Merge branch 'for-3.8/drivers' of git://git.kernel.dk/linux-block

Pull block driver update from Jens Axboe:
 "Now that the core bits are in, here are the driver bits for 3.8.  The
  branch contains:

   - A huge pile of drbd bits that were dumped from the 3.7 merge
     window.  Following that, it was both made perfectly clear that
     there is going to be no more over-the-wall pulls and how the
     situation on individual pulls can be improved.

   - A few cleanups from Akinobu Mita for drbd and cciss.

   - Queue improvement for loop from Lukas.  This grew into adding a
     generic interface for waiting/checking an even with a specific
     lock, allowing this to be pulled out of md and now loop and drbd is
     also using it.

   - A few fixes for xen back/front block driver from Roger Pau Monne.

   - Partition improvements from Stephen Warren, allowing partiion UUID
     to be used as an identifier."

* 'for-3.8/drivers' of git://git.kernel.dk/linux-block: (609 commits)
  drbd: update Kconfig to match current dependencies
  drbd: Fix drbdsetup wait-connect, wait-sync etc... commands
  drbd: close race between drbd_set_role and drbd_connect
  drbd: respect no-md-barriers setting also when changed online via disk-options
  drbd: Remove obsolete check
  drbd: fixup after wait_even_lock_irq() addition to generic code
  loop: Limit the number of requests in the bio list
  wait: add wait_event_lock_irq() interface
  xen-blkfront: free allocated page
  xen-blkback: move free persistent grants code
  block: partition: msdos: provide UUIDs for partitions
  init: reduce PARTUUID min length to 1 from 36
  block: store partition_meta_info.uuid as a string
  cciss: use check_signature()
  cciss: cleanup bitops usage
  drbd: use copy_highpage
  drbd: if the replication link breaks during handshake, keep retrying
  drbd: check return of kmalloc in receive_uuids
  drbd: Broadcast sync progress no more often than once per second
  drbd: don't try to clear bits once the disk has failed
  ...

26 files changed, 11141 insertions, 8086 deletions

diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c
index ca83f96756a..6526157edaf 100644
--- a/drivers/block/cciss.c
+++ b/drivers/block/cciss.c
@@ -41,8 +41,9 @@
 #include <linux/spinlock.h>
 #include <linux/compat.h>
 #include <linux/mutex.h>
+#include <linux/bitmap.h>
+#include <linux/io.h>
 #include <asm/uaccess.h>
-#include <asm/io.h>
 
 #include <linux/dma-mapping.h>
 #include <linux/blkdev.h>
@@ -978,8 +979,7 @@ static CommandList_struct *cmd_alloc(ctlr_info_t *h)
 		i = find_first_zero_bit(h->cmd_pool_bits, h->nr_cmds);
 		if (i == h->nr_cmds)
 			return NULL;
-	} while (test_and_set_bit(i & (BITS_PER_LONG - 1),
-		  h->cmd_pool_bits + (i / BITS_PER_LONG)) != 0);
+	} while (test_and_set_bit(i, h->cmd_pool_bits) != 0);
 	c = h->cmd_pool + i;
 	memset(c, 0, sizeof(CommandList_struct));
 	cmd_dma_handle = h->cmd_pool_dhandle + i * sizeof(CommandList_struct);
@@ -1046,8 +1046,7 @@ static void cmd_free(ctlr_info_t *h, CommandList_struct *c)
 	int i;
 
 	i = c - h->cmd_pool;
-	clear_bit(i & (BITS_PER_LONG - 1),
-		  h->cmd_pool_bits + (i / BITS_PER_LONG));
+	clear_bit(i, h->cmd_pool_bits);
 	h->nr_frees++;
 }
 
@@ -4268,10 +4267,7 @@ static void __devinit cciss_find_board_params(ctlr_info_t *h)
 
 static inline bool CISS_signature_present(ctlr_info_t *h)
 {
-	if ((readb(&h->cfgtable->Signature[0]) != 'C') ||
-	    (readb(&h->cfgtable->Signature[1]) != 'I') ||
-	    (readb(&h->cfgtable->Signature[2]) != 'S') ||
-	    (readb(&h->cfgtable->Signature[3]) != 'S')) {
+	if (!check_signature(h->cfgtable->Signature, "CISS", 4)) {
 		dev_warn(&h->pdev->dev, "not a valid CISS config table\n");
 		return false;
 	}
@@ -4812,8 +4808,7 @@ static __devinit int cciss_init_reset_devices(struct pci_dev *pdev)
 
 static __devinit int cciss_allocate_cmd_pool(ctlr_info_t *h)
 {
-	h->cmd_pool_bits = kmalloc(
-		DIV_ROUND_UP(h->nr_cmds, BITS_PER_LONG) *
+	h->cmd_pool_bits = kmalloc(BITS_TO_LONGS(h->nr_cmds) *
 		sizeof(unsigned long), GFP_KERNEL);
 	h->cmd_pool = pci_alloc_consistent(h->pdev,
 		h->nr_cmds * sizeof(CommandList_struct),
@@ -5068,9 +5063,7 @@ reinit_after_soft_reset:
 	pci_set_drvdata(pdev, h);
 	/* command and error info recs zeroed out before
 	   they are used */
-	memset(h->cmd_pool_bits, 0,
-	       DIV_ROUND_UP(h->nr_cmds, BITS_PER_LONG)
-			* sizeof(unsigned long));
+	bitmap_zero(h->cmd_pool_bits, h->nr_cmds);
 
 	h->num_luns = 0;
 	h->highest_lun = -1;
diff --git a/drivers/block/drbd/Kconfig b/drivers/block/drbd/Kconfig
index df098378739..7845bd6ee41 100644
--- a/drivers/block/drbd/Kconfig
+++ b/drivers/block/drbd/Kconfig
@@ -2,13 +2,14 @@
 # DRBD device driver configuration
 #
 
-comment "DRBD disabled because PROC_FS, INET or CONNECTOR not selected"
-	depends on PROC_FS='n' || INET='n' || CONNECTOR='n'
+comment "DRBD disabled because PROC_FS or INET not selected"
+	depends on PROC_FS='n' || INET='n'
 
 config BLK_DEV_DRBD
 	tristate "DRBD Distributed Replicated Block Device support"
-	depends on PROC_FS && INET && CONNECTOR
+	depends on PROC_FS && INET
 	select LRU_CACHE
+	select LIBCRC32C
 	default n
 	help
 
@@ -58,7 +59,8 @@ config DRBD_FAULT_INJECTION
 	  32	data read
 	  64	read ahead
 	  128	kmalloc of bitmap
-	  256	allocation of EE (epoch_entries)
+	  256	allocation of peer_requests
+	  512	insert data corruption on receiving side
 
 	  fault_devs: bitmask of minor numbers
 	  fault_rate: frequency in percent
diff --git a/drivers/block/drbd/Makefile b/drivers/block/drbd/Makefile
index 0d3f337ff5f..8b450338075 100644
--- a/drivers/block/drbd/Makefile
+++ b/drivers/block/drbd/Makefile
@@ -1,5 +1,7 @@
 drbd-y := drbd_bitmap.o drbd_proc.o
 drbd-y += drbd_worker.o drbd_receiver.o drbd_req.o drbd_actlog.o
 drbd-y += drbd_main.o drbd_strings.o drbd_nl.o
+drbd-y += drbd_interval.o drbd_state.o
+drbd-y += drbd_nla.o
 
 obj-$(CONFIG_BLK_DEV_DRBD)     += drbd.o
diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c
index 3fbef018ce5..92510f8ad01 100644
--- a/drivers/block/drbd/drbd_actlog.c
+++ b/drivers/block/drbd/drbd_actlog.c
@@ -24,21 +24,73 @@
  */
 
 #include <linux/slab.h>
+#include <linux/crc32c.h>
 #include <linux/drbd.h>
+#include <linux/drbd_limits.h>
+#include <linux/dynamic_debug.h>
 #include "drbd_int.h"
 #include "drbd_wrappers.h"
 
-/* We maintain a trivial checksum in our on disk activity log.
- * With that we can ensure correct operation even when the storage
- * device might do a partial (last) sector write while losing power.
- */
-struct __packed al_transaction {
-	u32       magic;
-	u32       tr_number;
-	struct __packed {
-		u32 pos;
-		u32 extent; } updates[1 + AL_EXTENTS_PT];
-	u32       xor_sum;
+
+enum al_transaction_types {
+	AL_TR_UPDATE = 0,
+	AL_TR_INITIALIZED = 0xffff
+};
+/* all fields on disc in big endian */
+struct __packed al_transaction_on_disk {
+	/* don't we all like magic */
+	__be32	magic;
+
+	/* to identify the most recent transaction block
+	 * in the on disk ring buffer */
+	__be32	tr_number;
+
+	/* checksum on the full 4k block, with this field set to 0. */
+	__be32	crc32c;
+
+	/* type of transaction, special transaction types like:
+	 * purge-all, set-all-idle, set-all-active, ... to-be-defined
+	 * see also enum al_transaction_types */
+	__be16	transaction_type;
+
+	/* we currently allow only a few thousand extents,
+	 * so 16bit will be enough for the slot number. */
+
+	/* how many updates in this transaction */
+	__be16	n_updates;
+
+	/* maximum slot number, "al-extents" in drbd.conf speak.
+	 * Having this in each transaction should make reconfiguration
+	 * of that parameter easier. */
+	__be16	context_size;
+
+	/* slot number the context starts with */
+	__be16	context_start_slot_nr;
+
+	/* Some reserved bytes.  Expected usage is a 64bit counter of
+	 * sectors-written since device creation, and other data generation tag
+	 * supporting usage */
+	__be32	__reserved[4];
+
+	/* --- 36 byte used --- */
+
+	/* Reserve space for up to AL_UPDATES_PER_TRANSACTION changes
+	 * in one transaction, then use the remaining byte in the 4k block for
+	 * context information.  "Flexible" number of updates per transaction
+	 * does not help, as we have to account for the case when all update
+	 * slots are used anyways, so it would only complicate code without
+	 * additional benefit.
+	 */
+	__be16	update_slot_nr[AL_UPDATES_PER_TRANSACTION];
+
+	/* but the extent number is 32bit, which at an extent size of 4 MiB
+	 * allows to cover device sizes of up to 2**54 Byte (16 PiB) */
+	__be32	update_extent_nr[AL_UPDATES_PER_TRANSACTION];
+
+	/* --- 420 bytes used (36 + 64*6) --- */
+
+	/* 4096 - 420 = 3676 = 919 * 4 */
+	__be32	context[AL_CONTEXT_PER_TRANSACTION];
 };
 
 struct update_odbm_work {
@@ -48,22 +100,11 @@ struct update_odbm_work {
 
 struct update_al_work {
 	struct drbd_work w;
-	struct lc_element *al_ext;
 	struct completion event;
-	unsigned int enr;
-	/* if old_enr != LC_FREE, write corresponding bitmap sector, too */
-	unsigned int old_enr;
-};
-
-struct drbd_atodb_wait {
-	atomic_t           count;
-	struct completion  io_done;
-	struct drbd_conf   *mdev;
-	int                error;
+	int err;
 };
 
-
-int w_al_write_transaction(struct drbd_conf *, struct drbd_work *, int);
+static int al_write_transaction(struct drbd_conf *mdev);
 
 void *drbd_md_get_buffer(struct drbd_conf *mdev)
 {
@@ -82,22 +123,24 @@ void drbd_md_put_buffer(struct drbd_conf *mdev)
 		wake_up(&mdev->misc_wait);
 }
 
-static bool md_io_allowed(struct drbd_conf *mdev)
-{
-	enum drbd_disk_state ds = mdev->state.disk;
-	return ds >= D_NEGOTIATING || ds == D_ATTACHING;
-}
-
-void wait_until_done_or_disk_failure(struct drbd_conf *mdev, struct drbd_backing_dev *bdev,
+void wait_until_done_or_force_detached(struct drbd_conf *mdev, struct drbd_backing_dev *bdev,
 				     unsigned int *done)
 {
-	long dt = bdev->dc.disk_timeout * HZ / 10;
+	long dt;
+
+	rcu_read_lock();
+	dt = rcu_dereference(bdev->disk_conf)->disk_timeout;
+	rcu_read_unlock();
+	dt = dt * HZ / 10;
 	if (dt == 0)
 		dt = MAX_SCHEDULE_TIMEOUT;
 
-	dt = wait_event_timeout(mdev->misc_wait, *done || !md_io_allowed(mdev), dt);
-	if (dt == 0)
+	dt = wait_event_timeout(mdev->misc_wait,
+			*done || test_bit(FORCE_DETACH, &mdev->flags), dt);
+	if (dt == 0) {
 		dev_err(DEV, "meta-data IO operation timed out\n");
+		drbd_chk_io_error(mdev, 1, DRBD_FORCE_DETACH);
+	}
 }
 
 static int _drbd_md_sync_page_io(struct drbd_conf *mdev,
@@ -106,7 +149,7 @@ static int _drbd_md_sync_page_io(struct drbd_conf *mdev,
 				 int rw, int size)
 {
 	struct bio *bio;
-	int ok;
+	int err;
 
 	mdev->md_io.done = 0;
 	mdev->md_io.error = -ENODEV;
@@ -118,8 +161,8 @@ static int _drbd_md_sync_page_io(struct drbd_conf *mdev,
 	bio = bio_alloc_drbd(GFP_NOIO);
 	bio->bi_bdev = bdev->md_bdev;
 	bio->bi_sector = sector;
-	ok = (bio_add_page(bio, page, size, 0) == size);
-	if (!ok)
+	err = -EIO;
+	if (bio_add_page(bio, page, size, 0) != size)
 		goto out;
 	bio->bi_private = &mdev->md_io;
 	bio->bi_end_io = drbd_md_io_complete;
@@ -127,7 +170,7 @@ static int _drbd_md_sync_page_io(struct drbd_conf *mdev,
 
 	if (!get_ldev_if_state(mdev, D_ATTACHING)) {  /* Corresponding put_ldev in drbd_md_io_complete() */
 		dev_err(DEV, "ASSERT FAILED: get_ldev_if_state() == 1 in _drbd_md_sync_page_io()\n");
-		ok = 0;
+		err = -ENODEV;
 		goto out;
 	}
 
@@ -137,86 +180,47 @@ static int _drbd_md_sync_page_io(struct drbd_conf *mdev,
 		bio_endio(bio, -EIO);
 	else
 		submit_bio(rw, bio);
-	wait_until_done_or_disk_failure(mdev, bdev, &mdev->md_io.done);
-	ok = bio_flagged(bio, BIO_UPTODATE) && mdev->md_io.error == 0;
+	wait_until_done_or_force_detached(mdev, bdev, &mdev->md_io.done);
+	if (bio_flagged(bio, BIO_UPTODATE))
+		err = mdev->md_io.error;
 
  out:
 	bio_put(bio);
-	return ok;
+	return err;
 }
 
 int drbd_md_sync_page_io(struct drbd_conf *mdev, struct drbd_backing_dev *bdev,
 			 sector_t sector, int rw)
 {
-	int logical_block_size, mask, ok;
-	int offset = 0;
+	int err;
 	struct page *iop = mdev->md_io_page;
 
 	D_ASSERT(atomic_read(&mdev->md_io_in_use) == 1);
 
 	BUG_ON(!bdev->md_bdev);
 
-	logical_block_size = bdev_logical_block_size(bdev->md_bdev);
-	if (logical_block_size == 0)
-		logical_block_size = MD_SECTOR_SIZE;
-
-	/* in case logical_block_size != 512 [ s390 only? ] */
-	if (logical_block_size != MD_SECTOR_SIZE) {
-		mask = (logical_block_size / MD_SECTOR_SIZE) - 1;
-		D_ASSERT(mask == 1 || mask == 3 || mask == 7);
-		D_ASSERT(logical_block_size == (mask+1) * MD_SECTOR_SIZE);
-		offset = sector & mask;
-		sector = sector & ~mask;
-		iop = mdev->md_io_tmpp;
-
-		if (rw & WRITE) {
-			/* these are GFP_KERNEL pages, pre-allocated
-			 * on device initialization */
-			void *p = page_address(mdev->md_io_page);
-			void *hp = page_address(mdev->md_io_tmpp);
-
-			ok = _drbd_md_sync_page_io(mdev, bdev, iop, sector,
-					READ, logical_block_size);
-
-			if (unlikely(!ok)) {
-				dev_err(DEV, "drbd_md_sync_page_io(,%llus,"
-				    "READ [logical_block_size!=512]) failed!\n",
-				    (unsigned long long)sector);
-				return 0;
-			}
-
-			memcpy(hp + offset*MD_SECTOR_SIZE, p, MD_SECTOR_SIZE);
-		}
-	}
+	dev_dbg(DEV, "meta_data io: %s [%d]:%s(,%llus,%s)\n",
+	     current->comm, current->pid, __func__,
+	     (unsigned long long)sector, (rw & WRITE) ? "WRITE" : "READ");
 
 	if (sector < drbd_md_first_sector(bdev) ||
-	    sector > drbd_md_last_sector(bdev))
+	    sector + 7 > drbd_md_last_sector(bdev))
 		dev_alert(DEV, "%s [%d]:%s(,%llus,%s) out of range md access!\n",
 		     current->comm, current->pid, __func__,
 		     (unsigned long long)sector, (rw & WRITE) ? "WRITE" : "READ");
 
-	ok = _drbd_md_sync_page_io(mdev, bdev, iop, sector, rw, logical_block_size);
-	if (unlikely(!ok)) {
-		dev_err(DEV, "drbd_md_sync_page_io(,%llus,%s) failed!\n",
-		    (unsigned long long)sector, (rw & WRITE) ? "WRITE" : "READ");
-		return 0;
-	}
-
-	if (logical_block_size != MD_SECTOR_SIZE && !(rw & WRITE)) {
-		void *p = page_address(mdev->md_io_page);
-		void *hp = page_address(mdev->md_io_tmpp);
-
-		memcpy(p, hp + offset*MD_SECTOR_SIZE, MD_SECTOR_SIZE);
+	err = _drbd_md_sync_page_io(mdev, bdev, iop, sector, rw, MD_BLOCK_SIZE);
+	if (err) {
+		dev_err(DEV, "drbd_md_sync_page_io(,%llus,%s) failed with error %d\n",
+		    (unsigned long long)sector, (rw & WRITE) ? "WRITE" : "READ", err);
 	}
-
-	return ok;
+	return err;
 }
 
 static struct lc_element *_al_get(struct drbd_conf *mdev, unsigned int enr)
 {
 	struct lc_element *al_ext;
 	struct lc_element *tmp;
-	unsigned long     al_flags = 0;
 	int wake;
 
 	spin_lock_irq(&mdev->al_lock);
@@ -231,76 +235,92 @@ static struct lc_element *_al_get(struct drbd_conf *mdev, unsigned int enr)
 			return NULL;
 		}
 	}
-	al_ext   = lc_get(mdev->act_log, enr);
-	al_flags = mdev->act_log->flags;
+	al_ext = lc_get(mdev->act_log, enr);
 	spin_unlock_irq(&mdev->al_lock);
-
-	/*
-	if (!al_ext) {
-		if (al_flags & LC_STARVING)
-			dev_warn(DEV, "Have to wait for LRU element (AL too small?)\n");
-		if (al_flags & LC_DIRTY)
-			dev_warn(DEV, "Ongoing AL update (AL device too slow?)\n");
-	}
-	*/
-
 	return al_ext;
 }
 
-void drbd_al_begin_io(struct drbd_conf *mdev, sector_t sector)
+void drbd_al_begin_io(struct drbd_conf *mdev, struct drbd_interval *i)
 {
-	unsigned int enr = (sector >> (AL_EXTENT_SHIFT-9));
-	struct lc_element *al_ext;
-	struct update_al_work al_work;
+	/* for bios crossing activity log extent boundaries,
+	 * we may need to activate two extents in one go */
+	unsigned first = i->sector >> (AL_EXTENT_SHIFT-9);
+	unsigned last = i->size == 0 ? first : (i->sector + (i->size >> 9) - 1) >> (AL_EXTENT_SHIFT-9);
+	unsigned enr;
+	bool locked = false;
 
+
+	D_ASSERT(first <= last);
 	D_ASSERT(atomic_read(&mdev->local_cnt) > 0);
 
-	wait_event(mdev->al_wait, (al_ext = _al_get(mdev, enr)));
+	for (enr = first; enr <= last; enr++)
+		wait_event(mdev->al_wait, _al_get(mdev, enr) != NULL);
+
+	/* Serialize multiple transactions.
+	 * This uses test_and_set_bit, memory barrier is implicit.
+	 */
+	wait_event(mdev->al_wait,
+			mdev->act_log->pending_changes == 0 ||
+			(locked = lc_try_lock_for_transaction(mdev->act_log)));
 
-	if (al_ext->lc_number != enr) {
+	if (locked) {
 		/* drbd_al_write_transaction(mdev,al_ext,enr);
 		 * recurses into generic_make_request(), which
 		 * disallows recursion, bios being serialized on the
 		 * current->bio_tail list now.
 		 * we have to delegate updates to the activity log
 		 * to the worker thread. */
-		init_completion(&al_work.event);
-		al_work.al_ext = al_ext;
-		al_work.enr = enr;
-		al_work.old_enr = al_ext->lc_number;
-		al_work.w.cb = w_al_write_transaction;
-		drbd_queue_work_front(&mdev->data.work, &al_work.w);
-		wait_for_completion(&al_work.event);
-
-		mdev->al_writ_cnt++;
-
-		spin_lock_irq(&mdev->al_lock);
-		lc_changed(mdev->act_log, al_ext);
-		spin_unlock_irq(&mdev->al_lock);
+
+		/* Double check: it may have been committed by someone else,
+		 * while we have been waiting for the lock. */
+		if (mdev->act_log->pending_changes) {
+			bool write_al_updates;
+
+			rcu_read_lock();
+			write_al_updates = rcu_dereference(mdev->ldev->disk_conf)->al_updates;
+			rcu_read_unlock();
+
+			if (write_al_updates) {
+				al_write_transaction(mdev);
+				mdev->al_writ_cnt++;
+			}
+
+			spin_lock_irq(&mdev->al_lock);
+			/* FIXME
+			if (err)
+				we need an "lc_cancel" here;
+			*/
+			lc_committed(mdev->act_log);
+			spin_unlock_irq(&mdev->al_lock);
+		}
+		lc_unlock(mdev->act_log);
 		wake_up(&mdev->al_wait);
 	}
 }
 
-void drbd_al_complete_io(struct drbd_conf *mdev, sector_t sector)
+void drbd_al_complete_io(struct drbd_conf *mdev, struct drbd_interval *i)
 {
-	unsigned int enr = (sector >> (AL_EXTENT_SHIFT-9));
+	/* for bios crossing activity log extent boundaries,
+	 * we may need to activate two extents in one go */
+	unsigned first = i->sector >> (AL_EXTENT_SHIFT-9);
+	unsigned last = i->size == 0 ? first : (i->sector + (i->size >> 9) - 1) >> (AL_EXTENT_SHIFT-9);
+	unsigned enr;
 	struct lc_element *extent;
 	unsigned long flags;
 
+	D_ASSERT(first <= last);
 	spin_lock_irqsave(&mdev->al_lock, flags);
 
-	extent = lc_find(mdev->act_log, enr);
-
-	if (!extent) {
-		spin_unlock_irqrestore(&mdev->al_lock, flags);
-		dev_err(DEV, "al_complete_io() called on inactive extent %u\n", enr);
-		return;
+	for (enr = first; enr <= last; enr++) {
+		extent = lc_find(mdev->act_log, enr);
+		if (!extent) {
+			dev_err(DEV, "al_complete_io() called on inactive extent %u\n", enr);
+			continue;
+		}
+		lc_put(mdev->act_log, extent);
 	}
-
-	if (lc_put(mdev->act_log, extent) == 0)
-		wake_up(&mdev->al_wait);
-
 	spin_unlock_irqrestore(&mdev->al_lock, flags);
+	wake_up(&mdev->al_wait);
 }
 
 #if (PAGE_SHIFT + 3) < (AL_EXTENT_SHIFT - BM_BLOCK_SHIFT)
@@ -326,296 +346,148 @@ static unsigned int rs_extent_to_bm_page(unsigned int rs_enr)
 	return rs_enr >>
 		/* bit to page */
 		((PAGE_SHIFT + 3) -
-		/* al extent number to bit */
+		/* resync extent number to bit */
 		 (BM_EXT_SHIFT - BM_BLOCK_SHIFT));
 }
 
-int
-w_al_write_transaction(struct drbd_conf *mdev, struct drbd_work *w, int unused)
+static int
+_al_write_transaction(struct drbd_conf *mdev)
 {
-	struct update_al_work *aw = container_of(w, struct update_al_work, w);
-	struct lc_element *updated = aw->al_ext;
-	const unsigned int new_enr = aw->enr;
-	const unsigned int evicted = aw->old_enr;
-	struct al_transaction *buffer;
+	struct al_transaction_on_disk *buffer;
+	struct lc_element *e;
 	sector_t sector;
-	int i, n, mx;
-	unsigned int extent_nr;
-	u32 xor_sum = 0;
+	int i, mx;
+	unsigned extent_nr;
+	unsigned crc = 0;
+	int err = 0;
 
 	if (!get_ldev(mdev)) {
-		dev_err(DEV,
-			"disk is %s, cannot start al transaction (-%d +%d)\n",
-			drbd_disk_str(mdev->state.disk), evicted, new_enr);
-		complete(&((struct update_al_work *)w)->event);
-		return 1;
+		dev_err(DEV, "disk is %s, cannot start al transaction\n",
+			drbd_disk_str(mdev->state.disk));
+		return -EIO;
 	}
-	/* do we have to do a bitmap write, first?
-	 * TODO reduce maximum latency:
-	 * submit both bios, then wait for both,
-	 * instead of doing two synchronous sector writes.
-	 * For now, we must not write the transaction,
-	 * if we cannot write out the bitmap of the evicted extent. */
-	if (mdev->state.conn < C_CONNECTED && evicted != LC_FREE)
-		drbd_bm_write_page(mdev, al_extent_to_bm_page(evicted));
 
 	/* The bitmap write may have failed, causing a state change. */
 	if (mdev->state.disk < D_INCONSISTENT) {
 		dev_err(DEV,
-			"disk is %s, cannot write al transaction (-%d +%d)\n",
-			drbd_disk_str(mdev->state.disk), evicted, new_enr);
-		complete(&((struct update_al_work *)w)->event);
+			"disk is %s, cannot write al transaction\n",
+			drbd_disk_str(mdev->state.disk));
 		put_ldev(mdev);
-		return 1;
+		return -EIO;
 	}
 
 	buffer = drbd_md_get_buffer(mdev); /* protects md_io_buffer, al_tr_cycle, ... */
 	if (!buffer) {
 		dev_err(DEV, "disk failed while waiting for md_io buffer\n");
-		complete(&((struct update_al_work *)w)->event);
 		put_ldev(mdev);
-		return 1;
+		return -ENODEV;
 	}
 
-	buffer->magic = __constant_cpu_to_be32(DRBD_MAGIC);
+	memset(buffer, 0, sizeof(*buffer));
+	buffer->magic = cpu_to_be32(DRBD_AL_MAGIC);
 	buffer->tr_number = cpu_to_be32(mdev->al_tr_number);
 
-	n = lc_index_of(mdev->act_log, updated);
+	i = 0;
+
+	/* Even though no one can start to change this list
+	 * once we set the LC_LOCKED -- from drbd_al_begin_io(),
+	 * lc_try_lock_for_transaction() --, someone may still
+	 * be in the process of changing it. */
+	spin_lock_irq(&mdev->al_lock);
+	list_for_each_entry(e, &mdev->act_log->to_be_changed, list) {
+		if (i == AL_UPDATES_PER_TRANSACTION) {
+			i++;
+			break;
+		}
+		buffer->update_slot_nr[i] = cpu_to_be16(e->lc_index);
+		buffer->update_extent_nr[i] = cpu_to_be32(e->lc_new_number);
+		if (e->lc_number != LC_FREE)
+			drbd_bm_mark_for_writeout(mdev,
+					al_extent_to_bm_page(e->lc_number));
+		i++;
+	}
+	spin_unlock_irq(&mdev->al_lock);
+	BUG_ON(i > AL_UPDATES_PER_TRANSACTION);
 
-	buffer->updates[0].pos = cpu_to_be32(n);
-	buffer->updates[0].extent = cpu_to_be32(new_enr);
+	buffer->n_updates = cpu_to_be16(i);
+	for ( ; i < AL_UPDATES_PER_TRANSACTION; i++) {
+		buffer->update_slot_nr[i] = cpu_to_be16(-1);
+		buffer->update_extent_nr[i] = cpu_to_be32(LC_FREE);
+	}
 
-	xor_sum ^= new_enr;
+	buffer->context_size = cpu_to_be16(mdev->act_log->nr_elements);
+	buffer->context_start_slot_nr = cpu_to_be16(mdev->al_tr_cycle);
 
-	mx = min_t(int, AL_EXTENTS_PT,
+	mx = min_t(int, AL_CONTEXT_PER_TRANSACTION,
 		   mdev->act_log->nr_elements - mdev->al_tr_cycle);
 	for (i = 0; i < mx; i++) {
 		unsigned idx = mdev->al_tr_cycle + i;
 		extent_nr = lc_element_by_index(mdev->act_log, idx)->lc_number;
-		buffer->updates[i+1].pos = cpu_to_be32(idx);
-		buffer->updates[i+1].extent = cpu_to_be32(extent_nr);
-		xor_sum ^= extent_nr;
-	}
-	for (; i < AL_EXTENTS_PT; i++) {
-		buffer->updates[i+1].pos = __constant_cpu_to_be32(-1);
-		buffer->updates[i+1].extent = __constant_cpu_to_be32(LC_FREE);
-		xor_sum ^= LC_FREE;
+		buffer->context[i] = cpu_to_be32(extent_nr);
 	}
-	mdev->al_tr_cycle += AL_EXTENTS_PT;
+	for (; i < AL_CONTEXT_PER_TRANSACTION; i++)
+		buffer->context[i] = cpu_to_be32(LC_FREE);
+
+	mdev->al_tr_cycle += AL_CONTEXT_PER_TRANSACTION;
 	if (mdev->al_tr_cycle >= mdev->act_log->nr_elements)
 		mdev->al_tr_cycle = 0;
 
-	buffer->xor_sum = cpu_to_be32(xor_sum);
-
 	sector =  mdev->ldev->md.md_offset
-		+ mdev->ldev->md.al_offset + mdev->al_tr_pos;
-
-	if (!drbd_md_sync_page_io(mdev, mdev->ldev, sector, WRITE))
-		drbd_chk_io_error(mdev, 1, DRBD_META_IO_ERROR);
+		+ mdev->ldev->md.al_offset
+		+ mdev->al_tr_pos * (MD_BLOCK_SIZE>>9);
 
-	if (++mdev->al_tr_pos >
-	    div_ceil(mdev->act_log->nr_elements, AL_EXTENTS_PT))
-		mdev->al_tr_pos = 0;
+	crc = crc32c(0, buffer, 4096);
+	buffer->crc32c = cpu_to_be32(crc);
 
-	D_ASSERT(mdev->al_tr_pos < MD_AL_MAX_SIZE);
-	mdev->al_tr_number++;
+	if (drbd_bm_write_hinted(mdev))
+		err = -EIO;
+		/* drbd_chk_io_error done already */
+	else if (drbd_md_sync_page_io(mdev, mdev->ldev, sector, WRITE)) {
+		err = -EIO;
+		drbd_chk_io_error(mdev, 1, DRBD_META_IO_ERROR);
+	} else {
+		/* advance ringbuffer position and transaction counter */
+		mdev->al_tr_pos = (mdev->al_tr_pos + 1) % (MD_AL_SECTORS*512/MD_BLOCK_SIZE);
+		mdev->al_tr_number++;
+	}
 
 	drbd_md_put_buffer(mdev);
-
-	complete(&((struct update_al_work *)w)->event);
 	put_ldev(mdev);
 
-	return 1;
+	return err;
 }
 
-/**
- * drbd_al_read_tr() - Read a single transaction from the on disk activity log
- * @mdev:	DRBD device.
- * @bdev:	Block device to read form.
- * @b:		pointer to an al_transaction.
- * @index:	On disk slot of the transaction to read.
- *
- * Returns -1 on IO error, 0 on checksum error and 1 upon success.
- */
-static int drbd_al_read_tr(struct drbd_conf *mdev,
-			   struct drbd_backing_dev *bdev,
-			   struct al_transaction *b,
-			   int index)
-{
-	sector_t sector;
-	int rv, i;
-	u32 xor_sum = 0;
-
-	sector = bdev->md.md_offset + bdev->md.al_offset + index;
-
-	/* Dont process error normally,
-	 * as this is done before disk is attached! */
-	if (!drbd_md_sync_page_io(mdev, bdev, sector, READ))
-		return -1;
-
-	rv = (be32_to_cpu(b->magic) == DRBD_MAGIC);
-
-	for (i = 0; i < AL_EXTENTS_PT + 1; i++)
-		xor_sum ^= be32_to_cpu(b->updates[i].extent);
-	rv &= (xor_sum == be32_to_cpu(b->xor_sum));
 
-	return rv;
-}
-
-/**
- * drbd_al_read_log() - Restores the activity log from its on disk representation.
- * @mdev:	DRBD device.
- * @bdev:	Block device to read form.
- *
- * Returns 1 on success, returns 0 when reading the log failed due to IO errors.
- */
-int drbd_al_read_log(struct drbd_conf *mdev, struct drbd_backing_dev *bdev)
+static int w_al_write_transaction(struct drbd_work *w, int unused)
 {
-	struct al_transaction *buffer;
-	int i;
-	int rv;
-	int mx;
-	int active_extents = 0;
-	int transactions = 0;
-	int found_valid = 0;
-	int from = 0;
-	int to = 0;
-	u32 from_tnr = 0;
-	u32 to_tnr = 0;
-	u32 cnr;
-
-	mx = div_ceil(mdev->act_log->nr_elements, AL_EXTENTS_PT);
-
-	/* lock out all other meta data io for now,
-	 * and make sure the page is mapped.
-	 */
-	buffer = drbd_md_get_buffer(mdev);
-	if (!buffer)
-		return 0;
-
-	/* Find the valid transaction in the log */
-	for (i = 0; i <= mx; i++) {
-		rv = drbd_al_read_tr(mdev, bdev, buffer, i);
-		if (rv == 0)
-			continue;
-		if (rv == -1) {
-			drbd_md_put_buffer(mdev);
-			return 0;
-		}
-		cnr = be32_to_cpu(buffer->tr_number);
-
-		if (++found_valid == 1) {
-			from = i;
-			to = i;
-			from_tnr = cnr;
-			to_tnr = cnr;
-			continue;
-		}
-		if ((int)cnr - (int)from_tnr < 0) {
-			D_ASSERT(from_tnr - cnr + i - from == mx+1);
-			from = i;
-			from_tnr = cnr;
-		}
-		if ((int)cnr - (int)to_tnr > 0) {
-			D_ASSERT(cnr - to_tnr == i - to);
-			to = i;
-			to_tnr = cnr;
-		}
-	}
-
-	if (!found_valid) {
-		dev_warn(DEV, "No usable activity log found.\n");
-		drbd_md_put_buffer(mdev);
-		return 1;
-	}
-
-	/* Read the valid transactions.
-	 * dev_info(DEV, "Reading from %d to %d.\n",from,to); */
-	i = from;
-	while (1) {
-		int j, pos;
-		unsigned int extent_nr;
-		unsigned int trn;
-
-		rv = drbd_al_read_tr(mdev, bdev, buffer, i);
-		ERR_IF(rv == 0) goto cancel;
-		if (rv == -1) {
-			drbd_md_put_buffer(mdev);
-			return 0;
-		}
-
-		trn = be32_to_cpu(buffer->tr_number);
-
-		spin_lock_irq(&mdev->al_lock);
-
-		/* This loop runs backwards because in the cyclic
-		   elements there might be an old version of the
-		   updated element (in slot 0). So the element in slot 0
-		   can overwrite old versions. */
-		for (j = AL_EXTENTS_PT; j >= 0; j--) {
-			pos = be32_to_cpu(buffer->updates[j].pos);
-			extent_nr = be32_to_cpu(buffer->updates[j].extent);
-
-			if (extent_nr == LC_FREE)
-				continue;
-
-			lc_set(mdev->act_log, extent_nr, pos);
-			active_extents++;