From: Steve Best sbest@redhat.com
redhat/configs: change so watchdog is module versus builtin
Description: Enable virtual watchdog timers provided by PAPR hypervisors.
Tested: make dist-configs
Signed-off-by: Steve Best sbest@redhat.com
diff --git a/redhat/configs/ark/generic/powerpc/CONFIG_PSERIES_WDT b/redhat/configs/ark/generic/powerpc/CONFIG_PSERIES_WDT index blahblah..blahblah 100644 --- a/redhat/configs/ark/generic/powerpc/CONFIG_PSERIES_WDT +++ b/redhat/configs/ark/generic/powerpc/CONFIG_PSERIES_WDT @@ -1 +1 @@ -CONFIG_PSERIES_WDT=y +CONFIG_PSERIES_WDT=m
-- https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044
From: Prarit Bhargava on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1127243...
@sfbest, no objection but just a question: The reason the watchdog is typically built in is so that it is immediately active on boot.
By making this a module, I'm wondering if this could expose a time where the watchdog should be active but isn't?
/me is 100% sure there's some valid reason for doing this but I don't have enough information to approve the change.
From: Steve Best on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1127245...
IBM requested it to be a module or an use case they have won't work.
From: Jarod Wilson on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1127494...
This merge request needs a rebase on latest os-build. It's got merge conflicts, and gitlab thinks there are 3202 patches and over 3000 files modified here.
From: Daniel P. Berrangé on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1129784...
That's a rather opaque justification - what's the actual problem being solved ?
From: Steve Best on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1129810...
The driver needs to be a module, otherwise LPAR termination style can't be configured.
From: Steve Best on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1130451...
I doubled checked with IBM and they still want this as a module.
We still need this as a module. We don't anticipate an issue, as the module is usually loaded very early in boot.
@prarit do you still want to block this or are you OK?
From: Prarit Bhargava on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1130702...
Nope, I didn't mean to block this. I think your reply turned it into a discussion.
From: Patrick Talbert on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2044#note_1153097...
@sfbest Just a reminder that this needs a rebase before it can be merged.
kernel@lists.fedoraproject.org
-
Daniel P. Berrangé (via Email Bridge) -
Jarod Wilson (via Email Bridge)
-
Patrick Talbert (via Email Bridge)
-
Prarit Bhargava (via Email Bridge)
-
Steve Best (via Email Bridge)