Bug#1065320: linux-image-6.1.0-18-amd64: 6.1.0-18 kernel enters ACPI Error loop during boot & requires power cycle
Control: severity -1 serious
Control: tags -1 + upstream fixed-upstream
Control: forwarded -1 https://lore.kernel.org/regressions/Zd2bsV8VsFJMlbFW@archie.me/ https://bugzilla.kernel.org/show_bug.cgi?id=218531
Control: found -1 6.6.15-1
Control: found -1 6.7.4-1~exp1
Hi Lee,
On Sat, Mar 02, 2024 at 05:33:55PM +0000, Lee Elliott wrote:
> Package: src:linux
> Version: 6.1.76-1
> Severity: critical
> Justification: breaks the whole system
> X-Debbugs-Cc: leeejobsaccount@mail.co.uk
>
> Dear Maintainer,
>
> * What led up to the situation?
>
> Trying to boot the system with the 6.1.0-18 kernel
>
> * What exactly did you do (or not do) that was effective (or
> ineffective)?
>
> I tried adding 'boot_delay=1000' boot option to slow the console
> scroll rate, to enable better recording of the error messages.
>
> I tried rebooting the previous 6.1.0-17 kernel.
>
> * What was the outcome of this action?
>
> After adding the 'boot_delay=1000' option the boot process
> progressed no further than "Loading initial ramdisk ..."
> (left for several minutes - required power cycle).
>
> The system boots sucessfully on the previous 6.1.0-17 kernel
>
> * What outcome did you expect instead?
>
> I expected the system to successfully boot.
>
> * Additional observations
>
> This system also normally includes 'hpet=disable' and
> 'acpi_enforce_resources=lax' boot options but removing these
> made no difference.
>
> Although I was not able to boot the system with the
> 'boot_delay=1000' option and obtain clear photographs of the
> console output - the ones I've attached suffer from
> 'overprinting' - it does seem clear that ACPI errors are
> being reported.
>
> There appear to be two distinct phases to this problem.
> Initially, ACPI seems to be reporting errors for "GPE", as
> shown in the first attached photograph, but after ~10 seconds
> or so, ACPI then switches to continuously reporting an error
> for PM_TIMER, as shown in the second attached photograph. At
> this point a power cycle is required.
>
> Purging and reinstalling the package made no difference. Atm,
> only three kernels are installed on this system but I have
> had more in the past as I normally compile my own kernels
> from the corresponding Debian source package. My own 6.1.76-1
> kernel also suffers from the same problem, whereas my own
> 6.1.69-1 kernel boots and runs Ok.
>
> Comparing the kernel configs for 6.1.0-17 and 6.1.0-18
> showed just one functional change - an additional
> Compile-time checks and compiler option, which did not seem
> relevant to this problem.
The images were missing but from the description I assume it's the
same regression as reported in
https://lore.kernel.org/regressions/Zd2bsV8VsFJMlbFW@archie.me/
https://bugzilla.kernel.org/show_bug.cgi?id=218531
It looks a fix has landed in mainline:
https://git.kernel.org/linus/aec7d25b497ce4a8d044e9496de0aa433f7f8f06
Regards,
Salvatore
Reply to: