[2.6 patch] MIPS: remove smp_tune_scheduling()

[Adrian Bunk <bunk@stusta.de>]

Since smp_tune_scheduling() didn't do anything we can simply remove it.

Signed-off-by: Adrian Bunk <bunk@stusta.de>

---

 arch/mips/kernel/smp.c |   28 ----------------------------
 1 file changed, 28 deletions(-)

--- linux-2.6.20-rc3-mm1/arch/mips/kernel/smp.c.old	2007-01-11 14:19:43.000000000 +0100
+++ linux-2.6.20-rc3-mm1/arch/mips/kernel/smp.c	2007-01-11 14:20:01.000000000 +0100
@@ -51,33 +51,6 @@
 EXPORT_SYMBOL(phys_cpu_present_map);
 EXPORT_SYMBOL(cpu_online_map);
 
-static void smp_tune_scheduling (void)
-{
-	struct cache_desc *cd = &current_cpu_data.scache;
-	unsigned long cachesize;       /* kB   */
-	unsigned long cpu_khz;
-
-	/*
-	 * Crude estimate until we actually meassure ...
-	 */
-	cpu_khz = loops_per_jiffy * 2 * HZ / 1000;
-
-	/*
-	 * Rough estimation for SMP scheduling, this is the number of
-	 * cycles it takes for a fully memory-limited process to flush
-	 * the SMP-local cache.
-	 *
-	 * (For a P5 this pretty much means we will choose another idle
-	 *  CPU almost always at wakeup time (this is due to the small
-	 *  L1 cache), on PIIs it's around 50-100 usecs, depending on
-	 *  the cache size)
-	 */
-	if (!cpu_khz)
-		return;
-
-	cachesize = cd->linesz * cd->sets * cd->ways;
-}
-
 extern void __init calibrate_delay(void);
 extern ATTRIB_NORET void cpu_idle(void);
 
@@ -245,7 +218,6 @@
 {
 	init_new_context(current, &init_mm);
 	current_thread_info()->cpu = 0;
-	smp_tune_scheduling();
 	plat_prepare_cpus(max_cpus);
 #ifndef CONFIG_HOTPLUG_CPU
 	cpu_present_map = cpu_possible_map;