Merge branch 'for-3.11/drivers' of git://git.kernel.dk/linux-block

Pull block IO driver bits from Jens Axboe:
 "As I mentioned in the core block pull request, due to real life
  circumstances the driver pull request would be late.  Now it looks
  like -rc2 late...  On the plus side, apart form the rsxx update, these
  are all things that I could argue could go in later in the cycle as
  they are fixes and not features.  So even though things are late, it's
  not ALL bad.

  The pull request contains:

   - Updates to bcache, all bug fixes, from Kent.

   - A pile of drbd bug fixes (no big features this time!).

   - xen blk front/back fixes.

   - rsxx driver updates, some of them deferred form 3.10.  So should be
     well cooked by now"

* 'for-3.11/drivers' of git://git.kernel.dk/linux-block: (63 commits)
  bcache: Allocation kthread fixes
  bcache: Fix GC_SECTORS_USED() calculation
  bcache: Journal replay fix
  bcache: Shutdown fix
  bcache: Fix a sysfs splat on shutdown
  bcache: Advertise that flushes are supported
  bcache: check for allocation failures
  bcache: Fix a dumb race
  bcache: Use standard utility code
  bcache: Update email address
  bcache: Delete fuzz tester
  bcache: Document shrinker reserve better
  bcache: FUA fixes
  drbd: Allow online change of al-stripes and al-stripe-size
  drbd: Constants should be UPPERCASE
  drbd: Ignore the exit code of a fence-peer handler if it returns too late
  drbd: Fix rcu_read_lock balance on error path
  drbd: fix error return code in drbd_init()
  drbd: Do not sleep inside rcu
  bcache: Refresh usage docs
  ...

6 files changed, 346 insertions, 110 deletions

diff --git a/include/linux/drbd.h b/include/linux/drbd.h
index 1b4d4ee1168f..de7d74ab3de6 100644
--- a/include/linux/drbd.h
+++ b/include/linux/drbd.h
@@ -177,7 +177,11 @@ enum drbd_ret_code {
 	ERR_NEED_APV_100	= 163,
 	ERR_NEED_ALLOW_TWO_PRI  = 164,
 	ERR_MD_UNCLEAN          = 165,
-
+	ERR_MD_LAYOUT_CONNECTED = 166,
+	ERR_MD_LAYOUT_TOO_BIG   = 167,
+	ERR_MD_LAYOUT_TOO_SMALL = 168,
+	ERR_MD_LAYOUT_NO_FIT    = 169,
+	ERR_IMPLICIT_SHRINK     = 170,
 	/* insert new ones above this line */
 	AFTER_LAST_ERR_CODE
 };
diff --git a/include/linux/drbd_genl.h b/include/linux/drbd_genl.h
index d0d8fac8a6e4..e8c44572b8cb 100644
--- a/include/linux/drbd_genl.h
+++ b/include/linux/drbd_genl.h
@@ -181,6 +181,8 @@ GENL_struct(DRBD_NLA_RESIZE_PARMS, 7, resize_parms,
 	__u64_field(1, DRBD_GENLA_F_MANDATORY,	resize_size)
 	__flg_field(2, DRBD_GENLA_F_MANDATORY,	resize_force)
 	__flg_field(3, DRBD_GENLA_F_MANDATORY,	no_resync)
+	__u32_field_def(4, 0 /* OPTIONAL */, al_stripes, DRBD_AL_STRIPES_DEF)
+	__u32_field_def(5, 0 /* OPTIONAL */, al_stripe_size, DRBD_AL_STRIPE_SIZE_DEF)
 )
 
 GENL_struct(DRBD_NLA_STATE_INFO, 8, state_info,
diff --git a/include/linux/drbd_limits.h b/include/linux/drbd_limits.h
index 1fedf2b17cc8..17e50bb00521 100644
--- a/include/linux/drbd_limits.h
+++ b/include/linux/drbd_limits.h
@@ -215,4 +215,13 @@
 #define DRBD_ALWAYS_ASBP_DEF	0
 #define DRBD_USE_RLE_DEF	1
 
+#define DRBD_AL_STRIPES_MIN     1
+#define DRBD_AL_STRIPES_MAX     1024
+#define DRBD_AL_STRIPES_DEF     1
+#define DRBD_AL_STRIPES_SCALE   '1'
+
+#define DRBD_AL_STRIPE_SIZE_MIN   4
+#define DRBD_AL_STRIPE_SIZE_MAX   16777216
+#define DRBD_AL_STRIPE_SIZE_DEF   32
+#define DRBD_AL_STRIPE_SIZE_SCALE 'k' /* kilobytes */
 #endif
diff --git a/include/trace/events/bcache.h b/include/trace/events/bcache.h
index 3cc5a0b278c3..5ebda976ea93 100644
--- a/include/trace/events/bcache.h
+++ b/include/trace/events/bcache.h
@@ -9,9 +9,7 @@
 struct search;
 
 DECLARE_EVENT_CLASS(bcache_request,
-
 	TP_PROTO(struct search *s, struct bio *bio),
-
 	TP_ARGS(s, bio),
 
 	TP_STRUCT__entry(
@@ -22,7 +20,6 @@ DECLARE_EVENT_CLASS(bcache_request,
 		__field(dev_t,		orig_sector		)
 		__field(unsigned int,	nr_sector		)
 		__array(char,		rwbs,	6		)
-		__array(char,		comm,	TASK_COMM_LEN	)
 	),
 
 	TP_fast_assign(
@@ -33,36 +30,66 @@ DECLARE_EVENT_CLASS(bcache_request,
 		__entry->orig_sector	= bio->bi_sector - 16;
 		__entry->nr_sector	= bio->bi_size >> 9;
 		blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size);
-		memcpy(__entry->comm, current->comm, TASK_COMM_LEN);
 	),
 
-	TP_printk("%d,%d %s %llu + %u [%s] (from %d,%d @ %llu)",
+	TP_printk("%d,%d %s %llu + %u (from %d,%d @ %llu)",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
-		  __entry->rwbs,
-		  (unsigned long long)__entry->sector,
-		  __entry->nr_sector, __entry->comm,
-		  __entry->orig_major, __entry->orig_minor,
+		  __entry->rwbs, (unsigned long long)__entry->sector,
+		  __entry->nr_sector, __entry->orig_major, __entry->orig_minor,
 		  (unsigned long long)__entry->orig_sector)
 );
 
-DEFINE_EVENT(bcache_request, bcache_request_start,
+DECLARE_EVENT_CLASS(bkey,
+	TP_PROTO(struct bkey *k),
+	TP_ARGS(k),
 
-	TP_PROTO(struct search *s, struct bio *bio),
+	TP_STRUCT__entry(
+		__field(u32,	size				)
+		__field(u32,	inode				)
+		__field(u64,	offset				)
+		__field(bool,	dirty				)
+	),
 
-	TP_ARGS(s, bio)
+	TP_fast_assign(
+		__entry->inode	= KEY_INODE(k);
+		__entry->offset	= KEY_OFFSET(k);
+		__entry->size	= KEY_SIZE(k);
+		__entry->dirty	= KEY_DIRTY(k);
+	),
+
+	TP_printk("%u:%llu len %u dirty %u", __entry->inode,
+		  __entry->offset, __entry->size, __entry->dirty)
 );
 
-DEFINE_EVENT(bcache_request, bcache_request_end,
+DECLARE_EVENT_CLASS(btree_node,
+	TP_PROTO(struct btree *b),
+	TP_ARGS(b),
+
+	TP_STRUCT__entry(
+		__field(size_t,		bucket			)
+	),
 
+	TP_fast_assign(
+		__entry->bucket	= PTR_BUCKET_NR(b->c, &b->key, 0);
+	),
+
+	TP_printk("bucket %zu", __entry->bucket)
+);
+
+/* request.c */
+
+DEFINE_EVENT(bcache_request, bcache_request_start,
 	TP_PROTO(struct search *s, struct bio *bio),
+	TP_ARGS(s, bio)
+);
 
+DEFINE_EVENT(bcache_request, bcache_request_end,
+	TP_PROTO(struct search *s, struct bio *bio),
 	TP_ARGS(s, bio)
 );
 
 DECLARE_EVENT_CLASS(bcache_bio,
-
 	TP_PROTO(struct bio *bio),
-
 	TP_ARGS(bio),
 
 	TP_STRUCT__entry(
@@ -70,7 +97,6 @@ DECLARE_EVENT_CLASS(bcache_bio,
 		__field(sector_t,	sector			)
 		__field(unsigned int,	nr_sector		)
 		__array(char,		rwbs,	6		)
-		__array(char,		comm,	TASK_COMM_LEN	)
 	),
 
 	TP_fast_assign(
@@ -78,191 +104,328 @@ DECLARE_EVENT_CLASS(bcache_bio,
 		__entry->sector		= bio->bi_sector;
 		__entry->nr_sector	= bio->bi_size >> 9;
 		blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size);
-		memcpy(__entry->comm, current->comm, TASK_COMM_LEN);
 	),
 
-	TP_printk("%d,%d  %s %llu + %u [%s]",
-		  MAJOR(__entry->dev), MINOR(__entry->dev),
-		  __entry->rwbs,
-		  (unsigned long long)__entry->sector,
-		  __entry->nr_sector, __entry->comm)
+	TP_printk("%d,%d  %s %llu + %u",
+		  MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs,
+		  (unsigned long long)__entry->sector, __entry->nr_sector)
 );
 
-
-DEFINE_EVENT(bcache_bio, bcache_passthrough,
-
+DEFINE_EVENT(bcache_bio, bcache_bypass_sequential,
 	TP_PROTO(struct bio *bio),
+	TP_ARGS(bio)
+);
 
+DEFINE_EVENT(bcache_bio, bcache_bypass_congested,
+	TP_PROTO(struct bio *bio),
 	TP_ARGS(bio)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_cache_hit,
+TRACE_EVENT(bcache_read,
+	TP_PROTO(struct bio *bio, bool hit, bool bypass),
+	TP_ARGS(bio, hit, bypass),
 
-	TP_PROTO(struct bio *bio),
+	TP_STRUCT__entry(
+		__field(dev_t,		dev			)
+		__field(sector_t,	sector			)
+		__field(unsigned int,	nr_sector		)
+		__array(char,		rwbs,	6		)
+		__field(bool,		cache_hit		)
+		__field(bool,		bypass			)
+	),
 
-	TP_ARGS(bio)
+	TP_fast_assign(
+		__entry->dev		= bio->bi_bdev->bd_dev;
+		__entry->sector		= bio->bi_sector;
+		__entry->nr_sector	= bio->bi_size >> 9;
+		blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size);
+		__entry->cache_hit = hit;
+		__entry->bypass = bypass;
+	),
+
+	TP_printk("%d,%d  %s %llu + %u hit %u bypass %u",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->rwbs, (unsigned long long)__entry->sector,
+		  __entry->nr_sector, __entry->cache_hit, __entry->bypass)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_cache_miss,
+TRACE_EVENT(bcache_write,
+	TP_PROTO(struct bio *bio, bool writeback, bool bypass),
+	TP_ARGS(bio, writeback, bypass),
 
-	TP_PROTO(struct bio *bio),
+	TP_STRUCT__entry(
+		__field(dev_t,		dev			)
+		__field(sector_t,	sector			)
+		__field(unsigned int,	nr_sector		)
+		__array(char,		rwbs,	6		)
+		__field(bool,		writeback		)
+		__field(bool,		bypass			)
+	),
 
-	TP_ARGS(bio)
+	TP_fast_assign(
+		__entry->dev		= bio->bi_bdev->bd_dev;
+		__entry->sector		= bio->bi_sector;
+		__entry->nr_sector	= bio->bi_size >> 9;
+		blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size);
+		__entry->writeback = writeback;
+		__entry->bypass = bypass;
+	),
+
+	TP_printk("%d,%d  %s %llu + %u hit %u bypass %u",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->rwbs, (unsigned long long)__entry->sector,
+		  __entry->nr_sector, __entry->writeback, __entry->bypass)
 );
 
 DEFINE_EVENT(bcache_bio, bcache_read_retry,
-
 	TP_PROTO(struct bio *bio),
-
 	TP_ARGS(bio)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_writethrough,
+DEFINE_EVENT(bkey, bcache_cache_insert,
+	TP_PROTO(struct bkey *k),
+	TP_ARGS(k)
+);
 
-	TP_PROTO(struct bio *bio),
+/* Journal */
 
-	TP_ARGS(bio)
-);
+DECLARE_EVENT_CLASS(cache_set,
+	TP_PROTO(struct cache_set *c),
+	TP_ARGS(c),
 
-DEFINE_EVENT(bcache_bio, bcache_writeback,
+	TP_STRUCT__entry(
+		__array(char,		uuid,	16 )
+	),
 
-	TP_PROTO(struct bio *bio),
+	TP_fast_assign(
+		memcpy(__entry->uuid, c->sb.set_uuid, 16);
+	),
 
-	TP_ARGS(bio)
+	TP_printk("%pU", __entry->uuid)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_write_skip,
-
-	TP_PROTO(struct bio *bio),
+DEFINE_EVENT(bkey, bcache_journal_replay_key,
+	TP_PROTO(struct bkey *k),
+	TP_ARGS(k)
+);
 
-	TP_ARGS(bio)
+DEFINE_EVENT(cache_set, bcache_journal_full,
+	TP_PROTO(struct cache_set *c),
+	TP_ARGS(c)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_btree_read,
+DEFINE_EVENT(cache_set, bcache_journal_entry_full,
+	TP_PROTO(struct cache_set *c),
+	TP_ARGS(c)
+);
 
+DEFINE_EVENT(bcache_bio, bcache_journal_write,
 	TP_PROTO(struct bio *bio),
-
 	TP_ARGS(bio)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_btree_write,
+/* Btree */
 
-	TP_PROTO(struct bio *bio),
+DEFINE_EVENT(cache_set, bcache_btree_cache_cannibalize,
+	TP_PROTO(struct cache_set *c),
+	TP_ARGS(c)
+);
 
-	TP_ARGS(bio)
+DEFINE_EVENT(btree_node, bcache_btree_read,
+	TP_PROTO(struct btree *b),
+	TP_ARGS(b)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_write_dirty,
+TRACE_EVENT(bcache_btree_write,
+	TP_PROTO(struct btree *b),
+	TP_ARGS(b),
 
-	TP_PROTO(struct bio *bio),
+	TP_STRUCT__entry(
+		__field(size_t,		bucket			)
+		__field(unsigned,	block			)
+		__field(unsigned,	keys			)
+	),
 
-	TP_ARGS(bio)
+	TP_fast_assign(
+		__entry->bucket	= PTR_BUCKET_NR(b->c, &b->key, 0);
+		__entry->block	= b->written;
+		__entry->keys	= b->sets[b->nsets].data->keys;
+	),
+
+	TP_printk("bucket %zu", __entry->bucket)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_read_dirty,
+DEFINE_EVENT(btree_node, bcache_btree_node_alloc,
+	TP_PROTO(struct btree *b),
+	TP_ARGS(b)
+);
 
-	TP_PROTO(struct bio *bio),
+DEFINE_EVENT(btree_node, bcache_btree_node_alloc_fail,
+	TP_PROTO(struct btree *b),
+	TP_ARGS(b)
+);
 
-	TP_ARGS(bio)
+DEFINE_EVENT(btree_node, bcache_btree_node_free,
+	TP_PROTO(struct btree *b),
+	TP_ARGS(b)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_write_moving,
+TRACE_EVENT(bcache_btree_gc_coalesce,
+	TP_PROTO(unsigned nodes),
+	TP_ARGS(nodes),
 
-	TP_PROTO(struct bio *bio),
+	TP_STRUCT__entry(
+		__field(unsigned,	nodes			)
+	),
 
-	TP_ARGS(bio)
+	TP_fast_assign(
+		__entry->nodes	= nodes;
+	),
+
+	TP_printk("coalesced %u nodes", __entry->nodes)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_read_moving,
+DEFINE_EVENT(cache_set, bcache_gc_start,
+	TP_PROTO(struct cache_set *c),
+	TP_ARGS(c)
+);
 
-	TP_PROTO(struct bio *bio),
+DEFINE_EVENT(cache_set, bcache_gc_end,
+	TP_PROTO(struct cache_set *c),
+	TP_ARGS(c)
+);
 
-	TP_ARGS(bio)
+DEFINE_EVENT(bkey, bcache_gc_copy,
+	TP_PROTO(struct bkey *k),
+	TP_ARGS(k)
 );
 
-DEFINE_EVENT(bcache_bio, bcache_journal_write,
+DEFINE_EVENT(bkey, bcache_gc_copy_collision,
+	TP_PROTO(struct bkey *k),
+	TP_ARGS(k)
+);
 
-	TP_PROTO(struct bio *bio),
+TRACE_EVENT(bcache_btree_insert_key,
+	TP_PROTO(struct btree *b, struct bkey *k, unsigned op, unsigned status),
+	TP_ARGS(b, k, op, status),
 
-	TP_ARGS(bio)
-);
+	TP_STRUCT__entry(
+		__field(u64,	btree_node			)
+		__field(u32,	btree_level			)
+		__field(u32,	inode				)
+		__field(u64,	offset				)
+		__field(u32,	size				)
+		__field(u8,	dirty				)
+		__field(u8,	op				)
+		__field(u8,	status				)
+	),
 
-DECLARE_EVENT_CLASS(bcache_cache_bio,
+	TP_fast_assign(
+		__entry->btree_node = PTR_BUCKET_NR(b->c, &b->key, 0);
+		__entry->btree_level = b->level;
+		__entry->inode	= KEY_INODE(k);
+		__entry->offset	= KEY_OFFSET(k);
+		__entry->size	= KEY_SIZE(k);
+		__entry->dirty	= KEY_DIRTY(k);
+		__entry->op = op;
+		__entry->status = status;
+	),
 
-	TP_PROTO(struct bio *bio,
-		 sector_t orig_sector,
-		 struct block_device* orig_bdev),
+	TP_printk("%u for %u at %llu(%u): %u:%llu len %u dirty %u",
+		  __entry->status, __entry->op,
+		  __entry->btree_node, __entry->btree_level,
+		  __entry->inode, __entry->offset,
+		  __entry->size, __entry->dirty)
+);
 
-	TP_ARGS(bio, orig_sector, orig_bdev),
+DECLARE_EVENT_CLASS(btree_split,
+	TP_PROTO(struct btree *b, unsigned keys),
+	TP_ARGS(b, keys),
 
 	TP_STRUCT__entry(
-		__field(dev_t,		dev			)
-		__field(dev_t,		orig_dev		)
-		__field(sector_t,	sector			)
-		__field(sector_t,	orig_sector		)
-		__field(unsigned int,	nr_sector		)
-		__array(char,		rwbs,	6		)
-		__array(char,		comm,	TASK_COMM_LEN	)
+		__field(size_t,		bucket			)
+		__field(unsigned,	keys			)
 	),
 
 	TP_fast_assign(
-		__entry->dev		= bio->bi_bdev->bd_dev;
-		__entry->orig_dev	= orig_bdev->bd_dev;
-		__entry->sector		= bio->bi_sector;
-		__entry->orig_sector	= orig_sector;
-		__entry->nr_sector	= bio->bi_size >> 9;
-		blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size);
-		memcpy(__entry->comm, current->comm, TASK_COMM_LEN);
+		__entry->bucket	= PTR_BUCKET_NR(b->c, &b->key, 0);
+		__entry->keys	= keys;
 	),
 
-	TP_printk("%d,%d  %s %llu + %u [%s] (from %d,%d %llu)",
-		  MAJOR(__entry->dev), MINOR(__entry->dev),
-		  __entry->rwbs,
-		  (unsigned long long)__entry->sector,
-		  __entry->nr_sector, __entry->comm,
-		  MAJOR(__entry->orig_dev), MINOR(__entry->orig_dev),
-		  (unsigned long long)__entry->orig_sector)
+	TP_printk("bucket %zu keys %u", __entry->bucket, __entry->keys)
 );
 
-DEFINE_EVENT(bcache_cache_bio, bcache_cache_insert,
-
-	TP_PROTO(struct bio *bio,
-		 sector_t orig_sector,
-		 struct block_device *orig_bdev),
+DEFINE_EVENT(btree_split, bcache_btree_node_split,
+	TP_PROTO(struct btree *b, unsigned keys),
+	TP_ARGS(b, keys)
+);
 
-	TP_ARGS(bio, orig_sector, orig_bdev)
+DEFINE_EVENT(btree_split, bcache_btree_node_compact,
+	TP_PROTO(struct btree *b, unsigned keys),
+	TP_ARGS(b, keys)
 );
 
-DECLARE_EVENT_CLASS(bcache_gc,
+DEFINE_EVENT(btree_node, bcache_btree_set_root,
+	TP_PROTO(struct btree *b),
+	TP_ARGS(b)
+);
 
-	TP_PROTO(uint8_t *uuid),
+/* Allocator */
 
-	TP_ARGS(uuid),
+TRACE_EVENT(bcache_alloc_invalidate,
+	TP_PROTO(struct cache *ca),
+	TP_ARGS(ca),
 
 	TP_STRUCT__entry(
-		__field(uint8_t *,	uuid)
+		__field(unsigned,	free			)
+		__field(unsigned,	free_inc		)
+		__field(unsigned,	free_inc_size		)
+		__field(unsigned,	unused			)
 	),
 
 	TP_fast_assign(
-		__entry->uuid		= uuid;
+		__entry->free		= fifo_used(&ca->free);
+		__entry->free_inc	= fifo_used(&ca->free_inc);
+		__entry->free_inc_size	= ca->free_inc.size;
+		__entry->unused		= fifo_used(&ca->unused);
 	),
 
-	TP_printk("%pU", __entry->uuid)
+	TP_printk("free %u free_inc %u/%u unused %u", __entry->free,
+		  __entry->free_inc, __entry->free_inc_size, __entry->unused)
 );
 
+TRACE_EVENT(bcache_alloc_fail,
+	TP_PROTO(struct cache *ca),
+	TP_ARGS(ca),
 
-DEFINE_EVENT(bcache_gc, bcache_gc_start,
+	TP_STRUCT__entry(
+		__field(unsigned,	free			)
+		__field(unsigned,	free_inc		)
+		__field(unsigned,	unused			)
+		__field(unsigned,	blocked			)
+	),
 
-	     TP_PROTO(uint8_t *uuid),
+	TP_fast_assign(
+		__entry->free		= fifo_used(&ca->free);
+		__entry->free_inc	= fifo_used(&ca->free_inc);
+		__entry->unused		= fifo_used(&ca->unused);
+		__entry->blocked	= atomic_read(&ca->set->prio_blocked);
+	),
 
-	     TP_ARGS(uuid)
+	TP_printk("free %u free_inc %u unused %u blocked %u", __entry->free,
+		  __entry->free_inc, __entry->unused, __entry->blocked)
 );
 
-DEFINE_EVENT(bcache_gc, bcache_gc_end,
+/* Background writeback */
 
-	     TP_PROTO(uint8_t *uuid),
+DEFINE_EVENT(bkey, bcache_writeback,
+	TP_PROTO(struct bkey *k),
+	TP_ARGS(k)
+);
 
-	     TP_ARGS(uuid)
+DEFINE_EVENT(bkey, bcache_writeback_collision,
+	TP_PROTO(struct bkey *k),
+	TP_ARGS(k)
 );
 
 #endif /* _TRACE_BCACHE_H */
diff --git a/include/xen/interface/io/blkif.h b/include/xen/interface/io/blkif.h
index ffd4652de91c..65e12099ef89 100644
--- a/include/xen/interface/io/blkif.h
+++ b/include/xen/interface/io/blkif.h
@@ -103,12 +103,46 @@ typedef uint64_t blkif_sector_t;
 #define BLKIF_OP_DISCARD           5
 
 /*
+ * Recognized if "feature-max-indirect-segments" in present in the backend
+ * xenbus info. The "feature-max-indirect-segments" node contains the maximum
+ * number of segments allowed by the backend per request. If the node is
+ * present, the frontend might use blkif_request_indirect structs in order to
+ * issue requests with more than BLKIF_MAX_SEGMENTS_PER_REQUEST (11). The
+ * maximum number of indirect segments is fixed by the backend, but the
+ * frontend can issue requests with any number of indirect segments as long as
+ * it's less than the number provided by the backend. The indirect_grefs field
+ * in blkif_request_indirect should be filled by the frontend with the
+ * grant references of the pages that are holding the indirect segments.
+ * This pages are filled with an array of blkif_request_segment_aligned
+ * that hold the information about the segments. The number of indirect
+ * pages to use is determined by the maximum number of segments
+ * a indirect request contains. Every indirect page can contain a maximum
+ * of 512 segments (PAGE_SIZE/sizeof(blkif_request_segment_aligned)),
+ * so to calculate the number of indirect pages to use we have to do
+ * ceil(indirect_segments/512).
+ *
+ * If a backend does not recognize BLKIF_OP_INDIRECT, it should *not*
+ * create the "feature-max-indirect-segments" node!
+ */
+#define BLKIF_OP_INDIRECT          6
+
+/*
  * Maximum scatter/gather segments per request.
  * This is carefully chosen so that sizeof(struct blkif_ring) <= PAGE_SIZE.
  * NB. This could be 12 if the ring indexes weren't stored in the same page.
  */
 #define BLKIF_MAX_SEGMENTS_PER_REQUEST 11
 
+#define BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST 8
+
+struct blkif_request_segment_aligned {
+	grant_ref_t gref;        /* reference to I/O buffer frame        */
+	/* @first_sect: first sector in frame to transfer (inclusive).   */
+	/* @last_sect: last sector in frame to transfer (inclusive).     */
+	uint8_t     first_sect, last_sect;
+	uint16_t    _pad; /* padding to make it 8 bytes, so it's cache-aligned */
+} __attribute__((__packed__));
+
 struct blkif_request_rw {
 	uint8_t        nr_segments;  /* number of segments                   */
 	blkif_vdev_t   handle;       /* only for read/write requests         */
@@ -147,12 +181,31 @@ struct blkif_request_other {
 	uint64_t     id;           /* private guest value, echoed in resp  */
 } __attribute__((__packed__));
 
+struct blkif_request_indirect {
+	uint8_t        indirect_op;
+	uint16_t       nr_segments;
+#ifdef CONFIG_X86_64
+	uint32_t       _pad1;        /* offsetof(blkif_...,u.indirect.id) == 8 */
+#endif
+	uint64_t       id;
+	blkif_sector_t sector_number;
+	blkif_vdev_t   handle;
+	uint16_t       _pad2;
+	grant_ref_t    indirect_grefs[BLKIF_MAX_INDIRECT_PAGES_PER_REQUEST];
+#ifdef CONFIG_X86_64
+	uint32_t      _pad3;         /* make it 64 byte aligned */
+#else
+	uint64_t      _pad3;         /* make it 64 byte aligned */
+#endif
+} __attribute__((__packed__));
+
 struct blkif_request {
 	uint8_t        operation;    /* BLKIF_OP_???                         */
 	union {
 		struct blkif_request_rw rw;
 		struct blkif_request_discard discard;
 		struct blkif_request_other other;
+		struct blkif_request_indirect indirect;
 	} u;
 } __attribute__((__packed__));
 
diff --git a/include/xen/interface/io/ring.h b/include/xen/interface/io/ring.h
index 75271b9a8f61..7d28aff605c7 100644
--- a/include/xen/interface/io/ring.h
+++ b/include/xen/interface/io/ring.h
@@ -188,6 +188,11 @@ struct __name##_back_ring {						\
 #define RING_REQUEST_CONS_OVERFLOW(_r, _cons)				\
     (((_cons) - (_r)->rsp_prod_pvt) >= RING_SIZE(_r))
 
+/* Ill-behaved frontend determination: Can there be this many requests? */
+#define RING_REQUEST_PROD_OVERFLOW(_r, _prod)               \
+    (((_prod) - (_r)->rsp_prod_pvt) > RING_SIZE(_r))
+
+
 #define RING_PUSH_REQUESTS(_r) do {					\
     wmb(); /* back sees requests /before/ updated producer index */	\
     (_r)->sring->req_prod = (_r)->req_prod_pvt;				\