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/sysctl.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>
+#include <linux/delay.h>
#define MAJOR_NR MD_MAJOR
-#define MD_DRIVER
/* 63 partitions with the alternate major number (mdp) */
#define MdpMinorShift 6
#ifndef MODULE
-static void autostart_arrays (int part);
+static void autostart_arrays(int part);
#endif
static LIST_HEAD(pers_list);
)
-static int md_fail_request (struct request_queue *q, struct bio *bio)
+static int md_fail_request(struct request_queue *q, struct bio *bio)
{
bio_io_error(bio);
return 0;
return mddev;
}
+static void mddev_delayed_delete(struct work_struct *ws);
+
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);
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;
}
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;
}
* 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;
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->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)
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. */
}
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);
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;
}
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;
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",
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)
err = 0;
}
- if (!err)
- sysfs_notify(&rdev->kobj, NULL, "state");
+ if (!err && rdev->sysfs_state)
+ sysfs_notify_dirent(rdev->sysfs_state);
return err ? err : len;
}
static struct rdev_sysfs_entry rdev_state =
md_wakeup_thread(rdev->mddev->thread);
} else if (rdev->mddev->pers) {
mdk_rdev_t *rdev2;
- struct list_head *tmp;
/* Activating a spare .. or possibly reactivating
* if we every get bitmaps working here.
*/
if (rdev->mddev->pers->hot_add_disk == NULL)
return -EINVAL;
- rdev_for_each(rdev2, tmp, rdev->mddev)
+ list_for_each_entry(rdev2, &rdev->mddev->disks, same_set)
if (rdev2->raid_disk == slot)
return -EEXIST;
rdev->raid_disk = -1;
return err;
} else
- sysfs_notify(&rdev->kobj, NULL, "state");
+ 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
clear_bit(Faulty, &rdev->flags);
clear_bit(WriteMostly, &rdev->flags);
set_bit(In_sync, &rdev->flags);
- sysfs_notify(&rdev->kobj, NULL, "state");
+ sysfs_notify_dirent(rdev->sysfs_state);
}
return len;
}
if (strict_strtoull(buf, 10, &size) < 0)
return -EINVAL;
- if (size < my_mddev->size)
- return -EINVAL;
if (my_mddev->pers && rdev->raid_disk >= 0) {
if (my_mddev->persistent) {
size = super_types[my_mddev->major_version].
size = (rdev->bdev->bd_inode->i_size >> 10);
size -= rdev->data_offset/2;
}
- if (size < my_mddev->size)
- return -EINVAL; /* component must fit device */
}
+ if (size < my_mddev->size)
+ return -EINVAL; /* component must fit device */
rdev->size = size;
if (size > oldsize && my_mddev->external) {
*/
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 &&
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;
}
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, 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, 0);
} else
break;
case read_auto:
if (mddev->pers) {
- if (mddev->ro != 1)
+ if (mddev->ro == 0)
err = do_md_stop(mddev, 1, 0);
- else
+ else if (mddev->ro == 1)
err = restart_array(mddev);
if (err == 0) {
mddev->ro = 2;
if (err)
return err;
else {
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ sysfs_notify_dirent(mddev->sysfs_state);
return len;
}
}
{
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")) {
}
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
- sysfs_notify(&mddev->kobj, NULL, "sync_action");
+ sysfs_notify_dirent(mddev->sysfs_action);
return len;
}
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 void mddev_delayed_delete(struct work_struct *ws)
+{
+ mddev_t *mddev = container_of(ws, mddev_t, del_work);
+
+ if (mddev->private == &md_redundancy_group) {
+ sysfs_remove_group(&mddev->kobj, &md_redundancy_group);
+ if (mddev->sysfs_action)
+ sysfs_put(mddev->sysfs_action);
+ mddev->sysfs_action = NULL;
+ mddev->private = NULL;
+ }
+ kobject_del(&mddev->kobj);
+ kobject_put(&mddev->kobj);
+}
+
+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;
+ }
+ /* 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_fail_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->dev.kobj,
- "%s", "md");
+ error = kobject_init_and_add(&mddev->kobj, &md_ktype,
+ &disk_to_dev(disk)->kobj, "%s", "md");
mutex_unlock(&disks_mutex);
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;
if (!atomic_read(&mddev->writes_pending)) {
mddev->safemode = 1;
if (mddev->external)
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ 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) {
}
}
-#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);
return -EINVAL;
}
}
- sysfs_notify(&rdev->kobj, NULL, "state");
+ sysfs_notify_dirent(rdev->sysfs_state);
}
md_probe(mddev->unit, NULL, NULL);
*/
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"
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);
+ set_capacity(disk, mddev->array_sectors);
/* If we call blk_queue_make_request here, it will
* re-initialise max_sectors etc which may have been
* 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);
- sysfs_notify(&mddev->kobj, NULL, "array_state");
- sysfs_notify(&mddev->kobj, NULL, "sync_action");
+ sysfs_notify_dirent(mddev->sysfs_state);
+ if (mddev->sysfs_action)
+ sysfs_notify_dirent(mddev->sysfs_action);
sysfs_notify(&mddev->kobj, NULL, "degraded");
- kobject_uevent(&mddev->gendisk->dev.kobj, KOBJ_CHANGE);
+ kobject_uevent(&disk_to_dev(mddev->gendisk)->kobj, KOBJ_CHANGE);
return 0;
}
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
md_wakeup_thread(mddev->sync_thread);
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ sysfs_notify_dirent(mddev->sysfs_state);
return 0;
}
int err = 0;
struct gendisk *disk = mddev->gendisk;
- if (atomic_read(&mddev->active) > 1 + is_open) {
+ if (atomic_read(&mddev->openers) > is_open) {
printk("md: %s still in use.\n",mdname(mddev));
return -EBUSY;
}
del_timer_sync(&mddev->safemode_timer);
- invalidate_partition(disk, 0);
-
switch(mode) {
case 1: /* readonly */
err = -ENXIO;
mddev->queue->merge_bvec_fn = NULL;
mddev->queue->unplug_fn = NULL;
mddev->queue->backing_dev_info.congested_fn = NULL;
- if (mddev->pers->sync_request)
- sysfs_remove_group(&mddev->kobj, &md_redundancy_group);
-
module_put(mddev->pers->owner);
+ if (mddev->pers->sync_request)
+ mddev->private = &md_redundancy_group;
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->array_sectors = 0;
mddev->size = 0;
mddev->raid_disks = 0;
mddev->recovery_cp = 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;
md_new_event(mddev);
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ 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, 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++;
if (err)
export_rdev(rdev);
else
- sysfs_notify(&rdev->kobj, NULL, "state");
+ sysfs_notify_dirent(rdev->sysfs_state);
md_update_sb(mddev, 1);
if (mddev->degraded)
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",
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",
* 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;
static int update_size(mddev_t *mddev, sector_t num_sectors)
{
- mdk_rdev_t * rdev;
+ mdk_rdev_t *rdev;
int rv;
- struct list_head *tmp;
int fit = (num_sectors == 0);
if (mddev->pers->resize == NULL)
*/
if (mddev->sync_thread)
return -EBUSY;
- rdev_for_each(rdev, tmp, mddev) {
+ 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;
avail = rdev->size * 2;
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 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, 1);
+ err = do_md_stop(mddev, 0, 1);
goto done_unlock;
case STOP_ARRAY_RO:
- err = do_md_stop (mddev, 1, 1);
+ err = do_md_stop(mddev, 1, 1);
goto done_unlock;
}
if (_IOC_TYPE(cmd) == MD_MAJOR && mddev->ro && mddev->pers) {
if (mddev->ro == 2) {
mddev->ro = 0;
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ sysfs_notify_dirent(mddev->sysfs_state);
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
} else {
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,
{
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 ",
{
mddev_t *mddev = v;
sector_t size;
- struct list_head *tmp2;
mdk_rdev_t *rdev;
struct mdstat_info *mi = seq->private;
struct bitmap *bitmap;
}
size = 0;
- rdev_for_each(rdev, tmp2, mddev) {
+ 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);
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)size);
}
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 ");
static int is_mddev_idle(mddev_t *mddev)
{
mdk_rdev_t * rdev;
- struct list_head *tmp;
int idle;
long 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]) -
+ curr_events = part_stat_read(&disk->part0, sectors[0]) +
+ 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
idle = 0;
}
}
+ rcu_read_unlock();
return idle;
}
spin_unlock_irq(&mddev->write_lock);
}
if (did_change)
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ 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));
mddev->safemode = 1;
spin_unlock_irq(&mddev->write_lock);
md_update_sb(mddev, 0);
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ sysfs_notify_dirent(mddev->sysfs_state);
} else
spin_unlock_irq(&mddev->write_lock);
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;
/* recovery follows the physical size of devices */
max_sectors = mddev->size << 1;
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) &&
} 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)
+ 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)) {
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);
mddev->safemode = 0;
spin_unlock_irq(&mddev->write_lock);
if (did_change)
- sysfs_notify(&mddev->kobj, NULL, "array_state");
+ sysfs_notify_dirent(mddev->sysfs_state);
}
if (mddev->flags)
md_update_sb(mddev, 0);
- rdev_for_each(rdev, rtmp, mddev)
+ list_for_each_entry(rdev, &mddev->disks, same_set)
if (test_and_clear_bit(StateChanged, &rdev->flags))
- sysfs_notify(&rdev->kobj, NULL, "state");
+ sysfs_notify_dirent(rdev->sysfs_state);
if (test_bit(MD_RECOVERY_RUNNING, &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 */
if (mddev->pers->spare_active(mddev))
* 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(&mddev->kobj, NULL, "sync_action");
+ sysfs_notify_dirent(mddev->sysfs_action);
md_new_event(mddev);
goto unlock;
}
} 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);
mddev->recovery = 0;
} else
md_wakeup_thread(mddev->sync_thread);
- sysfs_notify(&mddev->kobj, NULL, "sync_action");
+ sysfs_notify_dirent(mddev->sysfs_action);
md_new_event(mddev);
}
unlock:
clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
if (test_and_clear_bit(MD_RECOVERY_RECOVER,
&mddev->recovery))
- sysfs_notify(&mddev->kobj, NULL, "sync_action");
+ 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, 0);
+ /* 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);
}
/*
raid_table_header = register_sysctl_table(raid_root_table);
md_geninit();
- return (0);
+ return 0;
}
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