Re: Question about prefill_possible_map

From: Cao jin
Date: Wed Apr 08 2020 - 08:03:02 EST


Hi Thomas,

Seems you helped improve 2a51fe083eba7f, could you help with this question?

--
Sincerely,
Cao jin

On 3/27/20 2:38 PM, Cao jin wrote:
> Hi,
>
> At the beginning of this function, there is:
>
> /* No boot processor was found in mptable or ACPI MADT */
> if (!num_processors) {
> if (boot_cpu_has(X86_FEATURE_APIC)) {
> int apicid = boot_cpu_physical_apicid;
> int cpu = hard_smp_processor_id();
>
> pr_warn("Boot CPU (id %d) not listed by BIOS\n", cpu);
>
> /* Make sure boot cpu is enumerated */
> if (apic->cpu_present_to_apicid(0) == BAD_APICID &&
> apic->apic_id_valid(apicid))
> generic_processor_info(apicid, boot_cpu_apic_version);
> }
>
> if (!num_processors)
> num_processors = 1;
> }
>
> I see 3 cases will go through this code chunk,
>
> 1. UP, has no APCI MADT & MP table, but has APIC
> 2. 2a51fe083eba7f: kdump kernel running on the hot-plugged CPU
> 3. normal UP(No table, no APIC)
>
> And "if (boot_cpu_has(X86_FEATURE_APIC))" covers case 1 & 2 to my
> understanding.
>
> But both cases 1 & 2 already have boot_cpu_physical_apicid initialized
> from APIC ID register:
> 1. init_apic_mappings --> register_lapic_address
> 2. early_acpi_boot_init --> early_acpi_process_madt -->...-->
> register_lapic_address
>
> So my question is: is variable "cpu" redundant? Could it be:
>
> diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
> index 69881b2d446c..99fc03511568 100644
> --- a/arch/x86/kernel/smpboot.c
> +++ b/arch/x86/kernel/smpboot.c
> @@ -1456,9 +1456,9 @@ __init void prefill_possible_map(void)
> if (!num_processors) {
> if (boot_cpu_has(X86_FEATURE_APIC)) {
> int apicid = boot_cpu_physical_apicid;
> - int cpu = hard_smp_processor_id();
>
> - pr_warn("Boot CPU (id %d) not listed by BIOS\n", cpu);
> + pr_warn("Boot CPU (id %d) not listed by BIOS\n", \
> + boot_cpu_physical_apicid);
>
> /* Make sure boot cpu is enumerated */
> if (apic->cpu_present_to_apicid(0) == BAD_APICID &&
>