[PATCH] sched: cpu_sibling_map to cpu_mask

From: Nick Piggin <piggin@cyberone.com.au>

This is a (somewhat) trivial patch which converts cpu_sibling_map from an
array of CPUs to an array of cpumasks.  Needed for >2 siblings per package,
but it actually can simplify code as it allows the cpu_sibling_map to be
set up even when there is 1 sibling per package.  Intel want this, I use it
in the next patch to build scheduling domains for the P4 HT.

From: Thomas Schlichter <thomas.schlichter@web.de>

   Build fix

From: "Pallipadi, Venkatesh" <venkatesh.pallipadi@intel.com>

   Fix to handle more than 2 siblings per package.

5 files changed, 51 insertions, 50 deletions

diff --git a/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c b/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c
index 28d5e2243017..6e414b6a990d 100644
--- a/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c
+++ b/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c
@@ -57,8 +57,7 @@ static int cpufreq_p4_setdc(unsigned int cpu, unsigned int newstate)
 	u32 l, h;
 	cpumask_t cpus_allowed, affected_cpu_map;
 	struct cpufreq_freqs freqs;
-	int hyperthreading = 0;
-	int sibling = 0;
+	int j;
 
 	if (!cpu_online(cpu) || (newstate > DC_DISABLE) || 
 		(newstate == DC_RESV))
@@ -68,13 +67,10 @@ static int cpufreq_p4_setdc(unsigned int cpu, unsigned int newstate)
 	cpus_allowed = current->cpus_allowed;
 
 	/* only run on CPU to be set, or on its sibling */
-       affected_cpu_map = cpumask_of_cpu(cpu);
-#ifdef CONFIG_X86_HT
-	hyperthreading = ((cpu_has_ht) && (smp_num_siblings == 2));
-	if (hyperthreading) {
-		sibling = cpu_sibling_map[cpu];
-                cpu_set(sibling, affected_cpu_map);
-	}
+#ifdef CONFIG_SMP
+	affected_cpu_map = cpu_sibling_map[cpu];
+#else
+	affected_cpu_map = cpumask_of_cpu(cpu);
 #endif
 	set_cpus_allowed(current, affected_cpu_map);
         BUG_ON(!cpu_isset(smp_processor_id(), affected_cpu_map));
@@ -97,11 +93,11 @@ static int cpufreq_p4_setdc(unsigned int cpu, unsigned int newstate)
 	/* notifiers */
 	freqs.old = stock_freq * l / 8;
 	freqs.new = stock_freq * newstate / 8;
-	freqs.cpu = cpu;
-	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
-	if (hyperthreading) {
-		freqs.cpu = sibling;
-		cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	for_each_cpu(j) {
+		if (cpu_isset(j, affected_cpu_map)) {
+			freqs.cpu = j;
+			cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+		}
 	}
 
 	rdmsr(MSR_IA32_THERM_STATUS, l, h);
@@ -132,10 +128,11 @@ static int cpufreq_p4_setdc(unsigned int cpu, unsigned int newstate)
 	set_cpus_allowed(current, cpus_allowed);
 
 	/* notifiers */
-	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
-	if (hyperthreading) {
-		freqs.cpu = cpu;
-		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	for_each_cpu(j) {
+		if (cpu_isset(j, affected_cpu_map)) {
+			freqs.cpu = j;
+			cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+		}
 	}
 
 	return 0;
diff --git a/arch/i386/kernel/io_apic.c b/arch/i386/kernel/io_apic.c
index 92bf30a0bcb1..d17fa1891238 100644
--- a/arch/i386/kernel/io_apic.c
+++ b/arch/i386/kernel/io_apic.c
@@ -317,8 +317,7 @@ struct irq_cpu_info {
 
 #define IRQ_ALLOWED(cpu, allowed_mask)	cpu_isset(cpu, allowed_mask)
 
-#define CPU_TO_PACKAGEINDEX(i) \
-		((physical_balance && i > cpu_sibling_map[i]) ? cpu_sibling_map[i] : i)
+#define CPU_TO_PACKAGEINDEX(i) (first_cpu(cpu_sibling_map[i]))
 
 #define MAX_BALANCED_IRQ_INTERVAL	(5*HZ)
 #define MIN_BALANCED_IRQ_INTERVAL	(HZ/2)
@@ -401,6 +400,7 @@ static void do_irq_balance(void)
 	unsigned long max_cpu_irq = 0, min_cpu_irq = (~0);
 	unsigned long move_this_load = 0;
 	int max_loaded = 0, min_loaded = 0;
+	int load;
 	unsigned long useful_load_threshold = balanced_irq_interval + 10;
 	int selected_irq;
 	int tmp_loaded, first_attempt = 1;
@@ -452,7 +452,7 @@ static void do_irq_balance(void)
 	for (i = 0; i < NR_CPUS; i++) {
 		if (!cpu_online(i))
 			continue;
-		if (physical_balance && i > cpu_sibling_map[i])
+		if (i != CPU_TO_PACKAGEINDEX(i))
 			continue;
 		if (min_cpu_irq > CPU_IRQ(i)) {
 			min_cpu_irq = CPU_IRQ(i);
@@ -471,7 +471,7 @@ tryanothercpu:
 	for (i = 0; i < NR_CPUS; i++) {
 		if (!cpu_online(i))
 			continue;
-		if (physical_balance && i > cpu_sibling_map[i])
+		if (i != CPU_TO_PACKAGEINDEX(i))
 			continue;
 		if (max_cpu_irq <= CPU_IRQ(i)) 
 			continue;
@@ -551,9 +551,14 @@ tryanotherirq:
 	 * We seek the least loaded sibling by making the comparison
 	 * (A+B)/2 vs B
 	 */
-	if (physical_balance && (CPU_IRQ(min_loaded) >> 1) >
-					CPU_IRQ(cpu_sibling_map[min_loaded]))
-		min_loaded = cpu_sibling_map[min_loaded];
+	load = CPU_IRQ(min_loaded) >> 1;
+	for_each_cpu_mask(j, cpu_sibling_map[min_loaded]) {
+		if (load > CPU_IRQ(j)) {
+			/* This won't change cpu_sibling_map[min_loaded] */
+			load = CPU_IRQ(j);
+			min_loaded = j;
+		}
+	}
 
 	cpus_and(allowed_mask, cpu_online_map, irq_affinity[selected_irq]);
 	target_cpu_mask = cpumask_of_cpu(min_loaded);
diff --git a/arch/i386/kernel/smpboot.c b/arch/i386/kernel/smpboot.c
index 7baa4d420b73..77d4309d6c6e 100644
--- a/arch/i386/kernel/smpboot.c
+++ b/arch/i386/kernel/smpboot.c
@@ -936,7 +936,7 @@ static int boot_cpu_logical_apicid;
 /* Where the IO area was mapped on multiquad, always 0 otherwise */
 void *xquad_portio;
 
-int cpu_sibling_map[NR_CPUS] __cacheline_aligned;
+cpumask_t cpu_sibling_map[NR_CPUS] __cacheline_aligned;
 
 static void __init smp_boot_cpus(unsigned int max_cpus)
 {
@@ -1081,32 +1081,34 @@ static void __init smp_boot_cpus(unsigned int max_cpus)
 	Dprintk("Boot done.\n");
 
 	/*
-	 * If Hyper-Threading is avaialble, construct cpu_sibling_map[], so
-	 * that we can tell the sibling CPU efficiently.
+	 * construct cpu_sibling_map[], so that we can tell sibling CPUs
+	 * efficiently.
 	 */
-	if (cpu_has_ht && smp_num_siblings > 1) {
-		for (cpu = 0; cpu < NR_CPUS; cpu++)
-			cpu_sibling_map[cpu] = NO_PROC_ID;
-		
-		for (cpu = 0; cpu < NR_CPUS; cpu++) {
-			int 	i;
-			if (!cpu_isset(cpu, cpu_callout_map))
-				continue;
+	for (cpu = 0; cpu < NR_CPUS; cpu++)
+		cpu_sibling_map[cpu] = CPU_MASK_NONE;
+
+	for (cpu = 0; cpu < NR_CPUS; cpu++) {
+		int siblings = 0;
+		int i;
+		if (!cpu_isset(cpu, cpu_callout_map))
+			continue;
 
+		if (smp_num_siblings > 1) {
 			for (i = 0; i < NR_CPUS; i++) {
-				if (i == cpu || !cpu_isset(i, cpu_callout_map))
+				if (!cpu_isset(i, cpu_callout_map))
 					continue;
 				if (phys_proc_id[cpu] == phys_proc_id[i]) {
-					cpu_sibling_map[cpu] = i;
-					printk("cpu_sibling_map[%d] = %d\n", cpu, cpu_sibling_map[cpu]);
-					break;
+					siblings++;
+					cpu_set(i, cpu_sibling_map[cpu]);
 				}
 			}
-			if (cpu_sibling_map[cpu] == NO_PROC_ID) {
-				smp_num_siblings = 1;
-				printk(KERN_WARNING "WARNING: No sibling found for CPU %d.\n", cpu);
-			}
+		} else {
+			siblings++;
+			cpu_set(cpu, cpu_sibling_map[cpu]);
 		}
+
+		if (siblings != smp_num_siblings)
+			printk(KERN_WARNING "WARNING: %d siblings found for CPU%d, should be %d\n", siblings, cpu, smp_num_siblings);
 	}
 
 	smpboot_setup_io_apic();
diff --git a/arch/i386/oprofile/op_model_p4.c b/arch/i386/oprofile/op_model_p4.c
index 7e41dc9826a1..8b932b1b3f14 100644
--- a/arch/i386/oprofile/op_model_p4.c
+++ b/arch/i386/oprofile/op_model_p4.c
@@ -382,11 +382,8 @@ static struct p4_event_binding p4_events[NUM_EVENTS] = {
 static unsigned int get_stagger(void)
 {
 #ifdef CONFIG_SMP
-	int cpu;
-	if (smp_num_siblings > 1) {
-		cpu = smp_processor_id();
-		return (cpu_sibling_map[cpu] > cpu) ? 0 : 1;
-	}
+	int cpu = smp_processor_id();
+	return (cpu != first_cpu(cpu_sibling_map[cpu]));
 #endif	
 	return 0;
 }
diff --git a/include/asm-i386/smp.h b/include/asm-i386/smp.h
index 268dabd46653..88c461e5e107 100644
--- a/include/asm-i386/smp.h
+++ b/include/asm-i386/smp.h
@@ -34,7 +34,7 @@
 extern void smp_alloc_memory(void);
 extern int pic_mode;
 extern int smp_num_siblings;
-extern int cpu_sibling_map[];
+extern cpumask_t cpu_sibling_map[];
 
 extern void smp_flush_tlb(void);
 extern void smp_message_irq(int cpl, void *dev_id, struct pt_regs *regs);