan Horák from comment #19) > fix implemented in rawhide with > https://src.fedoraproject.org/rpms/s390utils/c/ > cc536faf67c5c1a363a5eb9ceda6161e0e6cc6e1?branch=rawhide, let's give it some > time for testing Seems there's a fix, so I'm reopening this. I can still reproduce this with a recent RHEL 8.8 compose. The text installation ends with Error The following error occurred while installing. This is a fatal error and installation will be aborted. DNF error: Error in POSTTRANS scriptlet in rpm package kernel-core On shell ... [anaconda root@localhost /]# kernel-install add $(uname -r) /usr/lib/modules/$(uname -r)/vmlinuz Could not determine the kernel command line parameters. Please specify the kernel command line in /etc/kernel/cmdline! # virt-install --virt-type kvm --name vm --ram 2048 --disk /var/lib/libvirt/images/graphical.qcow2,format=qcow2,size=20 --noautoconsole --os-type=linux --os-variant=rhel8-unknown --location /var/lib/libvirt/images/RHEL-8.8.0-20221216.0-s390x-dvd1.iso That's expected, the fix went only into Fedora (and ELN), thus is will be in RHEL 10. I haven't received much feedback, so I don't feel comfortable to introduce the change in RHEL 9 or even 8. Thanks for clarifying Dan. Setting CLOSED/DEFERRED. |