bio: allow individual slabs in the bio_set

Instead of having a global bio slab cache, add a reference to one
in each bio_set that is created. This allows for personalized slabs
in each bio_set, so that they can have bios of different sizes.

This means we can personalize the bios we return. File systems may
want to embed the bio inside another structure, to avoid allocation
more items (and stuffing them in ->bi_private) after the get a bio.
Or we may want to embed a number of bio_vecs directly at the end
of a bio, to avoid doing two allocations to return a bio. This is now
possible.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

6 files changed, 170 insertions, 35 deletions

diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index ce26c84af06..3326750ec02 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -1060,7 +1060,7 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 		goto bad_page_pool;
 	}
 
-	cc->bs = bioset_create(MIN_IOS, MIN_IOS);
+	cc->bs = bioset_create(MIN_IOS, 0);
 	if (!cc->bs) {
 		ti->error = "Cannot allocate crypt bioset";
 		goto bad_bs;
diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
index 2fd6d445063..a34338567a2 100644
--- a/drivers/md/dm-io.c
+++ b/drivers/md/dm-io.c
@@ -56,7 +56,7 @@ struct dm_io_client *dm_io_client_create(unsigned num_pages)
 	if (!client->pool)
 		goto bad;
 
-	client->bios = bioset_create(16, 16);
+	client->bios = bioset_create(16, 0);
 	if (!client->bios)
 		goto bad;
 
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 343094c3fee..421c9f02d8c 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -1093,7 +1093,7 @@ static struct mapped_device *alloc_dev(int minor)
 	if (!md->tio_pool)
 		goto bad_tio_pool;
 
-	md->bs = bioset_create(16, 16);
+	md->bs = bioset_create(16, 0);
 	if (!md->bs)
 		goto bad_no_bioset;
 
diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c
index 19caf7c962a..77ebc3c263d 100644
--- a/fs/bio-integrity.c
+++ b/fs/bio-integrity.c
@@ -111,7 +111,7 @@ void bio_integrity_free(struct bio *bio, struct bio_set *bs)
 	    && bip->bip_buf != NULL)
 		kfree(bip->bip_buf);
 
-	mempool_free(bip->bip_vec, bs->bvec_pools[bip->bip_pool]);
+	bvec_free_bs(bs, bip->bip_vec, bip->bip_pool);
 	mempool_free(bip, bs->bio_integrity_pool);
 
 	bio->bi_integrity = NULL;
diff --git a/fs/bio.c b/fs/bio.c
index 1ab8986b041..0146f80789e 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -31,8 +31,6 @@
 
 DEFINE_TRACE(block_split);
 
-static struct kmem_cache *bio_slab __read_mostly;
-
 static mempool_t *bio_split_pool __read_mostly;
 
 /*
@@ -40,9 +38,8 @@ static mempool_t *bio_split_pool __read_mostly;
  * break badly! cannot be bigger than what you can fit into an
  * unsigned short
  */
-
 #define BV(x) { .nr_vecs = x, .name = "biovec-"__stringify(x) }
-static struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = {
+struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = {
 	BV(1), BV(4), BV(16), BV(64), BV(128), BV(BIO_MAX_PAGES),
 };
 #undef BV
@@ -53,11 +50,119 @@ static struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = {
  */
 struct bio_set *fs_bio_set;
 
+/*
+ * Our slab pool management
+ */
+struct bio_slab {
+	struct kmem_cache *slab;
+	unsigned int slab_ref;
+	unsigned int slab_size;
+	char name[8];
+};
+static DEFINE_MUTEX(bio_slab_lock);
+static struct bio_slab *bio_slabs;
+static unsigned int bio_slab_nr, bio_slab_max;
+
+static struct kmem_cache *bio_find_or_create_slab(unsigned int extra_size)
+{
+	unsigned int sz = sizeof(struct bio) + extra_size;
+	struct kmem_cache *slab = NULL;
+	struct bio_slab *bslab;
+	unsigned int i, entry = -1;
+
+	mutex_lock(&bio_slab_lock);
+
+	i = 0;
+	while (i < bio_slab_nr) {
+		struct bio_slab *bslab = &bio_slabs[i];
+
+		if (!bslab->slab && entry == -1)
+			entry = i;
+		else if (bslab->slab_size == sz) {
+			slab = bslab->slab;
+			bslab->slab_ref++;
+			break;
+		}
+		i++;
+	}
+
+	if (slab)
+		goto out_unlock;
+
+	if (bio_slab_nr == bio_slab_max && entry == -1) {
+		bio_slab_max <<= 1;
+		bio_slabs = krealloc(bio_slabs,
+				     bio_slab_max * sizeof(struct bio_slab),
+				     GFP_KERNEL);
+		if (!bio_slabs)
+			goto out_unlock;
+	}
+	if (entry == -1)
+		entry = bio_slab_nr++;
+
+	bslab = &bio_slabs[entry];
+
+	snprintf(bslab->name, sizeof(bslab->name), "bio-%d", entry);
+	slab = kmem_cache_create(bslab->name, sz, 0, SLAB_HWCACHE_ALIGN, NULL);
+	if (!slab)
+		goto out_unlock;
+
+	printk("bio: create slab <%s> at %d\n", bslab->name, entry);
+	bslab->slab = slab;
+	bslab->slab_ref = 1;
+	bslab->slab_size = sz;
+out_unlock:
+	mutex_unlock(&bio_slab_lock);
+	return slab;
+}
+
+static void bio_put_slab(struct bio_set *bs)
+{
+	struct bio_slab *bslab = NULL;
+	unsigned int i;
+
+	mutex_lock(&bio_slab_lock);
+
+	for (i = 0; i < bio_slab_nr; i++) {
+		if (bs->bio_slab == bio_slabs[i].slab) {
+			bslab = &bio_slabs[i];
+			break;
+		}
+	}
+
+	if (WARN(!bslab, KERN_ERR "bio: unable to find slab!\n"))
+		goto out;
+
+	WARN_ON(!bslab->slab_ref);
+
+	if (--bslab->slab_ref)
+		goto out;
+
+	kmem_cache_destroy(bslab->slab);
+	bslab->slab = NULL;
+
+out:
+	mutex_unlock(&bio_slab_lock);
+}
+
 unsigned int bvec_nr_vecs(unsigned short idx)
 {
 	return bvec_slabs[idx].nr_vecs;
 }
 
+void bvec_free_bs(struct bio_set *bs, struct bio_vec *bv, unsigned int idx)
+{
+	BIO_BUG_ON(idx >= BIOVEC_NR_POOLS);
+
+	if (idx == BIOVEC_MAX_IDX)
+		mempool_free(bv, bs->bvec_pool);
+	else {
+		struct biovec_slab *bvs = bvec_slabs + idx;
+
+		kmem_cache_free(bvs->slab, bv);
+	}
+}
+
 struct bio_vec *bvec_alloc_bs(gfp_t gfp_mask, int nr, unsigned long *idx,
 			      struct bio_set *bs)
 {
@@ -134,24 +239,22 @@ fallback:
 
 void bio_free(struct bio *bio, struct bio_set *bs)
 {
-	if (bio->bi_io_vec) {
-		const int pool_idx = BIO_POOL_IDX(bio);
+	void *p;
 
-		BIO_BUG_ON(pool_idx >= BIOVEC_NR_POOLS);
-
-		if (pool_idx == BIOVEC_MAX_IDX)
-			mempool_free(bio->bi_io_vec, bs->bvec_pool);
-		else {
-			struct biovec_slab *bvs = bvec_slabs + pool_idx;
-
-			kmem_cache_free(bvs->slab, bio->bi_io_vec);
-		}
-	}
+	if (bio->bi_io_vec)
+		bvec_free_bs(bs, bio->bi_io_vec, BIO_POOL_IDX(bio));
 
 	if (bio_integrity(bio))
 		bio_integrity_free(bio, bs);
 
-	mempool_free(bio, bs->bio_pool);
+	/*
+	 * If we have front padding, adjust the bio pointer before freeing
+	 */
+	p = bio;
+	if (bs->front_pad)
+		p -= bs->front_pad;
+
+	mempool_free(p, bs->bio_pool);
 }
 
 /*
@@ -188,16 +291,20 @@ void bio_init(struct bio *bio)
  *   for a &struct bio to become free. If a %NULL @bs is passed in, we will
  *   fall back to just using @kmalloc to allocate the required memory.
  *
- *   allocate bio and iovecs from the memory pools specified by the
- *   bio_set structure, or @kmalloc if none given.
+ *   Note that the caller must set ->bi_destructor on succesful return
+ *   of a bio, to do the appropriate freeing of the bio once the reference
+ *   count drops to zero.
  **/
 struct bio *bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set *bs)
 {
-	struct bio *bio;
+	struct bio *bio = NULL;
+
+	if (bs) {
+		void *p = mempool_alloc(bs->bio_pool, gfp_mask);
 
-	if (bs)
-		bio = mempool_alloc(bs->bio_pool, gfp_mask);
-	else
+		if (p)
+			bio = p + bs->front_pad;
+	} else
 		bio = kmalloc(sizeof(*bio), gfp_mask);
 
 	if (likely(bio)) {
@@ -1398,11 +1505,25 @@ void bioset_free(struct bio_set *bs)
 
 	bioset_integrity_free(bs);
 	biovec_free_pools(bs);
+	bio_put_slab(bs);
 
 	kfree(bs);
 }
 
-struct bio_set *bioset_create(int bio_pool_size, int bvec_pool_size)
+/**
+ * bioset_create  - Create a bio_set
+ * @pool_size:	Number of bio and bio_vecs to cache in the mempool
+ * @front_pad:	Number of bytes to allocate in front of the returned bio
+ *
+ * Description:
+ *    Set up a bio_set to be used with @bio_alloc_bioset. Allows the caller
+ *    to ask for a number of bytes to be allocated in front of the bio.
+ *    Front pad allocation is useful for embedding the bio inside
+ *    another structure, to avoid allocating extra data to go with the bio.
+ *    Note that the bio must be embedded at the END of that structure always,
+ *    or things will break badly.
+ */
+struct bio_set *bioset_create(unsigned int pool_size, unsigned int front_pad)
 {
 	struct bio_set *bs;
 
@@ -1410,16 +1531,22 @@ struct bio_set *bioset_create(int bio_pool_size, int bvec_pool_size)
 	if (!bs)
 		return NULL;
 
-	bs->bio_slab = bio_slab;
+	bs->front_pad = front_pad;
 
-	bs->bio_pool = mempool_create_slab_pool(bio_pool_size, bs->bio_slab);
+	bs->bio_slab = bio_find_or_create_slab(front_pad);
+	if (!bs->bio_slab) {
+		kfree(bs);
+		return NULL;
+	}
+
+	bs->bio_pool = mempool_create_slab_pool(pool_size, bs->bio_slab);
 	if (!bs->bio_pool)
 		goto bad;
 
-	if (bioset_integrity_create(bs, bio_pool_size))
+	if (bioset_integrity_create(bs, pool_size))
 		goto bad;
 
-	if (!biovec_create_pools(bs, bvec_pool_size))
+	if (!biovec_create_pools(bs, pool_size))
 		return bs;
 
 bad:
@@ -1443,12 +1570,16 @@ static void __init biovec_init_slabs(void)
 
 static int __init init_bio(void)
 {
-	bio_slab = KMEM_CACHE(bio, SLAB_HWCACHE_ALIGN|SLAB_PANIC);
+	bio_slab_max = 2;
+	bio_slab_nr = 0;
+	bio_slabs = kzalloc(bio_slab_max * sizeof(struct bio_slab), GFP_KERNEL);
+	if (!bio_slabs)
+		panic("bio: can't allocate bios\n");
 
 	bio_integrity_init_slab();
 	biovec_init_slabs();
 
-	fs_bio_set = bioset_create(BIO_POOL_SIZE, 2);
+	fs_bio_set = bioset_create(BIO_POOL_SIZE, 0);
 	if (!fs_bio_set)
 		panic("bio: can't allocate bios\n");
 
diff --git a/include/linux/bio.h b/include/linux/bio.h
index 9340098d75d..4b80d3537f9 100644
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -334,7 +334,7 @@ struct bio_pair {
 extern struct bio_pair *bio_split(struct bio *bi, int first_sectors);
 extern void bio_pair_release(struct bio_pair *dbio);
 
-extern struct bio_set *bioset_create(int, int);
+extern struct bio_set *bioset_create(unsigned int, unsigned int);
 extern void bioset_free(struct bio_set *);
 
 extern struct bio *bio_alloc(gfp_t, int);
@@ -379,6 +379,7 @@ extern struct bio *bio_copy_user_iov(struct request_queue *,
 extern int bio_uncopy_user(struct bio *);
 void zero_fill_bio(struct bio *bio);
 extern struct bio_vec *bvec_alloc_bs(gfp_t, int, unsigned long *, struct bio_set *);
+extern void bvec_free_bs(struct bio_set *, struct bio_vec *, unsigned int);
 extern unsigned int bvec_nr_vecs(unsigned short idx);
 
 /*
@@ -401,6 +402,8 @@ static inline void bio_set_completion_cpu(struct bio *bio, unsigned int cpu)
 
 struct bio_set {
 	struct kmem_cache *bio_slab;
+	unsigned int front_pad;
+
 	mempool_t *bio_pool;
 #if defined(CONFIG_BLK_DEV_INTEGRITY)
 	mempool_t *bio_integrity_pool;
@@ -415,6 +418,7 @@ struct biovec_slab {
 };
 
 extern struct bio_set *fs_bio_set;
+extern struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly;
 
 /*
  * a small number of entries is fine, not going to be performance critical.