x86: Add x86_init platform override to fix up NUMA core numbering

Add an x86_init vector for handling inconsistent core numbering.
This is useful for multi-fabric platforms, such as Numascale
NumaConnect.

v2:
 - use struct x86_cpuinit_ops
 - provide default fall-back function to warn

Signed-off-by: Daniel J Blueman <daniel@numascale-asia.com>
Cc: Steffen Persvold <sp@numascale.com>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Link: http://lkml.kernel.org/r/1323073238-32686-2-git-send-email-daniel@numascale-asia.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index 0bab2b1..ef21bdc 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -353,6 +353,13 @@
 	if (node == NUMA_NO_NODE)
 		node = per_cpu(cpu_llc_id, cpu);
 
+	/*
+	 * If core numbers are inconsistent, it's likely a multi-fabric platform,
+	 * so invoke platform-specific handler
+	 */
+	if (c->phys_proc_id != node)
+		x86_cpuinit.fixup_cpu_id(c, node);
+
 	if (!node_online(node)) {
 		/*
 		 * Two possibilities here: