Bug 2136889
| Summary: | iTCO_wdt fails to fire | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 8 | Reporter: | David Teigland <teigland> |
| Component: | qemu-kvm | Assignee: | Michael S. Tsirkin <mst> |
| qemu-kvm sub component: | Devices | QA Contact: | Yiqian Wei <yiwei> |
| Status: | CLOSED NEXTRELEASE | Docs Contact: | |
| Severity: | unspecified | ||
| Priority: | high | CC: | ailan, berrange, coli, jinzhao, juzhang, mst, rjones, virt-maint, yiwei, ymankad |
| Version: | 8.5 | Keywords: | Triaged |
| 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: | 2023-01-31 09:42:24 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: | 2080207 | ||
| Bug Blocks: | |||
|
Description
David Teigland
2022-10-21 18:21:51 UTC
After some time debugging I found this problem is already known and reported as bug 2080207 in RHEL-9. Lets treat the RHEL-9 version as the primary one to investigate, and this one merely to track any possible backport once a solution is found. In the end QEMU's impl is correct and the problems lies in Linux >= 5.15
This commit causes a regression:
commit 1ae3e78c08209ac657c59f6f7ea21bbbd7f6a1d4
Author: Mika Westerberg <mika.westerberg.com>
Date: Tue Sep 21 13:29:00 2021 +0300
watchdog: iTCO_wdt: No need to stop the timer in probe
The watchdog core can handle pinging of the watchdog before userspace
opens the device. For this reason instead of stopping the timer, just
mark it as running and let the watchdog core take care of it.
Cc: Malin Jonsson <malin.jonsson>
Signed-off-by: Mika Westerberg <mika.westerberg.com>
Reviewed-by: Guenter Roeck <linux>
Link: https://lore.kernel.org/r/20210921102900.61586-1-mika.westerberg@linux.intel.com
Signed-off-by: Guenter Roeck <linux>
Signed-off-by: Wim Van Sebroeck <wim>
it marks the watchdog as running, but does NOT disable the "no reboot" flag.
This is reported to upstream maintainers listed in that commit, and a fix is in progress.
The problem doesn't exist in RHEL-8 kernel since that is way older.
@David can you confirm that the guest OS you were testing with has a Linux >= 5.15 kernel
Thanks for the update, yes the guest OS is 5.14.0-119.el9 > > @David can you confirm that the guest OS you were testing with has a Linux >= 5.15 kernel
> yes the guest OS is 5.14.0-119.el9
Rereading that, I'm not sure my answer made sense... the guest kernel I'm testing is 5.14.0-119.el9 and iTCO_wdt fails to reset the vm.
$ ./a.out /dev/watchdog
counting to timeout 30...
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31 failed to fire after 30 seconds
32 failed to fire after 30 seconds
^C
[localhost ~]$ uname -a
Linux localhost.localdomain 5.14.0-119.el9.x86_64 #1 SMP PREEMPT_DYNAMIC Fri Jun 24 06:37:48 EDT 2022 x86_64 x86_64 x86_64 GNU/Linux
[localhost ~]$ wdctl
Device: /dev/watchdog0
Identity: iTCO_wdt [version 0]
Timeout: 30 seconds
Pre-timeout: 0 seconds
Timeleft: 21 seconds
FLAG DESCRIPTION STATUS BOOT-STATUS
KEEPALIVEPING Keep alive ping reply 1 0
MAGICCLOSE Supports magic close char 0 0
SETTIMEOUT Set timeout (in seconds) 0 0
I expect you have not set the flag "-global ICH9-LPC.noreboot=false" for QEMU. This is something required to enable the watchdog in QEMU and tracked by bug 2137346 for libvirt integration. ok we are in agreement here. close/nextrelease this one. |