dm snapshot: rework writing to origin

[safe/jmp/linux-2.6] / drivers / md / md.c

diff --git a/drivers/md/md.c b/drivers/md/md.c
index f64b330..5f154ef 100644 (file)
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -98,44 +98,40 @@ static struct ctl_table_header *raid_table_header;
 
 static ctl_table raid_table[] = {
        {
-               .ctl_name       = DEV_RAID_SPEED_LIMIT_MIN,
                .procname       = "speed_limit_min",
                .data           = &sysctl_speed_limit_min,
                .maxlen         = sizeof(int),
                .mode           = S_IRUGO|S_IWUSR,
-               .proc_handler   = &proc_dointvec,
+               .proc_handler   = proc_dointvec,
        },
        {
-               .ctl_name       = DEV_RAID_SPEED_LIMIT_MAX,
                .procname       = "speed_limit_max",
                .data           = &sysctl_speed_limit_max,
                .maxlen         = sizeof(int),
                .mode           = S_IRUGO|S_IWUSR,
-               .proc_handler   = &proc_dointvec,
+               .proc_handler   = proc_dointvec,
        },
-       { .ctl_name = 0 }
+       { }
 };
 
 static ctl_table raid_dir_table[] = {
        {
-               .ctl_name       = DEV_RAID,
                .procname       = "raid",
                .maxlen         = 0,
                .mode           = S_IRUGO|S_IXUGO,
                .child          = raid_table,
        },
-       { .ctl_name = 0 }
+       { }
 };
 
 static ctl_table raid_root_table[] = {
        {
-               .ctl_name       = CTL_DEV,
                .procname       = "dev",
                .maxlen         = 0,
                .mode           = 0555,
                .child          = raid_dir_table,
        },
-       { .ctl_name = 0 }
+       {  }
 };
 
 static const struct block_device_operations md_fops;
@@ -262,6 +258,12 @@ static void mddev_resume(mddev_t *mddev)
        mddev->pers->quiesce(mddev, 0);
 }
 
+int mddev_congested(mddev_t *mddev, int bits)
+{
+       return mddev->suspended;
+}
+EXPORT_SYMBOL(mddev_congested);
+
 
 static inline mddev_t *mddev_get(mddev_t *mddev)
 {
@@ -938,6 +940,14 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
                            desc->raid_disk < mddev->raid_disks */) {
                        set_bit(In_sync, &rdev->flags);
                        rdev->raid_disk = desc->raid_disk;
+               } else if (desc->state & (1<<MD_DISK_ACTIVE)) {
+                       /* active but not in sync implies recovery up to
+                        * reshape position.  We don't know exactly where
+                        * that is, so set to zero for now */
+                       if (mddev->minor_version >= 91) {
+                               rdev->recovery_offset = 0;
+                               rdev->raid_disk = desc->raid_disk;
+                       }
                }
                if (desc->state & (1<<MD_DISK_WRITEMOSTLY))
                        set_bit(WriteMostly, &rdev->flags);
@@ -1026,8 +1036,19 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
        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)
-                   && !test_bit(Faulty, &rdev2->flags))
+               int is_active = test_bit(In_sync, &rdev2->flags);
+
+               if (rdev2->raid_disk >= 0 &&
+                   sb->minor_version >= 91)
+                       /* we have nowhere to store the recovery_offset,
+                        * but if it is not below the reshape_position,
+                        * we can piggy-back on that.
+                        */
+                       is_active = 1;
+               if (rdev2->raid_disk < 0 ||
+                   test_bit(Faulty, &rdev2->flags))
+                       is_active = 0;
+               if (is_active)
                        desc_nr = rdev2->raid_disk;
                else
                        desc_nr = next_spare++;
@@ -1037,16 +1058,16 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
                d->number = rdev2->desc_nr;
                d->major = MAJOR(rdev2->bdev->bd_dev);
                d->minor = MINOR(rdev2->bdev->bd_dev);
-               if (rdev2->raid_disk >= 0 && test_bit(In_sync, &rdev2->flags)
-                   && !test_bit(Faulty, &rdev2->flags))
+               if (is_active)
                        d->raid_disk = rdev2->raid_disk;
                else
                        d->raid_disk = rdev2->desc_nr; /* compatibility */
                if (test_bit(Faulty, &rdev2->flags))
                        d->state = (1<<MD_DISK_FAULTY);
-               else if (test_bit(In_sync, &rdev2->flags)) {
+               else if (is_active) {
                        d->state = (1<<MD_DISK_ACTIVE);
-                       d->state |= (1<<MD_DISK_SYNC);
+                       if (test_bit(In_sync, &rdev2->flags))
+                               d->state |= (1<<MD_DISK_SYNC);
                        active++;
                        working++;
                } else {
@@ -1376,8 +1397,6 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 
        if (rdev->raid_disk >= 0 &&
            !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);
@@ -1911,6 +1930,14 @@ static void sync_sbs(mddev_t * mddev, int nospares)
         */
        mdk_rdev_t *rdev;
 
+       /* First make sure individual recovery_offsets are correct */
+       list_for_each_entry(rdev, &mddev->disks, same_set) {
+               if (rdev->raid_disk >= 0 &&
+                   !test_bit(In_sync, &rdev->flags) &&
+                   mddev->curr_resync_completed > rdev->recovery_offset)
+                               rdev->recovery_offset = mddev->curr_resync_completed;
+
+       }       
        list_for_each_entry(rdev, &mddev->disks, same_set) {
                if (rdev->sb_events == mddev->events ||
                    (nospares &&
@@ -2625,7 +2652,7 @@ static void analyze_sbs(mddev_t * mddev)
                        rdev->desc_nr = i++;
                        rdev->raid_disk = rdev->desc_nr;
                        set_bit(In_sync, &rdev->flags);
-               } else if (rdev->raid_disk >= mddev->raid_disks) {
+               } else if (rdev->raid_disk >= (mddev->raid_disks - min(0, mddev->delta_disks))) {
                        rdev->raid_disk = -1;
                        clear_bit(In_sync, &rdev->flags);
                }
@@ -4218,7 +4245,7 @@ static int do_md_run(mddev_t * mddev)
                        set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
                        mddev->sync_thread = md_register_thread(md_do_sync,
                                                                mddev,
-                                                               "%s_resync");
+                                                               "resync");
                        if (!mddev->sync_thread) {
                                printk(KERN_ERR "%s: could not start resync"
                                       " thread...\n",
@@ -5631,7 +5658,10 @@ mdk_thread_t *md_register_thread(void (*run) (mddev_t *), mddev_t *mddev,
        thread->run = run;
        thread->mddev = mddev;
        thread->timeout = MAX_SCHEDULE_TIMEOUT;
-       thread->tsk = kthread_run(md_thread, thread, name, mdname(thread->mddev));
+       thread->tsk = kthread_run(md_thread, thread,
+                                 "%s_%s",
+                                 mdname(thread->mddev),
+                                 name ?: mddev->pers->name);
        if (IS_ERR(thread->tsk)) {
                kfree(thread);
                return NULL;
@@ -6495,8 +6525,9 @@ void md_do_sync(mddev_t *mddev)
  skip:
        mddev->curr_resync = 0;
        mddev->curr_resync_completed = 0;
-       mddev->resync_min = 0;
-       mddev->resync_max = MaxSector;
+       if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery))
+               /* We completed so max setting can be forgotten. */
+               mddev->resync_max = MaxSector;
        sysfs_notify(&mddev->kobj, NULL, "sync_completed");
        wake_up(&resync_wait);
        set_bit(MD_RECOVERY_DONE, &mddev->recovery);
@@ -6745,7 +6776,7 @@ void md_check_recovery(mddev_t *mddev)
                        }
                        mddev->sync_thread = md_register_thread(md_do_sync,
                                                                mddev,
-                                                               "%s_resync");
+                                                               "resync");
                        if (!mddev->sync_thread) {
                                printk(KERN_ERR "%s: could not start resync"
                                        " thread...\n",