Revision a7ba4bf5e7ff6bfe83e41c748b77b49297c1b5d9 authored by Linus Torvalds on 02 July 2015, 18:21:26 UTC, committed by Linus Torvalds on 02 July 2015, 18:21:26 UTC
Pull fuse updates from Miklos Szeredi:
 "This is the start of improving fuse scalability.

  An input queue and a processing queue is split out from the monolithic
  fuse connection, each of those having their own spinlock.  The end of
  the patchset adds the ability to clone a fuse connection.  This means,
  that instead of having to read/write requests/answers on a single fuse
  device fd, the fuse daemon can have multiple distinct file descriptors
  open.  Each of those can be used to receive requests and send answers,
  currently the only constraint is that a request must be answered on
  the same fd as it was read from.

  This can be extended further to allow binding a device clone to a
  specific CPU or NUMA node.

  Based on a patchset by Srinivas Eeda and Ashish Samant.  Thanks to
  Ashish for the review of this series"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse: (40 commits)
  fuse: update MAINTAINERS entry
  fuse: separate pqueue for clones
  fuse: introduce per-instance fuse_dev structure
  fuse: device fd clone
  fuse: abort: no fc->lock needed for request ending
  fuse: no fc->lock for pqueue parts
  fuse: no fc->lock in request_end()
  fuse: cleanup request_end()
  fuse: request_end(): do once
  fuse: add req flag for private list
  fuse: pqueue locking
  fuse: abort: group pqueue accesses
  fuse: cleanup fuse_dev_do_read()
  fuse: move list_del_init() from request_end() into callers
  fuse: duplicate ->connected in pqueue
  fuse: separate out processing queue
  fuse: simplify request_wait()
  fuse: no fc->lock for iqueue parts
  fuse: allow interrupt queuing without fc->lock
  fuse: iqueue locking
  ...
2 parent s a611fb7 + 0a30f61
Raw File
blk-mq-cpumap.c
/*
 * CPU <-> hardware queue mapping helpers
 *
 * Copyright (C) 2013-2014 Jens Axboe
 */
#include <linux/kernel.h>
#include <linux/threads.h>
#include <linux/module.h>
#include <linux/mm.h>
#include <linux/smp.h>
#include <linux/cpu.h>

#include <linux/blk-mq.h>
#include "blk.h"
#include "blk-mq.h"

static int cpu_to_queue_index(unsigned int nr_cpus, unsigned int nr_queues,
			      const int cpu)
{
	return cpu * nr_queues / nr_cpus;
}

static int get_first_sibling(unsigned int cpu)
{
	unsigned int ret;

	ret = cpumask_first(topology_sibling_cpumask(cpu));
	if (ret < nr_cpu_ids)
		return ret;

	return cpu;
}

int blk_mq_update_queue_map(unsigned int *map, unsigned int nr_queues)
{
	unsigned int i, nr_cpus, nr_uniq_cpus, queue, first_sibling;
	cpumask_var_t cpus;

	if (!alloc_cpumask_var(&cpus, GFP_ATOMIC))
		return 1;

	cpumask_clear(cpus);
	nr_cpus = nr_uniq_cpus = 0;
	for_each_online_cpu(i) {
		nr_cpus++;
		first_sibling = get_first_sibling(i);
		if (!cpumask_test_cpu(first_sibling, cpus))
			nr_uniq_cpus++;
		cpumask_set_cpu(i, cpus);
	}

	queue = 0;
	for_each_possible_cpu(i) {
		if (!cpu_online(i)) {
			map[i] = 0;
			continue;
		}

		/*
		 * Easy case - we have equal or more hardware queues. Or
		 * there are no thread siblings to take into account. Do
		 * 1:1 if enough, or sequential mapping if less.
		 */
		if (nr_queues >= nr_cpus || nr_cpus == nr_uniq_cpus) {
			map[i] = cpu_to_queue_index(nr_cpus, nr_queues, queue);
			queue++;
			continue;
		}

		/*
		 * Less then nr_cpus queues, and we have some number of
		 * threads per cores. Map sibling threads to the same
		 * queue.
		 */
		first_sibling = get_first_sibling(i);
		if (first_sibling == i) {
			map[i] = cpu_to_queue_index(nr_uniq_cpus, nr_queues,
							queue);
			queue++;
		} else
			map[i] = map[first_sibling];
	}

	free_cpumask_var(cpus);
	return 0;
}

unsigned int *blk_mq_make_queue_map(struct blk_mq_tag_set *set)
{
	unsigned int *map;

	/* If cpus are offline, map them to first hctx */
	map = kzalloc_node(sizeof(*map) * nr_cpu_ids, GFP_KERNEL,
				set->numa_node);
	if (!map)
		return NULL;

	if (!blk_mq_update_queue_map(map, set->nr_hw_queues))
		return map;

	kfree(map);
	return NULL;
}

/*
 * We have no quick way of doing reverse lookups. This is only used at
 * queue init time, so runtime isn't important.
 */
int blk_mq_hw_queue_to_node(unsigned int *mq_map, unsigned int index)
{
	int i;

	for_each_possible_cpu(i) {
		if (index == mq_map[i])
			return cpu_to_node(i);
	}

	return NUMA_NO_NODE;
}
back to top