Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
-#include <linux/module.h>
-#include <linux/kernel.h>
#include <linux/kthread.h>
-#include <linux/linkage.h>
-#include <linux/raid/md.h>
-#include <linux/raid/bitmap.h>
+#include <linux/blkdev.h>
#include <linux/sysctl.h>
+#include <linux/seq_file.h>
#include <linux/buffer_head.h> /* for invalidate_bdev */
#include <linux/poll.h>
-#include <linux/mutex.h>
#include <linux/ctype.h>
-#include <linux/freezer.h>
-
-#include <linux/init.h>
-
+#include <linux/hdreg.h>
+#include <linux/proc_fs.h>
+#include <linux/random.h>
+#include <linux/reboot.h>
#include <linux/file.h>
-
-#ifdef CONFIG_KMOD
-#include <linux/kmod.h>
-#endif
-
-#include <asm/unaligned.h>
-
-#define MAJOR_NR MD_MAJOR
-#define MD_DRIVER
-
-/* 63 partitions with the alternate major number (mdp) */
-#define MdpMinorShift 6
+#include <linux/delay.h>
+#include <linux/raid/md_p.h>
+#include <linux/raid/md_u.h>
+#include "md.h"
+#include "bitmap.h"
#define DEBUG 0
#define dprintk(x...) ((void)(DEBUG && printk(x)))
#ifndef MODULE
-static void autostart_arrays (int part);
+static void autostart_arrays(int part);
#endif
static LIST_HEAD(pers_list);
{
atomic_inc(&md_event_count);
wake_up(&md_event_waiters);
- sysfs_notify(&mddev->kobj, NULL, "sync_action");
}
EXPORT_SYMBOL_GPL(md_new_event);
)
-static int md_fail_request (struct request_queue *q, struct bio *bio)
+/* Rather than calling directly into the personality make_request function,
+ * IO requests come here first so that we can check if the device is
+ * being suspended pending a reconfiguration.
+ * We hold a refcount over the call to ->make_request. By the time that
+ * call has finished, the bio has been linked into some internal structure
+ * and so is visible to ->quiesce(), so we don't need the refcount any more.
+ */
+static int md_make_request(struct request_queue *q, struct bio *bio)
{
- bio_io_error(bio);
- return 0;
+ mddev_t *mddev = q->queuedata;
+ int rv;
+ if (mddev == NULL || mddev->pers == NULL) {
+ bio_io_error(bio);
+ return 0;
+ }
+ rcu_read_lock();
+ if (mddev->suspended) {
+ DEFINE_WAIT(__wait);
+ for (;;) {
+ prepare_to_wait(&mddev->sb_wait, &__wait,
+ TASK_UNINTERRUPTIBLE);
+ if (!mddev->suspended)
+ break;
+ rcu_read_unlock();
+ schedule();
+ rcu_read_lock();
+ }
+ finish_wait(&mddev->sb_wait, &__wait);
+ }
+ atomic_inc(&mddev->active_io);
+ rcu_read_unlock();
+ rv = mddev->pers->make_request(q, bio);
+ if (atomic_dec_and_test(&mddev->active_io) && mddev->suspended)
+ wake_up(&mddev->sb_wait);
+
+ return rv;
+}
+
+static void mddev_suspend(mddev_t *mddev)
+{
+ BUG_ON(mddev->suspended);
+ mddev->suspended = 1;
+ synchronize_rcu();
+ wait_event(mddev->sb_wait, atomic_read(&mddev->active_io) == 0);
+ mddev->pers->quiesce(mddev, 1);
+ md_unregister_thread(mddev->thread);
+ mddev->thread = NULL;
+ /* we now know that no code is executing in the personality module,
+ * except possibly the tail end of a ->bi_end_io function, but that
+ * is certain to complete before the module has a chance to get
+ * unloaded
+ */
}
+static void mddev_resume(mddev_t *mddev)
+{
+ mddev->suspended = 0;
+ wake_up(&mddev->sb_wait);
+ mddev->pers->quiesce(mddev, 0);
+}
+
+
static inline mddev_t *mddev_get(mddev_t *mddev)
{
atomic_inc(&mddev->active);
return mddev;
}
+static void mddev_delayed_delete(struct work_struct *ws)
+{
+ mddev_t *mddev = container_of(ws, mddev_t, del_work);
+ kobject_del(&mddev->kobj);
+ kobject_put(&mddev->kobj);
+}
+
static void mddev_put(mddev_t *mddev)
{
if (!atomic_dec_and_lock(&mddev->active, &all_mddevs_lock))
return;
- if (!mddev->raid_disks && list_empty(&mddev->disks)) {
+ if (!mddev->raid_disks && list_empty(&mddev->disks) &&
+ !mddev->hold_active) {
list_del(&mddev->all_mddevs);
- spin_unlock(&all_mddevs_lock);
- blk_cleanup_queue(mddev->queue);
- kobject_put(&mddev->kobj);
- } else
- spin_unlock(&all_mddevs_lock);
+ if (mddev->gendisk) {
+ /* we did a probe so need to clean up.
+ * Call schedule_work inside the spinlock
+ * so that flush_scheduled_work() after
+ * mddev_find will succeed in waiting for the
+ * work to be done.
+ */
+ INIT_WORK(&mddev->del_work, mddev_delayed_delete);
+ schedule_work(&mddev->del_work);
+ } else
+ kfree(mddev);
+ }
+ spin_unlock(&all_mddevs_lock);
}
static mddev_t * mddev_find(dev_t unit)
retry:
spin_lock(&all_mddevs_lock);
- list_for_each_entry(mddev, &all_mddevs, all_mddevs)
- if (mddev->unit == unit) {
- mddev_get(mddev);
+
+ if (unit) {
+ list_for_each_entry(mddev, &all_mddevs, all_mddevs)
+ if (mddev->unit == unit) {
+ mddev_get(mddev);
+ spin_unlock(&all_mddevs_lock);
+ kfree(new);
+ return mddev;
+ }
+
+ if (new) {
+ list_add(&new->all_mddevs, &all_mddevs);
spin_unlock(&all_mddevs_lock);
- kfree(new);
- return mddev;
+ new->hold_active = UNTIL_IOCTL;
+ return new;
}
-
- if (new) {
+ } else if (new) {
+ /* find an unused unit number */
+ static int next_minor = 512;
+ int start = next_minor;
+ int is_free = 0;
+ int dev = 0;
+ while (!is_free) {
+ dev = MKDEV(MD_MAJOR, next_minor);
+ next_minor++;
+ if (next_minor > MINORMASK)
+ next_minor = 0;
+ if (next_minor == start) {
+ /* Oh dear, all in use. */
+ spin_unlock(&all_mddevs_lock);
+ kfree(new);
+ return NULL;
+ }
+
+ is_free = 1;
+ list_for_each_entry(mddev, &all_mddevs, all_mddevs)
+ if (mddev->unit == dev) {
+ is_free = 0;
+ break;
+ }
+ }
+ new->unit = dev;
+ new->md_minor = MINOR(dev);
+ new->hold_active = UNTIL_STOP;
list_add(&new->all_mddevs, &all_mddevs);
spin_unlock(&all_mddevs_lock);
return new;
INIT_LIST_HEAD(&new->all_mddevs);
init_timer(&new->safemode_timer);
atomic_set(&new->active, 1);
+ atomic_set(&new->openers, 0);
+ atomic_set(&new->active_io, 0);
spin_lock_init(&new->write_lock);
init_waitqueue_head(&new->sb_wait);
init_waitqueue_head(&new->recovery_wait);
new->resync_max = MaxSector;
new->level = LEVEL_NONE;
- new->queue = blk_alloc_queue(GFP_KERNEL);
- if (!new->queue) {
- kfree(new);
- return NULL;
- }
- /* Can be unlocked because the queue is new: no concurrency */
- queue_flag_set_unlocked(QUEUE_FLAG_CLUSTER, new->queue);
-
- blk_queue_make_request(new->queue, md_fail_request);
-
goto retry;
}
return mutex_lock_interruptible(&mddev->reconfig_mutex);
}
+static inline int mddev_is_locked(mddev_t *mddev)
+{
+ return mutex_is_locked(&mddev->reconfig_mutex);
+}
+
static inline int mddev_trylock(mddev_t * mddev)
{
return mutex_trylock(&mddev->reconfig_mutex);
static mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr)
{
- mdk_rdev_t * rdev;
- struct list_head *tmp;
+ mdk_rdev_t *rdev;
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->desc_nr == nr)
return rdev;
- }
+
return NULL;
}
static mdk_rdev_t * find_rdev(mddev_t * mddev, dev_t dev)
{
- struct list_head *tmp;
mdk_rdev_t *rdev;
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->bdev->bd_dev == dev)
return rdev;
- }
+
return NULL;
}
return NULL;
}
+/* return the offset of the super block in 512byte sectors */
static inline sector_t calc_dev_sboffset(struct block_device *bdev)
{
- sector_t size = bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
- return MD_NEW_SIZE_BLOCKS(size);
+ sector_t num_sectors = bdev->bd_inode->i_size / 512;
+ return MD_NEW_SIZE_SECTORS(num_sectors);
}
-static sector_t calc_dev_size(mdk_rdev_t *rdev, unsigned chunk_size)
+static sector_t calc_num_sectors(mdk_rdev_t *rdev, unsigned chunk_size)
{
- sector_t size;
-
- size = rdev->sb_offset;
+ sector_t num_sectors = rdev->sb_start;
if (chunk_size)
- size &= ~((sector_t)chunk_size/1024 - 1);
- return size;
+ num_sectors &= ~((sector_t)chunk_size/512 - 1);
+ return num_sectors;
}
static int alloc_disk_sb(mdk_rdev_t * rdev)
rdev->sb_page = alloc_page(GFP_KERNEL);
if (!rdev->sb_page) {
printk(KERN_ALERT "md: out of memory.\n");
- return -EINVAL;
+ return -ENOMEM;
}
return 0;
put_page(rdev->sb_page);
rdev->sb_loaded = 0;
rdev->sb_page = NULL;
- rdev->sb_offset = 0;
- rdev->size = 0;
+ rdev->sb_start = 0;
+ rdev->sectors = 0;
}
}
* causes ENOTSUPP, we allocate a spare bio...
*/
struct bio *bio = bio_alloc(GFP_NOIO, 1);
- int rw = (1<<BIO_RW) | (1<<BIO_RW_SYNC);
+ int rw = (1<<BIO_RW) | (1<<BIO_RW_SYNCIO) | (1<<BIO_RW_UNPLUG);
bio->bi_bdev = rdev->bdev;
bio->bi_sector = sector;
struct completion event;
int ret;
- rw |= (1 << BIO_RW_SYNC);
+ rw |= (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG);
bio->bi_bdev = bdev;
bio->bi_sector = sector;
return 0;
- if (!sync_page_io(rdev->bdev, rdev->sb_offset<<1, size, rdev->sb_page, READ))
+ if (!sync_page_io(rdev->bdev, rdev->sb_start, size, rdev->sb_page, READ))
goto fail;
rdev->sb_loaded = 1;
return 0;
static int uuid_equal(mdp_super_t *sb1, mdp_super_t *sb2)
{
- if ( (sb1->set_uuid0 == sb2->set_uuid0) &&
- (sb1->set_uuid1 == sb2->set_uuid1) &&
- (sb1->set_uuid2 == sb2->set_uuid2) &&
- (sb1->set_uuid3 == sb2->set_uuid3))
-
- return 1;
-
- return 0;
+ return sb1->set_uuid0 == sb2->set_uuid0 &&
+ sb1->set_uuid1 == sb2->set_uuid1 &&
+ sb1->set_uuid2 == sb2->set_uuid2 &&
+ sb1->set_uuid3 == sb2->set_uuid3;
}
-
static int sb_equal(mdp_super_t *sb1, mdp_super_t *sb2)
{
int ret;
if (!tmp1 || !tmp2) {
ret = 0;
- printk(KERN_INFO "md.c: sb1 is not equal to sb2!\n");
+ printk(KERN_INFO "md.c sb_equal(): failed to allocate memory!\n");
goto abort;
}
tmp1->nr_disks = 0;
tmp2->nr_disks = 0;
- if (memcmp(tmp1, tmp2, MD_SB_GENERIC_CONSTANT_WORDS * 4))
- ret = 0;
- else
- ret = 1;
-
+ ret = (memcmp(tmp1, tmp2, MD_SB_GENERIC_CONSTANT_WORDS * 4) == 0);
abort:
kfree(tmp1);
kfree(tmp2);
*/
struct super_type {
- char *name;
- struct module *owner;
- int (*load_super)(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version);
- int (*validate_super)(mddev_t *mddev, mdk_rdev_t *rdev);
- void (*sync_super)(mddev_t *mddev, mdk_rdev_t *rdev);
+ char *name;
+ struct module *owner;
+ int (*load_super)(mdk_rdev_t *rdev, mdk_rdev_t *refdev,
+ int minor_version);
+ int (*validate_super)(mddev_t *mddev, mdk_rdev_t *rdev);
+ void (*sync_super)(mddev_t *mddev, mdk_rdev_t *rdev);
+ unsigned long long (*rdev_size_change)(mdk_rdev_t *rdev,
+ sector_t num_sectors);
};
/*
char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
mdp_super_t *sb;
int ret;
- sector_t sb_offset;
/*
- * Calculate the position of the superblock,
+ * Calculate the position of the superblock (512byte sectors),
* it's at the end of the disk.
*
* It also happens to be a multiple of 4Kb.
*/
- sb_offset = calc_dev_sboffset(rdev->bdev);
- rdev->sb_offset = sb_offset;
+ rdev->sb_start = calc_dev_sboffset(rdev->bdev);
ret = read_disk_sb(rdev, MD_SB_BYTES);
if (ret) return ret;
else
ret = 0;
}
- rdev->size = calc_dev_size(rdev, sb->chunk_size);
+ rdev->sectors = calc_num_sectors(rdev, sb->chunk_size);
- if (rdev->size < sb->size && sb->level > 1)
+ if (rdev->sectors < sb->size * 2 && sb->level > 1)
/* "this cannot possibly happen" ... */
ret = -EINVAL;
mddev->clevel[0] = 0;
mddev->layout = sb->layout;
mddev->raid_disks = sb->raid_disks;
- mddev->size = sb->size;
+ mddev->dev_sectors = sb->size * 2;
mddev->events = ev1;
mddev->bitmap_offset = 0;
mddev->default_bitmap_offset = MD_SB_BYTES >> 9;
static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
{
mdp_super_t *sb;
- struct list_head *tmp;
mdk_rdev_t *rdev2;
int next_spare = mddev->raid_disks;
sb->ctime = mddev->ctime;
sb->level = mddev->level;
- sb->size = mddev->size;
+ sb->size = mddev->dev_sectors / 2;
sb->raid_disks = mddev->raid_disks;
sb->md_minor = mddev->md_minor;
sb->not_persistent = 0;
sb->state |= (1<<MD_SB_BITMAP_PRESENT);
sb->disks[0].state = (1<<MD_DISK_REMOVED);
- rdev_for_each(rdev2, tmp, mddev) {
+ list_for_each_entry(rdev2, &mddev->disks, same_set) {
mdp_disk_t *d;
int desc_nr;
if (rdev2->raid_disk >= 0 && test_bit(In_sync, &rdev2->flags)
}
/*
+ * rdev_size_change for 0.90.0
+ */
+static unsigned long long
+super_90_rdev_size_change(mdk_rdev_t *rdev, sector_t num_sectors)
+{
+ if (num_sectors && num_sectors < rdev->mddev->dev_sectors)
+ return 0; /* component must fit device */
+ if (rdev->mddev->bitmap_offset)
+ return 0; /* can't move bitmap */
+ rdev->sb_start = calc_dev_sboffset(rdev->bdev);
+ if (!num_sectors || num_sectors > rdev->sb_start)
+ num_sectors = rdev->sb_start;
+ md_super_write(rdev->mddev, rdev, rdev->sb_start, rdev->sb_size,
+ rdev->sb_page);
+ md_super_wait(rdev->mddev);
+ return num_sectors / 2; /* kB for sysfs */
+}
+
+
+/*
* version 1 superblock
*/
{
struct mdp_superblock_1 *sb;
int ret;
- sector_t sb_offset;
+ sector_t sb_start;
char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
int bmask;
/*
- * Calculate the position of the superblock.
+ * Calculate the position of the superblock in 512byte sectors.
* It is always aligned to a 4K boundary and
* depeding on minor_version, it can be:
* 0: At least 8K, but less than 12K, from end of device
*/
switch(minor_version) {
case 0:
- sb_offset = rdev->bdev->bd_inode->i_size >> 9;
- sb_offset -= 8*2;
- sb_offset &= ~(sector_t)(4*2-1);
- /* convert from sectors to K */
- sb_offset /= 2;
+ sb_start = rdev->bdev->bd_inode->i_size >> 9;
+ sb_start -= 8*2;
+ sb_start &= ~(sector_t)(4*2-1);
break;
case 1:
- sb_offset = 0;
+ sb_start = 0;
break;
case 2:
- sb_offset = 4;
+ sb_start = 8;
break;
default:
return -EINVAL;
}
- rdev->sb_offset = sb_offset;
+ rdev->sb_start = sb_start;
/* superblock is rarely larger than 1K, but it can be larger,
* and it is safe to read 4k, so we do that
if (sb->magic != cpu_to_le32(MD_SB_MAGIC) ||
sb->major_version != cpu_to_le32(1) ||
le32_to_cpu(sb->max_dev) > (4096-256)/2 ||
- le64_to_cpu(sb->super_offset) != (rdev->sb_offset<<1) ||
+ le64_to_cpu(sb->super_offset) != rdev->sb_start ||
(le32_to_cpu(sb->feature_map) & ~MD_FEATURE_ALL) != 0)
return -EINVAL;
rdev->sb_size = (rdev->sb_size | bmask) + 1;
if (minor_version
- && rdev->data_offset < sb_offset + (rdev->sb_size/512))
+ && rdev->data_offset < sb_start + (rdev->sb_size/512))
return -EINVAL;
if (sb->level == cpu_to_le32(LEVEL_MULTIPATH))
ret = 0;
}
if (minor_version)
- rdev->size = ((rdev->bdev->bd_inode->i_size>>9) - le64_to_cpu(sb->data_offset)) / 2;
+ rdev->sectors = (rdev->bdev->bd_inode->i_size >> 9) -
+ le64_to_cpu(sb->data_offset);
else
- rdev->size = rdev->sb_offset;
- if (rdev->size < le64_to_cpu(sb->data_size)/2)
+ rdev->sectors = rdev->sb_start;
+ if (rdev->sectors < le64_to_cpu(sb->data_size))
return -EINVAL;
- rdev->size = le64_to_cpu(sb->data_size)/2;
+ rdev->sectors = le64_to_cpu(sb->data_size);
if (le32_to_cpu(sb->chunksize))
- rdev->size &= ~((sector_t)le32_to_cpu(sb->chunksize)/2 - 1);
+ rdev->sectors &= ~((sector_t)le32_to_cpu(sb->chunksize) - 1);
- if (le64_to_cpu(sb->size) > rdev->size*2)
+ if (le64_to_cpu(sb->size) > rdev->sectors)
return -EINVAL;
return ret;
}
mddev->clevel[0] = 0;
mddev->layout = le32_to_cpu(sb->layout);
mddev->raid_disks = le32_to_cpu(sb->raid_disks);
- mddev->size = le64_to_cpu(sb->size)/2;
+ mddev->dev_sectors = le64_to_cpu(sb->size);
mddev->events = ev1;
mddev->bitmap_offset = 0;
mddev->default_bitmap_offset = 1024 >> 9;
static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
{
struct mdp_superblock_1 *sb;
- struct list_head *tmp;
mdk_rdev_t *rdev2;
int max_dev, i;
/* make rdev->sb match mddev and rdev data. */
sb->cnt_corrected_read = cpu_to_le32(atomic_read(&rdev->corrected_errors));
sb->raid_disks = cpu_to_le32(mddev->raid_disks);
- sb->size = cpu_to_le64(mddev->size<<1);
+ sb->size = cpu_to_le64(mddev->dev_sectors);
if (mddev->bitmap && mddev->bitmap_file == NULL) {
sb->bitmap_offset = cpu_to_le32((__u32)mddev->bitmap_offset);
}
if (rdev->raid_disk >= 0 &&
- !test_bit(In_sync, &rdev->flags) &&
- rdev->recovery_offset > 0) {
- sb->feature_map |= cpu_to_le32(MD_FEATURE_RECOVERY_OFFSET);
- sb->recovery_offset = cpu_to_le64(rdev->recovery_offset);
+ !test_bit(In_sync, &rdev->flags)) {
+ if (mddev->curr_resync_completed > rdev->recovery_offset)
+ rdev->recovery_offset = mddev->curr_resync_completed;
+ if (rdev->recovery_offset > 0) {
+ sb->feature_map |=
+ cpu_to_le32(MD_FEATURE_RECOVERY_OFFSET);
+ sb->recovery_offset =
+ cpu_to_le64(rdev->recovery_offset);
+ }
}
if (mddev->reshape_position != MaxSector) {
}
max_dev = 0;
- rdev_for_each(rdev2, tmp, mddev)
+ list_for_each_entry(rdev2, &mddev->disks, same_set)
if (rdev2->desc_nr+1 > max_dev)
max_dev = rdev2->desc_nr+1;
for (i=0; i<max_dev;i++)
sb->dev_roles[i] = cpu_to_le16(0xfffe);
- rdev_for_each(rdev2, tmp, mddev) {
+ list_for_each_entry(rdev2, &mddev->disks, same_set) {
i = rdev2->desc_nr;
if (test_bit(Faulty, &rdev2->flags))
sb->dev_roles[i] = cpu_to_le16(0xfffe);
sb->sb_csum = calc_sb_1_csum(sb);
}
+static unsigned long long
+super_1_rdev_size_change(mdk_rdev_t *rdev, sector_t num_sectors)
+{
+ struct mdp_superblock_1 *sb;
+ sector_t max_sectors;
+ if (num_sectors && num_sectors < rdev->mddev->dev_sectors)
+ return 0; /* component must fit device */
+ if (rdev->sb_start < rdev->data_offset) {
+ /* minor versions 1 and 2; superblock before data */
+ max_sectors = rdev->bdev->bd_inode->i_size >> 9;
+ max_sectors -= rdev->data_offset;
+ if (!num_sectors || num_sectors > max_sectors)
+ num_sectors = max_sectors;
+ } else if (rdev->mddev->bitmap_offset) {
+ /* minor version 0 with bitmap we can't move */
+ return 0;
+ } else {
+ /* minor version 0; superblock after data */
+ sector_t sb_start;
+ sb_start = (rdev->bdev->bd_inode->i_size >> 9) - 8*2;
+ sb_start &= ~(sector_t)(4*2 - 1);
+ max_sectors = rdev->sectors + sb_start - rdev->sb_start;
+ if (!num_sectors || num_sectors > max_sectors)
+ num_sectors = max_sectors;
+ rdev->sb_start = sb_start;
+ }
+ sb = (struct mdp_superblock_1 *) page_address(rdev->sb_page);
+ sb->data_size = cpu_to_le64(num_sectors);
+ sb->super_offset = rdev->sb_start;
+ sb->sb_csum = calc_sb_1_csum(sb);
+ md_super_write(rdev->mddev, rdev, rdev->sb_start, rdev->sb_size,
+ rdev->sb_page);
+ md_super_wait(rdev->mddev);
+ return num_sectors / 2; /* kB for sysfs */
+}
static struct super_type super_types[] = {
[0] = {
.name = "0.90.0",
.owner = THIS_MODULE,
- .load_super = super_90_load,
- .validate_super = super_90_validate,
- .sync_super = super_90_sync,
+ .load_super = super_90_load,
+ .validate_super = super_90_validate,
+ .sync_super = super_90_sync,
+ .rdev_size_change = super_90_rdev_size_change,
},
[1] = {
.name = "md-1",
.owner = THIS_MODULE,
- .load_super = super_1_load,
- .validate_super = super_1_validate,
- .sync_super = super_1_sync,
+ .load_super = super_1_load,
+ .validate_super = super_1_validate,
+ .sync_super = super_1_sync,
+ .rdev_size_change = super_1_rdev_size_change,
},
};
static int match_mddev_units(mddev_t *mddev1, mddev_t *mddev2)
{
- struct list_head *tmp, *tmp2;
mdk_rdev_t *rdev, *rdev2;
- rdev_for_each(rdev, tmp, mddev1)
- rdev_for_each(rdev2, tmp2, mddev2)
+ rcu_read_lock();
+ rdev_for_each_rcu(rdev, mddev1)
+ rdev_for_each_rcu(rdev2, mddev2)
if (rdev->bdev->bd_contains ==
- rdev2->bdev->bd_contains)
+ rdev2->bdev->bd_contains) {
+ rcu_read_unlock();
return 1;
-
+ }
+ rcu_read_unlock();
return 0;
}
static LIST_HEAD(pending_raid_disks);
+static void md_integrity_check(mdk_rdev_t *rdev, mddev_t *mddev)
+{
+ struct mdk_personality *pers = mddev->pers;
+ struct gendisk *disk = mddev->gendisk;
+ struct blk_integrity *bi_rdev = bdev_get_integrity(rdev->bdev);
+ struct blk_integrity *bi_mddev = blk_get_integrity(disk);
+
+ /* Data integrity passthrough not supported on RAID 4, 5 and 6 */
+ if (pers && pers->level >= 4 && pers->level <= 6)
+ return;
+
+ /* If rdev is integrity capable, register profile for mddev */
+ if (!bi_mddev && bi_rdev) {
+ if (blk_integrity_register(disk, bi_rdev))
+ printk(KERN_ERR "%s: %s Could not register integrity!\n",
+ __func__, disk->disk_name);
+ else
+ printk(KERN_NOTICE "Enabling data integrity on %s\n",
+ disk->disk_name);
+ return;
+ }
+
+ /* Check that mddev and rdev have matching profiles */
+ if (blk_integrity_compare(disk, rdev->bdev->bd_disk) < 0) {
+ printk(KERN_ERR "%s: %s/%s integrity mismatch!\n", __func__,
+ disk->disk_name, rdev->bdev->bd_disk->disk_name);
+ printk(KERN_NOTICE "Disabling data integrity on %s\n",
+ disk->disk_name);
+ blk_integrity_unregister(disk);
+ }
+}
+
static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
{
char b[BDEVNAME_SIZE];
if (find_rdev(mddev, rdev->bdev->bd_dev))
return -EEXIST;
- /* make sure rdev->size exceeds mddev->size */
- if (rdev->size && (mddev->size == 0 || rdev->size < mddev->size)) {
+ /* make sure rdev->sectors exceeds mddev->dev_sectors */
+ if (rdev->sectors && (mddev->dev_sectors == 0 ||
+ rdev->sectors < mddev->dev_sectors)) {
if (mddev->pers) {
/* Cannot change size, so fail
* If mddev->level <= 0, then we don't care
if (mddev->level > 0)
return -ENOSPC;
} else
- mddev->size = rdev->size;
+ mddev->dev_sectors = rdev->sectors;
}
/* Verify rdev->desc_nr is unique.
if (find_rdev_nr(mddev, rdev->desc_nr))
return -EBUSY;
}
+ if (mddev->max_disks && rdev->desc_nr >= mddev->max_disks) {
+ printk(KERN_WARNING "md: %s: array is limited to %d devices\n",
+ mdname(mddev), mddev->max_disks);
+ return -EBUSY;
+ }
bdevname(rdev->bdev,b);
while ( (s=strchr(b, '/')) != NULL)
*s = '!';
if ((err = kobject_add(&rdev->kobj, &mddev->kobj, "dev-%s", b)))
goto fail;
- if (rdev->bdev->bd_part)
- ko = &rdev->bdev->bd_part->dev.kobj;
- else
- ko = &rdev->bdev->bd_disk->dev.kobj;
+ ko = &part_to_dev(rdev->bdev->bd_part)->kobj;
if ((err = sysfs_create_link(&rdev->kobj, ko, "block"))) {
kobject_del(&rdev->kobj);
goto fail;
}
- list_add(&rdev->same_set, &mddev->disks);
+ rdev->sysfs_state = sysfs_get_dirent(rdev->kobj.sd, "state");
+
+ list_add_rcu(&rdev->same_set, &mddev->disks);
bd_claim_by_disk(rdev->bdev, rdev->bdev->bd_holder, mddev->gendisk);
+
+ /* May as well allow recovery to be retried once */
+ mddev->recovery_disabled = 0;
+
+ md_integrity_check(rdev, mddev);
return 0;
fail:
return;
}
bd_release_from_disk(rdev->bdev, rdev->mddev->gendisk);
- list_del_init(&rdev->same_set);
+ list_del_rcu(&rdev->same_set);
printk(KERN_INFO "md: unbind<%s>\n", bdevname(rdev->bdev,b));
rdev->mddev = NULL;
sysfs_remove_link(&rdev->kobj, "block");
-
+ sysfs_put(rdev->sysfs_state);
+ rdev->sysfs_state = NULL;
/* We need to delay this, otherwise we can deadlock when
- * writing to 'remove' to "dev/state"
+ * writing to 'remove' to "dev/state". We also need
+ * to delay it due to rcu usage.
*/
+ synchronize_rcu();
INIT_WORK(&rdev->del_work, md_delayed_delete);
kobject_get(&rdev->kobj);
schedule_work(&rdev->del_work);
if (err) {
printk(KERN_ERR "md: could not bd_claim %s.\n",
bdevname(bdev, b));
- blkdev_put(bdev);
+ blkdev_put(bdev, FMODE_READ|FMODE_WRITE);
return err;
}
if (!shared)
if (!bdev)
MD_BUG();
bd_release(bdev);
- blkdev_put(bdev);
+ blkdev_put(bdev, FMODE_READ|FMODE_WRITE);
}
void md_autodetect_dev(dev_t dev);
if (rdev->mddev)
MD_BUG();
free_disk_sb(rdev);
- list_del_init(&rdev->same_set);
#ifndef MODULE
if (test_bit(AutoDetected, &rdev->flags))
md_autodetect_dev(rdev->bdev->bd_dev);
static void export_array(mddev_t *mddev)
{
- struct list_head *tmp;
- mdk_rdev_t *rdev;
+ mdk_rdev_t *rdev, *tmp;
rdev_for_each(rdev, tmp, mddev) {
if (!rdev->mddev) {
desc->major,desc->minor,desc->raid_disk,desc->state);
}
-static void print_sb(mdp_super_t *sb)
+static void print_sb_90(mdp_super_t *sb)
{
int i;
}
printk(KERN_INFO "md: THIS: ");
print_desc(&sb->this_disk);
-
}
-static void print_rdev(mdk_rdev_t *rdev)
+static void print_sb_1(struct mdp_superblock_1 *sb)
+{
+ __u8 *uuid;
+
+ uuid = sb->set_uuid;
+ printk(KERN_INFO "md: SB: (V:%u) (F:0x%08x) Array-ID:<%02x%02x%02x%02x"
+ ":%02x%02x:%02x%02x:%02x%02x:%02x%02x%02x%02x%02x%02x>\n"
+ KERN_INFO "md: Name: \"%s\" CT:%llu\n",
+ le32_to_cpu(sb->major_version),
+ le32_to_cpu(sb->feature_map),
+ uuid[0], uuid[1], uuid[2], uuid[3],
+ uuid[4], uuid[5], uuid[6], uuid[7],
+ uuid[8], uuid[9], uuid[10], uuid[11],
+ uuid[12], uuid[13], uuid[14], uuid[15],
+ sb->set_name,
+ (unsigned long long)le64_to_cpu(sb->ctime)
+ & MD_SUPERBLOCK_1_TIME_SEC_MASK);
+
+ uuid = sb->device_uuid;
+ printk(KERN_INFO "md: L%u SZ%llu RD:%u LO:%u CS:%u DO:%llu DS:%llu SO:%llu"
+ " RO:%llu\n"
+ KERN_INFO "md: Dev:%08x UUID: %02x%02x%02x%02x:%02x%02x:%02x%02x:%02x%02x"
+ ":%02x%02x%02x%02x%02x%02x\n"
+ KERN_INFO "md: (F:0x%08x) UT:%llu Events:%llu ResyncOffset:%llu CSUM:0x%08x\n"
+ KERN_INFO "md: (MaxDev:%u) \n",
+ le32_to_cpu(sb->level),
+ (unsigned long long)le64_to_cpu(sb->size),
+ le32_to_cpu(sb->raid_disks),
+ le32_to_cpu(sb->layout),
+ le32_to_cpu(sb->chunksize),
+ (unsigned long long)le64_to_cpu(sb->data_offset),
+ (unsigned long long)le64_to_cpu(sb->data_size),
+ (unsigned long long)le64_to_cpu(sb->super_offset),
+ (unsigned long long)le64_to_cpu(sb->recovery_offset),
+ le32_to_cpu(sb->dev_number),
+ uuid[0], uuid[1], uuid[2], uuid[3],
+ uuid[4], uuid[5], uuid[6], uuid[7],
+ uuid[8], uuid[9], uuid[10], uuid[11],
+ uuid[12], uuid[13], uuid[14], uuid[15],
+ sb->devflags,
+ (unsigned long long)le64_to_cpu(sb->utime) & MD_SUPERBLOCK_1_TIME_SEC_MASK,
+ (unsigned long long)le64_to_cpu(sb->events),
+ (unsigned long long)le64_to_cpu(sb->resync_offset),
+ le32_to_cpu(sb->sb_csum),
+ le32_to_cpu(sb->max_dev)
+ );
+}
+
+static void print_rdev(mdk_rdev_t *rdev, int major_version)
{
char b[BDEVNAME_SIZE];
- printk(KERN_INFO "md: rdev %s, SZ:%08llu F:%d S:%d DN:%u\n",
- bdevname(rdev->bdev,b), (unsigned long long)rdev->size,
+ printk(KERN_INFO "md: rdev %s, Sect:%08llu F:%d S:%d DN:%u\n",
+ bdevname(rdev->bdev, b), (unsigned long long)rdev->sectors,
test_bit(Faulty, &rdev->flags), test_bit(In_sync, &rdev->flags),
rdev->desc_nr);
if (rdev->sb_loaded) {
- printk(KERN_INFO "md: rdev superblock:\n");
- print_sb((mdp_super_t*)page_address(rdev->sb_page));
+ printk(KERN_INFO "md: rdev superblock (MJ:%d):\n", major_version);
+ switch (major_version) {
+ case 0:
+ print_sb_90((mdp_super_t*)page_address(rdev->sb_page));
+ break;
+ case 1:
+ print_sb_1((struct mdp_superblock_1 *)page_address(rdev->sb_page));
+ break;
+ }
} else
printk(KERN_INFO "md: no rdev superblock!\n");
}
static void md_print_devices(void)
{
- struct list_head *tmp, *tmp2;
+ struct list_head *tmp;
mdk_rdev_t *rdev;
mddev_t *mddev;
char b[BDEVNAME_SIZE];
bitmap_print_sb(mddev->bitmap);
else
printk("%s: ", mdname(mddev));
- rdev_for_each(rdev, tmp2, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
printk("<%s>", bdevname(rdev->bdev,b));
printk("\n");
- rdev_for_each(rdev, tmp2, mddev)
- print_rdev(rdev);
+ list_for_each_entry(rdev, &mddev->disks, same_set)
+ print_rdev(rdev, mddev->major_version);
}
printk("md: **********************************\n");
printk("\n");
* with the rest of the array)
*/
mdk_rdev_t *rdev;
- struct list_head *tmp;
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
if (rdev->sb_events == mddev->events ||
(nospares &&
rdev->raid_disk < 0 &&
static void md_update_sb(mddev_t * mddev, int force_change)
{
- struct list_head *tmp;
mdk_rdev_t *rdev;
int sync_req;
int nospares = 0;
mdname(mddev),mddev->in_sync);
bitmap_update_sb(mddev->bitmap);
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
char b[BDEVNAME_SIZE];
dprintk(KERN_INFO "md: ");
if (rdev->sb_loaded != 1)
dprintk("%s ", bdevname(rdev->bdev,b));
if (!test_bit(Faulty, &rdev->flags)) {
md_super_write(mddev,rdev,
- rdev->sb_offset<<1, rdev->sb_size,
+ rdev->sb_start, rdev->sb_size,
rdev->sb_page);
dprintk(KERN_INFO "(write) %s's sb offset: %llu\n",
bdevname(rdev->bdev,b),
- (unsigned long long)rdev->sb_offset);
+ (unsigned long long)rdev->sb_start);
rdev->sb_events = mddev->events;
} else
}
-/* words written to sysfs files may, or my not, be \n terminated.
+/* words written to sysfs files may, or may not, be \n terminated.
* We want to accept with case. For this we use cmd_match.
*/
static int cmd_match(const char *cmd, const char *str)
err = 0;
}
+ if (!err && rdev->sysfs_state)
+ sysfs_notify_dirent(rdev->sysfs_state);
return err ? err : len;
}
static struct rdev_sysfs_entry rdev_state =
slot = -1;
else if (e==buf || (*e && *e!= '\n'))
return -EINVAL;
- if (rdev->mddev->pers) {
+ if (rdev->mddev->pers && slot == -1) {
/* Setting 'slot' on an active array requires also
* updating the 'rd%d' link, and communicating
* with the personality with ->hot_*_disk.
* failed/spare devices. This normally happens automatically,
* but not when the metadata is externally managed.
*/
- if (slot != -1)
- return -EBUSY;
if (rdev->raid_disk == -1)
return -EEXIST;
/* personality does all needed checks */
sysfs_remove_link(&rdev->mddev->kobj, nm);
set_bit(MD_RECOVERY_NEEDED, &rdev->mddev->recovery);
md_wakeup_thread(rdev->mddev->thread);
+ } else if (rdev->mddev->pers) {
+ mdk_rdev_t *rdev2;
+ /* Activating a spare .. or possibly reactivating
+ * if we every get bitmaps working here.
+ */
+
+ if (rdev->raid_disk != -1)
+ return -EBUSY;
+
+ if (rdev->mddev->pers->hot_add_disk == NULL)
+ return -EINVAL;
+
+ list_for_each_entry(rdev2, &rdev->mddev->disks, same_set)
+ if (rdev2->raid_disk == slot)
+ return -EEXIST;
+
+ rdev->raid_disk = slot;
+ if (test_bit(In_sync, &rdev->flags))
+ rdev->saved_raid_disk = slot;
+ else
+ rdev->saved_raid_disk = -1;
+ err = rdev->mddev->pers->
+ hot_add_disk(rdev->mddev, rdev);
+ if (err) {
+ rdev->raid_disk = -1;
+ return err;
+ } else
+ sysfs_notify_dirent(rdev->sysfs_state);
+ sprintf(nm, "rd%d", rdev->raid_disk);
+ if (sysfs_create_link(&rdev->mddev->kobj, &rdev->kobj, nm))
+ printk(KERN_WARNING
+ "md: cannot register "
+ "%s for %s\n",
+ nm, mdname(rdev->mddev));
+
+ /* don't wakeup anyone, leave that to userspace. */
} else {
if (slot >= rdev->mddev->raid_disks)
return -ENOSPC;
clear_bit(Faulty, &rdev->flags);
clear_bit(WriteMostly, &rdev->flags);
set_bit(In_sync, &rdev->flags);
+ sysfs_notify_dirent(rdev->sysfs_state);
}
return len;
}
unsigned long long offset = simple_strtoull(buf, &e, 10);
if (e==buf || (*e && *e != '\n'))
return -EINVAL;
- if (rdev->mddev->pers)
+ if (rdev->mddev->pers && rdev->raid_disk >= 0)
return -EBUSY;
- if (rdev->size && rdev->mddev->external)
+ if (rdev->sectors && rdev->mddev->external)
/* Must set offset before size, so overlap checks
* can be sane */
return -EBUSY;
static ssize_t
rdev_size_show(mdk_rdev_t *rdev, char *page)
{
- return sprintf(page, "%llu\n", (unsigned long long)rdev->size);
+ return sprintf(page, "%llu\n", (unsigned long long)rdev->sectors / 2);
}
static int overlaps(sector_t s1, sector_t l1, sector_t s2, sector_t l2)
return 1;
}
+static int strict_blocks_to_sectors(const char *buf, sector_t *sectors)
+{
+ unsigned long long blocks;
+ sector_t new;
+
+ if (strict_strtoull(buf, 10, &blocks) < 0)
+ return -EINVAL;
+
+ if (blocks & 1ULL << (8 * sizeof(blocks) - 1))
+ return -EINVAL; /* sector conversion overflow */
+
+ new = blocks * 2;
+ if (new != blocks * 2)
+ return -EINVAL; /* unsigned long long to sector_t overflow */
+
+ *sectors = new;
+ return 0;
+}
+
static ssize_t
rdev_size_store(mdk_rdev_t *rdev, const char *buf, size_t len)
{
- char *e;
- unsigned long long size = simple_strtoull(buf, &e, 10);
- unsigned long long oldsize = rdev->size;
mddev_t *my_mddev = rdev->mddev;
+ sector_t oldsectors = rdev->sectors;
+ sector_t sectors;
- if (e==buf || (*e && *e != '\n'))
+ if (strict_blocks_to_sectors(buf, §ors) < 0)
return -EINVAL;
- if (my_mddev->pers)
- return -EBUSY;
- rdev->size = size;
- if (size > oldsize && rdev->mddev->external) {
+ if (my_mddev->pers && rdev->raid_disk >= 0) {
+ if (my_mddev->persistent) {
+ sectors = super_types[my_mddev->major_version].
+ rdev_size_change(rdev, sectors);
+ if (!sectors)
+ return -EBUSY;
+ } else if (!sectors)
+ sectors = (rdev->bdev->bd_inode->i_size >> 9) -
+ rdev->data_offset;
+ }
+ if (sectors < my_mddev->dev_sectors)
+ return -EINVAL; /* component must fit device */
+
+ rdev->sectors = sectors;
+ if (sectors > oldsectors && my_mddev->external) {
/* need to check that all other rdevs with the same ->bdev
* do not overlap. We need to unlock the mddev to avoid
- * a deadlock. We have already changed rdev->size, and if
+ * a deadlock. We have already changed rdev->sectors, and if
* we have to change it back, we will have the lock again.
*/
mddev_t *mddev;
int overlap = 0;
- struct list_head *tmp, *tmp2;
+ struct list_head *tmp;
mddev_unlock(my_mddev);
for_each_mddev(mddev, tmp) {
mdk_rdev_t *rdev2;
mddev_lock(mddev);
- rdev_for_each(rdev2, tmp2, mddev)
+ list_for_each_entry(rdev2, &mddev->disks, same_set)
if (test_bit(AllReserved, &rdev2->flags) ||
(rdev->bdev == rdev2->bdev &&
rdev != rdev2 &&
- overlaps(rdev->data_offset, rdev->size,
- rdev2->data_offset, rdev2->size))) {
+ overlaps(rdev->data_offset, rdev->sectors,
+ rdev2->data_offset,
+ rdev2->sectors))) {
overlap = 1;
break;
}
if (overlap) {
/* Someone else could have slipped in a size
* change here, but doing so is just silly.
- * We put oldsize back because we *know* it is
+ * We put oldsectors back because we *know* it is
* safe, and trust userspace not to race with
* itself
*/
- rdev->size = oldsize;
+ rdev->sectors = oldsectors;
return -EBUSY;
}
}
- if (size < my_mddev->size || my_mddev->size == 0)
- my_mddev->size = size;
return len;
}
static void analyze_sbs(mddev_t * mddev)
{
int i;
- struct list_head *tmp;
- mdk_rdev_t *rdev, *freshest;
+ mdk_rdev_t *rdev, *freshest, *tmp;
char b[BDEVNAME_SIZE];
freshest = NULL;
i = 0;
rdev_for_each(rdev, tmp, mddev) {
+ if (rdev->desc_nr >= mddev->max_disks ||
+ i > mddev->max_disks) {
+ printk(KERN_WARNING
+ "md: %s: %s: only %d devices permitted\n",
+ mdname(mddev), bdevname(rdev->bdev, b),
+ mddev->max_disks);
+ kick_rdev_from_array(rdev);
+ continue;
+ }
if (rdev != freshest)
if (super_types[mddev->major_version].
validate_super(mddev, rdev)) {
}
+static void md_safemode_timeout(unsigned long data);
+
static ssize_t
safe_delay_show(mddev_t *mddev, char *page)
{
int i;
unsigned long msec;
char buf[30];
- char *e;
+
/* remove a period, and count digits after it */
if (len >= sizeof(buf))
return -EINVAL;
- strlcpy(buf, cbuf, len);
- buf[len] = 0;
+ strlcpy(buf, cbuf, sizeof(buf));
for (i=0; i<len; i++) {
if (dot) {
if (isdigit(buf[i])) {
buf[i] = 0;
}
}
- msec = simple_strtoul(buf, &e, 10);
- if (e == buf || (*e && *e != '\n'))
+ if (strict_strtoul(buf, 10, &msec) < 0)
return -EINVAL;
msec = (msec * 1000) / scale;
if (msec == 0)
mddev->safemode_delay = 0;
else {
+ unsigned long old_delay = mddev->safemode_delay;
mddev->safemode_delay = (msec*HZ)/1000;
if (mddev->safemode_delay == 0)
mddev->safemode_delay = 1;
+ if (mddev->safemode_delay < old_delay)
+ md_safemode_timeout((unsigned long)mddev);
}
return len;
}
static ssize_t
level_store(mddev_t *mddev, const char *buf, size_t len)
{
+ char level[16];
ssize_t rv = len;
- if (mddev->pers)
+ struct mdk_personality *pers;
+ void *priv;
+
+ if (mddev->pers == NULL) {
+ if (len == 0)
+ return 0;
+ if (len >= sizeof(mddev->clevel))
+ return -ENOSPC;
+ strncpy(mddev->clevel, buf, len);
+ if (mddev->clevel[len-1] == '\n')
+ len--;
+ mddev->clevel[len] = 0;
+ mddev->level = LEVEL_NONE;
+ return rv;
+ }
+
+ /* request to change the personality. Need to ensure:
+ * - array is not engaged in resync/recovery/reshape
+ * - old personality can be suspended
+ * - new personality will access other array.
+ */
+
+ if (mddev->sync_thread || mddev->reshape_position != MaxSector)
return -EBUSY;
- if (len == 0)
- return 0;
- if (len >= sizeof(mddev->clevel))
- return -ENOSPC;
- strncpy(mddev->clevel, buf, len);
- if (mddev->clevel[len-1] == '\n')
+
+ if (!mddev->pers->quiesce) {
+ printk(KERN_WARNING "md: %s: %s does not support online personality change\n",
+ mdname(mddev), mddev->pers->name);
+ return -EINVAL;
+ }
+
+ /* Now find the new personality */
+ if (len == 0 || len >= sizeof(level))
+ return -EINVAL;
+ strncpy(level, buf, len);
+ if (level[len-1] == '\n')
len--;
- mddev->clevel[len] = 0;
- mddev->level = LEVEL_NONE;
+ level[len] = 0;
+
+ request_module("md-%s", level);
+ spin_lock(&pers_lock);
+ pers = find_pers(LEVEL_NONE, level);
+ if (!pers || !try_module_get(pers->owner)) {
+ spin_unlock(&pers_lock);
+ printk(KERN_WARNING "md: personality %s not loaded\n", level);
+ return -EINVAL;
+ }
+ spin_unlock(&pers_lock);
+
+ if (pers == mddev->pers) {
+ /* Nothing to do! */
+ module_put(pers->owner);
+ return rv;
+ }
+ if (!pers->takeover) {
+ module_put(pers->owner);
+ printk(KERN_WARNING "md: %s: %s does not support personality takeover\n",
+ mdname(mddev), level);
+ return -EINVAL;
+ }
+
+ /* ->takeover must set new_* and/or delta_disks
+ * if it succeeds, and may set them when it fails.
+ */
+ priv = pers->takeover(mddev);
+ if (IS_ERR(priv)) {
+ mddev->new_level = mddev->level;
+ mddev->new_layout = mddev->layout;
+ mddev->new_chunk = mddev->chunk_size;
+ mddev->raid_disks -= mddev->delta_disks;
+ mddev->delta_disks = 0;
+ module_put(pers->owner);
+ printk(KERN_WARNING "md: %s: %s would not accept array\n",
+ mdname(mddev), level);
+ return PTR_ERR(priv);
+ }
+
+ /* Looks like we have a winner */
+ mddev_suspend(mddev);
+ mddev->pers->stop(mddev);
+ module_put(mddev->pers->owner);
+ mddev->pers = pers;
+ mddev->private = priv;
+ strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
+ mddev->level = mddev->new_level;
+ mddev->layout = mddev->new_layout;
+ mddev->chunk_size = mddev->new_chunk;
+ mddev->delta_disks = 0;
+ pers->run(mddev);
+ mddev_resume(mddev);
+ set_bit(MD_CHANGE_DEVS, &mddev->flags);
+ set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+ md_wakeup_thread(mddev->thread);
return rv;
}
if (!*buf || (*e && *e != '\n'))
return -EINVAL;
- if (mddev->pers)
- return -EBUSY;
- if (mddev->reshape_position != MaxSector)
+ if (mddev->pers) {
+ int err;
+ if (mddev->pers->reconfig == NULL)
+ return -EBUSY;
+ err = mddev->pers->reconfig(mddev, n, -1);
+ if (err)
+ return err;
+ } else {
mddev->new_layout = n;
- else
- mddev->layout = n;
+ if (mddev->reshape_position == MaxSector)
+ mddev->layout = n;
+ }
return len;
}
static struct md_sysfs_entry md_layout =
static ssize_t
chunk_size_store(mddev_t *mddev, const char *buf, size_t len)
{
- /* can only set chunk_size if array is not yet active */
char *e;
unsigned long n = simple_strtoul(buf, &e, 10);
if (!*buf || (*e && *e != '\n'))
return -EINVAL;
- if (mddev->pers)
- return -EBUSY;
- else if (mddev->reshape_position != MaxSector)
+ if (mddev->pers) {
+ int err;
+ if (mddev->pers->reconfig == NULL)
+ return -EBUSY;
+ err = mddev->pers->reconfig(mddev, -1, n);
+ if (err)
+ return err;
+ } else {
mddev->new_chunk = n;
- else
- mddev->chunk_size = n;
+ if (mddev->reshape_position == MaxSector)
+ mddev->chunk_size = n;
+ }
return len;
}
static struct md_sysfs_entry md_chunk_size =
static ssize_t
resync_start_show(mddev_t *mddev, char *page)
{
+ if (mddev->recovery_cp == MaxSector)
+ return sprintf(page, "none\n");
return sprintf(page, "%llu\n", (unsigned long long)mddev->recovery_cp);
}
* When written, doesn't tear down array, but just stops it
* suspended (not supported yet)
* All IO requests will block. The array can be reconfigured.
- * Writing this, if accepted, will block until array is quiessent
+ * Writing this, if accepted, will block until array is quiescent
* readonly
* no resync can happen. no superblocks get written.
* write requests fail
else {
if (list_empty(&mddev->disks) &&
mddev->raid_disks == 0 &&
- mddev->size == 0)
+ mddev->dev_sectors == 0)
st = clear;
else
st = inactive;
return sprintf(page, "%s\n", array_states[st]);
}
-static int do_md_stop(mddev_t * mddev, int ro);
+static int do_md_stop(mddev_t * mddev, int ro, int is_open);
static int do_md_run(mddev_t * mddev);
static int restart_array(mddev_t *mddev);
break;
case clear:
/* stopping an active array */
- if (atomic_read(&mddev->active) > 1)
+ if (atomic_read(&mddev->openers) > 0)
return -EBUSY;
- err = do_md_stop(mddev, 0);
+ err = do_md_stop(mddev, 0, 0);
break;
case inactive:
/* stopping an active array */
if (mddev->pers) {
- if (atomic_read(&mddev->active) > 1)
+ if (atomic_read(&mddev->openers) > 0)
return -EBUSY;
- err = do_md_stop(mddev, 2);
+ err = do_md_stop(mddev, 2, 0);
} else
err = 0; /* already inactive */
break;
break; /* not supported yet */
case readonly:
if (mddev->pers)
- err = do_md_stop(mddev, 1);
+ err = do_md_stop(mddev, 1, 0);
else {
mddev->ro = 1;
set_disk_ro(mddev->gendisk, 1);
break;
case read_auto:
if (mddev->pers) {
- if (mddev->ro != 1)
- err = do_md_stop(mddev, 1);
- else
+ if (mddev->ro == 0)
+ err = do_md_stop(mddev, 1, 0);
+ else if (mddev->ro == 1)
err = restart_array(mddev);
if (err == 0) {
mddev->ro = 2;
}
if (err)
return err;
- else
+ else {
+ sysfs_notify_dirent(mddev->sysfs_state);
return len;
+ }
}
static struct md_sysfs_entry md_array_state =
__ATTR(array_state, S_IRUGO|S_IWUSR, array_state_show, array_state_store);
static ssize_t
size_show(mddev_t *mddev, char *page)
{
- return sprintf(page, "%llu\n", (unsigned long long)mddev->size);
+ return sprintf(page, "%llu\n",
+ (unsigned long long)mddev->dev_sectors / 2);
}
-static int update_size(mddev_t *mddev, unsigned long size);
+static int update_size(mddev_t *mddev, sector_t num_sectors);
static ssize_t
size_store(mddev_t *mddev, const char *buf, size_t len)
* not increase it (except from 0).
* If array is active, we can try an on-line resize
*/
- char *e;
- int err = 0;
- unsigned long long size = simple_strtoull(buf, &e, 10);
- if (!*buf || *buf == '\n' ||
- (*e && *e != '\n'))
- return -EINVAL;
+ sector_t sectors;
+ int err = strict_blocks_to_sectors(buf, §ors);
+ if (err < 0)
+ return err;
if (mddev->pers) {
- err = update_size(mddev, size);
+ err = update_size(mddev, sectors);
md_update_sb(mddev, 1);
} else {
- if (mddev->size == 0 ||
- mddev->size > size)
- mddev->size = size;
+ if (mddev->dev_sectors == 0 ||
+ mddev->dev_sectors > sectors)
+ mddev->dev_sectors = sectors;
else
err = -ENOSPC;
}
{
int major, minor;
char *e;
- if (!list_empty(&mddev->disks))
+ /* Changing the details of 'external' metadata is
+ * always permitted. Otherwise there must be
+ * no devices attached to the array.
+ */
+ if (mddev->external && strncmp(buf, "external:", 9) == 0)
+ ;
+ else if (!list_empty(&mddev->disks))
return -EBUSY;
if (cmd_match(buf, "none")) {
type = "check";
else
type = "repair";
- } else
+ } else if (test_bit(MD_RECOVERY_RECOVER, &mddev->recovery))
type = "recover";
}
return sprintf(page, "%s\n", type);
} else if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
test_bit(MD_RECOVERY_NEEDED, &mddev->recovery))
return -EBUSY;
- else if (cmd_match(page, "resync") || cmd_match(page, "recover"))
+ else if (cmd_match(page, "resync"))
+ set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+ else if (cmd_match(page, "recover")) {
+ set_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
- else if (cmd_match(page, "reshape")) {
+ } else if (cmd_match(page, "reshape")) {
int err;
if (mddev->pers->start_reshape == NULL)
return -EINVAL;
err = mddev->pers->start_reshape(mddev);
if (err)
return err;
+ sysfs_notify(&mddev->kobj, NULL, "degraded");
} else {
if (cmd_match(page, "check"))
set_bit(MD_RECOVERY_CHECK, &mddev->recovery);
}
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
+ sysfs_notify_dirent(mddev->sysfs_action);
return len;
}
sync_speed_show(mddev_t *mddev, char *page)
{
unsigned long resync, dt, db;
- resync = (mddev->curr_mark_cnt - atomic_read(&mddev->recovery_active));
- dt = ((jiffies - mddev->resync_mark) / HZ);
+ if (mddev->curr_resync == 0)
+ return sprintf(page, "none\n");
+ resync = mddev->curr_mark_cnt - atomic_read(&mddev->recovery_active);
+ dt = (jiffies - mddev->resync_mark) / HZ;
if (!dt) dt++;
- db = resync - (mddev->resync_mark_cnt);
- return sprintf(page, "%ld\n", db/dt/2); /* K/sec */
+ db = resync - mddev->resync_mark_cnt;
+ return sprintf(page, "%lu\n", db/dt/2); /* K/sec */
}
static struct md_sysfs_entry md_sync_speed = __ATTR_RO(sync_speed);
static ssize_t
sync_completed_show(mddev_t *mddev, char *page)
{
- unsigned long max_blocks, resync;
+ unsigned long max_sectors, resync;
if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
- max_blocks = mddev->resync_max_sectors;
+ max_sectors = mddev->resync_max_sectors;
else
- max_blocks = mddev->size << 1;
+ max_sectors = mddev->dev_sectors;
resync = (mddev->curr_resync - atomic_read(&mddev->recovery_active));
- return sprintf(page, "%lu / %lu\n", resync, max_blocks);
+ return sprintf(page, "%lu / %lu\n", resync, max_sectors);
}
static struct md_sysfs_entry md_sync_completed = __ATTR_RO(sync_completed);
__ATTR(reshape_position, S_IRUGO|S_IWUSR, reshape_position_show,
reshape_position_store);
+static ssize_t
+array_size_show(mddev_t *mddev, char *page)
+{
+ if (mddev->external_size)
+ return sprintf(page, "%llu\n",
+ (unsigned long long)mddev->array_sectors/2);
+ else
+ return sprintf(page, "default\n");
+}
+
+static ssize_t
+array_size_store(mddev_t *mddev, const char *buf, size_t len)
+{
+ sector_t sectors;
+
+ if (strncmp(buf, "default", 7) == 0) {
+ if (mddev->pers)
+ sectors = mddev->pers->size(mddev, 0, 0);
+ else
+ sectors = mddev->array_sectors;
+
+ mddev->external_size = 0;
+ } else {
+ if (strict_blocks_to_sectors(buf, §ors) < 0)
+ return -EINVAL;
+ if (mddev->pers && mddev->pers->size(mddev, 0, 0) < sectors)
+ return -EINVAL;
+
+ mddev->external_size = 1;
+ }
+
+ mddev->array_sectors = sectors;
+ set_capacity(mddev->gendisk, mddev->array_sectors);
+ if (mddev->pers) {
+ struct block_device *bdev = bdget_disk(mddev->gendisk, 0);
+
+ if (bdev) {
+ mutex_lock(&bdev->bd_inode->i_mutex);
+ i_size_write(bdev->bd_inode,
+ (loff_t)mddev->array_sectors << 9);
+ mutex_unlock(&bdev->bd_inode->i_mutex);
+ bdput(bdev);
+ }
+ }
+
+ return len;
+}
+
+static struct md_sysfs_entry md_array_size =
+__ATTR(array_size, S_IRUGO|S_IWUSR, array_size_show,
+ array_size_store);
static struct attribute *md_default_attrs[] = {
&md_level.attr,
&md_safe_delay.attr,
&md_array_state.attr,
&md_reshape_position.attr,
+ &md_array_size.attr,
NULL,
};
if (!capable(CAP_SYS_ADMIN))
return -EACCES;
rv = mddev_lock(mddev);
+ if (mddev->hold_active == UNTIL_IOCTL)
+ mddev->hold_active = 0;
if (!rv) {
rv = entry->store(mddev, page, length);
mddev_unlock(mddev);
static void md_free(struct kobject *ko)
{
mddev_t *mddev = container_of(ko, mddev_t, kobj);
+
+ if (mddev->sysfs_state)
+ sysfs_put(mddev->sysfs_state);
+
+ if (mddev->gendisk) {
+ del_gendisk(mddev->gendisk);
+ put_disk(mddev->gendisk);
+ }
+ if (mddev->queue)
+ blk_cleanup_queue(mddev->queue);
+
kfree(mddev);
}
int mdp_major = 0;
-static struct kobject *md_probe(dev_t dev, int *part, void *data)
+static int md_alloc(dev_t dev, char *name)
{
static DEFINE_MUTEX(disks_mutex);
mddev_t *mddev = mddev_find(dev);
struct gendisk *disk;
- int partitioned = (MAJOR(dev) != MD_MAJOR);
- int shift = partitioned ? MdpMinorShift : 0;
- int unit = MINOR(dev) >> shift;
+ int partitioned;
+ int shift;
+ int unit;
int error;
if (!mddev)
- return NULL;
+ return -ENODEV;
+
+ partitioned = (MAJOR(mddev->unit) != MD_MAJOR);
+ shift = partitioned ? MdpMinorShift : 0;
+ unit = MINOR(mddev->unit) >> shift;
+
+ /* wait for any previous instance if this device
+ * to be completed removed (mddev_delayed_delete).
+ */
+ flush_scheduled_work();
mutex_lock(&disks_mutex);
if (mddev->gendisk) {
mutex_unlock(&disks_mutex);
mddev_put(mddev);
- return NULL;
+ return -EEXIST;
+ }
+
+ if (name) {
+ /* Need to ensure that 'name' is not a duplicate.
+ */
+ mddev_t *mddev2;
+ spin_lock(&all_mddevs_lock);
+
+ list_for_each_entry(mddev2, &all_mddevs, all_mddevs)
+ if (mddev2->gendisk &&
+ strcmp(mddev2->gendisk->disk_name, name) == 0) {
+ spin_unlock(&all_mddevs_lock);
+ return -EEXIST;
+ }
+ spin_unlock(&all_mddevs_lock);
+ }
+
+ mddev->queue = blk_alloc_queue(GFP_KERNEL);
+ if (!mddev->queue) {
+ mutex_unlock(&disks_mutex);
+ mddev_put(mddev);
+ return -ENOMEM;
}
+ mddev->queue->queuedata = mddev;
+
+ /* Can be unlocked because the queue is new: no concurrency */
+ queue_flag_set_unlocked(QUEUE_FLAG_CLUSTER, mddev->queue);
+
+ blk_queue_make_request(mddev->queue, md_make_request);
+
disk = alloc_disk(1 << shift);
if (!disk) {
mutex_unlock(&disks_mutex);
+ blk_cleanup_queue(mddev->queue);
+ mddev->queue = NULL;
mddev_put(mddev);
- return NULL;
+ return -ENOMEM;
}
- disk->major = MAJOR(dev);
+ disk->major = MAJOR(mddev->unit);
disk->first_minor = unit << shift;
- if (partitioned)
+ if (name)
+ strcpy(disk->disk_name, name);
+ else if (partitioned)
sprintf(disk->disk_name, "md_d%d", unit);
else
sprintf(disk->disk_name, "md%d", unit);
disk->fops = &md_fops;
disk->private_data = mddev;
disk->queue = mddev->queue;
+ /* Allow extended partitions. This makes the
+ * 'mdp' device redundant, but we can't really
+ * remove it now.
+ */
+ disk->flags |= GENHD_FL_EXT_DEVT;
add_disk(disk);
mddev->gendisk = disk;
+ error = kobject_init_and_add(&mddev->kobj, &md_ktype,
+ &disk_to_dev(disk)->kobj, "%s", "md");
mutex_unlock(&disks_mutex);
- error = kobject_init_and_add(&mddev->kobj, &md_ktype, &disk->dev.kobj,
- "%s", "md");
if (error)
printk(KERN_WARNING "md: cannot register %s/md - name in use\n",
disk->disk_name);
- else
+ else {
kobject_uevent(&mddev->kobj, KOBJ_ADD);
+ mddev->sysfs_state = sysfs_get_dirent(mddev->kobj.sd, "array_state");
+ }
+ mddev_put(mddev);
+ return 0;
+}
+
+static struct kobject *md_probe(dev_t dev, int *part, void *data)
+{
+ md_alloc(dev, NULL);
return NULL;
}
+static int add_named_array(const char *val, struct kernel_param *kp)
+{
+ /* val must be "md_*" where * is not all digits.
+ * We allocate an array with a large free minor number, and
+ * set the name to val. val must not already be an active name.
+ */
+ int len = strlen(val);
+ char buf[DISK_NAME_LEN];
+
+ while (len && val[len-1] == '\n')
+ len--;
+ if (len >= DISK_NAME_LEN)
+ return -E2BIG;
+ strlcpy(buf, val, len+1);
+ if (strncmp(buf, "md_", 3) != 0)
+ return -EINVAL;
+ return md_alloc(0, buf);
+}
+
static void md_safemode_timeout(unsigned long data)
{
mddev_t *mddev = (mddev_t *) data;
- mddev->safemode = 1;
+ if (!atomic_read(&mddev->writes_pending)) {
+ mddev->safemode = 1;
+ if (mddev->external)
+ sysfs_notify_dirent(mddev->sysfs_state);
+ }
md_wakeup_thread(mddev->thread);
}
{
int err;
int chunk_size;
- struct list_head *tmp;
mdk_rdev_t *rdev;
struct gendisk *disk;
struct mdk_personality *pers;
return -EINVAL;
}
/*
- * chunk-size has to be a power of 2 and multiples of PAGE_SIZE
+ * chunk-size has to be a power of 2
*/
if ( (1 << ffz(~chunk_size)) != chunk_size) {
printk(KERN_ERR "chunk_size of %d not valid\n", chunk_size);
return -EINVAL;
}
- if (chunk_size < PAGE_SIZE) {
- printk(KERN_ERR "too small chunk_size: %d < %ld\n",
- chunk_size, PAGE_SIZE);
- return -EINVAL;
- }
/* devices must have minimum size of one chunk */
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
if (test_bit(Faulty, &rdev->flags))
continue;
- if (rdev->size < chunk_size / 1024) {
+ if (rdev->sectors < chunk_size / 512) {
printk(KERN_WARNING
"md: Dev %s smaller than chunk_size:"
- " %lluk < %dk\n",
+ " %llu < %d\n",
bdevname(rdev->bdev,b),
- (unsigned long long)rdev->size,
- chunk_size / 1024);
+ (unsigned long long)rdev->sectors,
+ chunk_size / 512);
return -EINVAL;
}
}
}
-#ifdef CONFIG_KMOD
if (mddev->level != LEVEL_NONE)
request_module("md-level-%d", mddev->level);
else if (mddev->clevel[0])
request_module("md-%s", mddev->clevel);
-#endif
/*
* Drop all container device buffers, from now on
* the only valid external interface is through the md
* device.
*/
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
if (test_bit(Faulty, &rdev->flags))
continue;
sync_blockdev(rdev->bdev);
/* perform some consistency tests on the device.
* We don't want the data to overlap the metadata,
- * Internal Bitmap issues has handled elsewhere.
+ * Internal Bitmap issues have been handled elsewhere.
*/
- if (rdev->data_offset < rdev->sb_offset) {
- if (mddev->size &&
- rdev->data_offset + mddev->size*2
- > rdev->sb_offset*2) {
+ if (rdev->data_offset < rdev->sb_start) {
+ if (mddev->dev_sectors &&
+ rdev->data_offset + mddev->dev_sectors
+ > rdev->sb_start) {
printk("md: %s: data overlaps metadata\n",
mdname(mddev));
return -EINVAL;
}
} else {
- if (rdev->sb_offset*2 + rdev->sb_size/512
+ if (rdev->sb_start + rdev->sb_size/512
> rdev->data_offset) {
printk("md: %s: metadata overlaps data\n",
mdname(mddev));
return -EINVAL;
}
}
+ sysfs_notify_dirent(rdev->sysfs_state);
}
md_probe(mddev->unit, NULL, NULL);
}
mddev->pers = pers;
spin_unlock(&pers_lock);
- mddev->level = pers->level;
+ if (mddev->level != pers->level) {
+ mddev->level = pers->level;
+ mddev->new_level = pers->level;
+ }
strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
+ if (pers->level >= 4 && pers->level <= 6)
+ /* Cannot support integrity (yet) */
+ blk_integrity_unregister(mddev->gendisk);
+
if (mddev->reshape_position != MaxSector &&
pers->start_reshape == NULL) {
/* This personality cannot handle reshaping... */
*/
char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
mdk_rdev_t *rdev2;
- struct list_head *tmp2;
int warned = 0;
- rdev_for_each(rdev, tmp, mddev) {
- rdev_for_each(rdev2, tmp2, mddev) {
+
+ list_for_each_entry(rdev, &mddev->disks, same_set)
+ list_for_each_entry(rdev2, &mddev->disks, same_set) {
if (rdev < rdev2 &&
rdev->bdev->bd_contains ==
rdev2->bdev->bd_contains) {
warned = 1;
}
}
- }
+
if (warned)
printk(KERN_WARNING
"True protection against single-disk"
}
mddev->recovery = 0;
- mddev->resync_max_sectors = mddev->size << 1; /* may be over-ridden by personality */
+ /* may be over-ridden by personality */
+ mddev->resync_max_sectors = mddev->dev_sectors;
+
mddev->barriers_work = 1;
mddev->ok_start_degraded = start_dirty_degraded;
mddev->ro = 2; /* read-only, but switch on first write */
err = mddev->pers->run(mddev);
- if (!err && mddev->pers->sync_request) {
+ if (err)
+ printk(KERN_ERR "md: pers->run() failed ...\n");
+ else if (mddev->pers->size(mddev, 0, 0) < mddev->array_sectors) {
+ WARN_ONCE(!mddev->external_size, "%s: default size too small,"
+ " but 'external_size' not in effect?\n", __func__);
+ printk(KERN_ERR
+ "md: invalid array_size %llu > default size %llu\n",
+ (unsigned long long)mddev->array_sectors / 2,
+ (unsigned long long)mddev->pers->size(mddev, 0, 0) / 2);
+ err = -EINVAL;
+ mddev->pers->stop(mddev);
+ }
+ if (err == 0 && mddev->pers->sync_request) {
err = bitmap_create(mddev);
if (err) {
printk(KERN_ERR "%s: failed to create bitmap (%d)\n",
}
}
if (err) {
- printk(KERN_ERR "md: pers->run() failed ...\n");
module_put(mddev->pers->owner);
mddev->pers = NULL;
bitmap_destroy(mddev);
printk(KERN_WARNING
"md: cannot register extra attributes for %s\n",
mdname(mddev));
+ mddev->sysfs_action = sysfs_get_dirent(mddev->kobj.sd, "sync_action");
} else if (mddev->ro == 2) /* auto-readonly not meaningful */
mddev->ro = 0;
mddev->safemode_delay = (200 * HZ)/1000 +1; /* 200 msec delay */
mddev->in_sync = 1;
- rdev_for_each(rdev, tmp, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->raid_disk >= 0) {
char nm[20];
sprintf(nm, "rd%d", rdev->raid_disk);
if (mddev->flags)
md_update_sb(mddev, 0);
- 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;
+ set_capacity(disk, mddev->array_sectors);
/* If there is a partially-recovered drive we need to
* start recovery here. If we leave it to md_check_recovery,
* it will remove the drives and not do the right thing
*/
if (mddev->degraded && !mddev->sync_thread) {
- struct list_head *rtmp;
int spares = 0;
- rdev_for_each(rdev, rtmp, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->raid_disk >= 0 &&
!test_bit(In_sync, &rdev->flags) &&
!test_bit(Faulty, &rdev->flags))
mddev->changed = 1;
md_new_event(mddev);
- kobject_uevent(&mddev->gendisk->dev.kobj, KOBJ_CHANGE);
+ sysfs_notify_dirent(mddev->sysfs_state);
+ if (mddev->sysfs_action)
+ sysfs_notify_dirent(mddev->sysfs_action);
+ sysfs_notify(&mddev->kobj, NULL, "degraded");
+ kobject_uevent(&disk_to_dev(mddev->gendisk)->kobj, KOBJ_CHANGE);
return 0;
}
static int restart_array(mddev_t *mddev)
{
struct gendisk *disk = mddev->gendisk;
- int err;
- /*
- * Complain if it has no devices
- */
- err = -ENXIO;
+ /* Complain if it has no devices */
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);
- md_wakeup_thread(mddev->sync_thread);
- err = 0;
- } else
- err = -EINVAL;
-
-out:
- return err;
+ return -ENXIO;
+ if (!mddev->pers)
+ return -EINVAL;
+ if (!mddev->ro)
+ return -EBUSY;
+ 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);
+ md_wakeup_thread(mddev->sync_thread);
+ sysfs_notify_dirent(mddev->sysfs_state);
+ return 0;
}
/* similar to deny_write_access, but accounts for our holding a reference
* 1 - switch to readonly
* 2 - stop but do not disassemble array
*/
-static int do_md_stop(mddev_t * mddev, int mode)
+static int do_md_stop(mddev_t * mddev, int mode, int is_open)
{
int err = 0;
struct gendisk *disk = mddev->gendisk;
+ if (atomic_read(&mddev->openers) > is_open) {
+ printk("md: %s still in use.\n",mdname(mddev));
+ return -EBUSY;
+ }
+
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_FROZEN, &mddev->recovery);
del_timer_sync(&mddev->safemode_timer);
- invalidate_partition(disk, 0);
-
switch(mode) {
case 1: /* readonly */
err = -ENXIO;
md_super_wait(mddev);
if (mddev->ro)
set_disk_ro(disk, 0);
- blk_queue_make_request(mddev->queue, md_fail_request);
+
mddev->pers->stop(mddev);
mddev->queue->merge_bvec_fn = NULL;
mddev->queue->unplug_fn = NULL;
mddev->queue->backing_dev_info.congested_fn = NULL;
- if (mddev->pers->sync_request)
+ if (mddev->pers->sync_request) {
sysfs_remove_group(&mddev->kobj, &md_redundancy_group);
-
+ if (mddev->sysfs_action)
+ sysfs_put(mddev->sysfs_action);
+ mddev->sysfs_action = NULL;
+ }
module_put(mddev->pers->owner);
mddev->pers = NULL;
/* tell userspace to handle 'inactive' */
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ sysfs_notify_dirent(mddev->sysfs_state);
set_capacity(disk, 0);
mddev->changed = 1;
*/
if (mode == 0) {
mdk_rdev_t *rdev;
- struct list_head *tmp;
printk(KERN_INFO "md: %s stopped.\n", mdname(mddev));
}
mddev->bitmap_offset = 0;
- rdev_for_each(rdev, tmp, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->raid_disk >= 0) {
char nm[20];
sprintf(nm, "rd%d", rdev->raid_disk);
export_array(mddev);
- mddev->array_size = 0;
- mddev->size = 0;
+ mddev->array_sectors = 0;
+ mddev->external_size = 0;
+ mddev->dev_sectors = 0;
mddev->raid_disks = 0;
mddev->recovery_cp = 0;
mddev->resync_min = 0;
mddev->degraded = 0;
mddev->barriers_work = 0;
mddev->safemode = 0;
+ kobject_uevent(&disk_to_dev(mddev->gendisk)->kobj, KOBJ_CHANGE);
+ if (mddev->hold_active == UNTIL_STOP)
+ mddev->hold_active = 0;
} else if (mddev->pers)
printk(KERN_INFO "md: %s switched to read-only mode.\n",
mdname(mddev));
err = 0;
+ blk_integrity_unregister(disk);
md_new_event(mddev);
+ sysfs_notify_dirent(mddev->sysfs_state);
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))
printk(KERN_INFO "md: running: ");
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
char b[BDEVNAME_SIZE];
printk("<%s>", bdevname(rdev->bdev,b));
}
printk("\n");
- err = do_md_run (mddev);
+ err = do_md_run(mddev);
if (err) {
printk(KERN_WARNING "md: do_md_run() returned %d\n", err);
- do_md_stop (mddev, 0);
+ do_md_stop(mddev, 0, 0);
}
}
*/
static void autorun_devices(int part)
{
- struct list_head *tmp;
- mdk_rdev_t *rdev0, *rdev;
+ mdk_rdev_t *rdev0, *rdev, *tmp;
mddev_t *mddev;
char b[BDEVNAME_SIZE];
printk(KERN_INFO "md: considering %s ...\n",
bdevname(rdev0->bdev,b));
INIT_LIST_HEAD(&candidates);
- rdev_for_each_list(rdev, tmp, pending_raid_disks)
+ rdev_for_each_list(rdev, tmp, &pending_raid_disks)
if (super_90_load(rdev, rdev0, 0) >= 0) {
printk(KERN_INFO "md: adding %s ...\n",
bdevname(rdev->bdev,b));
} else {
printk(KERN_INFO "md: created %s\n", mdname(mddev));
mddev->persistent = 1;
- rdev_for_each_list(rdev, tmp, candidates) {
+ rdev_for_each_list(rdev, tmp, &candidates) {
list_del_init(&rdev->same_set);
if (bind_rdev_to_array(rdev, mddev))
export_rdev(rdev);
/* on success, candidates will be empty, on error
* it won't...
*/
- rdev_for_each_list(rdev, tmp, candidates)
+ rdev_for_each_list(rdev, tmp, &candidates) {
+ list_del_init(&rdev->same_set);
export_rdev(rdev);
+ }
mddev_put(mddev);
}
printk(KERN_INFO "md: ... autorun DONE.\n");
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;
- rdev_for_each(rdev, tmp, mddev) {
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
nr++;
if (test_bit(Faulty, &rdev->flags))
failed++;
info.patch_version = MD_PATCHLEVEL_VERSION;
info.ctime = mddev->ctime;
info.level = mddev->level;
- info.size = mddev->size;
- if (info.size != mddev->size) /* overflow */
+ info.size = mddev->dev_sectors / 2;
+ if (info.size != mddev->dev_sectors / 2) /* overflow */
info.size = -1;
info.nr_disks = nr;
info.raid_disks = mddev->raid_disks;
char *ptr, *buf = NULL;
int err = -ENOMEM;
- md_allow_write(mddev);
+ if (md_allow_write(mddev))
+ file = kmalloc(sizeof(*file), GFP_NOIO);
+ else
+ file = kmalloc(sizeof(*file), GFP_KERNEL);
- file = kmalloc(sizeof(*file), GFP_KERNEL);
if (!file)
goto out;
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);
+ rdev = find_rdev_nr(mddev, info.number);
if (rdev) {
info.major = MAJOR(rdev->bdev->bd_dev);
info.minor = MINOR(rdev->bdev->bd_dev);
clear_bit(In_sync, &rdev->flags); /* just to be sure */
if (info->state & (1<<MD_DISK_WRITEMOSTLY))
set_bit(WriteMostly, &rdev->flags);
+ else
+ clear_bit(WriteMostly, &rdev->flags);
rdev->raid_disk = -1;
err = bind_rdev_to_array(rdev, mddev);
}
if (err)
export_rdev(rdev);
+ else
+ sysfs_notify_dirent(rdev->sysfs_state);
md_update_sb(mddev, 1);
+ if (mddev->degraded)
+ set_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
return err;
if (!(info->state & (1<<MD_DISK_FAULTY))) {
int err;
- rdev = md_import_device (dev, -1, 0);
+ rdev = md_import_device(dev, -1, 0);
if (IS_ERR(rdev)) {
printk(KERN_WARNING
"md: error, md_import_device() returned %ld\n",
if (!mddev->persistent) {
printk(KERN_INFO "md: nonpersistent superblock ...\n");
- rdev->sb_offset = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
+ rdev->sb_start = rdev->bdev->bd_inode->i_size / 512;
} else
- rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
- rdev->size = calc_dev_size(rdev, mddev->chunk_size);
+ rdev->sb_start = calc_dev_sboffset(rdev->bdev);
+ rdev->sectors = calc_num_sectors(rdev, mddev->chunk_size);
err = bind_rdev_to_array(rdev, mddev);
if (err) {
char b[BDEVNAME_SIZE];
mdk_rdev_t *rdev;
- if (!mddev->pers)
- return -ENODEV;
-
rdev = find_rdev(mddev, dev);
if (!rdev)
return -ENXIO;
{
char b[BDEVNAME_SIZE];
int err;
- unsigned int size;
mdk_rdev_t *rdev;
if (!mddev->pers)
return -EINVAL;
}
- rdev = md_import_device (dev, -1, 0);
+ rdev = md_import_device(dev, -1, 0);
if (IS_ERR(rdev)) {
printk(KERN_WARNING
"md: error, md_import_device() returned %ld\n",
}
if (mddev->persistent)
- rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
+ rdev->sb_start = calc_dev_sboffset(rdev->bdev);
else
- rdev->sb_offset =
- rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
+ rdev->sb_start = rdev->bdev->bd_inode->i_size / 512;
- size = calc_dev_size(rdev, mddev->chunk_size);
- rdev->size = size;
+ rdev->sectors = calc_num_sectors(rdev, mddev->chunk_size);
if (test_bit(Faulty, &rdev->flags)) {
printk(KERN_WARNING
* 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, 1);
md_new_event(mddev);
return 0;
-abort_unbind_export:
- unbind_rdev_from_array(rdev);
-
abort_export:
export_rdev(rdev);
return err;
mddev->level = info->level;
mddev->clevel[0] = 0;
- mddev->size = info->size;
+ mddev->dev_sectors = 2 * (sector_t)info->size;
mddev->raid_disks = info->raid_disks;
/* don't set md_minor, it is determined by which /dev/md* was
* openned
return 0;
}
-static int update_size(mddev_t *mddev, unsigned long size)
+void md_set_array_sectors(mddev_t *mddev, sector_t array_sectors)
{
- mdk_rdev_t * rdev;
+ WARN(!mddev_is_locked(mddev), "%s: unlocked mddev!\n", __func__);
+
+ if (mddev->external_size)
+ return;
+
+ mddev->array_sectors = array_sectors;
+}
+EXPORT_SYMBOL(md_set_array_sectors);
+
+static int update_size(mddev_t *mddev, sector_t num_sectors)
+{
+ mdk_rdev_t *rdev;
int rv;
- struct list_head *tmp;
- int fit = (size == 0);
+ int fit = (num_sectors == 0);
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 if 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.
+ /* The "num_sectors" is the number of sectors 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 this number if no resync or reconstruction is
+ * happening, and if the new size is acceptable. It must fit before the
+ * sb_start or, if that is <data_offset, it must fit before the size
+ * of each device. If num_sectors is zero, we find the largest size
+ * that fits.
+
*/
if (mddev->sync_thread)
return -EBUSY;
- rdev_for_each(rdev, tmp, mddev) {
- sector_t avail;
- avail = rdev->size * 2;
+ if (mddev->bitmap)
+ /* Sorry, cannot grow a bitmap yet, just remove it,
+ * grow, and re-add.
+ */
+ return -EBUSY;
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
+ sector_t avail = rdev->sectors;
- if (fit && (size == 0 || size > avail/2))
- size = avail/2;
- if (avail < ((sector_t)size << 1))
+ if (fit && (num_sectors == 0 || num_sectors > avail))
+ num_sectors = avail;
+ if (avail < num_sectors)
return -ENOSPC;
}
- rv = mddev->pers->resize(mddev, (sector_t)size *2);
+ rv = mddev->pers->resize(mddev, num_sectors);
if (!rv) {
struct block_device *bdev;
bdev = bdget_disk(mddev->gendisk, 0);
if (bdev) {
mutex_lock(&bdev->bd_inode->i_mutex);
- i_size_write(bdev->bd_inode, (loff_t)mddev->array_size << 10);
+ i_size_write(bdev->bd_inode,
+ (loff_t)mddev->array_sectors << 9);
mutex_unlock(&bdev->bd_inode->i_mutex);
bdput(bdev);
}
)
return -EINVAL;
/* Check there is only one change */
- if (info->size >= 0 && mddev->size != info->size) cnt++;
- if (mddev->raid_disks != info->raid_disks) cnt++;
- if (mddev->layout != info->layout) cnt++;
- if ((state ^ info->state) & (1<<MD_SB_BITMAP_PRESENT)) cnt++;
- if (cnt == 0) return 0;
- if (cnt > 1) return -EINVAL;
+ if (info->size >= 0 && mddev->dev_sectors / 2 != info->size)
+ cnt++;
+ if (mddev->raid_disks != info->raid_disks)
+ cnt++;
+ if (mddev->layout != info->layout)
+ cnt++;
+ if ((state ^ info->state) & (1<<MD_SB_BITMAP_PRESENT))
+ cnt++;
+ if (cnt == 0)
+ return 0;
+ if (cnt > 1)
+ return -EINVAL;
if (mddev->layout != info->layout) {
/* Change layout
else
return mddev->pers->reconfig(mddev, info->layout, -1);
}
- if (info->size >= 0 && mddev->size != info->size)
- rv = update_size(mddev, info->size);
+ if (info->size >= 0 && mddev->dev_sectors / 2 != info->size)
+ rv = update_size(mddev, (sector_t)info->size * 2);
if (mddev->raid_disks != info->raid_disks)
rv = update_raid_disks(mddev, info->raid_disks);
return 0;
}
+/*
+ * 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... ;-)
+ */
static int md_getgeo(struct block_device *bdev, struct hd_geometry *geo)
{
mddev_t *mddev = bdev->bd_disk->private_data;
return 0;
}
-static int md_ioctl(struct inode *inode, struct file *file,
+static int md_ioctl(struct block_device *bdev, fmode_t mode,
unsigned int cmd, unsigned long arg)
{
int err = 0;
* Commands creating/starting a new array:
*/
- mddev = inode->i_bdev->bd_disk->private_data;
+ mddev = bdev->bd_disk->private_data;
if (!mddev) {
BUG();
goto done_unlock;
case STOP_ARRAY:
- err = do_md_stop (mddev, 0);
+ err = do_md_stop(mddev, 0, 1);
goto done_unlock;
case STOP_ARRAY_RO:
- err = do_md_stop (mddev, 1);
+ err = do_md_stop(mddev, 1, 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... ;-)
- */
}
/*
* here and hit the 'default' below, so only disallow
* 'md' ioctls, and switch to rw mode if started auto-readonly.
*/
- if (_IOC_TYPE(cmd) == MD_MAJOR &&
- mddev->ro && mddev->pers) {
+ if (_IOC_TYPE(cmd) == MD_MAJOR && mddev->ro && mddev->pers) {
if (mddev->ro == 2) {
mddev->ro = 0;
- set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
- md_wakeup_thread(mddev->thread);
-
+ sysfs_notify_dirent(mddev->sysfs_state);
+ set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+ md_wakeup_thread(mddev->thread);
} else {
err = -EROFS;
goto abort_unlock;
goto done_unlock;
case RUN_ARRAY:
- err = do_md_run (mddev);
+ err = do_md_run(mddev);
goto done_unlock;
case SET_BITMAP_FILE:
done_unlock:
abort_unlock:
+ if (mddev->hold_active == UNTIL_IOCTL &&
+ err != -EINVAL)
+ mddev->hold_active = 0;
mddev_unlock(mddev);
return err;
return err;
}
-static int md_open(struct inode *inode, struct file *file)
+static int md_open(struct block_device *bdev, fmode_t mode)
{
/*
* 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;
+ mddev_t *mddev = mddev_find(bdev->bd_dev);
int err;
+ if (mddev->gendisk != bdev->bd_disk) {
+ /* we are racing with mddev_put which is discarding this
+ * bd_disk.
+ */
+ mddev_put(mddev);
+ /* Wait until bdev->bd_disk is definitely gone */
+ flush_scheduled_work();
+ /* Then retry the open from the top */
+ return -ERESTARTSYS;
+ }
+ BUG_ON(mddev != bdev->bd_disk->private_data);
+
if ((err = mutex_lock_interruptible_nested(&mddev->reconfig_mutex, 1)))
goto out;
err = 0;
- mddev_get(mddev);
+ atomic_inc(&mddev->openers);
mddev_unlock(mddev);
- check_disk_change(inode->i_bdev);
+ check_disk_change(bdev);
out:
return err;
}
-static int md_release(struct inode *inode, struct file * file)
+static int md_release(struct gendisk *disk, fmode_t mode)
{
- mddev_t *mddev = inode->i_bdev->bd_disk->private_data;
+ mddev_t *mddev = disk->private_data;
BUG_ON(!mddev);
+ atomic_dec(&mddev->openers);
mddev_put(mddev);
return 0;
.owner = THIS_MODULE,
.open = md_open,
.release = md_release,
- .ioctl = md_ioctl,
+ .locked_ioctl = md_ioctl,
.getgeo = md_getgeo,
.media_changed = md_media_changed,
.revalidate_disk= md_revalidate,
void md_unregister_thread(mdk_thread_t *thread)
{
+ if (!thread)
+ return;
dprintk("interrupting MD-thread pid %d\n", task_pid_nr(thread->tsk));
kthread_stop(thread->tsk);
if (!mddev->pers->error_handler)
return;
mddev->pers->error_handler(mddev,rdev);
+ if (mddev->degraded)
+ set_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
+ set_bit(StateChanged, &rdev->flags);
set_bit(MD_RECOVERY_INTR, &mddev->recovery);
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
{
int i = 0;
mdk_rdev_t *rdev;
- struct list_head *tmp;
seq_printf(seq, "unused devices: ");
- rdev_for_each_list(rdev, tmp, pending_raid_disks) {
+ list_for_each_entry(rdev, &pending_raid_disks, same_set) {
char b[BDEVNAME_SIZE];
i++;
seq_printf(seq, "%s ",
if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
max_blocks = mddev->resync_max_sectors >> 1;
else
- max_blocks = mddev->size;
+ max_blocks = mddev->dev_sectors / 2;
/*
* Should not happen.
static int md_seq_show(struct seq_file *seq, void *v)
{
mddev_t *mddev = v;
- sector_t size;
- struct list_head *tmp2;
+ sector_t sectors;
mdk_rdev_t *rdev;
struct mdstat_info *mi = seq->private;
struct bitmap *bitmap;
seq_printf(seq, " %s", mddev->pers->name);
}
- size = 0;
- rdev_for_each(rdev, tmp2, mddev) {
+ sectors = 0;
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
char b[BDEVNAME_SIZE];
seq_printf(seq, " %s[%d]",
bdevname(rdev->bdev,b), rdev->desc_nr);
continue;
} else if (rdev->raid_disk < 0)
seq_printf(seq, "(S)"); /* spare */
- size += rdev->size;
+ sectors += rdev->sectors;
}
if (!list_empty(&mddev->disks)) {
if (mddev->pers)
seq_printf(seq, "\n %llu blocks",
- (unsigned long long)mddev->array_size);
+ (unsigned long long)
+ mddev->array_sectors / 2);
else
seq_printf(seq, "\n %llu blocks",
- (unsigned long long)size);
+ (unsigned long long)sectors / 2);
}
if (mddev->persistent) {
if (mddev->major_version != 0 ||
seq_printf(seq, " super non-persistent");
if (mddev->pers) {
- mddev->pers->status (seq, mddev);
+ mddev->pers->status(seq, mddev);
seq_printf(seq, "\n ");
if (mddev->pers->sync_request) {
if (mddev->curr_resync > 2) {
- status_resync (seq, mddev);
+ status_resync(seq, mddev);
seq_printf(seq, "\n ");
} else if (mddev->curr_resync == 1 || mddev->curr_resync == 2)
seq_printf(seq, "\tresync=DELAYED\n ");
return 0;
}
-static int is_mddev_idle(mddev_t *mddev)
+static int is_mddev_idle(mddev_t *mddev, int init)
{
mdk_rdev_t * rdev;
- struct list_head *tmp;
int idle;
- long curr_events;
+ int curr_events;
idle = 1;
- rdev_for_each(rdev, tmp, mddev) {
+ rcu_read_lock();
+ rdev_for_each_rcu(rdev, mddev) {
struct gendisk *disk = rdev->bdev->bd_contains->bd_disk;
- curr_events = disk_stat_read(disk, sectors[0]) +
- disk_stat_read(disk, sectors[1]) -
- atomic_read(&disk->sync_io);
+ curr_events = (int)part_stat_read(&disk->part0, sectors[0]) +
+ (int)part_stat_read(&disk->part0, sectors[1]) -
+ atomic_read(&disk->sync_io);
/* sync IO will cause sync_io to increase before the disk_stats
* as sync_io is counted when a request starts, and
* disk_stats is counted when it completes.
* always make curr_events less than last_events.
*
*/
- if (curr_events - rdev->last_events > 4096) {
+ if (init || curr_events - rdev->last_events > 64) {
rdev->last_events = curr_events;
idle = 0;
}
}
+ rcu_read_unlock();
return idle;
}
*/
void md_write_start(mddev_t *mddev, struct bio *bi)
{
+ int did_change = 0;
if (bio_data_dir(bi) != WRITE)
return;
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
md_wakeup_thread(mddev->sync_thread);
+ did_change = 1;
}
atomic_inc(&mddev->writes_pending);
if (mddev->safemode == 1)
mddev->in_sync = 0;
set_bit(MD_CHANGE_CLEAN, &mddev->flags);
md_wakeup_thread(mddev->thread);
+ did_change = 1;
}
spin_unlock_irq(&mddev->write_lock);
- sysfs_notify(&mddev->kobj, NULL, "array_state");
}
+ if (did_change)
+ sysfs_notify_dirent(mddev->sysfs_state);
wait_event(mddev->sb_wait,
!test_bit(MD_CHANGE_CLEAN, &mddev->flags) &&
!test_bit(MD_CHANGE_PENDING, &mddev->flags));
* may proceed without blocking. It is important to call this before
* attempting a GFP_KERNEL allocation while holding the mddev lock.
* Must be called with mddev_lock held.
+ *
+ * In the ->external case MD_CHANGE_CLEAN can not be cleared until mddev->lock
+ * is dropped, so return -EAGAIN after notifying userspace.
*/
-void md_allow_write(mddev_t *mddev)
+int md_allow_write(mddev_t *mddev)
{
if (!mddev->pers)
- return;
+ return 0;
if (mddev->ro)
- return;
+ return 0;
+ if (!mddev->pers->sync_request)
+ return 0;
spin_lock_irq(&mddev->write_lock);
if (mddev->in_sync) {
mddev->safemode = 1;
spin_unlock_irq(&mddev->write_lock);
md_update_sb(mddev, 0);
-
- sysfs_notify(&mddev->kobj, NULL, "array_state");
- /* wait for the dirty state to be recorded in the metadata */
- wait_event(mddev->sb_wait,
- !test_bit(MD_CHANGE_CLEAN, &mddev->flags) &&
- !test_bit(MD_CHANGE_PENDING, &mddev->flags));
+ sysfs_notify_dirent(mddev->sysfs_state);
} else
spin_unlock_irq(&mddev->write_lock);
+
+ if (test_bit(MD_CHANGE_CLEAN, &mddev->flags))
+ return -EAGAIN;
+ else
+ return 0;
}
EXPORT_SYMBOL_GPL(md_allow_write);
struct list_head *tmp;
sector_t last_check;
int skipped = 0;
- struct list_head *rtmp;
mdk_rdev_t *rdev;
char *desc;
* time 'round when curr_resync == 2
*/
continue;
- prepare_to_wait(&resync_wait, &wq, TASK_UNINTERRUPTIBLE);
+ /* We need to wait 'interruptible' so as not to
+ * contribute to the load average, and not to
+ * be caught by 'softlockup'
+ */
+ prepare_to_wait(&resync_wait, &wq, TASK_INTERRUPTIBLE);
if (!kthread_should_stop() &&
mddev2->curr_resync >= mddev->curr_resync) {
printk(KERN_INFO "md: delaying %s of %s"
" share one or more physical units)\n",
desc, mdname(mddev), mdname(mddev2));
mddev_put(mddev2);
+ if (signal_pending(current))
+ flush_signals(current);
schedule();
finish_wait(&resync_wait, &wq);
goto try_again;
j = mddev->recovery_cp;
} else if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
- max_sectors = mddev->size << 1;
+ max_sectors = mddev->dev_sectors;
else {
/* recovery follows the physical size of devices */
- max_sectors = mddev->size << 1;
+ max_sectors = mddev->dev_sectors;
j = MaxSector;
- rdev_for_each(rdev, rtmp, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->raid_disk >= 0 &&
!test_bit(Faulty, &rdev->flags) &&
!test_bit(In_sync, &rdev->flags) &&
"(but not more than %d KB/sec) for %s.\n",
speed_max(mddev), desc);
- is_mddev_idle(mddev); /* this also initializes IO event counters */
+ is_mddev_idle(mddev, 1); /* this initializes IO event counters */
io_sectors = 0;
for (m = 0; m < SYNC_MARKS; m++) {
}
if (kthread_should_stop())
goto interrupted;
+
+ if (mddev->curr_resync > mddev->curr_resync_completed &&
+ (mddev->curr_resync - mddev->curr_resync_completed)
+ > (max_sectors >> 4)) {
+ /* time to update curr_resync_completed */
+ blk_unplug(mddev->queue);
+ wait_event(mddev->recovery_wait,
+ atomic_read(&mddev->recovery_active) == 0);
+ mddev->curr_resync_completed =
+ mddev->curr_resync;
+ set_bit(MD_CHANGE_CLEAN, &mddev->flags);
+ }
sectors = mddev->pers->sync_request(mddev, j, &skipped,
currspeed < speed_min(mddev));
if (sectors == 0) {
if (currspeed > speed_min(mddev)) {
if ((currspeed > speed_max(mddev)) ||
- !is_mddev_idle(mddev)) {
+ !is_mddev_idle(mddev, 0)) {
msleep(500);
goto repeat;
}
} else {
if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery))
mddev->curr_resync = MaxSector;
- rdev_for_each(rdev, rtmp, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->raid_disk >= 0 &&
!test_bit(Faulty, &rdev->flags) &&
!test_bit(In_sync, &rdev->flags) &&
static int remove_and_add_spares(mddev_t *mddev)
{
mdk_rdev_t *rdev;
- struct list_head *rtmp;
int spares = 0;
- rdev_for_each(rdev, rtmp, mddev)
+ mddev->curr_resync_completed = 0;
+
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (rdev->raid_disk >= 0 &&
!test_bit(Blocked, &rdev->flags) &&
(test_bit(Faulty, &rdev->flags) ||
}
}
- if (mddev->degraded) {
- rdev_for_each(rdev, rtmp, mddev) {
+ if (mddev->degraded && ! mddev->ro && !mddev->recovery_disabled) {
+ list_for_each_entry(rdev, &mddev->disks, same_set) {
if (rdev->raid_disk >= 0 &&
- !test_bit(In_sync, &rdev->flags))
+ !test_bit(In_sync, &rdev->flags) &&
+ !test_bit(Blocked, &rdev->flags))
spares++;
if (rdev->raid_disk < 0
&& !test_bit(Faulty, &rdev->flags)) {
rdev->recovery_offset = 0;
- if (mddev->pers->hot_add_disk(mddev,rdev)) {
+ if (mddev->pers->
+ hot_add_disk(mddev, rdev) == 0) {
char nm[20];
sprintf(nm, "rd%d", rdev->raid_disk);
if (sysfs_create_link(&mddev->kobj,
void md_check_recovery(mddev_t *mddev)
{
mdk_rdev_t *rdev;
- struct list_head *rtmp;
if (mddev->bitmap)
flush_signals(current);
}
+ if (mddev->ro && !test_bit(MD_RECOVERY_NEEDED, &mddev->recovery))
+ return;
if ( ! (
(mddev->flags && !mddev->external) ||
test_bit(MD_RECOVERY_NEEDED, &mddev->recovery) ||
if (mddev_trylock(mddev)) {
int spares = 0;
+ if (mddev->ro) {
+ /* Only thing we do on a ro array is remove
+ * failed devices.
+ */
+ remove_and_add_spares(mddev);
+ clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+ goto unlock;
+ }
+
if (!mddev->external) {
+ int did_change = 0;
spin_lock_irq(&mddev->write_lock);
if (mddev->safemode &&
!atomic_read(&mddev->writes_pending) &&
!mddev->in_sync &&
mddev->recovery_cp == MaxSector) {
mddev->in_sync = 1;
+ did_change = 1;
if (mddev->persistent)
set_bit(MD_CHANGE_CLEAN, &mddev->flags);
}
if (mddev->safemode == 1)
mddev->safemode = 0;
spin_unlock_irq(&mddev->write_lock);
+ if (did_change)
+ sysfs_notify_dirent(mddev->sysfs_state);
}
if (mddev->flags)
md_update_sb(mddev, 0);
+ list_for_each_entry(rdev, &mddev->disks, same_set)
+ if (test_and_clear_bit(StateChanged, &rdev->flags))
+ sysfs_notify_dirent(rdev->sysfs_state);
+
if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) &&
!test_bit(MD_RECOVERY_DONE, &mddev->recovery)) {
/* resync has finished, collect result */
md_unregister_thread(mddev->sync_thread);
mddev->sync_thread = NULL;
- if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
+ if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery) &&
+ !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
/* success...*/
/* activate any spares */
- mddev->pers->spare_active(mddev);
+ if (mddev->pers->spare_active(mddev))
+ sysfs_notify(&mddev->kobj, NULL,
+ "degraded");
}
+ if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery) &&
+ mddev->pers->finish_reshape)
+ mddev->pers->finish_reshape(mddev);
md_update_sb(mddev, 1);
/* if array is no-longer degraded, then any saved_raid_disk
* information must be scrapped
*/
if (!mddev->degraded)
- rdev_for_each(rdev, rtmp, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
rdev->saved_raid_disk = -1;
mddev->recovery = 0;
/* flag recovery needed just to double check */
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+ sysfs_notify_dirent(mddev->sysfs_action);
md_new_event(mddev);
goto unlock;
}
+ /* Set RUNNING before clearing NEEDED to avoid
+ * any transients in the value of "sync_action".
+ */
+ set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
+ clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
/* Clear some bits that don't mean anything, but
* might be left set
*/
- clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
clear_bit(MD_RECOVERY_INTR, &mddev->recovery);
clear_bit(MD_RECOVERY_DONE, &mddev->recovery);
/* Cannot proceed */
goto unlock;
set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
+ clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
} else if ((spares = remove_and_add_spares(mddev))) {
clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
+ clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
+ set_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
} else if (mddev->recovery_cp < MaxSector) {
set_bit(MD_RECOVERY_SYNC, &mddev->recovery);
+ clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
} else if (!test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
/* nothing to be done ... */
goto unlock;
if (mddev->pers->sync_request) {
- set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
if (spares && mddev->bitmap && ! mddev->bitmap->file) {
/* We are adding a device or devices to an array
* which has the bitmap stored on all devices.
mddev->recovery = 0;
} else
md_wakeup_thread(mddev->sync_thread);
+ sysfs_notify_dirent(mddev->sysfs_action);
md_new_event(mddev);
}
unlock:
+ if (!mddev->sync_thread) {
+ clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
+ if (test_and_clear_bit(MD_RECOVERY_RECOVER,
+ &mddev->recovery))
+ if (mddev->sysfs_action)
+ sysfs_notify_dirent(mddev->sysfs_action);
+ }
mddev_unlock(mddev);
}
}
void md_wait_for_blocked_rdev(mdk_rdev_t *rdev, mddev_t *mddev)
{
- sysfs_notify(&rdev->kobj, NULL, "state");
+ sysfs_notify_dirent(rdev->sysfs_state);
wait_event_timeout(rdev->blocked_wait,
!test_bit(Blocked, &rdev->flags),
msecs_to_jiffies(5000));
for_each_mddev(mddev, tmp)
if (mddev_trylock(mddev)) {
- do_md_stop (mddev, 1);
+ /* Force a switch to readonly even array
+ * appears to still be in use. Hence
+ * the '100'.
+ */
+ do_md_stop(mddev, 1, 100);
mddev_unlock(mddev);
}
/*
static int __init md_init(void)
{
- if (register_blkdev(MAJOR_NR, "md"))
+ if (register_blkdev(MD_MAJOR, "md"))
return -1;
if ((mdp_major=register_blkdev(0, "mdp"))<=0) {
- unregister_blkdev(MAJOR_NR, "md");
+ unregister_blkdev(MD_MAJOR, "md");
return -1;
}
- blk_register_region(MKDEV(MAJOR_NR, 0), 1UL<<MINORBITS, THIS_MODULE,
+ blk_register_region(MKDEV(MD_MAJOR, 0), 1UL<<MINORBITS, THIS_MODULE,
md_probe, NULL, NULL);
blk_register_region(MKDEV(mdp_major, 0), 1UL<<MINORBITS, THIS_MODULE,
md_probe, NULL, NULL);
raid_table_header = register_sysctl_table(raid_root_table);
md_geninit();
- return (0);
+ return 0;
}
mddev_t *mddev;
struct list_head *tmp;
- blk_unregister_region(MKDEV(MAJOR_NR,0), 1U << MINORBITS);
+ blk_unregister_region(MKDEV(MD_MAJOR,0), 1U << MINORBITS);
blk_unregister_region(MKDEV(mdp_major,0), 1U << MINORBITS);
- unregister_blkdev(MAJOR_NR,"md");
+ unregister_blkdev(MD_MAJOR,"md");
unregister_blkdev(mdp_major, "mdp");
unregister_reboot_notifier(&md_notifier);
unregister_sysctl_table(raid_table_header);
remove_proc_entry("mdstat", NULL);
for_each_mddev(mddev, tmp) {
- struct gendisk *disk = mddev->gendisk;
- if (!disk)
- continue;
export_array(mddev);
- del_gendisk(disk);
- put_disk(disk);
- mddev->gendisk = NULL;
- mddev_put(mddev);
+ mddev->hold_active = 0;
}
}
module_param_call(start_ro, set_ro, get_ro, NULL, S_IRUSR|S_IWUSR);
module_param(start_dirty_degraded, int, S_IRUGO|S_IWUSR);
+module_param_call(new_array, add_named_array, NULL, NULL, S_IWUSR);
EXPORT_SYMBOL(register_md_personality);
EXPORT_SYMBOL(unregister_md_personality);
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff