Mark,

On Wed, Sep 6, 2023 at 9:06 AM Douglas Anderson <diand...@chromium.org> wrote:
>
> +#ifdef CONFIG_ARM64_ACPI_PARKING_PROTOCOL
> +void arch_send_wakeup_ipi(unsigned int cpu)
> +{
> +       /*
> +        * We use a scheduler IPI to wake the CPU as this avoids the need for 
> a
> +        * dedicated IPI and we can safely handle spurious scheduler IPIs.
> +        */
> +       arch_smp_send_reschedule(cpu);

I was backporting this to our ChromeOS kernels and our build test bot
noticed that arch_smp_send_reschedule() didn't exist in older kernels.
That's fine--I can always adjust this patch when backporting or
cherry-pick extra patches, but it made me wonder. Is there a reason
you chose to use arch_smp_send_reschedule() directly here instead of
smp_send_reschedule()? I guess the only difference is that you're
bypassing the tracing. Is that on purpose? Should we add a comment
about it, or change this to smp_send_reschedule()?

Thanks!

-Doug


_______________________________________________
Kgdb-bugreport mailing list
Kgdb-bugreport@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/kgdb-bugreport

Reply via email to