[PATCH] schedule_on_each_cpu(): reduce kmalloc() size (b6136773) · Commits · 戴 / test

schedule_on_each_cpu() presently does a large kmalloc - 96 kbytes on 1024 CPU 64-bit. Rework it so that we do one 8192-byte allocation and then a pile of tiny ones, via alloc_percpu(). This has a much higher chance of success (100% in the current VM). This also has the effect of reducing the memory requirements from NR_CPUS*n to num_possible_cpus()*n. Cc: Christoph Lameter <clameter@engr.sgi.com> Cc: Andi Kleen <ak@muc.de> Signed-off-by:

Andrew Morton <akpm@osdl.org> Signed-off-by:

Linus Torvalds <torvalds@osdl.org>

kernel/workqueue.c

+20 −8

Original line number	Diff line number	Diff line
		@@ -428,22 +428,34 @@ int schedule_delayed_work_on(int cpu,
		return ret;
		}

		/**
		* schedule_on_each_cpu - call a function on each online CPU from keventd
		* @func: the function to call
		* @info: a pointer to pass to func()
		*
		* Returns zero on success.
		* Returns -ve errno on failure.
		*
		* Appears to be racy against CPU hotplug.
		*
		* schedule_on_each_cpu() is very slow.
		*/
		int schedule_on_each_cpu(void (func)(void info), void *info)
		{
		int cpu;
		struct work_struct *work;
		struct work_struct *works;

		work = kmalloc(NR_CPUS * sizeof(struct work_struct), GFP_KERNEL);

		if (!work)
		works = alloc_percpu(struct work_struct);
		if (!works)
		return -ENOMEM;

		for_each_online_cpu(cpu) {
		INIT_WORK(work + cpu, func, info);
		INIT_WORK(per_cpu_ptr(works, cpu), func, info);
		__queue_work(per_cpu_ptr(keventd_wq->cpu_wq, cpu),
		work + cpu);
		per_cpu_ptr(works, cpu));
		}
		flush_workqueue(keventd_wq);
		kfree(work);
		free_percpu(works);
		return 0;
		}

Admin message