kernek/fork.c: allocate idle task for a CPU always on its local node

commit 725fc629 upstream.

Linux preallocates the task structs of the idle tasks for all possible
CPUs.  This currently means they all end up on node 0.  This also
implies that the cache line of MWAIT, which is around the flags field in
the task struct, are all located in node 0.

We see a noticeable performance improvement on Knights Landing CPUs when
the cache lines used for MWAIT are located in the local nodes of the
CPUs using them.  I would expect this to give a (likely slight)
improvement on other systems too.

The patch implements placing the idle task in the node of its CPUs, by
passing the right target node to copy_process()

[js] to tls in copy_process in 3.12 yet

[akpm@linux-foundation.org: use NUMA_NO_NODE, not a bare -1]
Link: http://lkml.kernel.org/r/1463492694-15833-1-git-send-email-andi@firstfloor.orgSigned-off-by: Andi Kleen <ak@linux.intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>

kernel/fork.c

@@ -292,14 +292,15 @@ int __attribute__((weak)) arch_dup_task_struct(struct task_struct *dst,
 	return 0;
 }
 
-static struct task_struct *dup_task_struct(struct task_struct *orig)
+static struct task_struct *dup_task_struct(struct task_struct *orig, int node)
 {
 	struct task_struct *tsk;
 	struct thread_info *ti;
 	unsigned long *stackend;
-	int node = tsk_fork_get_node(orig);
 	int err;
 
+	if (node == NUMA_NO_NODE)
+		node = tsk_fork_get_node(orig);
 	tsk = alloc_task_struct_node(node);
 	if (!tsk)
 		return NULL;
@@ -1142,7 +1143,8 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 					unsigned long stack_size,
 					int __user *child_tidptr,
 					struct pid *pid,
-					int trace)
+					int trace,
+					int node)
 {
 	int retval;
 	struct task_struct *p;
@@ -1195,7 +1197,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 		goto fork_out;
 
 	retval = -ENOMEM;
-	p = dup_task_struct(current);
+	p = dup_task_struct(current, node);
 	if (!p)
 		goto fork_out;
 
@@ -1565,7 +1567,8 @@ static inline void init_idle_pids(struct pid_link *links)
 struct task_struct *fork_idle(int cpu)
 {
 	struct task_struct *task;
-	task = copy_process(CLONE_VM, 0, 0, NULL, &init_struct_pid, 0);
+	task = copy_process(CLONE_VM, 0, 0, NULL, &init_struct_pid, 0,
+			    cpu_to_node(cpu));
 	if (!IS_ERR(task)) {
 		init_idle_pids(task->pids);
 		init_idle(task, cpu);
@@ -1609,7 +1612,7 @@ long do_fork(unsigned long clone_flags,
 	}
 
 	p = copy_process(clone_flags, stack_start, stack_size,
-			 child_tidptr, NULL, trace);
+			 child_tidptr, NULL, trace, NUMA_NO_NODE);
 	/*
 	 * Do this prior waking up the new thread - the thread pointer
 	 * might get invalid after that point, if the thread exits quickly.