Re: [PATCH v9 0/8] Parallel CPU bringup for x86_64

[David Woodhouse]

On 21 February 2023 09:49:51 GMT, Oleksandr Natalenko <oleksandr@xxxxxxxxxxxxxx> wrote:
>On 21.02.2023 10:06, David Woodhouse wrote:
>> Why does arch/x86/kernel/acpi/sleep.c::x86_acpi_suspend_lowlevel() set
>> 
>>     initial_gs = per_cpu_offset(smp_processor_id()) ?
>> 
>> Would it not be CPU#0 that comes back up, and should it not get
>> per_cpu_offset(0) ?
>
>Wanna me try `initial_gs = per_cpu_offset(0);` too?

Hm, yes please. There's another one to make zero on the next line up, I think?

>> Or maybe we should just set up smpboot_control for the CPU to find its
>> own stuff, *even* on waking. Since the structures are already set up,
>> it isn't like a clean boot.
>> 
>> If you let it boot in parallel mode, what if you just *remove* the line
>> that sets smpboot_control=0 ?
>
>If the `smpboot_control = 0;` line in arch/x86/kernel/acpi/sleep.c::x86_acpi_suspend_lowlevel() is commented out, and the system is booted in parallel mode, then suspend/resume works.

Well that's entertaining. Now, can we come up with any theory which doesn't leave us wondering why it ever worked in the first place...?