x86-32, NUMA: Automatically set apicid -> node in setup_local_APIC()

Some x86-32 NUMA implementations (NUMAQ) don't initialize apicid ->
node mapping using set_apicid_to_node() during NUMA init but implement
custom apic->x86_32_numa_cpu_node() instead.

This patch automatically initializes the default apic -> node mapping
table from apic->x86_32_numa_cpu_node() from setup_local_APIC() such
that the mapping table is in sync with the actual mapping.

As the table isn't used by custom implementations, this doesn't make
any difference at this point.  This is in preparation of unifying
numa_cpu_node() between x86-32 and 64.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>

arch/x86/kernel/apic/apic.c

@@ -1237,6 +1237,16 @@ void __cpuinit setup_local_APIC(void)
 	/* always use the value from LDR */
 	early_per_cpu(x86_cpu_to_logical_apicid, cpu) =
 		logical_smp_processor_id();
+
+	/*
+	 * Some NUMA implementations (NUMAQ) don't initialize apicid to
+	 * node mapping during NUMA init.  Now that logical apicid is
+	 * guaranteed to be known, give it another chance.  This is already
+	 * a bit too late - percpu allocation has already happened without
+	 * proper NUMA affinity.
+	 */
+	set_apicid_to_node(early_per_cpu(x86_cpu_to_apicid, cpu),
+			   apic->x86_32_numa_cpu_node(cpu));
 #endif
 
 	/*