Revision d7831a0bdf06b9f722b947bb0c205ff7d77cebd8 authored by Richard Kennedy on 30 June 2009, 18:41:35 UTC, committed by Linus Torvalds on 01 July 2009, 01:56:01 UTC
balance_dirty_pages can overreact and move all of the dirty pages to
writeback unnecessarily.

balance_dirty_pages makes its decision to throttle based on the number of
dirty plus writeback pages that are over the calculated limit,so it will
continue to move pages even when there are plenty of pages in writeback
and less than the threshold still dirty.

This allows it to overshoot its limits and move all the dirty pages to
writeback while waiting for the drives to catch up and empty the writeback
list.

A simple fio test easily demonstrates this problem.

fio --name=f1 --directory=/disk1 --size=2G -rw=write --name=f2 --directory=/disk2 --size=1G --rw=write --startdelay=10

This is the simplest fix I could find, but I'm not entirely sure that it
alone will be enough for all cases.  But it certainly is an improvement on
my desktop machine writing to 2 disks.

Do we need something more for machines with large arrays where
bdi_threshold * number_of_drives is greater than the dirty_ratio ?

Signed-off-by: Richard Kennedy <richard@rsk.demon.co.uk>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent df279ca
Raw File
dm-linear.c
/*
 * Copyright (C) 2001-2003 Sistina Software (UK) Limited.
 *
 * This file is released under the GPL.
 */

#include "dm.h"
#include <linux/module.h>
#include <linux/init.h>
#include <linux/blkdev.h>
#include <linux/bio.h>
#include <linux/slab.h>
#include <linux/device-mapper.h>

#define DM_MSG_PREFIX "linear"

/*
 * Linear: maps a linear range of a device.
 */
struct linear_c {
	struct dm_dev *dev;
	sector_t start;
};

/*
 * Construct a linear mapping: <dev_path> <offset>
 */
static int linear_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
	struct linear_c *lc;
	unsigned long long tmp;

	if (argc != 2) {
		ti->error = "Invalid argument count";
		return -EINVAL;
	}

	lc = kmalloc(sizeof(*lc), GFP_KERNEL);
	if (lc == NULL) {
		ti->error = "dm-linear: Cannot allocate linear context";
		return -ENOMEM;
	}

	if (sscanf(argv[1], "%llu", &tmp) != 1) {
		ti->error = "dm-linear: Invalid device sector";
		goto bad;
	}
	lc->start = tmp;

	if (dm_get_device(ti, argv[0], lc->start, ti->len,
			  dm_table_get_mode(ti->table), &lc->dev)) {
		ti->error = "dm-linear: Device lookup failed";
		goto bad;
	}

	ti->num_flush_requests = 1;
	ti->private = lc;
	return 0;

      bad:
	kfree(lc);
	return -EINVAL;
}

static void linear_dtr(struct dm_target *ti)
{
	struct linear_c *lc = (struct linear_c *) ti->private;

	dm_put_device(ti, lc->dev);
	kfree(lc);
}

static sector_t linear_map_sector(struct dm_target *ti, sector_t bi_sector)
{
	struct linear_c *lc = ti->private;

	return lc->start + (bi_sector - ti->begin);
}

static void linear_map_bio(struct dm_target *ti, struct bio *bio)
{
	struct linear_c *lc = ti->private;

	bio->bi_bdev = lc->dev->bdev;
	if (bio_sectors(bio))
		bio->bi_sector = linear_map_sector(ti, bio->bi_sector);
}

static int linear_map(struct dm_target *ti, struct bio *bio,
		      union map_info *map_context)
{
	linear_map_bio(ti, bio);

	return DM_MAPIO_REMAPPED;
}

static int linear_status(struct dm_target *ti, status_type_t type,
			 char *result, unsigned int maxlen)
{
	struct linear_c *lc = (struct linear_c *) ti->private;

	switch (type) {
	case STATUSTYPE_INFO:
		result[0] = '\0';
		break;

	case STATUSTYPE_TABLE:
		snprintf(result, maxlen, "%s %llu", lc->dev->name,
				(unsigned long long)lc->start);
		break;
	}
	return 0;
}

static int linear_ioctl(struct dm_target *ti, unsigned int cmd,
			unsigned long arg)
{
	struct linear_c *lc = (struct linear_c *) ti->private;
	return __blkdev_driver_ioctl(lc->dev->bdev, lc->dev->mode, cmd, arg);
}

static int linear_merge(struct dm_target *ti, struct bvec_merge_data *bvm,
			struct bio_vec *biovec, int max_size)
{
	struct linear_c *lc = ti->private;
	struct request_queue *q = bdev_get_queue(lc->dev->bdev);

	if (!q->merge_bvec_fn)
		return max_size;

	bvm->bi_bdev = lc->dev->bdev;
	bvm->bi_sector = linear_map_sector(ti, bvm->bi_sector);

	return min(max_size, q->merge_bvec_fn(q, bvm, biovec));
}

static int linear_iterate_devices(struct dm_target *ti,
				  iterate_devices_callout_fn fn, void *data)
{
	struct linear_c *lc = ti->private;

	return fn(ti, lc->dev, lc->start, data);
}

static struct target_type linear_target = {
	.name   = "linear",
	.version = {1, 1, 0},
	.module = THIS_MODULE,
	.ctr    = linear_ctr,
	.dtr    = linear_dtr,
	.map    = linear_map,
	.status = linear_status,
	.ioctl  = linear_ioctl,
	.merge  = linear_merge,
	.iterate_devices = linear_iterate_devices,
};

int __init dm_linear_init(void)
{
	int r = dm_register_target(&linear_target);

	if (r < 0)
		DMERR("register failed %d", r);

	return r;
}

void dm_linear_exit(void)
{
	dm_unregister_target(&linear_target);
}
back to top