raid1.c

来自「Linux Kernel 2.6.9 for OMAP1710」· C语言 代码 · 共 1,426 行 · 第 1/3 页

				bio->bi_rw = READ;
				unplug = 1;
				generic_make_request(bio);
			}
		}
	}
	spin_unlock_irqrestore(&retry_list_lock, flags);
	if (unplug)
		unplug_slaves(mddev);
}


static int init_resync(conf_t *conf)
{
	int buffs;

	buffs = RESYNC_WINDOW / RESYNC_BLOCK_SIZE;
	if (conf->r1buf_pool)
		BUG();
	conf->r1buf_pool = mempool_create(buffs, r1buf_pool_alloc, r1buf_pool_free,
					  conf->poolinfo);
	if (!conf->r1buf_pool)
		return -ENOMEM;
	conf->next_resync = 0;
	return 0;
}

/*
 * perform a "sync" on one "block"
 *
 * We need to make sure that no normal I/O request - particularly write
 * requests - conflict with active sync requests.
 *
 * This is achieved by tracking pending requests and a 'barrier' concept
 * that can be installed to exclude normal IO requests.
 */

static int sync_request(mddev_t *mddev, sector_t sector_nr, int go_faster)
{
	conf_t *conf = mddev_to_conf(mddev);
	mirror_info_t *mirror;
	r1bio_t *r1_bio;
	struct bio *bio;
	sector_t max_sector, nr_sectors;
	int disk;
	int i;
	int write_targets = 0;

	if (!conf->r1buf_pool)
		if (init_resync(conf))
			return -ENOMEM;

	max_sector = mddev->size << 1;
	if (sector_nr >= max_sector) {
		close_sync(conf);
		return 0;
	}

	/*
	 * If there is non-resync activity waiting for us then
	 * put in a delay to throttle resync.
	 */
	if (!go_faster && waitqueue_active(&conf->wait_resume))
		schedule_timeout(HZ);
	device_barrier(conf, sector_nr + RESYNC_SECTORS);

	/*
	 * If reconstructing, and >1 working disc,
	 * could dedicate one to rebuild and others to
	 * service read requests ..
	 */
	disk = conf->last_used;
	/* make sure disk is operational */
	spin_lock_irq(&conf->device_lock);
	while (conf->mirrors[disk].rdev == NULL ||
	       !conf->mirrors[disk].rdev->in_sync) {
		if (disk <= 0)
			disk = conf->raid_disks;
		disk--;
		if (disk == conf->last_used)
			break;
	}
	conf->last_used = disk;
	atomic_inc(&conf->mirrors[disk].rdev->nr_pending);
	spin_unlock_irq(&conf->device_lock);

	mirror = conf->mirrors + disk;

	r1_bio = mempool_alloc(conf->r1buf_pool, GFP_NOIO);

	spin_lock_irq(&conf->resync_lock);
	conf->nr_pending++;
	spin_unlock_irq(&conf->resync_lock);

	r1_bio->mddev = mddev;
	r1_bio->sector = sector_nr;
	set_bit(R1BIO_IsSync, &r1_bio->state);
	r1_bio->read_disk = disk;

	for (i=0; i < conf->raid_disks; i++) {
		bio = r1_bio->bios[i];

		/* take from bio_init */
		bio->bi_next = NULL;
		bio->bi_flags |= 1 << BIO_UPTODATE;
		bio->bi_rw = 0;
		bio->bi_vcnt = 0;
		bio->bi_idx = 0;
		bio->bi_phys_segments = 0;
		bio->bi_hw_segments = 0;
		bio->bi_size = 0;
		bio->bi_end_io = NULL;
		bio->bi_private = NULL;

		if (i == disk) {
			bio->bi_rw = READ;
			bio->bi_end_io = end_sync_read;
		} else if (conf->mirrors[i].rdev &&
			   !conf->mirrors[i].rdev->faulty &&
			   (!conf->mirrors[i].rdev->in_sync ||
			    sector_nr + RESYNC_SECTORS > mddev->recovery_cp)) {
			bio->bi_rw = WRITE;
			bio->bi_end_io = end_sync_write;
			write_targets ++;
		} else
			continue;
		bio->bi_sector = sector_nr + conf->mirrors[i].rdev->data_offset;
		bio->bi_bdev = conf->mirrors[i].rdev->bdev;
		bio->bi_private = r1_bio;
	}
	if (write_targets == 0) {
		/* There is nowhere to write, so all non-sync
		 * drives must be failed - so we are finished
		 */
		int rv = max_sector - sector_nr;
		md_done_sync(mddev, rv, 1);
		put_buf(r1_bio);
		atomic_dec(&conf->mirrors[disk].rdev->nr_pending);
		return rv;
	}

	nr_sectors = 0;
	do {
		struct page *page;
		int len = PAGE_SIZE;
		if (sector_nr + (len>>9) > max_sector)
			len = (max_sector - sector_nr) << 9;
		if (len == 0)
			break;
		for (i=0 ; i < conf->raid_disks; i++) {
			bio = r1_bio->bios[i];
			if (bio->bi_end_io) {
				page = r1_bio->bios[0]->bi_io_vec[bio->bi_vcnt].bv_page;
				if (bio_add_page(bio, page, len, 0) == 0) {
					/* stop here */
					r1_bio->bios[0]->bi_io_vec[bio->bi_vcnt].bv_page = page;
					while (i > 0) {
						i--;
						bio = r1_bio->bios[i];
						if (bio->bi_end_io==NULL) continue;
						/* remove last page from this bio */
						bio->bi_vcnt--;
						bio->bi_size -= len;
						bio->bi_flags &= ~(1<< BIO_SEG_VALID);
					}
					goto bio_full;
				}
			}
		}
		nr_sectors += len>>9;
		sector_nr += len>>9;
	} while (r1_bio->bios[disk]->bi_vcnt < RESYNC_PAGES);
 bio_full:
	bio = r1_bio->bios[disk];
	r1_bio->sectors = nr_sectors;

	md_sync_acct(mirror->rdev->bdev, nr_sectors);

	generic_make_request(bio);

	return nr_sectors;
}

static int run(mddev_t *mddev)
{
	conf_t *conf;
	int i, j, disk_idx;
	mirror_info_t *disk;
	mdk_rdev_t *rdev;
	struct list_head *tmp;

	if (mddev->level != 1) {
		printk("raid1: %s: raid level not set to mirroring (%d)\n",
		       mdname(mddev), mddev->level);
		goto out;
	}
	/*
	 * copy the already verified devices into our private RAID1
	 * bookkeeping area. [whatever we allocate in run(),
	 * should be freed in stop()]
	 */
	conf = kmalloc(sizeof(conf_t), GFP_KERNEL);
	mddev->private = conf;
	if (!conf)
		goto out_no_mem;

	memset(conf, 0, sizeof(*conf));
	conf->mirrors = kmalloc(sizeof(struct mirror_info)*mddev->raid_disks, 
				 GFP_KERNEL);
	if (!conf->mirrors)
		goto out_no_mem;

	memset(conf->mirrors, 0, sizeof(struct mirror_info)*mddev->raid_disks);

	conf->poolinfo = kmalloc(sizeof(*conf->poolinfo), GFP_KERNEL);
	if (!conf->poolinfo)
		goto out_no_mem;
	conf->poolinfo->mddev = mddev;
	conf->poolinfo->raid_disks = mddev->raid_disks;
	conf->r1bio_pool = mempool_create(NR_RAID1_BIOS, r1bio_pool_alloc,
					  r1bio_pool_free,
					  conf->poolinfo);
	if (!conf->r1bio_pool)
		goto out_no_mem;

	mddev->queue->unplug_fn = raid1_unplug;

	mddev->queue->issue_flush_fn = raid1_issue_flush;

	ITERATE_RDEV(mddev, rdev, tmp) {
		disk_idx = rdev->raid_disk;
		if (disk_idx >= mddev->raid_disks
		    || disk_idx < 0)
			continue;
		disk = conf->mirrors + disk_idx;

		disk->rdev = rdev;

		blk_queue_stack_limits(mddev->queue,
				       rdev->bdev->bd_disk->queue);
		/* as we don't honour merge_bvec_fn, we must never risk
		 * violating it, so limit ->max_sector to one PAGE, as
		 * a one page request is never in violation.
		 */
		if (rdev->bdev->bd_disk->queue->merge_bvec_fn &&
		    mddev->queue->max_sectors > (PAGE_SIZE>>9))
			blk_queue_max_sectors(mddev->queue, PAGE_SIZE>>9);

		disk->head_position = 0;
		if (!rdev->faulty && rdev->in_sync)
			conf->working_disks++;
	}
	conf->raid_disks = mddev->raid_disks;
	conf->mddev = mddev;
	conf->device_lock = SPIN_LOCK_UNLOCKED;
	if (conf->working_disks == 1)
		mddev->recovery_cp = MaxSector;

	conf->resync_lock = SPIN_LOCK_UNLOCKED;
	init_waitqueue_head(&conf->wait_idle);
	init_waitqueue_head(&conf->wait_resume);

	if (!conf->working_disks) {
		printk(KERN_ERR "raid1: no operational mirrors for %s\n",
			mdname(mddev));
		goto out_free_conf;
	}

	mddev->degraded = 0;
	for (i = 0; i < conf->raid_disks; i++) {

		disk = conf->mirrors + i;

		if (!disk->rdev) {
			disk->head_position = 0;
			mddev->degraded++;
		}
	}

	/*
	 * find the first working one and use it as a starting point
	 * to read balancing.
	 */
	for (j = 0; j < conf->raid_disks &&
		     (!conf->mirrors[j].rdev ||
		      !conf->mirrors[j].rdev->in_sync) ; j++)
		/* nothing */;
	conf->last_used = j;



	{
		mddev->thread = md_register_thread(raid1d, mddev, "%s_raid1");
		if (!mddev->thread) {
			printk(KERN_ERR 
				"raid1: couldn't allocate thread for %s\n", 
				mdname(mddev));
			goto out_free_conf;
		}
	}
	printk(KERN_INFO 
		"raid1: raid set %s active with %d out of %d mirrors\n",
		mdname(mddev), mddev->raid_disks - mddev->degraded, 
		mddev->raid_disks);
	/*
	 * Ok, everything is just fine now
	 */
	mddev->array_size = mddev->size;

	return 0;

out_no_mem:
	printk(KERN_ERR "raid1: couldn't allocate memory for %s\n",
	       mdname(mddev));

out_free_conf:
	if (conf) {
		if (conf->r1bio_pool)
			mempool_destroy(conf->r1bio_pool);
		if (conf->mirrors)
			kfree(conf->mirrors);
		if (conf->poolinfo)
			kfree(conf->poolinfo);
		kfree(conf);
		mddev->private = NULL;
	}
out:
	return -EIO;
}

static int stop(mddev_t *mddev)
{
	conf_t *conf = mddev_to_conf(mddev);

	md_unregister_thread(mddev->thread);
	mddev->thread = NULL;
	if (conf->r1bio_pool)
		mempool_destroy(conf->r1bio_pool);
	if (conf->mirrors)
		kfree(conf->mirrors);
	if (conf->poolinfo)
		kfree(conf->poolinfo);
	kfree(conf);
	mddev->private = NULL;
	return 0;
}

static int raid1_resize(mddev_t *mddev, sector_t sectors)
{
	/* no resync is happening, and there is enough space
	 * on all devices, so we can resize.
	 * We need to make sure resync covers any new space.
	 * If the array is shrinking we should possibly wait until
	 * any io in the removed space completes, but it hardly seems
	 * worth it.
	 */
	mddev->array_size = sectors>>1;
	set_capacity(mddev->gendisk, mddev->array_size << 1);
	mddev->changed = 1;
	if (mddev->array_size > mddev->size && mddev->recovery_cp == MaxSector) {
		mddev->recovery_cp = mddev->size << 1;
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	}
	mddev->size = mddev->array_size;
	return 0;
}

static int raid1_reshape(mddev_t *mddev, int raid_disks)
{
	/* We need to:
	 * 1/ resize the r1bio_pool
	 * 2/ resize conf->mirrors
	 *
	 * We allocate a new r1bio_pool if we can.
	 * Then raise a device barrier and wait until all IO stops.
	 * Then resize conf->mirrors and swap in the new r1bio pool.
	 */
	mempool_t *newpool, *oldpool;
	struct pool_info *newpoolinfo;
	mirror_info_t *newmirrors;
	conf_t *conf = mddev_to_conf(mddev);

	int d;

	for (d= raid_disks; d < conf->raid_disks; d++)
		if (conf->mirrors[d].rdev)
			return -EBUSY;

	newpoolinfo = kmalloc(sizeof(newpoolinfo), GFP_KERNEL);
	if (!newpoolinfo)
		return -ENOMEM;
	newpoolinfo->mddev = mddev;
	newpoolinfo->raid_disks = raid_disks;

	newpool = mempool_create(NR_RAID1_BIOS, r1bio_pool_alloc,
				 r1bio_pool_free, newpoolinfo);
	if (!newpool) {
		kfree(newpoolinfo);
		return -ENOMEM;
	}
	newmirrors = kmalloc(sizeof(struct mirror_info) * raid_disks, GFP_KERNEL);
	if (!newmirrors) {
		kfree(newpoolinfo);
		mempool_destroy(newpool);
		return -ENOMEM;
	}
	memset(newmirrors, 0, sizeof(struct mirror_info)*raid_disks);

	spin_lock_irq(&conf->resync_lock);
	conf->barrier++;
	wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
			    conf->resync_lock, unplug_slaves(mddev));
	spin_unlock_irq(&conf->resync_lock);

	/* ok, everything is stopped */
	oldpool = conf->r1bio_pool;
	conf->r1bio_pool = newpool;
	for (d=0; d < raid_disks && d < conf->raid_disks; d++)
		newmirrors[d] = conf->mirrors[d];
	kfree(conf->mirrors);
	conf->mirrors = newmirrors;
	kfree(conf->poolinfo);
	conf->poolinfo = newpoolinfo;

	mddev->degraded += (raid_disks - conf->raid_disks);
	conf->raid_disks = mddev->raid_disks = raid_disks;

	spin_lock_irq(&conf->resync_lock);
	conf->barrier--;
	spin_unlock_irq(&conf->resync_lock);
	wake_up(&conf->wait_resume);
	wake_up(&conf->wait_idle);


	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	md_wakeup_thread(mddev->thread);

	mempool_destroy(oldpool);
	return 0;
}


static mdk_personality_t raid1_personality =
{
	.name		= "raid1",
	.owner		= THIS_MODULE,
	.make_request	= make_request,
	.run		= run,
	.stop		= stop,
	.status		= status,
	.error_handler	= error,
	.hot_add_disk	= raid1_add_disk,
	.hot_remove_disk= raid1_remove_disk,
	.spare_active	= raid1_spare_active,
	.sync_request	= sync_request,
	.resize		= raid1_resize,
	.reshape	= raid1_reshape,
};

static int __init raid_init(void)
{
	return register_md_personality(RAID1, &raid1_personality);
}

static void raid_exit(void)
{
	unregister_md_personality(RAID1);
}

module_init(raid_init);
module_exit(raid_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS("md-personality-3"); /* RAID1 */