md: add 'size' as a personality method

In preparation for giving userspace control over ->array_sectors we need
to be able to retrieve the 'default' size, and the 'anticipated' size
when a reshape is requested.  For personalities that do not reshape emit
a warning if anything but the default size is requested.

In the raid5 case we need to update ->previous_raid_disks to make the
new 'default' size available.

Reviewed-by: Andre Noll <maan@systemlinux.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>

diff --git a/drivers/md/faulty.c b/drivers/md/faulty.c
index 18793c137278113b818ed875940e960d9ef63a72..d76c87f2a8318f0ef44100ae8f8d65e7843d311c 100644 (file)
--- a/drivers/md/faulty.c
+++ b/drivers/md/faulty.c
@@ -283,6 +283,17 @@ static int reconfig(mddev_t *mddev, int layout, int chunk_size)
        return 0;
 }
 
+static sector_t faulty_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+{
+       WARN_ONCE(raid_disks,
+                 "%s does not support generic reshape\n", __func__);
+
+       if (sectors == 0)
+               return mddev->dev_sectors;
+
+       return sectors;
+}
+
 static int run(mddev_t *mddev)
 {
        mdk_rdev_t *rdev;
@@ -301,7 +312,7 @@ static int run(mddev_t *mddev)
        list_for_each_entry(rdev, &mddev->disks, same_set)
                conf->rdev = rdev;
 
-       mddev->array_sectors = mddev->dev_sectors;
+       mddev->array_sectors = faulty_size(mddev, 0, 0);
        mddev->private = conf;
 
        reconfig(mddev, mddev->layout, -1);
@@ -328,6 +339,7 @@ static struct mdk_personality faulty_personality =
        .stop           = stop,
        .status         = status,
        .reconfig       = reconfig,
+       .size           = faulty_size,
 };
 
 static int __init raid_init(void)

diff --git a/drivers/md/linear.c b/drivers/md/linear.c
index d5d99290041a82840b0c61768df1271eb2e78587..b6bb976123daaa0cf765818f07e5c6a1549570a0 100644 (file)
--- a/drivers/md/linear.c
+++ b/drivers/md/linear.c
@@ -101,6 +101,16 @@ static int linear_congested(void *data, int bits)
        return ret;
 }
 
+static sector_t linear_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+{
+       linear_conf_t *conf = mddev_to_conf(mddev);
+
+       WARN_ONCE(sectors || raid_disks,
+                 "%s does not support generic reshape\n", __func__);
+
+       return conf->array_sectors;
+}
+
 static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
 {
        linear_conf_t *conf;
@@ -253,7 +263,7 @@ static int linear_run (mddev_t *mddev)
        if (!conf)
                return 1;
        mddev->private = conf;
-       mddev->array_sectors = conf->array_sectors;
+       mddev->array_sectors = linear_size(mddev, 0, 0);
 
        blk_queue_merge_bvec(mddev->queue, linear_mergeable_bvec);
        mddev->queue->unplug_fn = linear_unplug;
@@ -287,7 +297,7 @@ static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
        newconf->prev = mddev_to_conf(mddev);
        mddev->private = newconf;
        mddev->raid_disks++;
-       mddev->array_sectors = newconf->array_sectors;
+       mddev->array_sectors = linear_size(mddev, 0, 0);
        set_capacity(mddev->gendisk, mddev->array_sectors);
        return 0;
 }
@@ -385,6 +395,7 @@ static struct mdk_personality linear_personality =
        .stop           = linear_stop,
        .status         = linear_status,
        .hot_add_disk   = linear_add,
+       .size           = linear_size,
 };
 
 static int __init linear_init (void)

diff --git a/drivers/md/md.h b/drivers/md/md.h
index 8034f62a9d28a207e0e69fabd907b74c65f916a3..d2c50da1ae7fd47f874a071258b6859498addba8 100644 (file)
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -312,6 +312,7 @@ struct mdk_personality
        int (*spare_active) (mddev_t *mddev);
        sector_t (*sync_request)(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster);
        int (*resize) (mddev_t *mddev, sector_t sectors);
+       sector_t (*size) (mddev_t *mddev, sector_t sectors, int raid_disks);
        int (*check_reshape) (mddev_t *mddev);
        int (*start_reshape) (mddev_t *mddev);
        int (*reconfig) (mddev_t *mddev, int layout, int chunk_size);

diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c
index 87accf74e4b4d12cafab31428e4d5810656f3fbe..a3f6d8668bb35af437514ea851dfee8173e3f96b 100644 (file)
--- a/drivers/md/multipath.c
+++ b/drivers/md/multipath.c
@@ -406,6 +406,14 @@ static void multipathd (mddev_t *mddev)
        spin_unlock_irqrestore(&conf->device_lock, flags);
 }
 
+static sector_t multipath_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+{
+       WARN_ONCE(sectors || raid_disks,
+                 "%s does not support generic reshape\n", __func__);
+
+       return mddev->dev_sectors;
+}
+
 static int multipath_run (mddev_t *mddev)
 {
        multipath_conf_t *conf;
@@ -502,7 +510,7 @@ static int multipath_run (mddev_t *mddev)
        /*
         * Ok, everything is just fine now
         */
-       mddev->array_sectors = mddev->dev_sectors;
+       mddev->array_sectors = multipath_size(mddev, 0, 0);
 
        mddev->queue->unplug_fn = multipath_unplug;
        mddev->queue->backing_dev_info.congested_fn = multipath_congested;
@@ -547,6 +555,7 @@ static struct mdk_personality multipath_personality =
        .error_handler  = multipath_error,
        .hot_add_disk   = multipath_add_disk,
        .hot_remove_disk= multipath_remove_disk,
+       .size           = multipath_size,
 };
 
 static int __init multipath_init (void)

diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index 9aebb4c93b60424edee8d6cd60301eeae7467534..bb0df6a18443c4291235f96406668ff8cb3bb94a 100644 (file)
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -263,12 +263,25 @@ static int raid0_mergeable_bvec(struct request_queue *q,
                return max;
 }
 
+static sector_t raid0_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+{
+       sector_t array_sectors = 0;
+       mdk_rdev_t *rdev;
+
+       WARN_ONCE(sectors || raid_disks,
+                 "%s does not support generic reshape\n", __func__);
+
+       list_for_each_entry(rdev, &mddev->disks, same_set)
+               array_sectors += rdev->sectors;
+
+       return array_sectors;
+}
+
 static int raid0_run (mddev_t *mddev)
 {
        unsigned  cur=0, i=0, nb_zone;
        s64 sectors;
        raid0_conf_t *conf;
-       mdk_rdev_t *rdev;
 
        if (mddev->chunk_size == 0) {
                printk(KERN_ERR "md/raid0: non-zero chunk size required.\n");
@@ -293,9 +306,7 @@ static int raid0_run (mddev_t *mddev)
                goto out_free_conf;
 
        /* calculate array device size */
-       mddev->array_sectors = 0;
-       list_for_each_entry(rdev, &mddev->disks, same_set)
-               mddev->array_sectors += rdev->sectors;
+       mddev->array_sectors = raid0_size(mddev, 0, 0);
 
        printk(KERN_INFO "raid0 : md_size is %llu sectors.\n",
                (unsigned long long)mddev->array_sectors);
@@ -511,6 +522,7 @@ static struct mdk_personality raid0_personality=
        .run            = raid0_run,
        .stop           = raid0_stop,
        .status         = raid0_status,
+       .size           = raid0_size,
 };
 
 static int __init raid0_init (void)

diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 7eaca32093649be95f3855cefdf51706ec6a1346..a2a83b0d5a95fbcf3e320de16d77a2bbea851329 100644 (file)
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1922,6 +1922,14 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
        return nr_sectors;
 }
 
+static sector_t raid1_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+{
+       if (sectors)
+               return sectors;
+
+       return mddev->dev_sectors;
+}
+
 static int run(mddev_t *mddev)
 {
        conf_t *conf;
@@ -2051,7 +2059,7 @@ static int run(mddev_t *mddev)
        /*
         * Ok, everything is just fine now
         */
-       mddev->array_sectors = mddev->dev_sectors;
+       mddev->array_sectors = raid1_size(mddev, 0, 0);
 
        mddev->queue->unplug_fn = raid1_unplug;
        mddev->queue->backing_dev_info.congested_fn = raid1_congested;
@@ -2116,7 +2124,7 @@ static int raid1_resize(mddev_t *mddev, sector_t sectors)
         * any io in the removed space completes, but it hardly seems
         * worth it.
         */
-       mddev->array_sectors = sectors;
+       mddev->array_sectors = raid1_size(mddev, sectors, 0);
        set_capacity(mddev->gendisk, mddev->array_sectors);
        mddev->changed = 1;
        if (mddev->array_sectors > mddev->dev_sectors &&
@@ -2270,6 +2278,7 @@ static struct mdk_personality raid1_personality =
        .spare_active   = raid1_spare_active,
        .sync_request   = sync_request,
        .resize         = raid1_resize,
+       .size           = raid1_size,
        .check_reshape  = raid1_reshape,
        .quiesce        = raid1_quiesce,
 };

diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index c2059e25d03f1ed28db23e1f59b877a512e04dc1..5bf1b24ae802d9e64e5fe82534268d9865b5059d 100644 (file)
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -2023,6 +2023,25 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
        goto skipped;
 }
 
+static sector_t
+raid10_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+{
+       sector_t size;
+       conf_t *conf = mddev_to_conf(mddev);
+
+       if (!raid_disks)
+               raid_disks = mddev->raid_disks;
+       if (!sectors)
+               sectors = mddev->dev_sectors;
+
+       size = sectors >> conf->chunk_shift;
+       sector_div(size, conf->far_copies);
+       size = size * raid_disks;
+       sector_div(size, conf->near_copies);
+
+       return size << conf->chunk_shift;
+}
+
 static int run(mddev_t *mddev)
 {
        conf_t *conf;
@@ -2174,8 +2193,8 @@ static int run(mddev_t *mddev)
        /*
         * Ok, everything is just fine now
         */
-       mddev->array_sectors = size << conf->chunk_shift;
-       mddev->resync_max_sectors = size << conf->chunk_shift;
+       mddev->array_sectors = raid10_size(mddev, 0, 0);
+       mddev->resync_max_sectors = mddev->array_sectors;
 
        mddev->queue->unplug_fn = raid10_unplug;
        mddev->queue->backing_dev_info.congested_fn = raid10_congested;
@@ -2261,6 +2280,7 @@ static struct mdk_personality raid10_personality =
        .spare_active   = raid10_spare_active,
        .sync_request   = sync_request,
        .quiesce        = raid10_quiesce,
+       .size           = raid10_size,
 };
 
 static int __init raid_init(void)

diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 70b50af2bcd7d33491aa4dbf73185cb29c22c319..2cd619ff07614cef56d781f31ebd23c76bf76a54 100644 (file)
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -4166,6 +4166,20 @@ static struct attribute_group raid5_attrs_group = {
        .attrs = raid5_attrs,
 };
 
+static sector_t
+raid5_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+{
+       raid5_conf_t *conf = mddev_to_conf(mddev);
+
+       if (!sectors)
+               sectors = mddev->dev_sectors;
+       if (!raid_disks)
+               raid_disks = conf->previous_raid_disks;
+
+       sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
+       return sectors * (raid_disks - conf->max_degraded);
+}
+
 static raid5_conf_t *setup_conf(mddev_t *mddev)
 {
        raid5_conf_t *conf;
@@ -4460,8 +4474,7 @@ static int run(mddev_t *mddev)
        mddev->queue->backing_dev_info.congested_data = mddev;
        mddev->queue->backing_dev_info.congested_fn = raid5_congested;
 
-       mddev->array_sectors = mddev->dev_sectors *
-               (conf->previous_raid_disks - conf->max_degraded);
+       mddev->array_sectors = raid5_size(mddev, 0, 0);
 
        blk_queue_merge_bvec(mddev->queue, raid5_mergeable_bvec);
 
@@ -4684,11 +4697,8 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors)
         * any io in the removed space completes, but it hardly seems
         * worth it.
         */
-       raid5_conf_t *conf = mddev_to_conf(mddev);
-
        sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
-       mddev->array_sectors = sectors * (mddev->raid_disks
-                                         - conf->max_degraded);
+       mddev->array_sectors = raid5_size(mddev, sectors, mddev->raid_disks);
        set_capacity(mddev->gendisk, mddev->array_sectors);
        mddev->changed = 1;
        if (sectors > mddev->dev_sectors && mddev->recovery_cp == MaxSector) {
@@ -4824,10 +4834,12 @@ static void end_reshape(raid5_conf_t *conf)
        struct block_device *bdev;
 
        if (!test_bit(MD_RECOVERY_INTR, &conf->mddev->recovery)) {
-               conf->mddev->array_sectors = conf->mddev->dev_sectors *
-                       (conf->raid_disks - conf->max_degraded);
-               set_capacity(conf->mddev->gendisk, conf->mddev->array_sectors);
-               conf->mddev->changed = 1;
+               mddev_t *mddev = conf->mddev;
+
+               mddev->array_sectors = raid5_size(mddev, 0, conf->raid_disks);
+               set_capacity(mddev->gendisk, mddev->array_sectors);
+               mddev->changed = 1;
+               conf->previous_raid_disks = conf->raid_disks;
 
                bdev = bdget_disk(conf->mddev->gendisk, 0);
                if (bdev) {
@@ -5076,6 +5088,7 @@ static struct mdk_personality raid6_personality =
        .spare_active   = raid5_spare_active,
        .sync_request   = sync_request,
        .resize         = raid5_resize,
+       .size           = raid5_size,
 #ifdef CONFIG_MD_RAID5_RESHAPE
        .check_reshape  = raid5_check_reshape,
        .start_reshape  = raid5_start_reshape,
@@ -5098,6 +5111,7 @@ static struct mdk_personality raid5_personality =
        .spare_active   = raid5_spare_active,
        .sync_request   = sync_request,
        .resize         = raid5_resize,
+       .size           = raid5_size,
 #ifdef CONFIG_MD_RAID5_RESHAPE
        .check_reshape  = raid5_check_reshape,
        .start_reshape  = raid5_start_reshape,
@@ -5122,6 +5136,7 @@ static struct mdk_personality raid4_personality =
        .spare_active   = raid5_spare_active,
        .sync_request   = sync_request,
        .resize         = raid5_resize,
+       .size           = raid5_size,
 #ifdef CONFIG_MD_RAID5_RESHAPE
        .check_reshape  = raid5_check_reshape,
        .start_reshape  = raid5_start_reshape,