md.c

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

	printk(KERN_WARNING "md: cannot remove active disk %s from %s ... \n",
		bdevname(rdev->bdev,b), mdname(mddev));
	return -EBUSY;
}

static int hot_add_disk(mddev_t * mddev, dev_t dev)
{
	char b[BDEVNAME_SIZE];
	int err;
	unsigned int size;
	mdk_rdev_t *rdev;

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

	if (mddev->major_version != 0) {
		printk(KERN_WARNING "%s: HOT_ADD may only be used with"
			" version-0 superblocks.\n",
			mdname(mddev));
		return -EINVAL;
	}
	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, -1, 0);
	if (IS_ERR(rdev)) {
		printk(KERN_WARNING 
			"md: error, md_import_device() returned %ld\n",
			PTR_ERR(rdev));
		return -EINVAL;
	}

	if (mddev->persistent)
		rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
	else
		rdev->sb_offset =
			rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;

	size = calc_dev_size(rdev, mddev->chunk_size);
	rdev->size = size;

	if (size < mddev->size) {
		printk(KERN_WARNING 
			"%s: disk size %llu blocks < array size %llu\n",
			mdname(mddev), (unsigned long long)size,
			(unsigned long long)mddev->size);
		err = -ENOSPC;
		goto abort_export;
	}

	if (rdev->faulty) {
		printk(KERN_WARNING 
			"md: can not hot-add faulty %s disk to %s!\n",
			bdevname(rdev->bdev,b), mdname(mddev));
		err = -EINVAL;
		goto abort_export;
	}
	rdev->in_sync = 0;
	rdev->desc_nr = -1;
	bind_rdev_to_array(rdev, mddev);

	/*
	 * The rest should better be atomic, we can have disk failures
	 * noticed in interrupt contexts ...
	 */

	if (rdev->desc_nr == mddev->max_disks) {
		printk(KERN_WARNING "%s: can not hot-add to full array!\n",
			mdname(mddev));
		err = -EBUSY;
		goto abort_unbind_export;
	}

	rdev->raid_disk = -1;

	md_update_sb(mddev);

	/*
	 * Kick recovery, maybe this spare has to be added to the
	 * array immediately.
	 */
	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	md_wakeup_thread(mddev->thread);

	return 0;

abort_unbind_export:
	unbind_rdev_from_array(rdev);

abort_export:
	export_rdev(rdev);
	return err;
}

/*
 * set_array_info is used two different ways
 * The original usage is when creating a new array.
 * In this usage, raid_disks is > 0 and it together with
 *  level, size, not_persistent,layout,chunksize determine the
 *  shape of the array.
 *  This will always create an array with a type-0.90.0 superblock.
 * The newer usage is when assembling an array.
 *  In this case raid_disks will be 0, and the major_version field is
 *  use to determine which style super-blocks are to be found on the devices.
 *  The minor and patch _version numbers are also kept incase the
 *  super_block handler wishes to interpret them.
 */
static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
{

	if (info->raid_disks == 0) {
		/* just setting version number for superblock loading */
		if (info->major_version < 0 ||
		    info->major_version >= sizeof(super_types)/sizeof(super_types[0]) ||
		    super_types[info->major_version].name == NULL) {
			/* maybe try to auto-load a module? */
			printk(KERN_INFO 
				"md: superblock version %d not known\n",
				info->major_version);
			return -EINVAL;
		}
		mddev->major_version = info->major_version;
		mddev->minor_version = info->minor_version;
		mddev->patch_version = info->patch_version;
		return 0;
	}
	mddev->major_version = MD_MAJOR_VERSION;
	mddev->minor_version = MD_MINOR_VERSION;
	mddev->patch_version = MD_PATCHLEVEL_VERSION;
	mddev->ctime         = get_seconds();

	mddev->level         = info->level;
	mddev->size          = info->size;
	mddev->raid_disks    = info->raid_disks;
	/* don't set md_minor, it is determined by which /dev/md* was
	 * openned
	 */
	if (info->state & (1<<MD_SB_CLEAN))
		mddev->recovery_cp = MaxSector;
	else
		mddev->recovery_cp = 0;
	mddev->persistent    = ! info->not_persistent;

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

	mddev->max_disks     = MD_SB_DISKS;

	mddev->sb_dirty      = 1;

	/*
	 * Generate a 128 bit UUID
	 */
	get_random_bytes(mddev->uuid, 16);

	return 0;
}

/*
 * update_array_info is used to change the configuration of an
 * on-line array.
 * The version, ctime,level,size,raid_disks,not_persistent, layout,chunk_size
 * fields in the info are checked against the array.
 * Any differences that cannot be handled will cause an error.
 * Normally, only one change can be managed at a time.
 */
static int update_array_info(mddev_t *mddev, mdu_array_info_t *info)
{
	int rv = 0;
	int cnt = 0;

	if (mddev->major_version != info->major_version ||
	    mddev->minor_version != info->minor_version ||
/*	    mddev->patch_version != info->patch_version || */
	    mddev->ctime         != info->ctime         ||
	    mddev->level         != info->level         ||
	    mddev->layout        != info->layout        ||
	    !mddev->persistent	 != info->not_persistent||
	    mddev->chunk_size    != info->chunk_size    )
		return -EINVAL;
	/* Check there is only one change */
	if (mddev->size != info->size) cnt++;
	if (mddev->raid_disks != info->raid_disks) cnt++;
	if (cnt == 0) return 0;
	if (cnt > 1) return -EINVAL;

	if (mddev->size != info->size) {
		mdk_rdev_t * rdev;
		struct list_head *tmp;
		if (mddev->pers->resize == NULL)
			return -EINVAL;
		/* The "size" is the amount of each device that is used.
		 * This can only make sense for arrays with redundancy.
		 * linear and raid0 always use whatever space is available
		 * We can only consider changing the size of no resync
		 * or reconstruction is happening, and if the new size
		 * is acceptable. It must fit before the sb_offset or,
		 * if that is <data_offset, it must fit before the
		 * size of each device.
		 * If size is zero, we find the largest size that fits.
		 */
		if (mddev->sync_thread)
			return -EBUSY;
		ITERATE_RDEV(mddev,rdev,tmp) {
			sector_t avail;
			int fit = (info->size == 0);
			if (rdev->sb_offset > rdev->data_offset)
				avail = (rdev->sb_offset*2) - rdev->data_offset;
			else
				avail = get_capacity(rdev->bdev->bd_disk)
					- rdev->data_offset;
			if (fit && (info->size == 0 || info->size > avail/2))
				info->size = avail/2;
			if (avail < ((sector_t)info->size << 1))
				return -ENOSPC;
		}
		rv = mddev->pers->resize(mddev, (sector_t)info->size *2);
		if (!rv) {
			struct block_device *bdev;

			bdev = bdget_disk(mddev->gendisk, 0);
			if (bdev) {
				down(&bdev->bd_inode->i_sem);
				i_size_write(bdev->bd_inode, mddev->array_size << 10);
				up(&bdev->bd_inode->i_sem);
				bdput(bdev);
			}
		}
	}
	if (mddev->raid_disks    != info->raid_disks) {
		/* change the number of raid disks */
		if (mddev->pers->reshape == NULL)
			return -EINVAL;
		if (info->raid_disks <= 0 ||
		    info->raid_disks >= mddev->max_disks)
			return -EINVAL;
		if (mddev->sync_thread)
			return -EBUSY;
		rv = mddev->pers->reshape(mddev, info->raid_disks);
		if (!rv) {
			struct block_device *bdev;

			bdev = bdget_disk(mddev->gendisk, 0);
			if (bdev) {
				down(&bdev->bd_inode->i_sem);
				i_size_write(bdev->bd_inode, mddev->array_size << 10);
				up(&bdev->bd_inode->i_sem);
				bdput(bdev);
			}
		}
	}
	md_update_sb(mddev);
	return rv;
}

static int set_disk_faulty(mddev_t *mddev, dev_t dev)
{
	mdk_rdev_t *rdev;

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

	md_error(mddev, rdev);
	return 0;
}

static int md_ioctl(struct inode *inode, struct file *file,
			unsigned int cmd, unsigned long arg)
{
	int err = 0;
	void __user *argp = (void __user *)arg;
	struct hd_geometry __user *loc = argp;
	mddev_t *mddev = NULL;

	if (!capable(CAP_SYS_ADMIN))
		return -EACCES;

	/*
	 * Commands dealing with the RAID driver but not any
	 * particular array:
	 */
	switch (cmd)
	{
		case RAID_VERSION:
			err = get_version(argp);
			goto done;

		case PRINT_RAID_DEBUG:
			err = 0;
			md_print_devices();
			goto done;

#ifndef MODULE
		case RAID_AUTORUN:
			err = 0;
			autostart_arrays(arg);
			goto done;
#endif
		default:;
	}

	/*
	 * Commands creating/starting a new array:
	 */

	mddev = inode->i_bdev->bd_disk->private_data;

	if (!mddev) {
		BUG();
		goto abort;
	}


	if (cmd == START_ARRAY) {
		/* START_ARRAY doesn't need to lock the array as autostart_array
		 * does the locking, and it could even be a different array
		 */
		static int cnt = 3;
		if (cnt > 0 ) {
			printk(KERN_WARNING
			       "md: %s(pid %d) used deprecated START_ARRAY ioctl. "
			       "This will not be supported beyond 2.6\n",
			       current->comm, current->pid);
			cnt--;
		}
		err = autostart_array(new_decode_dev(arg));
		if (err) {
			printk(KERN_WARNING "md: autostart failed!\n");
			goto abort;
		}
		goto done;
	}

	err = mddev_lock(mddev);
	if (err) {
		printk(KERN_INFO 
			"md: ioctl lock interrupted, reason %d, cmd %d\n",
			err, cmd);
		goto abort;
	}

	switch (cmd)
	{
		case SET_ARRAY_INFO:
			{
				mdu_array_info_t info;
				if (!arg)
					memset(&info, 0, sizeof(info));
				else if (copy_from_user(&info, argp, sizeof(info))) {
					err = -EFAULT;
					goto abort_unlock;
				}
				if (mddev->pers) {
					err = update_array_info(mddev, &info);
					if (err) {
						printk(KERN_WARNING "md: couldn't update"
						       " array info. %d\n", err);
						goto abort_unlock;
					}
					goto done_unlock;
				}
				if (!list_empty(&mddev->disks)) {
					printk(KERN_WARNING
					       "md: array %s already has disks!\n",
					       mdname(mddev));
					err = -EBUSY;
					goto abort_unlock;
				}
				if (mddev->raid_disks) {
					printk(KERN_WARNING
					       "md: array %s already initialised!\n",
					       mdname(mddev));
					err = -EBUSY;
					goto abort_unlock;
				}
				err = set_array_info(mddev, &info);
				if (err) {
					printk(KERN_WARNING "md: couldn't set"
					       " array info. %d\n", err);
					goto abort_unlock;
				}
			}
			goto done_unlock;

		default:;
	}

	/*
	 * Commands querying/configuring an existing array:
	 */
	/* if we are initialised yet, only ADD_NEW_DISK or STOP_ARRAY is allowed */
	if (!mddev->raid_disks && cmd != ADD_NEW_DISK && cmd != STOP_ARRAY && cmd != RUN_ARRAY) {
		err = -ENODEV;
		goto abort_unlock;
	}

	/*
	 * Commands even a read-only array can execute:
	 */
	switch (cmd)
	{
		case GET_ARRAY_INFO:
			err = get_array_info(mddev, argp);
			goto done_unlock;

		case GET_DISK_INFO:
			err = get_disk_info(mddev, argp);
			goto done_unlock;

		case RESTART_ARRAY_RW:
			err = restart_array(mddev);
			goto done_unlock;

		case STOP_ARRAY:
			err = do_md_stop (mddev, 0);
			goto done_unlock;

		case STOP_ARRAY_RO:
			err = do_md_stop (mddev, 1);
			goto done_unlock;

	/*
	 * We have a problem here : there is no easy way to give a CHS
	 * virtual geometry. We currently pretend that we have a 2 heads
	 * 4 sectors (with a BIG number of cylinders...). This drives
	 * dosfs just mad... ;-)
	 */
		case HDIO_GETGEO:
			if (!loc) {
				err = -EINVAL;
				goto abort_unlock;
			}
			err = put_user (2, (char __user *) &loc->heads);
			if (err)
				goto abort_unlock;
			err = put_user (4, (char __user *) &loc->sectors);
			if (err)
				goto abort_unlock;
			err = put_user(get_capacity(mddev->gendisk)/8,
					(short __user *) &loc->cylinders);
			if (err)
				goto abort_unlock;
			err = put_user (get_start_sect(inode->i_bdev),
						(long __user *) &loc->start);
			goto done_unlock;
	}

	/*
	 * The remaining ioctls are changing the state of the
	 * superblock, so we do not allow read-only arrays
	 * here:
	 */
	if (mddev->ro) {
		err = -EROFS;
		goto abort_unlock;
	}

	switch (cmd)
	{
		case ADD_NEW_DISK:
		{
			mdu_disk_info_t info;
			if (copy_from_user(&info, argp, sizeof(info)))
				err = -EFAULT;
			else
				err = add_new_disk(mddev, &info);
			goto done_unlock;
		}

		case HOT_REMOVE_DISK:
			err = hot_remove_disk(mddev, new_decode_dev(arg));
			goto done_unlock;

		case HOT_ADD_DISK:
			err = hot_add_disk(mddev, new_decode_dev(arg));
			goto done_unlock;

		case SET_DISK_FAULTY:
			err = set_disk_faulty(mddev, new_decode_dev(arg));
			goto done_unlock;

		case RUN_ARRAY:
			err = do_md_run (mddev);
			goto done_unlock;

		default:
			if (_IOC_TYPE(cmd) == MD_MAJOR)
				printk(KERN_WARNING "md: %s(pid %d) used"
					" obsolete MD ioctl, upgrade your"
					" software to use new ictls.\n",
					current->comm, current->pid);
			err = -EINVAL;
			goto abort_unlock;
	}

done_unlock:
abort_unlock:
	mddev_unlock(mddev);

	return err;
done:
	if (err)
		MD_BUG();
abort:
	return err;
}

static int md_open(struct inode *inode, struct file *file)
{
	/*
	 * Succeed if we can lock the mddev, which confirms that
	 * it isn't being stopped right now.
	 */
	mddev_t *mddev = inode->i_bdev->bd_disk->private_data;
	int err;

	if ((err = mddev_lock(mddev)))
		goto out;

	err = 0;
	mddev_get(mddev);
	mddev_unlock(mddev);

	check_disk_change(inode->i_bdev);
 out:
	return err;
}

static int md_release(struct inode *inode, struct file * file)
{
 	mddev_t *mddev = inode->i_bdev->bd_disk->private_data;

	if (!mddev)
		BUG();
	mddev_put(mddev);

	return 0;
}

static int md_media_changed(struct gendisk *disk)
{
	mddev_t *mddev = disk->private_data;

	return mddev->changed;
}

static int md_revalidate(struct gendisk *disk)
{
	mddev_t *mddev = disk->private_data;

	mddev->changed =