Kai-Heng Feng kai.heng.feng at canonical.com
Tue Apr 14 09:11:56 UTC 2020

> On Apr 10, 2020, at 02:13, Seth Forshee <seth.forshee at canonical.com> wrote:
> On Thu, Apr 09, 2020 at 10:38:46PM +0800, Kai-Heng Feng wrote:
>> BugLink: https://bugs.launchpad.net/bugs/1871819
>> Pstore with EFI backend is really useful for debugging kernel oops.
>> Make it builtin to always collect logs.
>> Signed-off-by: Kai-Heng Feng <kai.heng.feng at canonical.com>
> I'm a little hesitant about doing this, in large part because there have
> been cases in the past of machines being bricked by writing to NVRAM.
> Hopefully those problems are a thing of the past, but that history still
> makes me wary of doing this globally.

Ok, that does sound risky.

> Why is it so important to have this loaded by default, rather than
> loading the module when trying to collect crash data?

Because this needs to be loaded before crashing.
Combine with systemd-pstore, we can upload these crash logs via Apport.

If we can't make it builtin, another approach is to have it in /etc/modules.


More information about the kernel-team mailing list