[PATCH] percpu: convert RCU

Patch from Dipankar Sarma <dipankar@in.ibm.com>

This patch convers RCU per_cpu data to use per_cpu data area
and makes it safe for cpu_possible allocation by using CPU
notifiers.

include/linux/rcupdate.h

@@ -39,6 +39,7 @@
 #include <linux/list.h>
 #include <linux/spinlock.h>
 #include <linux/threads.h>
+#include <linux/percpu.h>
 
 /**
  * struct rcu_head - callback structure for use with RCU
@@ -94,16 +95,16 @@ struct rcu_data {
         long  	       	batch;           /* Batch # for current RCU batch */
         struct list_head  nxtlist;
         struct list_head  curlist;
-} ____cacheline_aligned_in_smp;
+};
 
-extern struct rcu_data rcu_data[NR_CPUS];
+DECLARE_PER_CPU(struct rcu_data, rcu_data);
 extern struct rcu_ctrlblk rcu_ctrlblk;
 
-#define RCU_qsctr(cpu) 		(rcu_data[(cpu)].qsctr)
-#define RCU_last_qsctr(cpu) 	(rcu_data[(cpu)].last_qsctr)
-#define RCU_batch(cpu) 		(rcu_data[(cpu)].batch)
-#define RCU_nxtlist(cpu) 	(rcu_data[(cpu)].nxtlist)
-#define RCU_curlist(cpu) 	(rcu_data[(cpu)].curlist)
+#define RCU_qsctr(cpu) 		(per_cpu(rcu_data, (cpu)).qsctr)
+#define RCU_last_qsctr(cpu) 	(per_cpu(rcu_data, (cpu)).last_qsctr)
+#define RCU_batch(cpu) 		(per_cpu(rcu_data, (cpu)).batch)
+#define RCU_nxtlist(cpu) 	(per_cpu(rcu_data, (cpu)).nxtlist)
+#define RCU_curlist(cpu) 	(per_cpu(rcu_data, (cpu)).curlist)
 
 #define RCU_QSCTR_INVALID	0
 

kernel/rcupdate.c

@@ -41,13 +41,14 @@
 #include <linux/module.h>
 #include <linux/completion.h>
 #include <linux/percpu.h>
+#include <linux/notifier.h>
 #include <linux/rcupdate.h>
 
 /* Definition for rcupdate control block. */
 struct rcu_ctrlblk rcu_ctrlblk = 
 	{ .mutex = SPIN_LOCK_UNLOCKED, .curbatch = 1, 
 	  .maxbatch = 1, .rcu_cpu_mask = 0 };
-struct rcu_data rcu_data[NR_CPUS] __cacheline_aligned;
+DEFINE_PER_CPU(struct rcu_data, rcu_data) = { 0L };
 
 /* Fake initialization required by compiler */
 static DEFINE_PER_CPU(struct tasklet_struct, rcu_tasklet) = {NULL};
@@ -198,6 +199,33 @@ void rcu_check_callbacks(int cpu, int user)
 	tasklet_schedule(&RCU_tasklet(cpu));
 }
 
+static void __devinit rcu_online_cpu(int cpu)
+{
+	memset(&per_cpu(rcu_data, cpu), 0, sizeof(struct rcu_data));
+	tasklet_init(&RCU_tasklet(cpu), rcu_process_callbacks, 0UL);
+	INIT_LIST_HEAD(&RCU_nxtlist(cpu));
+	INIT_LIST_HEAD(&RCU_curlist(cpu));
+}
+
+static int __devinit rcu_cpu_notify(struct notifier_block *self, 
+				unsigned long action, void *hcpu)
+{
+	long cpu = (long)hcpu;
+	switch (action) {
+	case CPU_UP_PREPARE:
+		rcu_online_cpu(cpu);
+		break;
+	/* Space reserved for CPU_OFFLINE :) */
+	default:
+		break;
+	}
+	return NOTIFY_OK;
+}
+
+static struct notifier_block __devinitdata rcu_nb = {
+	.notifier_call	= rcu_cpu_notify,
+};
+
 /*
  * Initializes rcu mechanism.  Assumed to be called early.
  * That is before local timer(SMP) or jiffie timer (uniproc) is setup.
@@ -206,16 +234,13 @@ void rcu_check_callbacks(int cpu, int user)
  */
 void __init rcu_init(void)
 {
-	int i;
-
-	memset(&rcu_data[0], 0, sizeof(rcu_data));
-	for (i = 0; i < NR_CPUS; i++) {
-		tasklet_init(&RCU_tasklet(i), rcu_process_callbacks, 0UL);
-		INIT_LIST_HEAD(&RCU_nxtlist(i));
-		INIT_LIST_HEAD(&RCU_curlist(i));
-	}
+	rcu_cpu_notify(&rcu_nb, CPU_UP_PREPARE,
+			(void *)(long)smp_processor_id());
+	/* Register notifier for non-boot CPUs */
+	register_cpu_notifier(&rcu_nb);
 }
 
+
 /* Because of FASTCALL declaration of complete, we use this wrapper */
 static void wakeme_after_rcu(void *completion)
 {