md.c

来自「Linux Kernel 2.6.9 for OMAP1710」· C语言 代码 · 共 2,796 行 · 第 1/5 页

	mddev->pers = pers[pnum];
	spin_unlock(&pers_lock);

	mddev->resync_max_sectors = mddev->size << 1; /* may be over-ridden by personality */

	err = mddev->pers->run(mddev);
	if (err) {
		printk(KERN_ERR "md: pers->run() failed ...\n");
		module_put(mddev->pers->owner);
		mddev->pers = NULL;
		return -EINVAL;
	}
 	atomic_set(&mddev->writes_pending,0);
	mddev->safemode = 0;
	mddev->safemode_timer.function = md_safemode_timeout;
	mddev->safemode_timer.data = (unsigned long) mddev;
	mddev->safemode_delay = (20 * HZ)/1000 +1; /* 20 msec delay */
	mddev->in_sync = 1;
	
	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	
	if (mddev->sb_dirty)
		md_update_sb(mddev);

	set_capacity(disk, mddev->array_size<<1);

	/* If we call blk_queue_make_request here, it will
	 * re-initialise max_sectors etc which may have been
	 * refined inside -> run.  So just set the bits we need to set.
	 * Most initialisation happended when we called
	 * blk_queue_make_request(..., md_fail_request)
	 * earlier.
	 */
	mddev->queue->queuedata = mddev;
	mddev->queue->make_request_fn = mddev->pers->make_request;
	mddev->queue->issue_flush_fn = md_flush_all;

	mddev->changed = 1;
	return 0;
}

static int restart_array(mddev_t *mddev)
{
	struct gendisk *disk = mddev->gendisk;
	int err;

	/*
	 * Complain if it has no devices
	 */
	err = -ENXIO;
	if (list_empty(&mddev->disks))
		goto out;

	if (mddev->pers) {
		err = -EBUSY;
		if (!mddev->ro)
			goto out;

		mddev->safemode = 0;
		mddev->ro = 0;
		set_disk_ro(disk, 0);

		printk(KERN_INFO "md: %s switched to read-write mode.\n",
			mdname(mddev));
		/*
		 * Kick recovery or resync if necessary
		 */
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
		md_wakeup_thread(mddev->thread);
		err = 0;
	} else {
		printk(KERN_ERR "md: %s has no personality assigned.\n",
			mdname(mddev));
		err = -EINVAL;
	}

out:
	return err;
}

static int do_md_stop(mddev_t * mddev, int ro)
{
	int err = 0;
	struct gendisk *disk = mddev->gendisk;

	if (mddev->pers) {
		if (atomic_read(&mddev->active)>2) {
			printk("md: %s still in use.\n",mdname(mddev));
			return -EBUSY;
		}

		if (mddev->sync_thread) {
			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
			md_unregister_thread(mddev->sync_thread);
			mddev->sync_thread = NULL;
		}

		del_timer_sync(&mddev->safemode_timer);

		invalidate_partition(disk, 0);

		if (ro) {
			err  = -ENXIO;
			if (mddev->ro)
				goto out;
			mddev->ro = 1;
		} else {
			if (mddev->ro)
				set_disk_ro(disk, 0);
			blk_queue_make_request(mddev->queue, md_fail_request);
			mddev->pers->stop(mddev);
			module_put(mddev->pers->owner);
			mddev->pers = NULL;
			if (mddev->ro)
				mddev->ro = 0;
		}
		if (!mddev->in_sync) {
			/* mark array as shutdown cleanly */
			mddev->in_sync = 1;
			md_update_sb(mddev);
		}
		if (ro)
			set_disk_ro(disk, 1);
	}
	/*
	 * Free resources if final stop
	 */
	if (!ro) {
		struct gendisk *disk;
		printk(KERN_INFO "md: %s stopped.\n", mdname(mddev));

		export_array(mddev);

		mddev->array_size = 0;
		disk = mddev->gendisk;
		if (disk)
			set_capacity(disk, 0);
		mddev->changed = 1;
	} else
		printk(KERN_INFO "md: %s switched to read-only mode.\n",
			mdname(mddev));
	err = 0;
out:
	return err;
}

static void autorun_array(mddev_t *mddev)
{
	mdk_rdev_t *rdev;
	struct list_head *tmp;
	int err;

	if (list_empty(&mddev->disks)) {
		MD_BUG();
		return;
	}

	printk(KERN_INFO "md: running: ");

	ITERATE_RDEV(mddev,rdev,tmp) {
		char b[BDEVNAME_SIZE];
		printk("<%s>", bdevname(rdev->bdev,b));
	}
	printk("\n");

	err = do_md_run (mddev);
	if (err) {
		printk(KERN_WARNING "md :do_md_run() returned %d\n", err);
		do_md_stop (mddev, 0);
	}
}

/*
 * lets try to run arrays based on all disks that have arrived
 * until now. (those are in pending_raid_disks)
 *
 * the method: pick the first pending disk, collect all disks with
 * the same UUID, remove all from the pending list and put them into
 * the 'same_array' list. Then order this list based on superblock
 * update time (freshest comes first), kick out 'old' disks and
 * compare superblocks. If everything's fine then run it.
 *
 * If "unit" is allocated, then bump its reference count
 */
static void autorun_devices(int part)
{
	struct list_head candidates;
	struct list_head *tmp;
	mdk_rdev_t *rdev0, *rdev;
	mddev_t *mddev;
	char b[BDEVNAME_SIZE];

	printk(KERN_INFO "md: autorun ...\n");
	while (!list_empty(&pending_raid_disks)) {
		dev_t dev;
		rdev0 = list_entry(pending_raid_disks.next,
					 mdk_rdev_t, same_set);

		printk(KERN_INFO "md: considering %s ...\n",
			bdevname(rdev0->bdev,b));
		INIT_LIST_HEAD(&candidates);
		ITERATE_RDEV_PENDING(rdev,tmp)
			if (super_90_load(rdev, rdev0, 0) >= 0) {
				printk(KERN_INFO "md:  adding %s ...\n",
					bdevname(rdev->bdev,b));
				list_move(&rdev->same_set, &candidates);
			}
		/*
		 * now we have a set of devices, with all of them having
		 * mostly sane superblocks. It's time to allocate the
		 * mddev.
		 */
		if (rdev0->preferred_minor < 0 || rdev0->preferred_minor >= MAX_MD_DEVS) {
			printk(KERN_INFO "md: unit number in %s is bad: %d\n",
			       bdevname(rdev0->bdev, b), rdev0->preferred_minor);
			break;
		}
		if (part)
			dev = MKDEV(mdp_major,
				    rdev0->preferred_minor << MdpMinorShift);
		else
			dev = MKDEV(MD_MAJOR, rdev0->preferred_minor);

		md_probe(dev, NULL, NULL);
		mddev = mddev_find(dev);
		if (!mddev) {
			printk(KERN_ERR 
				"md: cannot allocate memory for md drive.\n");
			break;
		}
		if (mddev_lock(mddev)) 
			printk(KERN_WARNING "md: %s locked, cannot run\n",
			       mdname(mddev));
		else if (mddev->raid_disks || mddev->major_version
			 || !list_empty(&mddev->disks)) {
			printk(KERN_WARNING 
				"md: %s already running, cannot run %s\n",
				mdname(mddev), bdevname(rdev0->bdev,b));
			mddev_unlock(mddev);
		} else {
			printk(KERN_INFO "md: created %s\n", mdname(mddev));
			ITERATE_RDEV_GENERIC(candidates,rdev,tmp) {
				list_del_init(&rdev->same_set);
				if (bind_rdev_to_array(rdev, mddev))
					export_rdev(rdev);
			}
			autorun_array(mddev);
			mddev_unlock(mddev);
		}
		/* on success, candidates will be empty, on error
		 * it won't...
		 */
		ITERATE_RDEV_GENERIC(candidates,rdev,tmp)
			export_rdev(rdev);
		mddev_put(mddev);
	}
	printk(KERN_INFO "md: ... autorun DONE.\n");
}

/*
 * import RAID devices based on one partition
 * if possible, the array gets run as well.
 */

static int autostart_array(dev_t startdev)
{
	char b[BDEVNAME_SIZE];
	int err = -EINVAL, i;
	mdp_super_t *sb = NULL;
	mdk_rdev_t *start_rdev = NULL, *rdev;

	start_rdev = md_import_device(startdev, 0, 0);
	if (IS_ERR(start_rdev))
		return err;


	/* NOTE: this can only work for 0.90.0 superblocks */
	sb = (mdp_super_t*)page_address(start_rdev->sb_page);
	if (sb->major_version != 0 ||
	    sb->minor_version != 90 ) {
		printk(KERN_WARNING "md: can only autostart 0.90.0 arrays\n");
		export_rdev(start_rdev);
		return err;
	}

	if (start_rdev->faulty) {
		printk(KERN_WARNING 
			"md: can not autostart based on faulty %s!\n",
			bdevname(start_rdev->bdev,b));
		export_rdev(start_rdev);
		return err;
	}
	list_add(&start_rdev->same_set, &pending_raid_disks);

	for (i = 0; i < MD_SB_DISKS; i++) {
		mdp_disk_t *desc = sb->disks + i;
		dev_t dev = MKDEV(desc->major, desc->minor);

		if (!dev)
			continue;
		if (dev == startdev)
			continue;
		if (MAJOR(dev) != desc->major || MINOR(dev) != desc->minor)
			continue;
		rdev = md_import_device(dev, 0, 0);
		if (IS_ERR(rdev))
			continue;

		list_add(&rdev->same_set, &pending_raid_disks);
	}

	/*
	 * possibly return codes
	 */
	autorun_devices(0);
	return 0;

}


static int get_version(void __user * arg)
{
	mdu_version_t ver;

	ver.major = MD_MAJOR_VERSION;
	ver.minor = MD_MINOR_VERSION;
	ver.patchlevel = MD_PATCHLEVEL_VERSION;

	if (copy_to_user(arg, &ver, sizeof(ver)))
		return -EFAULT;

	return 0;
}

static int get_array_info(mddev_t * mddev, void __user * arg)
{
	mdu_array_info_t info;
	int nr,working,active,failed,spare;
	mdk_rdev_t *rdev;
	struct list_head *tmp;

	nr=working=active=failed=spare=0;
	ITERATE_RDEV(mddev,rdev,tmp) {
		nr++;
		if (rdev->faulty)
			failed++;
		else {
			working++;
			if (rdev->in_sync)
				active++;	
			else
				spare++;
		}
	}

	info.major_version = mddev->major_version;
	info.minor_version = mddev->minor_version;
	info.patch_version = 1;
	info.ctime         = mddev->ctime;
	info.level         = mddev->level;
	info.size          = mddev->size;
	info.nr_disks      = nr;
	info.raid_disks    = mddev->raid_disks;
	info.md_minor      = mddev->md_minor;
	info.not_persistent= !mddev->persistent;

	info.utime         = mddev->utime;
	info.state         = 0;
	if (mddev->in_sync)
		info.state = (1<<MD_SB_CLEAN);
	info.active_disks  = active;
	info.working_disks = working;
	info.failed_disks  = failed;
	info.spare_disks   = spare;

	info.layout        = mddev->layout;
	info.chunk_size    = mddev->chunk_size;

	if (copy_to_user(arg, &info, sizeof(info)))
		return -EFAULT;

	return 0;
}

static int get_disk_info(mddev_t * mddev, void __user * arg)
{
	mdu_disk_info_t info;
	unsigned int nr;
	mdk_rdev_t *rdev;

	if (copy_from_user(&info, arg, sizeof(info)))
		return -EFAULT;

	nr = info.number;

	rdev = find_rdev_nr(mddev, nr);
	if (rdev) {
		info.major = MAJOR(rdev->bdev->bd_dev);
		info.minor = MINOR(rdev->bdev->bd_dev);
		info.raid_disk = rdev->raid_disk;
		info.state = 0;
		if (rdev->faulty)
			info.state |= (1<<MD_DISK_FAULTY);
		else if (rdev->in_sync) {
			info.state |= (1<<MD_DISK_ACTIVE);
			info.state |= (1<<MD_DISK_SYNC);
		}
	} else {
		info.major = info.minor = 0;
		info.raid_disk = -1;
		info.state = (1<<MD_DISK_REMOVED);
	}

	if (copy_to_user(arg, &info, sizeof(info)))
		return -EFAULT;

	return 0;
}

static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
{
	char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
	mdk_rdev_t *rdev;
	dev_t dev = MKDEV(info->major,info->minor);

	if (info->major != MAJOR(dev) || info->minor != MINOR(dev))
		return -EOVERFLOW;

	if (!mddev->raid_disks) {
		int err;
		/* expecting a device which has a superblock */
		rdev = md_import_device(dev, mddev->major_version, mddev->minor_version);
		if (IS_ERR(rdev)) {
			printk(KERN_WARNING 
				"md: md_import_device returned %ld\n",
				PTR_ERR(rdev));
			return PTR_ERR(rdev);
		}
		if (!list_empty(&mddev->disks)) {
			mdk_rdev_t *rdev0 = list_entry(mddev->disks.next,
							mdk_rdev_t, same_set);
			int err = super_types[mddev->major_version]
				.load_super(rdev, rdev0, mddev->minor_version);
			if (err < 0) {
				printk(KERN_WARNING 
					"md: %s has different UUID to %s\n",
					bdevname(rdev->bdev,b), 
					bdevname(rdev0->bdev,b2));
				export_rdev(rdev);
				return -EINVAL;
			}
		}
		err = bind_rdev_to_array(rdev, mddev);
		if (err)
			export_rdev(rdev);
		return err;
	}

	/*
	 * add_new_disk can be used once the array is assembled
	 * to add "hot spares".  They must already have a superblock
	 * written
	 */
	if (mddev->pers) {
		int err;
		if (!mddev->pers->hot_add_disk) {
			printk(KERN_WARNING 
				"%s: personality does not support diskops!\n",
			       mdname(mddev));
			return -EINVAL;
		}
		rdev = md_import_device(dev, mddev->major_version,
					mddev->minor_version);
		if (IS_ERR(rdev)) {
			printk(KERN_WARNING 
				"md: md_import_device returned %ld\n",
				PTR_ERR(rdev));
			return PTR_ERR(rdev);
		}
		rdev->in_sync = 0; /* just to be sure */
		rdev->raid_disk = -1;
		err = bind_rdev_to_array(rdev, mddev);
		if (err)
			export_rdev(rdev);
		if (mddev->thread)
			md_wakeup_thread(mddev->thread);
		return err;
	}

	/* otherwise, add_new_disk is only allowed
	 * for major_version==0 superblocks
	 */
	if (mddev->major_version != 0) {
		printk(KERN_WARNING "%s: ADD_NEW_DISK not supported\n",
		       mdname(mddev));
		return -EINVAL;
	}

	if (!(info->state & (1<<MD_DISK_FAULTY))) {
		int err;
		rdev = md_import_device (dev, -1, 0);
		if (IS_ERR(rdev)) {
			printk(KERN_WARNING 
				"md: error, md_import_device() returned %ld\n",
				PTR_ERR(rdev));
			return PTR_ERR(rdev);
		}
		rdev->desc_nr = info->number;
		if (info->raid_disk < mddev->raid_disks)
			rdev->raid_disk = info->raid_disk;
		else
			rdev->raid_disk = -1;

		rdev->faulty = 0;
		if (rdev->raid_disk < mddev->raid_disks)
			rdev->in_sync = (info->state & (1<<MD_DISK_SYNC));
		else
			rdev->in_sync = 0;

		err = bind_rdev_to_array(rdev, mddev);
		if (err) {
			export_rdev(rdev);
			return err;
		}

		if (!mddev->persistent) {
			printk(KERN_INFO "md: nonpersistent superblock ...\n");
			rdev->sb_offset = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
		} else 
			rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
		rdev->size = calc_dev_size(rdev, mddev->chunk_size);

		if (!mddev->size || (mddev->size > rdev->size))
			mddev->size = rdev->size;
	}

	return 0;
}

static int hot_remove_disk(mddev_t * mddev, dev_t dev)
{
	char b[BDEVNAME_SIZE];
	mdk_rdev_t *rdev;

	if (!mddev->pers)
		return -ENODEV;

	rdev = find_rdev(mddev, dev);
	if (!rdev)
		return -ENXIO;

	if (rdev->raid_disk >= 0)
		goto busy;

	kick_rdev_from_array(rdev);
	md_update_sb(mddev);

	return 0;
busy: