md/raid5: allow the stripe_cache to grow and shrink.

The default setting of 256 stripe_heads is probably
much too small for many configurations.  So it is best to make it
auto-configure.

Shrinking the cache under memory pressure is easy.  The only
interesting part here is that we put a fairly high cost
('seeks') on shrinking the cache as the cost is greater than
just having to read more data, it reduces parallelism.

Growing the cache on demand needs to be done carefully.  If we allow
fast growth, that can upset memory balance as lots of dirty memory can
quickly turn into lots of memory queued in the stripe_cache.
It is important for the raid5 block device to appear congested to
allow write-throttling to work.

So we only add stripes slowly. We set a flag when an allocation
fails because all stripes are in use, allocate at a convenient
time when that flag is set, and don't allow it to be set again
until at least one stripe_head has been released for re-use.

This means that a spurt of requests will only cause one stripe_head
to be allocated, but a steady stream of requests will slowly
increase the cache size - until memory pressure puts it back again.

It could take hours to reach a steady state.

The value written to, and displayed in, stripe_cache_size is
used as a minimum.  The cache can grow above this and shrink back
down to it.  The actual size is not directly visible, though it can
be deduced to some extent by watching stripe_cache_active.

Signed-off-by: NeilBrown <neilb@suse.de>

diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index b7cd32e7f29ecfb18c7bfd03c184797c43661c85..9716319cc477b8e665eca5577a821f7ce10c04a8 100644 (file)
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -672,8 +672,13 @@ get_active_stripe(struct r5conf *conf, sector_t sector,
                                    *(conf->hash_locks + hash));
                sh = __find_stripe(conf, sector, conf->generation - previous);
                if (!sh) {
-                       if (!test_bit(R5_INACTIVE_BLOCKED, &conf->cache_state))
+                       if (!test_bit(R5_INACTIVE_BLOCKED, &conf->cache_state)) {
                                sh = get_free_stripe(conf, hash);
+                               if (!sh && llist_empty(&conf->released_stripes) &&
+                                   !test_bit(R5_DID_ALLOC, &conf->cache_state))
+                                       set_bit(R5_ALLOC_MORE,
+                                               &conf->cache_state);
+                       }
                        if (noblock && sh == NULL)
                                break;
                        if (!sh) {
@@ -5761,6 +5766,8 @@ static void raid5d(struct md_thread *thread)
                int batch_size, released;
 
                released = release_stripe_list(conf, conf->temp_inactive_list);
+               if (released)
+                       clear_bit(R5_DID_ALLOC, &conf->cache_state);
 
                if (
                    !list_empty(&conf->bitmap_list)) {
@@ -5799,6 +5806,13 @@ static void raid5d(struct md_thread *thread)
        pr_debug("%d stripes handled\n", handled);
 
        spin_unlock_irq(&conf->device_lock);
+       if (test_and_clear_bit(R5_ALLOC_MORE, &conf->cache_state)) {
+               grow_one_stripe(conf, __GFP_NOWARN);
+               /* Set flag even if allocation failed.  This helps
+                * slow down allocation requests when mem is short
+                */
+               set_bit(R5_DID_ALLOC, &conf->cache_state);
+       }
 
        async_tx_issue_pending_all();
        blk_finish_plug(&plug);
@@ -5814,7 +5828,7 @@ raid5_show_stripe_cache_size(struct mddev *mddev, char *page)
        spin_lock(&mddev->lock);
        conf = mddev->private;
        if (conf)
-               ret = sprintf(page, "%d\n", conf->max_nr_stripes);
+               ret = sprintf(page, "%d\n", conf->min_nr_stripes);
        spin_unlock(&mddev->lock);
        return ret;
 }
@@ -5828,10 +5842,12 @@ raid5_set_cache_size(struct mddev *mddev, int size)
        if (size <= 16 || size > 32768)
                return -EINVAL;
 
+       conf->min_nr_stripes = size;
        while (size < conf->max_nr_stripes &&
               drop_one_stripe(conf))
                ;
 
+
        err = md_allow_write(mddev);
        if (err)
                return err;
@@ -5947,7 +5963,7 @@ raid5_store_preread_threshold(struct mddev *mddev, const char *page, size_t len)
        conf = mddev->private;
        if (!conf)
                err = -ENODEV;
-       else if (new > conf->max_nr_stripes)
+       else if (new > conf->min_nr_stripes)
                err = -EINVAL;
        else
                conf->bypass_threshold = new;
@@ -6228,6 +6244,8 @@ static void raid5_free_percpu(struct r5conf *conf)
 
 static void free_conf(struct r5conf *conf)
 {
+       if (conf->shrinker.seeks)
+               unregister_shrinker(&conf->shrinker);
        free_thread_groups(conf);
        shrink_stripes(conf);
        raid5_free_percpu(conf);
@@ -6295,6 +6313,30 @@ static int raid5_alloc_percpu(struct r5conf *conf)
        return err;
 }
 
+static unsigned long raid5_cache_scan(struct shrinker *shrink,
+                                     struct shrink_control *sc)
+{
+       struct r5conf *conf = container_of(shrink, struct r5conf, shrinker);
+       int ret = 0;
+       while (ret < sc->nr_to_scan) {
+               if (drop_one_stripe(conf) == 0)
+                       return SHRINK_STOP;
+               ret++;
+       }
+       return ret;
+}
+
+static unsigned long raid5_cache_count(struct shrinker *shrink,
+                                      struct shrink_control *sc)
+{
+       struct r5conf *conf = container_of(shrink, struct r5conf, shrinker);
+
+       if (conf->max_nr_stripes < conf->min_nr_stripes)
+               /* unlikely, but not impossible */
+               return 0;
+       return conf->max_nr_stripes - conf->min_nr_stripes;
+}
+
 static struct r5conf *setup_conf(struct mddev *mddev)
 {
        struct r5conf *conf;
@@ -6445,10 +6487,11 @@ static struct r5conf *setup_conf(struct mddev *mddev)
                conf->prev_algo = mddev->layout;
        }
 
-       memory = NR_STRIPES * (sizeof(struct stripe_head) +
+       conf->min_nr_stripes = NR_STRIPES;
+       memory = conf->min_nr_stripes * (sizeof(struct stripe_head) +
                 max_disks * ((sizeof(struct bio) + PAGE_SIZE))) / 1024;
        atomic_set(&conf->empty_inactive_list_nr, NR_STRIPE_HASH_LOCKS);
-       if (grow_stripes(conf, NR_STRIPES)) {
+       if (grow_stripes(conf, conf->min_nr_stripes)) {
                printk(KERN_ERR
                       "md/raid:%s: couldn't allocate %dkB for buffers\n",
                       mdname(mddev), memory);
@@ -6456,6 +6499,17 @@ static struct r5conf *setup_conf(struct mddev *mddev)
        } else
                printk(KERN_INFO "md/raid:%s: allocated %dkB\n",
                       mdname(mddev), memory);
+       /*
+        * Losing a stripe head costs more than the time to refill it,
+        * it reduces the queue depth and so can hurt throughput.
+        * So set it rather large, scaled by number of devices.
+        */
+       conf->shrinker.seeks = DEFAULT_SEEKS * conf->raid_disks * 4;
+       conf->shrinker.scan_objects = raid5_cache_scan;
+       conf->shrinker.count_objects = raid5_cache_count;
+       conf->shrinker.batch = 128;
+       conf->shrinker.flags = 0;
+       register_shrinker(&conf->shrinker);
 
        sprintf(pers_name, "raid%d", mddev->new_level);
        conf->thread = md_register_thread(raid5d, mddev, pers_name);
@@ -7097,9 +7151,9 @@ static int check_stripe_cache(struct mddev *mddev)
         */
        struct r5conf *conf = mddev->private;
        if (((mddev->chunk_sectors << 9) / STRIPE_SIZE) * 4
-           > conf->max_nr_stripes ||
+           > conf->min_nr_stripes ||
            ((mddev->new_chunk_sectors << 9) / STRIPE_SIZE) * 4
-           > conf->max_nr_stripes) {
+           > conf->min_nr_stripes) {
                printk(KERN_WARNING "md/raid:%s: reshape: not enough stripes.  Needed %lu\n",
                       mdname(mddev),
                       ((max(mddev->chunk_sectors, mddev->new_chunk_sectors) << 9)

diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
index ebe4e24bc14d3dc82185ac60cfaee4277ef7730f..7dc0dd86074b1702276ccb51ba166a38d5d0f7e3 100644 (file)
--- a/drivers/md/raid5.h
+++ b/drivers/md/raid5.h
@@ -433,6 +433,7 @@ struct r5conf {
        int                     max_degraded;
        int                     raid_disks;
        int                     max_nr_stripes;
+       int                     min_nr_stripes;
 
        /* reshape_progress is the leading edge of a 'reshape'
         * It has value MaxSector when no reshape is happening
@@ -513,7 +514,15 @@ struct r5conf {
 #define R5_INACTIVE_BLOCKED    1       /* release of inactive stripes blocked,
                                         * waiting for 25% to be free
                                         */
-
+#define R5_ALLOC_MORE          2       /* It might help to allocate another
+                                        * stripe.
+                                        */
+#define R5_DID_ALLOC           4       /* A stripe was allocated, don't allocate
+                                        * more until at least one has been
+                                        * released.  This avoids flooding
+                                        * the cache.
+                                        */
+       struct shrinker         shrinker;
        int                     pool_size; /* number of disks in stripeheads in pool */
        spinlock_t              device_lock;
        struct disk_info        *disks;