SGI UV cpumask: use static temp cpumask in flush_tlb

Impact: Improve tlb flush performance for UV

Calling alloc_cpumask_var a zillion times a second does affect
performance.  Replace with static cpumask.

Note: when CONFIG_X86_UV is defined, this extra PER_CPU memory
will be optimized out for non-UV configs as is_uv_system() will
then return a constant 0.
Signed-off-by: Mike Travis <travis@sgi.com>

arch/x86/kernel/tlb_64.c

@@ -202,17 +202,18 @@ void native_flush_tlb_others(const struct cpumask *cpumask,
 			     struct mm_struct *mm, unsigned long va)
 {
 	if (is_uv_system()) {
-		cpumask_var_t after_uv_flush;
+		/* FIXME: could be an percpu_alloc'd thing */
+		static DEFINE_PER_CPU(cpumask_t, flush_tlb_mask);
+		struct cpumask *after_uv_flush = &get_cpu_var(flush_tlb_mask);
 
-		if (alloc_cpumask_var(&after_uv_flush, GFP_ATOMIC)) {
-			cpumask_andnot(after_uv_flush,
-				       cpumask, cpumask_of(smp_processor_id()));
+		cpumask_andnot(after_uv_flush, cpumask,
+			       cpumask_of(smp_processor_id()));
 		if (!uv_flush_tlb_others(after_uv_flush, mm, va))
 			flush_tlb_others_ipi(after_uv_flush, mm, va);
-			free_cpumask_var(after_uv_flush);
+
+		put_cpu_var(flush_tlb_uv_cpumask);
 		return;
 	}
-	}
 	flush_tlb_others_ipi(cpumask, mm, va);
 }