Merge commit 'linux-2.6/master' (early part) into oprofile/core

This branch depends on these apic patches:

      apic, x86: Use BIOS settings for IBS and MCE threshold interrupt LVT offsets
      apic, x86: Check if EILVT APIC registers are available (AMD only)

Signed-off-by: Robert Richter <robert.richter@amd.com>

53 files changed, 3353 insertions, 1127 deletions

diff --git a/drivers/acpi/acpi_pad.c b/drivers/acpi/acpi_pad.c
index 6b115f6c431..6afceb3d403 100644
--- a/drivers/acpi/acpi_pad.c
+++ b/drivers/acpi/acpi_pad.c
@@ -30,18 +30,13 @@
 #include <linux/slab.h>
 #include <acpi/acpi_bus.h>
 #include <acpi/acpi_drivers.h>
+#include <asm/mwait.h>
 
 #define ACPI_PROCESSOR_AGGREGATOR_CLASS	"acpi_pad"
 #define ACPI_PROCESSOR_AGGREGATOR_DEVICE_NAME "Processor Aggregator"
 #define ACPI_PROCESSOR_AGGREGATOR_NOTIFY 0x80
 static DEFINE_MUTEX(isolated_cpus_lock);
 
-#define MWAIT_SUBSTATE_MASK	(0xf)
-#define MWAIT_CSTATE_MASK	(0xf)
-#define MWAIT_SUBSTATE_SIZE	(4)
-#define CPUID_MWAIT_LEAF (5)
-#define CPUID5_ECX_EXTENSIONS_SUPPORTED (0x1)
-#define CPUID5_ECX_INTERRUPT_BREAK	(0x2)
 static unsigned long power_saving_mwait_eax;
 
 static unsigned char tsc_detected_unstable;
diff --git a/drivers/base/topology.c b/drivers/base/topology.c
index 9fc630ce1dd..f6f37a05a0c 100644
--- a/drivers/base/topology.c
+++ b/drivers/base/topology.c
@@ -45,7 +45,8 @@ static ssize_t show_##name(struct sys_device *dev,		\
 	return sprintf(buf, "%d\n", topology_##name(cpu));	\
 }
 
-#if defined(topology_thread_cpumask) || defined(topology_core_cpumask)
+#if defined(topology_thread_cpumask) || defined(topology_core_cpumask) || \
+    defined(topology_book_cpumask)
 static ssize_t show_cpumap(int type, const struct cpumask *mask, char *buf)
 {
 	ptrdiff_t len = PTR_ALIGN(buf + PAGE_SIZE - 1, PAGE_SIZE) - buf;
@@ -114,6 +115,14 @@ define_siblings_show_func(core_cpumask);
 define_one_ro_named(core_siblings, show_core_cpumask);
 define_one_ro_named(core_siblings_list, show_core_cpumask_list);
 
+#ifdef CONFIG_SCHED_BOOK
+define_id_show_func(book_id);
+define_one_ro(book_id);
+define_siblings_show_func(book_cpumask);
+define_one_ro_named(book_siblings, show_book_cpumask);
+define_one_ro_named(book_siblings_list, show_book_cpumask_list);
+#endif
+
 static struct attribute *default_attrs[] = {
 	&attr_physical_package_id.attr,
 	&attr_core_id.attr,
@@ -121,6 +130,11 @@ static struct attribute *default_attrs[] = {
 	&attr_thread_siblings_list.attr,
 	&attr_core_siblings.attr,
 	&attr_core_siblings_list.attr,
+#ifdef CONFIG_SCHED_BOOK
+	&attr_book_id.attr,
+	&attr_book_siblings.attr,
+	&attr_book_siblings_list.attr,
+#endif
 	NULL
 };
 
diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
index de277689da6..4b9359a6f6c 100644
--- a/drivers/block/Kconfig
+++ b/drivers/block/Kconfig
@@ -488,4 +488,21 @@ config BLK_DEV_HD
 
 	  If unsure, say N.
 
+config BLK_DEV_RBD
+	tristate "Rados block device (RBD)"
+	depends on INET && EXPERIMENTAL && BLOCK
+	select CEPH_LIB
+	select LIBCRC32C
+	select CRYPTO_AES
+	select CRYPTO
+	default n
+	help
+	  Say Y here if you want include the Rados block device, which stripes
+	  a block device over objects stored in the Ceph distributed object
+	  store.
+
+	  More information at http://ceph.newdream.net/.
+
+	  If unsure, say N.
+
 endif # BLK_DEV
diff --git a/drivers/block/Makefile b/drivers/block/Makefile
index aff5ac925c3..d7f463d6312 100644
--- a/drivers/block/Makefile
+++ b/drivers/block/Makefile
@@ -37,5 +37,6 @@ obj-$(CONFIG_BLK_DEV_HD)	+= hd.o
 
 obj-$(CONFIG_XEN_BLKDEV_FRONTEND)	+= xen-blkfront.o
 obj-$(CONFIG_BLK_DEV_DRBD)     += drbd/
+obj-$(CONFIG_BLK_DEV_RBD)     += rbd.o
 
 swim_mod-objs	:= swim.o swim_asm.o
diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
new file mode 100644
index 00000000000..6ec9d53806c
--- /dev/null
+++ b/drivers/block/rbd.c
@@ -0,0 +1,1841 @@
+/*
+   rbd.c -- Export ceph rados objects as a Linux block device
+
+
+   based on drivers/block/osdblk.c:
+
+   Copyright 2009 Red Hat, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program; see the file COPYING.  If not, write to
+   the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+
+
+
+   Instructions for use
+   --------------------
+
+   1) Map a Linux block device to an existing rbd image.
+
+      Usage: <mon ip addr> <options> <pool name> <rbd image name> [snap name]
+
+      $ echo "192.168.0.1 name=admin rbd foo" > /sys/class/rbd/add
+
+      The snapshot name can be "-" or omitted to map the image read/write.
+
+   2) List all active blkdev<->object mappings.
+
+      In this example, we have performed step #1 twice, creating two blkdevs,
+      mapped to two separate rados objects in the rados rbd pool
+
+      $ cat /sys/class/rbd/list
+      #id     major   client_name     pool    name    snap    KB
+      0       254     client4143      rbd     foo     -      1024000
+
+      The columns, in order, are:
+      - blkdev unique id
+      - blkdev assigned major
+      - rados client id
+      - rados pool name
+      - rados block device name
+      - mapped snapshot ("-" if none)
+      - device size in KB
+
+
+   3) Create a snapshot.
+
+      Usage: <blkdev id> <snapname>
+
+      $ echo "0 mysnap" > /sys/class/rbd/snap_create
+
+
+   4) Listing a snapshot.
+
+      $ cat /sys/class/rbd/snaps_list
+      #id     snap    KB
+      0       -       1024000 (*)
+      0       foo     1024000
+
+      The columns, in order, are:
+      - blkdev unique id
+      - snapshot name, '-' means none (active read/write version)
+      - size of device at time of snapshot
+      - the (*) indicates this is the active version
+
+   5) Rollback to snapshot.
+
+      Usage: <blkdev id> <snapname>
+
+      $ echo "0 mysnap" > /sys/class/rbd/snap_rollback
+
+
+   6) Mapping an image using snapshot.
+
+      A snapshot mapping is read-only. This is being done by passing
+      snap=<snapname> to the options when adding a device.
+
+      $ echo "192.168.0.1 name=admin,snap=mysnap rbd foo" > /sys/class/rbd/add
+
+
+   7) Remove an active blkdev<->rbd image mapping.
+
+      In this example, we remove the mapping with blkdev unique id 1.
+
+      $ echo 1 > /sys/class/rbd/remove
+
+
+   NOTE:  The actual creation and deletion of rados objects is outside the scope
+   of this driver.
+
+ */
+
+#include <linux/ceph/libceph.h>
+#include <linux/ceph/osd_client.h>
+#include <linux/ceph/mon_client.h>
+#include <linux/ceph/decode.h>
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/blkdev.h>
+
+#include "rbd_types.h"
+
+#define DRV_NAME "rbd"
+#define DRV_NAME_LONG "rbd (rados block device)"
+
+#define RBD_MINORS_PER_MAJOR	256		/* max minors per blkdev */
+
+#define RBD_MAX_MD_NAME_LEN	(96 + sizeof(RBD_SUFFIX))
+#define RBD_MAX_POOL_NAME_LEN	64
+#define RBD_MAX_SNAP_NAME_LEN	32
+#define RBD_MAX_OPT_LEN		1024
+
+#define RBD_SNAP_HEAD_NAME	"-"
+
+#define DEV_NAME_LEN		32
+
+/*
+ * block device image metadata (in-memory version)
+ */
+struct rbd_image_header {
+	u64 image_size;
+	char block_name[32];
+	__u8 obj_order;
+	__u8 crypt_type;
+	__u8 comp_type;
+	struct rw_semaphore snap_rwsem;
+	struct ceph_snap_context *snapc;
+	size_t snap_names_len;
+	u64 snap_seq;
+	u32 total_snaps;
+
+	char *snap_names;
+	u64 *snap_sizes;
+};
+
+/*
+ * an instance of the client.  multiple devices may share a client.
+ */
+struct rbd_client {
+	struct ceph_client	*client;
+	struct kref		kref;
+	struct list_head	node;
+};
+
+/*
+ * a single io request
+ */
+struct rbd_request {
+	struct request		*rq;		/* blk layer request */
+	struct bio		*bio;		/* cloned bio */
+	struct page		**pages;	/* list of used pages */
+	u64			len;
+};
+
+/*
+ * a single device
+ */
+struct rbd_device {
+	int			id;		/* blkdev unique id */
+
+	int			major;		/* blkdev assigned major */
+	struct gendisk		*disk;		/* blkdev's gendisk and rq */
+	struct request_queue	*q;
+
+	struct ceph_client	*client;
+	struct rbd_client	*rbd_client;
+
+	char			name[DEV_NAME_LEN]; /* blkdev name, e.g. rbd3 */
+
+	spinlock_t		lock;		/* queue lock */
+
+	struct rbd_image_header	header;
+	char			obj[RBD_MAX_OBJ_NAME_LEN]; /* rbd image name */
+	int			obj_len;
+	char			obj_md_name[RBD_MAX_MD_NAME_LEN]; /* hdr nm. */
+	char			pool_name[RBD_MAX_POOL_NAME_LEN];
+	int			poolid;
+
+	char                    snap_name[RBD_MAX_SNAP_NAME_LEN];
+	u32 cur_snap;	/* index+1 of current snapshot within snap context
+			   0 - for the head */
+	int read_only;
+
+	struct list_head	node;
+};
+
+static spinlock_t node_lock;      /* protects client get/put */
+
+static struct class *class_rbd;	  /* /sys/class/rbd */
+static DEFINE_MUTEX(ctl_mutex);	  /* Serialize open/close/setup/teardown */
+static LIST_HEAD(rbd_dev_list);    /* devices */
+static LIST_HEAD(rbd_client_list);      /* clients */
+
+
+static int rbd_open(struct block_device *bdev, fmode_t mode)
+{
+	struct gendisk *disk = bdev->bd_disk;
+	struct rbd_device *rbd_dev = disk->private_data;
+
+	set_device_ro(bdev, rbd_dev->read_only);
+
+	if ((mode & FMODE_WRITE) && rbd_dev->read_only)
+		return -EROFS;
+
+	return 0;
+}
+
+static const struct block_device_operations rbd_bd_ops = {
+	.owner			= THIS_MODULE,
+	.open			= rbd_open,
+};
+
+/*
+ * Initialize an rbd client instance.
+ * We own *opt.
+ */
+static struct rbd_client *rbd_client_create(struct ceph_options *opt)
+{
+	struct rbd_client *rbdc;
+	int ret = -ENOMEM;
+
+	dout("rbd_client_create\n");
+	rbdc = kmalloc(sizeof(struct rbd_client), GFP_KERNEL);
+	if (!rbdc)
+		goto out_opt;
+
+	kref_init(&rbdc->kref);
+	INIT_LIST_HEAD(&rbdc->node);
+
+	rbdc->client = ceph_create_client(opt, rbdc);
+	if (IS_ERR(rbdc->client))
+		goto out_rbdc;
+	opt = NULL; /* Now rbdc->client is responsible for opt */
+
+	ret = ceph_open_session(rbdc->client);
+	if (ret < 0)
+		goto out_err;
+
+	spin_lock(&node_lock);
+	list_add_tail(&rbdc->node, &rbd_client_list);
+	spin_unlock(&node_lock);
+
+	dout("rbd_client_create created %p\n", rbdc);
+	return rbdc;
+
+out_err:
+	ceph_destroy_client(rbdc->client);
+out_rbdc:
+	kfree(rbdc);
+out_opt:
+	if (opt)
+		ceph_destroy_options(opt);
+	return ERR_PTR(ret);
+}
+
+/*
+ * Find a ceph client with specific addr and configuration.
+ */
+static struct rbd_client *__rbd_client_find(struct ceph_options *opt)
+{
+	struct rbd_client *client_node;
+
+	if (opt->flags & CEPH_OPT_NOSHARE)
+		return NULL;
+
+	list_for_each_entry(client_node, &rbd_client_list, node)
+		if (ceph_compare_options(opt, client_node->client) == 0)
+			return client_node;
+	return NULL;
+}
+
+/*
+ * Get a ceph client with specific addr and configuration, if one does
+ * not exist create it.
+ */
+static int rbd_get_client(struct rbd_device *rbd_dev, const char *mon_addr,
+			  char *options)
+{
+	struct rbd_client *rbdc;
+	struct ceph_options *opt;
+	int ret;
+
+	ret = ceph_parse_options(&opt, options, mon_addr,
+				 mon_addr + strlen(mon_addr), NULL, NULL);
+	if (ret < 0)
+		return ret;
+
+	spin_lock(&node_lock);
+	rbdc = __rbd_client_find(opt);
+	if (rbdc) {
+		ceph_destroy_options(opt);
+
+		/* using an existing client */
+		kref_get(&rbdc->kref);
+		rbd_dev->rbd_client = rbdc;
+		rbd_dev->client = rbdc->client;
+		spin_unlock(&node_lock);
+		return 0;
+	}
+	spin_unlock(&node_lock);
+
+	rbdc = rbd_client_create(opt);
+	if (IS_ERR(rbdc))
+		return PTR_ERR(rbdc);
+
+	rbd_dev->rbd_client = rbdc;
+	rbd_dev->client = rbdc->client;
+	return 0;
+}
+
+/*
+ * Destroy ceph client
+ */
+static void rbd_client_release(struct kref *kref)
+{
+	struct rbd_client *rbdc = container_of(kref, struct rbd_client, kref);
+
+	dout("rbd_release_client %p\n", rbdc);
+	spin_lock(&node_lock);
+	list_del(&rbdc->node);
+	spin_unlock(&node_lock);
+
+	ceph_destroy_client(rbdc->client);
+	kfree(rbdc);
+}
+
+/*
+ * Drop reference to ceph client node. If it's not referenced anymore, release
+ * it.
+ */
+static void rbd_put_client(struct rbd_device *rbd_dev)
+{
+	kref_put(&rbd_dev->rbd_client->kref, rbd_client_release);
+	rbd_dev->rbd_client = NULL;
+	rbd_dev->client = NULL;
+}
+
+
+/*
+ * Create a new header structure, translate header format from the on-disk
+ * header.
+ */
+static int rbd_header_from_disk(struct rbd_image_header *header,
+				 struct rbd_image_header_ondisk *ondisk,
+				 int allocated_snaps,
+				 gfp_t gfp_flags)
+{
+	int i;
+	u32 snap_count = le32_to_cpu(ondisk->snap_count);
+	int ret = -ENOMEM;
+
+	init_rwsem(&header->snap_rwsem);
+
+	header->snap_names_len = le64_to_cpu(ondisk->snap_names_len);
+	header->snapc = kmalloc(sizeof(struct ceph_snap_context) +
+				snap_count *
+				 sizeof(struct rbd_image_snap_ondisk),
+				gfp_flags);
+	if (!header->snapc)
+		return -ENOMEM;
+	if (snap_count) {
+		header->snap_names = kmalloc(header->snap_names_len,
+					     GFP_KERNEL);
+		if (!header->snap_names)
+			goto err_snapc;
+		header->snap_sizes = kmalloc(snap_count * sizeof(u64),
+					     GFP_KERNEL);
+		if (!header->snap_sizes)
+			goto err_names;
+	} else {
+		header->snap_names = NULL;
+		header->snap_sizes = NULL;
+	}
+	memcpy(header->block_name, ondisk->block_name,
+	       sizeof(ondisk->block_name));
+
+	header->image_size = le64_to_cpu(ondisk->image_size);
+	header->obj_order = ondisk->options.order;
+	header->crypt_type = ondisk->options.crypt_type;
+	header->comp_type = ondisk->options.comp_type;
+
+	atomic_set(&header->snapc->nref, 1);
+	header->snap_seq = le64_to_cpu(ondisk->snap_seq);
+	header->snapc->num_snaps = snap_count;
+	header->total_snaps = snap_count;
+
+	if (snap_count &&
+	    allocated_snaps == snap_count) {
+		for (i = 0; i < snap_count; i++) {
+			header->snapc->snaps[i] =
+				le64_to_cpu(ondisk->snaps[i].id);
+			header->snap_sizes[i] =
+				le64_to_cpu(ondisk->snaps[i].image_size);
+		}
+
+		/* copy snapshot names */
+		memcpy(header->snap_names, &ondisk->snaps[i],
+			header->snap_names_len);
+	}
+
+	return 0;
+
+err_names:
+	kfree(header->snap_names);
+err_snapc:
+	kfree(header->snapc);
+	return ret;
+}
+
+static int snap_index(struct rbd_image_header *header, int snap_num)
+{
+	return header->total_snaps - snap_num;
+}
+
+static u64 cur_snap_id(struct rbd_device *rbd_dev)
+{
+	struct rbd_image_header *header = &rbd_dev->header;
+
+	if (!rbd_dev->cur_snap)
+		return 0;
+
+	return header->snapc->snaps[snap_index(header, rbd_dev->cur_snap)];
+}
+
+static int snap_by_name(struct rbd_image_header *header, const char *snap_name,
+			u64 *seq, u64 *size)
+{
+	int i;
+	char *p = header->snap_names;
+
+	for (i = 0; i < header->total_snaps; i++, p += strlen(p) + 1) {
+		if (strcmp(snap_name, p) == 0)
+			break;
+	}
+	if (i == header->total_snaps)
+		return -ENOENT;
+	if (seq)
+		*seq = header->snapc->snaps[i];
+
+	if (size)
+		*size = header->snap_sizes[i];
+
+	return i;
+}
+
+static int rbd_header_set_snap(struct rbd_device *dev,
+			       const char *snap_name,
+			       u64 *size)
+{
+	struct rbd_image_header *header = &dev->header;
+	struct ceph_snap_context *snapc = header->snapc;
+	int ret = -ENOENT;
+
+	down_write(&header->snap_rwsem);
+
+	if (!snap_name ||
+	    !*snap_name ||
+	    strcmp(snap_name, "-") == 0 ||
+	    strcmp(snap_name, RBD_SNAP_HEAD_NAME) == 0) {
+		if (header->total_snaps)
+			snapc->seq = header->snap_seq;
+		else
+			snapc->seq = 0;
+		dev->cur_snap = 0;
+		dev->read_only = 0;
+		if (size)
+			*size = header->image_size;
+	} else {
+		ret = snap_by_name(header, snap_name, &snapc->seq, size);
+		if (ret < 0)
+			goto done;
+
+		dev->cur_snap = header->total_snaps - ret;
+		dev->read_only = 1;
+	}
+
+	ret = 0;
+done:
+	up_write(&header->snap_rwsem);
+	return ret;
+}
+
+static void rbd_header_free(struct rbd_image_header *header)
+{
+	kfree(header->snapc);
+	kfree(header->snap_names);
+	kfree(header->snap_sizes);
+}
+
+/*
+ * get the actual striped segment name, offset and length
+ */
+static u64 rbd_get_segment(struct rbd_image_header *header,
+			   const char *block_name,
+			   u64 ofs, u64 len,
+			   char *seg_name, u64 *segofs)
+{
+	u64 seg = ofs >> header->obj_order;
+
+	if (seg_name)
+		snprintf(seg_name, RBD_MAX_SEG_NAME_LEN,
+			 "%s.%012llx", block_name, seg);
+
+	ofs = ofs & ((1 << header->obj_order) - 1);
+	len = min_t(u64, len, (1 << header->obj_order) - ofs);
+
+	if (segofs)
+		*segofs = ofs;
+
+	return len;
+}
+
+/*
+ * bio helpers
+ */
+
+static void bio_chain_put(struct bio *chain)
+{
+	struct bio *tmp;
+
+	while (chain) {
+		tmp = chain;
+		chain = chain->bi_next;
+		bio_put(tmp);
+	}
+}
+
+/*
+ * zeros a bio chain, starting at specific offset
+ */
+static void zero_bio_chain(struct bio *chain, int start_ofs)
+{
+	struct bio_vec *bv;
+	unsigned long flags;
+	void *buf;
+	int i;
+	int pos = 0;
+
+	while (chain) {
+		bio_for_each_segment(bv, chain, i) {
+			if (pos + bv->bv_len > start_ofs) {
+				int remainder = max(start_ofs - pos, 0);
+				buf = bvec_kmap_irq(bv, &flags);
+				memset(buf + remainder, 0,
+				       bv->bv_len - remainder);
+				bvec_kunmap_irq(buf, &flags);
+			}
+			pos += bv->bv_len;
+		}
+
+		chain = chain->bi_next;
+	}
+}
+
+/*
+ * bio_chain_clone - clone a chain of bios up to a certain length.
+ * might return a bio_pair that will need to be released.
+ */
+static struct bio *bio_chain_clone(struct bio **old, struct bio **next,
+				   struct bio_pair **bp,
+				   int len, gfp_t gfpmask)
+{
+	struct bio *tmp, *old_chain = *old, *new_chain = NULL, *tail = NULL;
+	int total = 0;
+
+	if (*bp) {
+		bio_pair_release(*bp);
+		*bp = NULL;
+	}
+
+	while (old_chain && (total < len)) {
+		tmp = bio_kmalloc(gfpmask, old_chain->bi_max_vecs);
+		if (!tmp)
+			goto err_out;
+
+		if (total + old_chain->bi_size > len) {
+			struct bio_pair *bp;
+
+			/*
+			 * this split can only happen with a single paged bio,
+			 * split_bio will BUG_ON if this is not the case
+			 */
+			dout("bio_chain_clone split! total=%d remaining=%d"
+			     "bi_size=%d\n",
+			     (int)total, (int)len-total,
+			     (int)old_chain->bi_size);
+
+			/* split the bio. We'll release it either in the next
+			   call, or it will have to be released outside */
+			bp = bio_split(old_chain, (len - total) / 512ULL);
+			if (!bp)
+				goto err_out;
+
+			__bio_clone(tmp, &bp->bio1);
+
+			*next = &bp->bio2;
+		} else {
+			__bio_clone(tmp, old_chain);
+			*next = old_chain->bi_next;
+		}
+
+		tmp->bi_bdev = NULL;
+		gfpmask &= ~__GFP_WAIT;
+		tmp->bi_next = NULL;
+
+		if (!new_chain) {
+			new_chain = tail = tmp;
+		} else {
+			tail->bi_next = tmp;
+			tail = tmp;
+		}
+		old_chain = old_chain->bi_next;
+
+		total += tmp->bi_size;
+	}
+
+	BUG_ON(total < len);
+
+	if (tail)
+		tail->bi_next = NULL;
+
+	*old = old_chain;
+
+	return new_chain;
+
+err_out:
+	dout("bio_chain_clone with err\n");
+	bio_chain_put(new_chain);
+	return NULL;
+}
+
+/*
+ * helpers for osd request op vectors.
+ */
+static int rbd_create_rw_ops(struct ceph_osd_req_op **ops,
+			    int num_ops,
+			    int opcode,
+			    u32 payload_len)
+{
+	*ops = kzalloc(sizeof(struct ceph_osd_req_op) * (num_ops + 1),
+		       GFP_NOIO);
+	if (!*ops)
+		return -ENOMEM;
+	(*ops)[0].op = opcode;
+	/*
+	 * op extent offset and length will be set later on
+	 * in calc_raw_layout()
+	 */
+	(*ops)[0].payload_len = payload_len;
+	return 0;
+}
+
+static void rbd_destroy_ops(struct ceph_osd_req_op *ops)
+{
+	kfree(ops);
+}
+
+/*
+ * Send ceph osd request
+ */
+static int rbd_do_request(struct request *rq,
+			  struct rbd_device *dev,
+			  struct ceph_snap_context *snapc,
+			  u64 snapid,
+			  const char *obj, u64 ofs, u64 len,
+			  struct bio *bio,
+			  struct page **pages,
+			  int num_pages,
+			  int flags,
+			  struct ceph_osd_req_op *ops,
+			  int num_reply,
+			  void (*rbd_cb)(struct ceph_osd_request *req,
+					 struct ceph_msg *msg))
+{
+	struct ceph_osd_request *req;
+	struct ceph_file_layout *layout;
+	int ret;
+	u64 bno;
+	struct timespec mtime = CURRENT_TIME;
+	struct rbd_request *req_data;
+	struct ceph_osd_request_head *reqhead;
+	struct rbd_image_header *header = &dev->header;
+
+	ret = -ENOMEM;
+	req_data = kzalloc(sizeof(*req_data), GFP_NOIO);
+	if (!req_data)
+		goto done;
+
+	dout("rbd_do_request len=%lld ofs=%lld\n", len, ofs);
+
+	down_read(&header->snap_rwsem);
+
+	req = ceph_osdc_alloc_request(&dev->client->osdc, flags,
+				      snapc,
+				      ops,
+				      false,
+				      GFP_NOIO, pages, bio);
+	if (IS_ERR(req)) {
+		up_read(&header->snap_rwsem);
+		ret = PTR_ERR(req);
+		goto done_pages;
+	}
+
+	req->r_callback = rbd_cb;
+
+	req_data->rq = rq;
+	req_data->bio = bio;
+	req_data->pages = pages;
+	req_data->len = len;
+
+	req->r_priv = req_data;
+
+	reqhead = req->r_request->front.iov_base;
+	reqhead->snapid = cpu_to_le64(CEPH_NOSNAP);
+
+	strncpy(req->r_oid, obj, sizeof(req->r_oid));
+	req->r_oid_len = strlen(req->r_oid);
+
+	layout = &req->r_file_layout;
+	memset(layout, 0, sizeof(*layout));
+	layout->fl_stripe_unit = cpu_to_le32(1 << RBD_MAX_OBJ_ORDER);
+	layout->fl_stripe_count = cpu_to_le32(1);
+	layout->fl_object_size = cpu_to_le32(1 << RBD_MAX_OBJ_ORDER);
+	layout->fl_pg_preferred = cpu_to_le32(-1);
+	layout->fl_pg_pool = cpu_to_le32(dev->poolid);
+	ceph_calc_raw_layout(&dev->client->osdc, layout, snapid,
+			     ofs, &len, &bno, req, ops);
+
+	ceph_osdc_build_request(req, ofs, &len,
+				ops,
+				snapc,
+				&mtime,
+				req->r_oid, req->r_oid_len);
+	up_read(&header->snap_rwsem);
+
+	ret = ceph_osdc_start_request(&dev->client->osdc, req, false);
+	if (ret < 0)
+		goto done_err;
+
+	if (!rbd_cb) {
+		ret = ceph_osdc_wait_request(&dev->client->osdc, req);
+		ceph_osdc_put_request(req);
+	}
+	return ret;
+
+done_err:
+	bio_chain_put(req_data->bio);
+	ceph_osdc_put_request(req);
+done_pages:
+	kfree(req_data);
+done:
+	if (rq)
+		blk_end_request(rq, ret, len);
+	return ret;
+}
+
+/*
+ * Ceph osd op callback
+ */
+static void rbd_req_cb(struct ceph_osd_request *req, struct ceph_msg *msg)
+{
+	struct rbd_request *req_data = req->r_priv;
+	struct ceph_osd_reply_head *replyhead;
+	struct ceph_osd_op *op;
+	__s32 rc;
+	u64 bytes;
+	int read_op;
+
+	/* parse reply */
+	replyhead = msg->front.iov_base;
+	WARN_ON(le32_to_cpu(replyhead->num_ops) == 0);
+	op = (void *)(replyhead + 1);
+	rc = le32_to_cpu(replyhead->result);
+	bytes = le64_to_cpu(op->extent.length);
+	read_op = (le32_to_cpu(op->op) == CEPH_OSD_OP_READ);
+
+	dout("rbd_req_cb bytes=%lld readop=%d rc=%d\n", bytes, read_op, rc);
+
+	if (rc == -ENOENT && read_op) {
+		zero_bio_chain(req_data->bio, 0);
+		rc = 0;
+	} else if (rc == 0 && read_op && bytes < req_data->len) {
+		zero_bio_chain(req_data->bio, bytes);
+		bytes = req_data->len;
+	}
+
+	blk_end_request(req_data->rq, rc, bytes);
+
+	if (req_data->bio)
+		bio_chain_put(req_data->bio);
+
+	ceph_osdc_put_request(req);
+	kfree(req_data);
+}
+
+/*
+ * Do a synchronous ceph osd operation
+ */
+static int rbd_req_sync_op(struct rbd_device *dev,
+			   struct ceph_snap_context *snapc,
+			   u64 snapid,
+			   int opcode,
+			   int flags,
+			   struct ceph_osd_req_op *orig_ops,
+			   int num_reply,
+			   const char *obj,
+			   u64 ofs, u64 len,
+			   char *buf)
+{
+	int ret;
+	struct page **pages;
+	int num_pages;
+	struct ceph_osd_req_op *ops = orig_ops;
+	u32 payload_len;
+
+	num_pages = calc_pages_for(ofs , len);
+	pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL);
+	if (IS_ERR(pages))
+		return PTR_ERR(pages);
+
+	if (!orig_ops) {
+		payload_len = (flags & CEPH_OSD_FLAG_WRITE ? len : 0);
+		ret = rbd_create_rw_ops(&ops, 1, opcode, payload_len);
+		if (ret < 0)
+			goto done;
+
+		if ((flags & CEPH_OSD_FLAG_WRITE) && buf) {
+			ret = ceph_copy_to_page_vector(pages, buf, ofs, len);
+			if (ret < 0)
+				goto done_ops;
+		}
+	}
+
+	ret = rbd_do_request(NULL, dev, snapc, snapid,
+			  obj, ofs, len, NULL,
+			  pages, num_pages,
+			  flags,
+			  ops,
+			  2,
+			  NULL);
+	if (ret < 0)
+		goto done_ops;
+
+	if ((flags & CEPH_OSD_FLAG_READ) && buf)
+		ret = ceph_copy_from_page_vector(pages, buf, ofs, ret);
+
+done_ops:
+	if (!orig_ops)
+		rbd_destroy_ops(ops);
+done:
+	ceph_release_page_vector(pages, num_pages);
+	return ret;
+}
+
+/*
+ * Do an asynchronous ceph osd operation
+ */
+static int rbd_do_op(struct request *rq,
+		     struct rbd_device *rbd_dev ,
+		     struct ceph_snap_context *snapc,
+		     u64 snapid,
+		     int opcode, int flags, int num_reply,
+		     u64 ofs, u64 len,
+		     struct bio *bio)
+{
+	char *seg_name;
+	u64 seg_ofs;
+	u64 seg_len;
+	int ret;
+	struct ceph_osd_req_op *ops;
+	u32 payload_len;
+
+	seg_name = kmalloc(RBD_MAX_SEG_NAME_LEN + 1, GFP_NOIO);
+	if (!seg_name)
+		return -ENOMEM;
+
+	seg_len = rbd_get_segment(&rbd_dev->header,
+				  rbd_dev->header.block_name,
+				  ofs, len,
+				  seg_name, &seg_ofs);
+
+	payload_len = (flags & CEPH_OSD_FLAG_WRITE ? seg_len : 0);
+
+	ret = rbd_create_rw_ops(&ops, 1, opcode, payload_len);
+	if (ret < 0)
+		goto done;
+
+	/* we've taken care of segment sizes earlier when we
+	   cloned the bios. We should never have a segment
+	   truncated at this point */
+	BUG_ON(seg_len < len);
+
+	ret = rbd_do_request(rq, rbd_dev, snapc, snapid,
+			     seg_name, seg_ofs, seg_len,
+			     bio,
+			     NULL, 0,
+			     flags,
+			     ops,
+			     num_reply,
+			     rbd_req_cb);
+done:
+	kfree(seg_name);
+	return ret;
+}
+
+/*
+ * Request async osd write
+ */
+static int rbd_req_write(struct request *rq,
+			 struct rbd_device *rbd_dev,
+			 struct ceph_snap_context *snapc,
+			 u64 ofs, u64 len,
+			 struct bio *bio)
+{
+	return rbd_do_op(rq, rbd_dev, snapc, CEPH_NOSNAP,
+			 CEPH_OSD_OP_WRITE,
+			 CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK,
+			 2,
+			 ofs, len, bio);
+}
+
+/*
+ * Request async osd read
+ */
+static int rbd_req_read(struct request *rq,
+			 struct rbd_device *rbd_dev,
+			 u64 snapid,
+			 u64 ofs, u64 len,
+			 struct bio *bio)
+{
+	return rbd_do_op(rq, rbd_dev, NULL,
+			 (snapid ? snapid : CEPH_NOSNAP),
+			 CEPH_OSD_OP_READ,
+			 CEPH_OSD_FLAG_READ,
+			 2,
+			 ofs, len, bio);
+}
+
+/*
+ * Request sync osd read
+ */
+static int rbd_req_sync_read(struct rbd_device *dev,
+			  struct ceph_snap_context *snapc,
+			  u64 snapid,
+			  const char *obj,
+			  u64 ofs, u64 len,
+			  char *buf)
+{
+	return rbd_req_sync_op(dev, NULL,
+			       (snapid ? snapid : CEPH_NOSNAP),
+			       CEPH_OSD_OP_READ,
+			       CEPH_OSD_FLAG_READ,
+			       NULL,
+			       1, obj, ofs, len, buf);
+}
+
+/*
+ * Request sync osd read
+ */
+static int rbd_req_sync_rollback_obj(struct rbd_device *dev,
+				     u64 snapid,
+				     const char *obj)
+{
+	struct ceph_osd_req_op *ops;
+	int ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_ROLLBACK, 0);
+	if (ret < 0)
+		return ret;
+
+	ops[0].snap.snapid = snapid;
+
+	ret = rbd_req_sync_op(dev, NULL,
+			       CEPH_NOSNAP,
+			       0,
+			       CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK,
+			       ops,
+			       1, obj, 0, 0, NULL);
+
+	rbd_destroy_ops(ops);
+
+	if (ret < 0)
+		return ret;
+
+	return ret;
+}
+
+/*
+ * Request sync osd read
+ */
+static int rbd_req_sync_exec(struct rbd_device *dev,
+			     const char *obj,
+			     const char *cls,
+			     const char *method,
+			     const char *data,
+			     int len)
+{
+	struct ceph_osd_req_op *ops;
+	int cls_len = strlen(cls);
+	int method_len = strlen(method);
+	int ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_CALL,
+				    cls_len + method_len + len);
+	if (ret < 0)
+		return ret;
+
+	ops[0].cls.class_name = cls;
+	ops[0].cls.class_len = (__u8)cls_len;
+	ops[0].cls.method_name = method;
+	ops[0].cls.method_len = (__u8)method_len;
+	ops[0].cls.argc = 0;
+	ops[0].cls.indata = data;
+	ops[0].cls.indata_len = len;
+
+	ret = rbd_req_sync_op(dev, NULL,
+			       CEPH_NOSNAP,
+			       0,
+			       CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK,
+			       ops,
+			       1, obj, 0, 0, NULL);
+
+	rbd_destroy_ops(ops);
+
+	dout("cls_exec returned %d\n", ret);
+	return ret;
+}
+
+/*
+ * block device queue callback
+ */
+static void rbd_rq_fn(struct request_queue *q)
+{
+	struct rbd_device *rbd_dev = q->queuedata;
+	struct request *rq;
+	struct bio_pair *bp = NULL;
+
+	rq = blk_fetch_request(q);
+
+	while (1) {
+		struct bio *bio;
+		struct bio *rq_bio, *next_bio = NULL;
+		bool do_write;
+		int size, op_size = 0;
+		u64 ofs;
+
+		/* peek at request from block layer */
+		if (!rq)
+			break;
+
+		dout("fetched request\n");
+
+		/* filter out block requests we don't understand */
+		if ((rq->cmd_type != REQ_TYPE_FS)) {
+			__blk_end_request_all(rq, 0);
+			goto next;
+		}
+
+		/* deduce our operation (read, write) */
+		do_write = (rq_data_dir(rq) == WRITE);
+
+		size = blk_rq_bytes(rq);
+		ofs = blk_rq_pos(rq) * 512ULL;
+		rq_bio = rq->bio;
+		if (do_write && rbd_dev->read_only) {
+			__blk_end_request_all(rq, -EROFS);
+			goto next;
+		}
+
+		spin_unlock_irq(q->queue_lock);
+
+		dout("%s 0x%x bytes at 0x%llx\n",
+		     do_write ? "write" : "read",
+		     size, blk_rq_pos(rq) * 512ULL);
+
+		do {
+			/* a bio clone to be passed down to OSD req */
+			dout("rq->bio->bi_vcnt=%d\n", rq->bio->bi_vcnt);
+			op_size = rbd_get_segment(&rbd_dev->header,
+						  rbd_dev->header.block_name,
+						  ofs, size,
+						  NULL, NULL);
+			bio = bio_chain_clone(&rq_bio, &next_bio, &bp,
+					      op_size, GFP_ATOMIC);
+			if (!bio) {
+				spin_lock_irq(q->queue_lock);
+				__blk_end_request_all(rq, -ENOMEM);
+				goto next;
+			}
+
+			/* init OSD command: write or read */
+			if (do_write)
+				rbd_req_write(rq, rbd_dev,
+					      rbd_dev->header.snapc,
+					      ofs,
+					      op_size, bio);
+			else
+				rbd_req_read(rq, rbd_dev,
+					     cur_snap_id(rbd_dev),
+					     ofs,
+					     op_size, bio);
+
+			size -= op_size;
+			ofs += op_size;
+
+			rq_bio = next_bio;
+		} while (size > 0);
+
+		if (bp)
+			bio_pair_release(bp);
+
+		spin_lock_irq(q->queue_lock);
+next:
+		rq = blk_fetch_request(q);
+	}
+}
+
+/*
+ * a queue callback. Makes sure that we don't create a bio that spans across
+ * multiple osd objects. One exception would be with a single page bios,
+ * which we handle later at bio_chain_clone
+ */
+static int rbd_merge_bvec(struct request_queue *q, struct bvec_merge_data *bmd,
+			  struct bio_vec *bvec)
+{
+	struct rbd_device *rbd_dev = q->queuedata;
+	unsigned int chunk_sectors = 1 << (rbd_dev->header.obj_order - 9);
+	sector_t sector = bmd->bi_sector + get_start_sect(bmd->bi_bdev);
+	unsigned int bio_sectors = bmd->bi_size >> 9;
+	int max;
+
+	max =  (chunk_sectors - ((sector & (chunk_sectors - 1))
+				 + bio_sectors)) << 9;
+	if (max < 0)
+		max = 0; /* bio_add cannot handle a negative return */
+	if (max <= bvec->bv_len && bio_sectors == 0)
+		return bvec->bv_len;
+	return max;
+}
+
+static void rbd_free_disk(struct rbd_device *rbd_dev)
+{
+	struct gendisk *disk = rbd_dev->disk;
+
+	if (!disk)
+		return;
+
+	rbd_header_free(&rbd_dev->header);
+
+	if (disk->flags & GENHD_FL_UP)
+		del_gendisk(disk);
+	if (disk->queue)
+		blk_cleanup_queue(disk->queue);
+	put_disk(disk);
+}
+
+/*
+ * reload the ondisk the header 
+ */
+static int rbd_read_header(struct rbd_device *rbd_dev,
+			   struct rbd_image_header *header)
+{
+	ssize_t rc;
+	struct rbd_image_header_ondisk *dh;
+	int snap_count = 0;
+	u64 snap_names_len = 0;
+
+	while (1) {
+		int len = sizeof(*dh) +
+			  snap_count * sizeof(struct rbd_image_snap_ondisk) +
+			  snap_names_len;
+
+		rc = -ENOMEM;
+		dh = kmalloc(len, GFP_KERNEL);
+		if (!dh)
+			return -ENOMEM;
+
+		rc = rbd_req_sync_read(rbd_dev,
+				       NULL, CEPH_NOSNAP,
+				       rbd_dev->obj_md_name,
+				       0, len,
+				       (char *)dh);
+		if (rc < 0)
+			goto out_dh;
+
+		rc = rbd_header_from_disk(header, dh, snap_count, GFP_KERNEL);
+		if (rc < 0)
+			goto out_dh;
+
+		if (snap_count != header->total_snaps) {
+			snap_count = header->total_snaps;
+			snap_names_len = header->snap_names_len;
+			rbd_header_free(header);
+			kfree(dh);
+			continue;
+		}
+		break;
+	}
+
+out_dh:
+	kfree(dh);
+	return rc;
+}
+
+/*
+ * create a snapshot
+ */
+static int rbd_header_add_snap(struct rbd_device *dev,
+			       const char *snap_name,
+			       gfp_t gfp_flags)
+{
+	int name_len = strlen(snap_name);
+	u64 new_snapid;
+	int ret;
+	void *data, *data_start, *data_end;
+
+	/* we should create a snapshot only if we're pointing at the head */
+	if (dev->cur_snap)
+		return -EINVAL;
+
+	ret = ceph_monc_create_snapid(&dev->client->monc, dev->poolid,
+				      &new_snapid);
+	dout("created snapid=%lld\n", new_snapid);
+	if (ret < 0)
+		return ret;
+
+	data = kmalloc(name_len + 16, gfp_flags);
+	if (!data)
+		return -ENOMEM;
+
+	data_start = data;
+	data_end = data + name_len + 16;
+
+	ceph_encode_string_safe(&data, data_end, snap_name, name_len, bad);
+	ceph_encode_64_safe(&data, data_end, new_snapid, bad);
+
+	ret = rbd_req_sync_exec(dev, dev->obj_md_name, "rbd", "snap_add",
+				data_start, data - data_start);
+
+	kfree(data_start);
+
+	if (ret < 0)
+		return ret;
+
+	dev->header.snapc->seq =  new_snapid;
+
+	return 0;
+bad:
+	return -ERANGE;
+}
+
+/*
+ * only read the first part of the ondisk header, without the snaps info
+ */
+static int rbd_update_snaps(struct rbd_device *rbd_dev)
+{
+	int ret;
+	struct rbd_image_header h;
+	u64 snap_seq;
+
+	ret = rbd_read_header(rbd_dev, &h);
+	if (ret < 0)
+		return ret;
+
+	down_write(&rbd_dev->header.snap_rwsem);
+
+	snap_seq = rbd_dev->header.snapc->seq;
+
+	kfree(rbd_dev->header.snapc);
+	kfree(rbd_dev->header.snap_names);
+	kfree(rbd_dev->header.snap_sizes);
+
+	rbd_dev->header.total_snaps = h.total_snaps;
+	rbd_dev->header.snapc = h.snapc;
+	rbd_dev->header.snap_names = h.snap_names;
+	rbd_dev->header.snap_sizes = h.snap_sizes;
+	rbd_dev->header.snapc->seq = snap_seq;
+
+	up_write(&rbd_dev->header.snap_rwsem);
+
+	return 0;
+}
+
+static int rbd_init_disk(struct rbd_device *rbd_dev)
+{
+	struct gendisk *disk;
+	struct request_queue *q;
+	int rc;
+	u64 total_size = 0;
+
+	/* contact OSD, request size info about the object being mapped */
+	rc = rbd_read_header(rbd_dev, &rbd_dev->header);
+	if (rc)
+		return rc;
+
+	rc = rbd_header_set_snap(rbd_dev, rbd_dev->snap_name, &total_size);
+	if (rc)
+		return rc;
+
+	/* create gendisk info */
+	rc = -ENOMEM;
+	disk = alloc_disk(RBD_MINORS_PER_MAJOR);
+	if (!disk)
+		goto out;
+
+	sprintf(disk->disk_name, DRV_NAME "%d", rbd_dev->id);
+	disk->major = rbd_dev->major;
+	disk->first_minor = 0;
+	disk->fops = &rbd_bd_ops;
+	disk->private_data = rbd_dev;
+
+	/* init rq */
+	rc = -ENOMEM;
+	q = blk_init_queue(rbd_rq_fn, &rbd_dev->lock);
+	if (!q)
+		goto out_disk;
+	blk_queue_merge_bvec(q, rbd_merge_bvec);
+	disk->queue = q;
+
+	q->queuedata = rbd_dev;
+
+	rbd_dev->disk = disk;
+	rbd_dev->q = q;
+
+	/* finally, announce the disk to the world */
+	set_capacity(disk, total_size / 512ULL);
+	add_disk(disk);
+
+	pr_info("%s: added with size 0x%llx\n",
+		disk->disk_name, (unsigned long long)total_size);
+	return 0;
+
+out_disk:
+	put_disk(disk);
+out:
+	return rc;
+}
+
+/********************************************************************
+ * /sys/class/rbd/
+ *                   add	map rados objects to blkdev
+ *                   remove	unmap rados objects
+ *                   list	show mappings
+ *******************************************************************/
+
+static void class_rbd_release(struct class *cls)
+{
+	kfree(cls);
+}
+
+static ssize_t class_rbd_list(struct class *c,
+			      struct class_attribute *attr,
+			      char *data)
+{
+	int n = 0;
+	struct list_head *tmp;
+	int max = PAGE_SIZE;
+
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	n += snprintf(data, max,
+		      "#id\tmajor\tclient_name\tpool\tname\tsnap\tKB\n");
+
+	list_for_each(tmp, &rbd_dev_list) {
+		struct rbd_device *rbd_dev;
+
+		rbd_dev = list_entry(tmp, struct rbd_device, node);
+		n += snprintf(data+n, max-n,
+			      "%d\t%d\tclient%lld\t%s\t%s\t%s\t%lld\n",
+			      rbd_dev->id,
+			      rbd_dev->major,
+			      ceph_client_id(rbd_dev->client),
+			      rbd_dev->pool_name,
+			      rbd_dev->obj, rbd_dev->snap_name,
+			      rbd_dev->header.image_size >> 10);
+		if (n == max)
+			break;
+	}
+
+	mutex_unlock(&ctl_mutex);
+	return n;
+}
+
+static ssize_t class_rbd_add(struct class *c,
+			     struct class_attribute *attr,
+			     const char *buf, size_t count)
+{
+	struct ceph_osd_client *osdc;
+	struct rbd_device *rbd_dev;
+	ssize_t rc = -ENOMEM;
+	int irc, new_id = 0;
+	struct list_head *tmp;
+	char *mon_dev_name;
+	char *options;
+
+	if (!try_module_get(THIS_MODULE))
+		return -ENODEV;
+
+	mon_dev_name = kmalloc(RBD_MAX_OPT_LEN, GFP_KERNEL);
+	if (!mon_dev_name)
+		goto err_out_mod;
+
+	options = kmalloc(RBD_MAX_OPT_LEN, GFP_KERNEL);
+	if (!options)
+		goto err_mon_dev;
+
+	/* new rbd_device object */
+	rbd_dev = kzalloc(sizeof(*rbd_dev), GFP_KERNEL);
+	if (!rbd_dev)
+		goto err_out_opt;
+
+	/* static rbd_device initialization */
+	spin_lock_init(&rbd_dev->lock);
+	INIT_LIST_HEAD(&rbd_dev->node);
+
+	/* generate unique id: find highest unique id, add one */
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	list_for_each(tmp, &rbd_dev_list) {
+		struct rbd_device *rbd_dev;
+
+		rbd_dev = list_entry(tmp, struct rbd_device, node);
+		if (rbd_dev->id >= new_id)
+			new_id = rbd_dev->id + 1;
+	}
+
+	rbd_dev->id = new_id;
+
+	/* add to global list */
+	list_add_tail(&rbd_dev->node, &rbd_dev_list);
+
+	/* parse add command */
+	if (sscanf(buf, "%" __stringify(RBD_MAX_OPT_LEN) "s "
+		   "%" __stringify(RBD_MAX_OPT_LEN) "s "
+		   "%" __stringify(RBD_MAX_POOL_NAME_LEN) "s "
+		   "%" __stringify(RBD_MAX_OBJ_NAME_LEN) "s"
+		   "%" __stringify(RBD_MAX_SNAP_NAME_LEN) "s",
+		   mon_dev_name, options, rbd_dev->pool_name,
+		   rbd_dev->obj, rbd_dev->snap_name) < 4) {
+		rc = -EINVAL;
+		goto err_out_slot;
+	}
+
+	if (rbd_dev->snap_name[0] == 0)
+		rbd_dev->snap_name[0] = '-';
+
+	rbd_dev->obj_len = strlen(rbd_dev->obj);
+	snprintf(rbd_dev->obj_md_name, sizeof(rbd_dev->obj_md_name), "%s%s",
+		 rbd_dev->obj, RBD_SUFFIX);
+
+	/* initialize rest of new object */
+	snprintf(rbd_dev->name, DEV_NAME_LEN, DRV_NAME "%d", rbd_dev->id);
+	rc = rbd_get_client(rbd_dev, mon_dev_name, options);
+	if (rc < 0)
+		goto err_out_slot;
+
+	mutex_unlock(&ctl_mutex);
+
+	/* pick the pool */
+	osdc = &rbd_dev->client->osdc;
+	rc = ceph_pg_poolid_by_name(osdc->osdmap, rbd_dev->pool_name);
+	if (rc < 0)
+		goto err_out_client;
+	rbd_dev->poolid = rc;
+
+	/* register our block device */
+	irc = register_blkdev(0, rbd_dev->name);
+	if (irc < 0) {
+		rc = irc;
+		goto err_out_client;
+	}
+	rbd_dev->major = irc;
+
+	/* set up and announce blkdev mapping */
+	rc = rbd_init_disk(rbd_dev);
+	if (rc)
+		goto err_out_blkdev;
+
+	return count;
+
+err_out_blkdev:
+	unregister_blkdev(rbd_dev->major, rbd_dev->name);
+err_out_client:
+	rbd_put_client(rbd_dev);
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+err_out_slot:
+	list_del_init(&rbd_dev->node);
+	mutex_unlock(&ctl_mutex);
+
+	kfree(rbd_dev);
+err_out_opt:
+	kfree(options);
+err_mon_dev:
+	kfree(mon_dev_name);
+err_out_mod:
+	dout("Error adding device %s\n", buf);
+	module_put(THIS_MODULE);
+	return rc;
+}
+
+static struct rbd_device *__rbd_get_dev(unsigned long id)
+{
+	struct list_head *tmp;
+	struct rbd_device *rbd_dev;
+
+	list_for_each(tmp, &rbd_dev_list) {
+		rbd_dev = list_entry(tmp, struct rbd_device, node);
+		if (rbd_dev->id == id)
+			return rbd_dev;
+	}
+	return NULL;
+}
+
+static ssize_t class_rbd_remove(struct class *c,
+				struct class_attribute *attr,
+				const char *buf,
+				size_t count)
+{
+	struct rbd_device *rbd_dev = NULL;
+	int target_id, rc;
+	unsigned long ul;
+
+	rc = strict_strtoul(buf, 10, &ul);
+	if (rc)
+		return rc;
+
+	/* convert to int; abort if we lost anything in the conversion */
+	target_id = (int) ul;
+	if (target_id != ul)
+		return -EINVAL;
+
+	/* remove object from list immediately */
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	rbd_dev = __rbd_get_dev(target_id);
+	if (rbd_dev)
+		list_del_init(&rbd_dev->node);
+
+	mutex_unlock(&ctl_mutex);
+
+	if (!rbd_dev)
+		return -ENOENT;
+
+	rbd_put_client(rbd_dev);
+
+	/* clean up and free blkdev */
+	rbd_free_disk(rbd_dev);
+	unregister_blkdev(rbd_dev->major, rbd_dev->name);
+	kfree(rbd_dev);
+
+	/* release module ref */
+	module_put(THIS_MODULE);
+
+	return count;
+}
+
+static ssize_t class_rbd_snaps_list(struct class *c,
+			      struct class_attribute *attr,
+			      char *data)
+{
+	struct rbd_device *rbd_dev = NULL;
+	struct list_head *tmp;
+	struct rbd_image_header *header;
+	int i, n = 0, max = PAGE_SIZE;
+	int ret;
+
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	n += snprintf(data, max, "#id\tsnap\tKB\n");
+
+	list_for_each(tmp, &rbd_dev_list) {
+		char *names, *p;
+		struct ceph_snap_context *snapc;
+
+		rbd_dev = list_entry(tmp, struct rbd_device, node);
+		header = &rbd_dev->header;
+
+		down_read(&header->snap_rwsem);
+
+		names = header->snap_names;
+		snapc = header->snapc;
+
+		n += snprintf(data + n, max - n, "%d\t%s\t%lld%s\n",
+			      rbd_dev->id, RBD_SNAP_HEAD_NAME,
+			      header->image_size >> 10,
+			      (!rbd_dev->cur_snap ? " (*)" : ""));
+		if (n == max)
+			break;
+
+		p = names;
+		for (i = 0; i < header->total_snaps; i++, p += strlen(p) + 1) {
+			n += snprintf(data + n, max - n, "%d\t%s\t%lld%s\n",
+			      rbd_dev->id, p, header->snap_sizes[i] >> 10,
+			      (rbd_dev->cur_snap &&
+			       (snap_index(header, i) == rbd_dev->cur_snap) ?
+			       " (*)" : ""));
+			if (n == max)
+				break;
+		}
+
+		up_read(&header->snap_rwsem);
+	}
+
+
+	ret = n;
+	mutex_unlock(&ctl_mutex);
+	return ret;
+}
+
+static ssize_t class_rbd_snaps_refresh(struct class *c,
+				struct class_attribute *attr,
+				const char *buf,
+				size_t count)
+{
+	struct rbd_device *rbd_dev = NULL;
+	int target_id, rc;
+	unsigned long ul;
+	int ret = count;
+
+	rc = strict_strtoul(buf, 10, &ul);
+	if (rc)
+		return rc;
+
+	/* convert to int; abort if we lost anything in the conversion */
+	target_id = (int) ul;
+	if (target_id != ul)
+		return -EINVAL;
+
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	rbd_dev = __rbd_get_dev(target_id);
+	if (!rbd_dev) {
+		ret = -ENOENT;
+		goto done;
+	}
+
+	rc = rbd_update_snaps(rbd_dev);
+	if (rc < 0)
+		ret = rc;
+
+done:
+	mutex_unlock(&ctl_mutex);
+	return ret;
+}
+
+static ssize_t class_rbd_snap_create(struct class *c,
+				struct class_attribute *attr,
+				const char *buf,
+				size_t count)
+{
+	struct rbd_device *rbd_dev = NULL;
+	int target_id, ret;
+	char *name;
+
+	name = kmalloc(RBD_MAX_SNAP_NAME_LEN + 1, GFP_KERNEL);
+	if (!name)
+		return -ENOMEM;
+
+	/* parse snaps add command */
+	if (sscanf(buf, "%d "
+		   "%" __stringify(RBD_MAX_SNAP_NAME_LEN) "s",
+		   &target_id,
+		   name) != 2) {
+		ret = -EINVAL;
+		goto done;
+	}
+
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	rbd_dev = __rbd_get_dev(target_id);
+	if (!rbd_dev) {
+		ret = -ENOENT;
+		goto done_unlock;
+	}
+
+	ret = rbd_header_add_snap(rbd_dev,
+				  name, GFP_KERNEL);
+	if (ret < 0)
+		goto done_unlock;
+
+	ret = rbd_update_snaps(rbd_dev);
+	if (ret < 0)
+		goto done_unlock;
+
+	ret = count;
+done_unlock:
+	mutex_unlock(&ctl_mutex);
+done:
+	kfree(name);
+	return ret;
+}
+
+static ssize_t class_rbd_rollback(struct class *c,
+				struct class_attribute *attr,
+				const char *buf,
+				size_t count)
+{
+	struct rbd_device *rbd_dev = NULL;
+	int target_id, ret;
+	u64 snapid;
+	char snap_name[RBD_MAX_SNAP_NAME_LEN];
+	u64 cur_ofs;
+	char *seg_name;
+
+	/* parse snaps add command */
+	if (sscanf(buf, "%d "
+		   "%" __stringify(RBD_MAX_SNAP_NAME_LEN) "s",
+		   &target_id,
+		   snap_name) != 2) {
+		return -EINVAL;
+	}
+
+	ret = -ENOMEM;
+	seg_name = kmalloc(RBD_MAX_SEG_NAME_LEN + 1, GFP_NOIO);
+	if (!seg_name)
+		return ret;
+
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	rbd_dev = __rbd_get_dev(target_id);
+	if (!rbd_dev) {
+		ret = -ENOENT;
+		goto done_unlock;
+	}
+
+	ret = snap_by_name(&rbd_dev->header, snap_name, &snapid, NULL);
+	if (ret < 0)
+		goto done_unlock;
+
+	dout("snapid=%lld\n", snapid);
+
+	cur_ofs = 0;
+	while (cur_ofs < rbd_dev->header.image_size) {
+		cur_ofs += rbd_get_segment(&rbd_dev->header,
+					   rbd_dev->obj,
+					   cur_ofs, (u64)-1,
+					   seg_name, NULL);
+		dout("seg_name=%s\n", seg_name);
+
+		ret = rbd_req_sync_rollback_obj(rbd_dev, snapid, seg_name);
+		if (ret < 0)
+			pr_warning("could not roll back obj %s err=%d\n",
+				   seg_name, ret);
+	}
+
+	ret = rbd_update_snaps(rbd_dev);
+	if (ret < 0)
+		goto done_unlock;
+
+	ret = count;
+
+done_unlock:
+	mutex_unlock(&ctl_mutex);
+	kfree(seg_name);
+
+	return ret;
+}
+
+static struct class_attribute class_rbd_attrs[] = {
+	__ATTR(add,		0200, NULL, class_rbd_add),
+	__ATTR(remove,		0200, NULL, class_rbd_remove),
+	__ATTR(list,		0444, class_rbd_list, NULL),
+	__ATTR(snaps_refresh,	0200, NULL, class_rbd_snaps_refresh),
+	__ATTR(snap_create,	0200, NULL, class_rbd_snap_create),
+	__ATTR(snaps_list,	0444, class_rbd_snaps_list, NULL),
+	__ATTR(snap_rollback,	0200, NULL, class_rbd_rollback),
+	__ATTR_NULL
+};
+
+/*
+ * create control files in sysfs
+ * /sys/class/rbd/...
+ */
+static int rbd_sysfs_init(void)
+{
+	int ret = -ENOMEM;
+
+	class_rbd = kzalloc(sizeof(*class_rbd), GFP_KERNEL);
+	if (!class_rbd)
+		goto out;
+
+	class_rbd->name = DRV_NAME;
+	class_rbd->owner = THIS_MODULE;
+	class_rbd->class_release = class_rbd_release;
+	class_rbd->class_attrs = class_rbd_attrs;
+
+	ret = class_register(class_rbd);
+	if (ret)
+		goto out_class;
+	return 0;
+
+out_class:
+	kfree(class_rbd);
+	class_rbd = NULL;
+	pr_err(DRV_NAME ": failed to create class rbd\n");
+out:
+	return ret;
+}
+
+static void rbd_sysfs_cleanup(void)
+{
+	if (class_rbd)
+		class_destroy(class_rbd);
+	class_rbd = NULL;
+}
+
+int __init rbd_init(void)
+{
+	int rc;
+
+	rc = rbd_sysfs_init();
+	if (rc)
+		return rc;
+	spin_lock_init(&node_lock);
+	pr_info("loaded " DRV_NAME_LONG "\n");
+	return 0;
+}
+
+void __exit rbd_exit(void)
+{
+	rbd_sysfs_cleanup();
+}
+
+module_init(rbd_init);
+module_exit(rbd_exit);
+
+MODULE_AUTHOR("Sage Weil <sage@newdream.net>");
+MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>");
+MODULE_DESCRIPTION("rados block device");
+
+/* following authorship retained from original osdblk.c */
+MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>");
+
+MODULE_LICENSE("GPL");
diff --git a/drivers/block/rbd_types.h b/drivers/block/rbd_types.h
new file mode 100644
index 00000000000..fc6c678aa2c
--- /dev/null
+++ b/drivers/block/rbd_types.h
@@ -0,0 +1,73 @@
+/*
+ * Ceph - scalable distributed file system
+ *
+ * Copyright (C) 2004-2010 Sage Weil <sage@newdream.net>
+ *
+ * This is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License version 2.1, as published by the Free Software
+ * Foundation.  See file COPYING.
+ *
+ */
+
+#ifndef CEPH_RBD_TYPES_H
+#define CEPH_RBD_TYPES_H
+
+#include <linux/types.h>
+
+/*
+ * rbd image 'foo' consists of objects
+ *   foo.rbd      - image metadata
+ *   foo.00000000
+ *   foo.00000001
+ *   ...          - data
+ */
+
+#define RBD_SUFFIX		".rbd"
+#define RBD_DIRECTORY           "rbd_directory"
+#define RBD_INFO                "rbd_info"
+
+#define RBD_DEFAULT_OBJ_ORDER	22   /* 4MB */
+#define RBD_MIN_OBJ_ORDER       16
+#define RBD_MAX_OBJ_ORDER       30
+
+#define RBD_MAX_OBJ_NAME_LEN	96
+#define RBD_MAX_SEG_NAME_LEN	128
+
+#define RBD_COMP_NONE		0
+#define RBD_CRYPT_NONE		0
+
+#define RBD_HEADER_TEXT		"<<< Rados Block Device Image >>>\n"
+#define RBD_HEADER_SIGNATURE	"RBD"
+#define RBD_HEADER_VERSION	"001.005"
+
+struct rbd_info {
+	__le64 max_id;
+} __attribute__ ((packed));
+
+struct rbd_image_snap_ondisk {
+	__le64 id;
+	__le64 image_size;
+} __attribute__((packed));
+
+struct rbd_image_header_ondisk {
+	char text[40];
+	char block_name[24];
+	char signature[4];
+	char version[8];
+	struct {
+		__u8 order;
+		__u8 crypt_type;
+		__u8 comp_type;
+		__u8 unused;
+	} __attribute__((packed)) options;
+	__le64 image_size;
+	__le64 snap_seq;
+	__le32 snap_count;
+	__le32 reserved;
+	__le64 snap_names_len;
+	struct rbd_image_snap_ondisk snaps[0];
+} __attribute__((packed));
+
+
+#endif
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
index 1101e251a62..8320490226b 100644
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -2,7 +2,6 @@
 #include <linux/spinlock.h>
 #include <linux/slab.h>
 #include <linux/blkdev.h>
-#include <linux/smp_lock.h>
 #include <linux/hdreg.h>
 #include <linux/virtio.h>
 #include <linux/virtio_blk.h>
@@ -222,8 +221,8 @@ static int virtblk_get_id(struct gendisk *disk, char *id_str)
 	return err;
 }
 
-static int virtblk_locked_ioctl(struct block_device *bdev, fmode_t mode,
-			 unsigned cmd, unsigned long data)
+static int virtblk_ioctl(struct block_device *bdev, fmode_t mode,
+			     unsigned int cmd, unsigned long data)
 {
 	struct gendisk *disk = bdev->bd_disk;
 	struct virtio_blk *vblk = disk->private_data;
@@ -238,18 +237,6 @@ static int virtblk_locked_ioctl(struct block_device *bdev, fmode_t mode,
 			      (void __user *)data);
 }
 
-static int virtblk_ioctl(struct block_device *bdev, fmode_t mode,
-			     unsigned int cmd, unsigned long param)
-{
-	int ret;
-
-	lock_kernel();
-	ret = virtblk_locked_ioctl(bdev, mode, cmd, param);
-	unlock_kernel();
-
-	return ret;
-}
-
 /* We provide getgeo only to please some old bootloader/partitioning tools */
 static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo)
 {
diff --git a/drivers/char/agp/Kconfig b/drivers/char/agp/Kconfig
index 4b66c69eaf5..5ddf67e76f8 100644
--- a/drivers/char/agp/Kconfig
+++ b/drivers/char/agp/Kconfig
@@ -57,7 +57,7 @@ config AGP_AMD
 
 config AGP_AMD64
 	tristate "AMD Opteron/Athlon64 on-CPU GART support"
-	depends on AGP && X86 && K8_NB
+	depends on AGP && X86 && AMD_NB
 	help
 	  This option gives you AGP support for the GLX component of
 	  X using the on-CPU northbridge of the AMD Athlon64/Opteron CPUs.
diff --git a/drivers/char/agp/amd64-agp.c b/drivers/char/agp/amd64-agp.c
index 70312da4c96..42396df5555 100644
--- a/drivers/char/agp/amd64-agp.c
+++ b/drivers/char/agp/amd64-agp.c
@@ -15,7 +15,7 @@
 #include <linux/mmzone.h>
 #include <asm/page.h>		/* PAGE_SIZE */
 #include <asm/e820.h>
-#include <asm/k8.h>
+#include <asm/amd_nb.h>
 #include <asm/gart.h>
 #include "agp.h"
 
@@ -124,7 +124,7 @@ static int amd64_fetch_size(void)
 	u32 temp;
 	struct aper_size_info_32 *values;
 
-	dev = k8_northbridges[0];
+	dev = k8_northbridges.nb_misc[0];
 	if (dev==NULL)
 		return 0;
 
@@ -181,10 +181,14 @@ static int amd_8151_configure(void)
 	unsigned long gatt_bus = virt_to_phys(agp_bridge->gatt_table_real);
 	int i;
 
+	if (!k8_northbridges.gart_supported)
+		return 0;
+
 	/* Configure AGP regs in each x86-64 host bridge. */
-        for (i = 0; i < num_k8_northbridges; i++) {
+	for (i = 0; i < k8_northbridges.num; i++) {
 		agp_bridge->gart_bus_addr =
-				amd64_configure(k8_northbridges[i], gatt_bus);
+				amd64_configure(k8_northbridges.nb_misc[i],
+						gatt_bus);
 	}
 	k8_flush_garts();
 	return 0;
@@ -195,11 +199,15 @@ static void amd64_cleanup(void)
 {
 	u32 tmp;
 	int i;
-        for (i = 0; i < num_k8_northbridges; i++) {
-		struct pci_dev *dev = k8_northbridges[i];
+
+	if (!k8_northbridges.gart_supported)
+		return;
+
+	for (i = 0; i < k8_northbridges.num; i++) {
+		struct pci_dev *dev = k8_northbridges.nb_misc[i];
 		/* disable gart translation */
 		pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &tmp);
-		tmp &= ~AMD64_GARTEN;
+		tmp &= ~GARTEN;
 		pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, tmp);
 	}
 }
@@ -313,22 +321,25 @@ static __devinit int fix_northbridge(struct pci_dev *nb, struct pci_dev *agp,
 	if (order < 0 || !agp_aperture_valid(aper, (32*1024*1024)<<order))
 		return -1;
 
-	pci_write_config_dword(nb, AMD64_GARTAPERTURECTL, order << 1);
+	gart_set_size_and_enable(nb, order);
 	pci_write_config_dword(nb, AMD64_GARTAPERTUREBASE, aper >> 25);
 
 	return 0;
 }
 
-static __devinit int cache_nbs (struct pci_dev *pdev, u32 cap_ptr)
+static __devinit int cache_nbs(struct pci_dev *pdev, u32 cap_ptr)
 {
 	int i;
 
 	if (cache_k8_northbridges() < 0)
 		return -ENODEV;
 
+	if (!k8_northbridges.gart_supported)
+		return -ENODEV;
+
 	i = 0;
-	for (i = 0; i < num_k8_northbridges; i++) {
-		struct pci_dev *dev = k8_northbridges[i];
+	for (i = 0; i < k8_northbridges.num; i++) {
+		struct pci_dev *dev = k8_northbridges.nb_misc[i];
 		if (fix_northbridge(dev, pdev, cap_ptr) < 0) {
 			dev_err(&dev->dev, "no usable aperture found\n");
 #ifdef __x86_64__
@@ -405,7 +416,8 @@ static int __devinit uli_agp_init(struct pci_dev *pdev)
 	}
 
 	/* shadow x86-64 registers into ULi registers */
-	pci_read_config_dword (k8_northbridges[0], AMD64_GARTAPERTUREBASE, &httfea);
+	pci_read_config_dword (k8_northbridges.nb_misc[0], AMD64_GARTAPERTUREBASE,
+			       &httfea);
 
 	/* if x86-64 aperture base is beyond 4G, exit here */
 	if ((httfea & 0x7fff) >> (32 - 25)) {
@@ -472,7 +484,8 @@ static int nforce3_agp_init(struct pci_dev *pdev)
 	pci_write_config_dword(dev1, NVIDIA_X86_64_1_APSIZE, tmp);
 
 	/* shadow x86-64 registers into NVIDIA registers */
-	pci_read_config_dword (k8_northbridges[0], AMD64_GARTAPERTUREBASE, &apbase);
+	pci_read_config_dword (k8_northbridges.nb_misc[0], AMD64_GARTAPERTUREBASE,
+			       &apbase);
 
 	/* if x86-64 aperture base is beyond 4G, exit here */
 	if ( (apbase & 0x7fff) >> (32 - 25) ) {
diff --git a/drivers/char/agp/generic.c b/drivers/char/agp/generic.c
index d2abf514398..64255cef8a7 100644
--- a/drivers/char/agp/generic.c
+++ b/drivers/char/agp/generic.c
@@ -984,7 +984,9 @@ int agp_generic_create_gatt_table(struct agp_bridge_data *bridge)
 
 	bridge->driver->cache_flush();
 #ifdef CONFIG_X86
-	set_memory_uc((unsigned long)table, 1 << page_order);
+	if (set_memory_uc((unsigned long)table, 1 << page_order))
+		printk(KERN_WARNING "Could not set GATT table memory to UC!");
+
 	bridge->gatt_table = (void *)table;
 #else
 	bridge->gatt_table = ioremap_nocache(virt_to_phys(table),
diff --git a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c
index 05ad4a17a28..7c4133582db 100644
--- a/drivers/char/tpm/tpm.c
+++ b/drivers/char/tpm/tpm.c
@@ -47,6 +47,16 @@ enum tpm_duration {
 #define TPM_MAX_PROTECTED_ORDINAL 12
 #define TPM_PROTECTED_ORDINAL_MASK 0xFF
 
+/*
+ * Bug workaround - some TPM's don't flush the most
+ * recently changed pcr on suspend, so force the flush
+ * with an extend to the selected _unused_ non-volatile pcr.
+ */
+static int tpm_suspend_pcr;
+module_param_named(suspend_pcr, tpm_suspend_pcr, uint, 0644);
+MODULE_PARM_DESC(suspend_pcr,
+		 "PCR to use for dummy writes to faciltate flush on suspend.");
+
 static LIST_HEAD(tpm_chip_list);
 static DEFINE_SPINLOCK(driver_lock);
 static DECLARE_BITMAP(dev_mask, TPM_NUM_DEVICES);
@@ -1077,18 +1087,6 @@ static struct tpm_input_header savestate_header = {
 	.ordinal = TPM_ORD_SAVESTATE
 };
 
-/* Bug workaround - some TPM's don't flush the most
- * recently changed pcr on suspend, so force the flush
- * with an extend to the selected _unused_ non-volatile pcr.
- */
-static int tpm_suspend_pcr;
-static int __init tpm_suspend_setup(char *str)
-{
-	get_option(&str, &tpm_suspend_pcr);
-	return 1;
-}
-__setup("tpm_suspend_pcr=", tpm_suspend_setup);
-
 /*
  * We are about to suspend. Save the TPM state
  * so that it can be restored.
diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c
index 0f69c5ec0ec..6c1b676643a 100644
--- a/drivers/char/virtio_console.c
+++ b/drivers/char/virtio_console.c
@@ -48,6 +48,9 @@ struct ports_driver_data {
 	/* Used for exporting per-port information to debugfs */
 	struct dentry *debugfs_dir;
 
+	/* List of all the devices we're handling */
+	struct list_head portdevs;
+
 	/* Number of devices this driver is handling */
 	unsigned int index;
 
@@ -108,6 +111,9 @@ struct port_buffer {
  * ports for that device (vdev->priv).
  */
 struct ports_device {
+	/* Next portdev in the list, head is in the pdrvdata struct */
+	struct list_head list;
+
 	/*
 	 * Workqueue handlers where we process deferred work after
 	 * notification
@@ -178,15 +184,21 @@ struct port {
 	struct console cons;
 
 	/* Each port associates with a separate char device */
-	struct cdev cdev;
+	struct cdev *cdev;
 	struct device *dev;
 
+	/* Reference-counting to handle port hot-unplugs and file operations */
+	struct kref kref;
+
 	/* A waitqueue for poll() or blocking read operations */
 	wait_queue_head_t waitqueue;
 
 	/* The 'name' of the port that we expose via sysfs properties */
 	char *name;
 
+	/* We can notify apps of host connect / disconnect events via SIGIO */
+	struct fasync_struct *async_queue;
+
 	/* The 'id' to identify the port with the Host */
 	u32 id;
 
@@ -221,6 +233,41 @@ out:
 	return port;
 }
 
+static struct port *find_port_by_devt_in_portdev(struct ports_device *portdev,
+						 dev_t dev)
+{
+	struct port *port;
+	unsigned long flags;
+
+	spin_lock_irqsave(&portdev->ports_lock, flags);
+	list_for_each_entry(port, &portdev->ports, list)
+		if (port->cdev->dev == dev)
+			goto out;
+	port = NULL;
+out:
+	spin_unlock_irqrestore(&portdev->ports_lock, flags);
+
+	return port;
+}
+
+static struct port *find_port_by_devt(dev_t dev)
+{
+	struct ports_device *portdev;
+	struct port *port;
+	unsigned long flags;
+
+	spin_lock_irqsave(&pdrvdata_lock, flags);
+	list_for_each_entry(portdev, &pdrvdata.portdevs, list) {
+		port = find_port_by_devt_in_portdev(portdev, dev);
+		if (port)
+			goto out;
+	}
+	port = NULL;
+out:
+	spin_unlock_irqrestore(&pdrvdata_lock, flags);
+	return port;
+}
+
 static struct port *find_port_by_id(struct ports_device *portdev, u32 id)
 {
 	struct port *port;
@@ -410,7 +457,10 @@ static ssize_t __send_control_msg(struct ports_device *portdev, u32 port_id,
 static ssize_t send_control_msg(struct port *port, unsigned int event,
 				unsigned int value)
 {
-	return __send_control_msg(port->portdev, port->id, event, value);
+	/* Did the port get unplugged before userspace closed it? */
+	if (port->portdev)
+		return __send_control_msg(port->portdev, port->id, event, value);
+	return 0;
 }
 
 /* Callers must take the port->outvq_lock */
@@ -525,6 +575,10 @@ static ssize_t fill_readbuf(struct port *port, char *out_buf, size_t out_count,
 /* The condition that must be true for polling to end */
 static bool will_read_block(struct port *port)
 {
+	if (!port->guest_connected) {
+		/* Port got hot-unplugged. Let's exit. */
+		return false;
+	}
 	return !port_has_data(port) && port->host_connected;
 }
 
@@ -575,6 +629,9 @@ static ssize_t port_fops_read(struct file *filp, char __user *ubuf,
 		if (ret < 0)
 			return ret;
 	}
+	/* Port got hot-unplugged. */
+	if (!port->guest_connected)
+		return -ENODEV;
 	/*
 	 * We could've received a disconnection message while we were
 	 * waiting for more data.
@@ -616,6 +673,9 @@ static ssize_t port_fops_write(struct file *filp, const char __user *ubuf,
 		if (ret < 0)
 			return ret;
 	}
+	/* Port got hot-unplugged. */
+	if (!port->guest_connected)
+		return -ENODEV;
 
 	count = min((size_t)(32 * 1024), count);
 
@@ -656,6 +716,10 @@ static unsigned int port_fops_poll(struct file *filp, poll_table *wait)
 	port = filp->private_data;
 	poll_wait(filp, &port->waitqueue, wait);
 
+	if (!port->guest_connected) {
+		/* Port got unplugged */
+		return POLLHUP;
+	}
 	ret = 0;
 	if (!will_read_block(port))
 		ret |= POLLIN | POLLRDNORM;
@@ -667,6 +731,8 @@ static unsigned int port_fops_poll(struct file *filp, poll_table *wait)
 	return ret;
 }
 
+static void remove_port(struct kref *kref);
+
 static int port_fops_release(struct inode *inode, struct file *filp)
 {
 	struct port *port;
@@ -687,6 +753,16 @@ static int port_fops_release(struct inode *inode, struct file *filp)
 	reclaim_consumed_buffers(port);
 	spin_unlock_irq(&port->outvq_lock);
 
+	/*
+	 * Locks aren't necessary here as a port can't be opened after
+	 * unplug, and if a port isn't unplugged, a kref would already
+	 * exist for the port.  Plus, taking ports_lock here would
+	 * create a dependency on other locks taken by functions
+	 * inside remove_port if we're the last holder of the port,
+	 * creating many problems.
+	 */
+	kref_put(&port->kref, remove_port);
+
 	return 0;
 }
 
@@ -694,22 +770,31 @@ static int port_fops_open(struct inode *inode, struct file *filp)
 {
 	struct cdev *cdev = inode->i_cdev;
 	struct port *port;
+	int ret;
 
-	port = container_of(cdev, struct port, cdev);
+	port = find_port_by_devt(cdev->dev);
 	filp->private_data = port;
 
+	/* Prevent against a port getting hot-unplugged at the same time */
+	spin_lock_irq(&port->portdev->ports_lock);
+	kref_get(&port->kref);
+	spin_unlock_irq(&port->portdev->ports_lock);
+
 	/*
 	 * Don't allow opening of console port devices -- that's done
 	 * via /dev/hvc
 	 */
-	if (is_console_port(port))
-		return -ENXIO;
+	if (is_console_port(port)) {
+		ret = -ENXIO;
+		goto out;
+	}
 
 	/* Allow only one process to open a particular port at a time */
 	spin_lock_irq(&port->inbuf_lock);
 	if (port->guest_connected) {
 		spin_unlock_irq(&port->inbuf_lock);
-		return -EMFILE;
+		ret = -EMFILE;
+		goto out;
 	}
 
 	port->guest_connected = true;
@@ -724,10 +809,23 @@ static int port_fops_open(struct inode *inode, struct file *filp)
 	reclaim_consumed_buffers(port);
 	spin_unlock_irq(&port->outvq_lock);
 
+	nonseekable_open(inode, filp);
+
 	/* Notify host of port being opened */
 	send_control_msg(filp->private_data, VIRTIO_CONSOLE_PORT_OPEN, 1);
 
 	return 0;
+out:
+	kref_put(&port->kref, remove_port);
+	return ret;
+}
+
+static int port_fops_fasync(int fd, struct file *filp, int mode)
+{
+	struct port *port;
+
+	port = filp->private_data;
+	return fasync_helper(fd, filp, mode, &port->async_queue);
 }
 
 /*
@@ -743,6 +841,8 @@ static const struct file_operations port_fops = {
 	.write = port_fops_write,
 	.poll  = port_fops_poll,
 	.release = port_fops_release,
+	.fasync = port_fops_fasync,
+	.llseek = no_llseek,
 };
 
 /*
@@ -1001,6 +1101,12 @@ static unsigned int fill_queue(struct virtqueue *vq, spinlock_t *lock)
 	return nr_added_bufs;
 }
 
+static void send_sigio_to_port(struct port *port)
+{
+	if (port->async_queue && port->guest_connected)
+		kill_fasync(&port->async_queue, SIGIO, POLL_OUT);
+}
+
 static int add_port(struct ports_device *portdev, u32 id)
 {
 	char debugfs_name[16];
@@ -1015,6 +1121,7 @@ static int add_port(struct ports_device *portdev, u32 id)
 		err = -ENOMEM;
 		goto fail;
 	}
+	kref_init(&port->kref);
 
 	port->portdev = portdev;
 	port->id = id;
@@ -1022,6 +1129,7 @@ static int add_port(struct ports_device *portdev, u32 id)
 	port->name = NULL;
 	port->inbuf = NULL;
 	port->cons.hvc = NULL;
+	port->async_queue = NULL;
 
 	port->cons.ws.ws_row = port->cons.ws.ws_col = 0;
 
@@ -1032,14 +1140,20 @@ static int add_port(struct ports_device *portdev, u32 id)
 	port->in_vq = portdev->in_vqs[port->id];
 	port->out_vq = portdev->out_vqs[port->id];
 
-	cdev_init(&port->cdev, &port_fops);
+	port->cdev = cdev_alloc();
+	if (!port->cdev) {
+		dev_err(&port->portdev->vdev->dev, "Error allocating cdev\n");
+		err = -ENOMEM;
+		goto free_port;
+	}
+	port->cdev->ops = &port_fops;
 
 	devt = MKDEV(portdev->chr_major, id);
-	err = cdev_add(&port->cdev, devt, 1);
+	err = cdev_add(port->cdev, devt, 1);
 	if (err < 0) {
 		dev_err(&port->portdev->vdev->dev,
 			"Error %d adding cdev for port %u\n", err, id);
-		goto free_port;
+		goto free_cdev;
 	}
 	port->dev = device_create(pdrvdata.class, &port->portdev->vdev->dev,
 				  devt, port, "vport%up%u",
@@ -1104,7 +1218,7 @@ free_inbufs:
 free_device:
 	device_destroy(pdrvdata.class, port->dev->devt);
 free_cdev:
-	cdev_del(&port->cdev);
+	cdev_del(port->cdev);
 free_port:
 	kfree(port);
 fail:
@@ -1113,21 +1227,45 @@ fail:
 	return err;
 }
 
-/* Remove all port-specific data. */
-static int remove_port(struct port *port)
+/* No users remain, remove all port-specific data. */
+static void remove_port(struct kref *kref)
+{
+	struct port *port;
+
+	port = container_of(kref, struct port, kref);
+
+	sysfs_remove_group(&port->dev->kobj, &port_attribute_group);
+	device_destroy(pdrvdata.class, port->dev->devt);
+	cdev_del(port->cdev);
+
+	kfree(port->name);
+
+	debugfs_remove(port->debugfs_file);
+
+	kfree(port);
+}
+
+/*
+ * Port got unplugged.  Remove port from portdev's list and drop the
+ * kref reference.  If no userspace has this port opened, it will
+ * result in immediate removal the port.
+ */
+static void unplug_port(struct port *port)
 {
 	struct port_buffer *buf;
 
+	spin_lock_irq(&port->portdev->ports_lock);
+	list_del(&port->list);
+	spin_unlock_irq(&port->portdev->ports_lock);
+
 	if (port->guest_connected) {
 		port->guest_connected = false;
 		port->host_connected = false;
 		wake_up_interruptible(&port->waitqueue);
-		send_control_msg(port, VIRTIO_CONSOLE_PORT_OPEN, 0);
-	}
 
-	spin_lock_irq(&port->portdev->ports_lock);
-	list_del(&port->list);
-	spin_unlock_irq(&port->portdev->ports_lock);
+		/* Let the app know the port is going down. */
+		send_sigio_to_port(port);
+	}
 
 	if (is_console_port(port)) {
 		spin_lock_irq(&pdrvdata_lock);
@@ -1146,9 +1284,6 @@ static int remove_port(struct port *port)
 		hvc_remove(port->cons.hvc);
 #endif
 	}
-	sysfs_remove_group(&port->dev->kobj, &port_attribute_group);
-	device_destroy(pdrvdata.class, port->dev->devt);
-	cdev_del(&port->cdev);
 
 	/* Remove unused data this port might have received. */
 	discard_port_data(port);
@@ -1159,12 +1294,19 @@ static int remove_port(struct port *port)
 	while ((buf = virtqueue_detach_unused_buf(port->in_vq)))
 		free_buf(buf);
 
-	kfree(port->name);
-
-	debugfs_remove(port->debugfs_file);
+	/*
+	 * We should just assume the device itself has gone off --
+	 * else a close on an open port later will try to send out a
+	 * control message.
+	 */
+	port->portdev = NULL;
 
-	kfree(port);
-	return 0;
+	/*
+	 * Locks around here are not necessary - a port can't be
+	 * opened after we removed the port struct from ports_list
+	 * above.
+	 */
+	kref_put(&port->kref, remove_port);
 }
 
 /* Any private messages that the Host and Guest want to share */
@@ -1203,7 +1345,7 @@ static void handle_control_message(struct ports_device *portdev,
 		add_port(portdev, cpkt->id);
 		break;
 	case VIRTIO_CONSOLE_PORT_REMOVE:
-		remove_port(port);
+		unplug_port(port);
 		break;
 	case VIRTIO_CONSOLE_CONSOLE_PORT:
 		if (!cpkt->value)
@@ -1245,6 +1387,12 @@ static void handle_control_message(struct ports_device *portdev,
 		spin_lock_irq(&port->outvq_lock);
 		reclaim_consumed_buffers(port);
 		spin_unlock_irq(&port->outvq_lock);
+
+		/*
+		 * If the guest is connected, it'll be interested in
+		 * knowing the host connection state changed.
+		 */
+		send_sigio_to_port(port);
 		break;
 	case VIRTIO_CONSOLE_PORT_NAME:
 		/*
@@ -1341,6 +1489,9 @@ static void in_intr(struct virtqueue *vq)
 
 	wake_up_interruptible(&port->waitqueue);
 
+	/* Send a SIGIO indicating new data in case the process asked for it */
+	send_sigio_to_port(port);
+
 	if (is_console_port(port) && hvc_poll(port->cons.hvc))
 		hvc_kick();
 }
@@ -1577,6 +1728,10 @@ static int __devinit virtcons_probe(struct virtio_device *vdev)
 		add_port(portdev, 0);
 	}
 
+	spin_lock_irq(&pdrvdata_lock);
+	list_add_tail(&portdev->list, &pdrvdata.portdevs);
+	spin_unlock_irq(&pdrvdata_lock);
+
 	__send_control_msg(portdev, VIRTIO_CONSOLE_BAD_ID,
 			   VIRTIO_CONSOLE_DEVICE_READY, 1);
 	return 0;
@@ -1600,23 +1755,41 @@ static void virtcons_remove(struct virtio_device *vdev)
 {
 	struct ports_device *portdev;
 	struct port *port, *port2;
-	struct port_buffer *buf;
-	unsigned int len;
 
 	portdev = vdev->priv;
 
+	spin_lock_irq(&pdrvdata_lock);
+	list_del(&portdev->list);
+	spin_unlock_irq(&pdrvdata_lock);
+
+	/* Disable interrupts for vqs */
+	vdev->config->reset(vdev);
+	/* Finish up work that's lined up */
 	cancel_work_sync(&portdev->control_work);
 
 	list_for_each_entry_safe(port, port2, &portdev->ports, list)
-		remove_port(port);
+		unplug_port(port);
 
 	unregister_chrdev(portdev->chr_major, "virtio-portsdev");
 
-	while ((buf = virtqueue_get_buf(portdev->c_ivq, &len)))
-		free_buf(buf);
+	/*
+	 * When yanking out a device, we immediately lose the
+	 * (device-side) queues.  So there's no point in keeping the
+	 * guest side around till we drop our final reference.  This
+	 * also means that any ports which are in an open state will
+	 * have to just stop using the port, as the vqs are going
+	 * away.
+	 */
+	if (use_multiport(portdev)) {
+		struct port_buffer *buf;
+		unsigned int len;
 
-	while ((buf = virtqueue_detach_unused_buf(portdev->c_ivq)))
-		free_buf(buf);
+		while ((buf = virtqueue_get_buf(portdev->c_ivq, &len)))
+			free_buf(buf);
+
+		while ((buf = virtqueue_detach_unused_buf(portdev->c_ivq)))
+			free_buf(buf);
+	}
 
 	vdev->config->del_vqs(vdev);
 	kfree(portdev->in_vqs);
@@ -1663,6 +1836,7 @@ static int __init init(void)
 			   PTR_ERR(pdrvdata.debugfs_dir));
 	}
 	INIT_LIST_HEAD(&pdrvdata.consoles);
+	INIT_LIST_HEAD(&pdrvdata.portdevs);
 
 	return register_virtio_driver(&virtio_console);
 }
diff --git a/drivers/edac/Kconfig b/drivers/edac/Kconfig
index 70bb350de99..9dbb28b9559 100644
--- a/drivers/edac/Kconfig
+++ b/drivers/edac/Kconfig
@@ -39,7 +39,7 @@ config EDAC_DEBUG
 	  there're four debug levels (x=0,1,2,3 from low to high).
 	  Usually you should select 'N'.
 
- config EDAC_DECODE_MCE
+config EDAC_DECODE_MCE
 	tristate "Decode MCEs in human-readable form (only on AMD for now)"
 	depends on CPU_SUP_AMD && X86_MCE
 	default y
@@ -51,6 +51,16 @@ config EDAC_DEBUG
 	  which occur really early upon boot, before the module infrastructure
 	  has been initialized.
 
+config EDAC_MCE_INJ
+	tristate "Simple MCE injection interface over /sysfs"
+	depends on EDAC_DECODE_MCE
+	default n
+	help
+	  This is a simple interface to inject MCEs over /sysfs and test
+	  the MCE decoding code in EDAC.
+
+	  This is currently AMD-only.
+
 config EDAC_MM_EDAC
 	tristate "Main Memory EDAC (Error Detection And Correction) reporting"
 	help
@@ -66,13 +76,13 @@ config EDAC_MCE
 
 config EDAC_AMD64
 	tristate "AMD64 (Opteron, Athlon64) K8, F10h, F11h"
-	depends on EDAC_MM_EDAC && K8_NB && X86_64 && PCI && EDAC_DECODE_MCE
+	depends on EDAC_MM_EDAC && AMD_NB && X86_64 && PCI && EDAC_DECODE_MCE
 	help
 	  Support for error detection and correction on the AMD 64
 	  Families of Memory Controllers (K8, F10h and F11h)
 
 config EDAC_AMD64_ERROR_INJECTION
-	bool "Sysfs Error Injection facilities"
+	bool "Sysfs HW Error injection facilities"
 	depends on EDAC_AMD64
 	help
 	  Recent Opterons (Family 10h and later) provide for Memory Error
diff --git a/drivers/edac/Makefile b/drivers/edac/Makefile
index ca6b1bb24cc..32c7bc93c52 100644
--- a/drivers/edac/Makefile
+++ b/drivers/edac/Makefile
@@ -17,6 +17,9 @@ ifdef CONFIG_PCI
 edac_core-objs	+= edac_pci.o edac_pci_sysfs.o
 endif
 
+obj-$(CONFIG_EDAC_MCE_INJ)		+= mce_amd_inj.o
+
+edac_mce_amd-objs			:= mce_amd.o
 obj-$(CONFIG_EDAC_DECODE_MCE)		+= edac_mce_amd.o
 
 obj-$(CONFIG_EDAC_AMD76X)		+= amd76x_edac.o
diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c
index e7d5d6b5dcf..8521401bbd7 100644
--- a/drivers/edac/amd64_edac.c
+++ b/drivers/edac/amd64_edac.c
@@ -1,5 +1,5 @@
 #include "amd64_edac.h"
-#include <asm/k8.h>
+#include <asm/amd_nb.h>
 
 static struct edac_pci_ctl_info *amd64_ctl_pci;
 
@@ -2073,11 +2073,18 @@ static inline void __amd64_decode_bus_error(struct mem_ctl_info *mci,
 		amd64_handle_ue(mci, info);
 }
 
-void amd64_decode_bus_error(int node_id, struct err_regs *regs)
+void amd64_decode_bus_error(int node_id, struct mce *m, u32 nbcfg)
 {
 	struct mem_ctl_info *mci = mci_lookup[node_id];
+	struct err_regs regs;
 
-	__amd64_decode_bus_error(mci, regs);
+	regs.nbsl  = (u32) m->status;
+	regs.nbsh  = (u32)(m->status >> 32);
+	regs.nbeal = (u32) m->addr;
+	regs.nbeah = (u32)(m->addr >> 32);
+	regs.nbcfg = nbcfg;
+
+	__amd64_decode_bus_error(mci, &regs);
 
 	/*
 	 * Check the UE bit of the NB status high register, if set generate some
@@ -2086,7 +2093,7 @@ void amd64_decode_bus_error(int node_id, struct err_regs *regs)
 	 *
 	 * FIXME: this should go somewhere else, if at all.
 	 */
-	if (regs->nbsh & K8_NBSH_UC_ERR && !report_gart_errors)
+	if (regs.nbsh & K8_NBSH_UC_ERR && !report_gart_errors)
 		edac_mc_handle_ue_no_info(mci, "UE bit is set");
 
 }
@@ -2927,7 +2934,7 @@ static int __init amd64_edac_init(void)
 	 * to finish initialization of the MC instances.
 	 */
 	err = -ENODEV;
-	for (nb = 0; nb < num_k8_northbridges; nb++) {
+	for (nb = 0; nb < k8_northbridges.num; nb++) {
 		if (!pvt_lookup[nb])
 			continue;
 
diff --git a/drivers/edac/amd64_edac.h b/drivers/edac/amd64_edac.h
index 613b9381e71..044aee4f944 100644
--- a/drivers/edac/amd64_edac.h
+++ b/drivers/edac/amd64_edac.h
@@ -72,7 +72,7 @@
 #include <linux/edac.h>
 #include <asm/msr.h>
 #include "edac_core.h"
-#include "edac_mce_amd.h"
+#include "mce_amd.h"
 
 #define amd64_printk(level, fmt, arg...) \
 	edac_printk(level, "amd64", fmt, ##arg)
@@ -482,11 +482,10 @@ extern const char *rrrr_msgs[16];
 extern const char *to_msgs[2];
 extern const char *pp_msgs[4];
 extern const char *ii_msgs[4];
-extern const char *ext_msgs[32];
 extern const char *htlink_msgs[8];
 
 #ifdef CONFIG_EDAC_DEBUG
-#define NUM_DBG_ATTRS 9
+#define NUM_DBG_ATTRS 5
 #else
 #define NUM_DBG_ATTRS 0
 #endif
diff --git a/drivers/edac/amd64_edac_dbg.c b/drivers/edac/amd64_edac_dbg.c
index 59cf2cf6e11..e3562288f4c 100644
--- a/drivers/edac/amd64_edac_dbg.c
+++ b/drivers/edac/amd64_edac_dbg.c
@@ -1,167 +1,16 @@
 #include "amd64_edac.h"
 
-/*
- * accept a hex value and store it into the virtual error register file, field:
- * nbeal and nbeah. Assume virtual error values have already been set for: NBSL,
- * NBSH and NBCFG. Then proceed to map the error values to a MC, CSROW and
- * CHANNEL
- */
-static ssize_t amd64_nbea_store(struct mem_ctl_info *mci, const char *data,
-				size_t count)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-	unsigned long long value;
-	int ret = 0;
-
-	ret = strict_strtoull(data, 16, &value);
-	if (ret != -EINVAL) {
-		debugf0("received NBEA= 0x%llx\n", value);
-
-		/* place the value into the virtual error packet */
-		pvt->ctl_error_info.nbeal = (u32) value;
-		value >>= 32;
-		pvt->ctl_error_info.nbeah = (u32) value;
-
-		/* Process the Mapping request */
-		/* TODO: Add race prevention */
-		amd_decode_nb_mce(pvt->mc_node_id, &pvt->ctl_error_info, 1);
-
-		return count;
-	}
-	return ret;
-}
-
-/* display back what the last NBEA (MCA NB Address (MC4_ADDR)) was written */
-static ssize_t amd64_nbea_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-	u64 value;
-
-	value = pvt->ctl_error_info.nbeah;
-	value <<= 32;
-	value |= pvt->ctl_error_info.nbeal;
-
-	return sprintf(data, "%llx\n", value);
-}
-
-/* store the NBSL (MCA NB Status Low (MC4_STATUS)) value user desires */
-static ssize_t amd64_nbsl_store(struct mem_ctl_info *mci, const char *data,
-				size_t count)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-	unsigned long value;
-	int ret = 0;
-
-	ret = strict_strtoul(data, 16, &value);
-	if (ret != -EINVAL) {
-		debugf0("received NBSL= 0x%lx\n", value);
-
-		pvt->ctl_error_info.nbsl = (u32) value;
-
-		return count;
-	}
-	return ret;
-}
-
-/* display back what the last NBSL value written */
-static ssize_t amd64_nbsl_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-	u32 value;
-
-	value = pvt->ctl_error_info.nbsl;
-
-	return sprintf(data, "%x\n", value);
-}
-
-/* store the NBSH (MCA NB Status High) value user desires */
-static ssize_t amd64_nbsh_store(struct mem_ctl_info *mci, const char *data,
-				size_t count)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-	unsigned long value;
-	int ret = 0;
-
-	ret = strict_strtoul(data, 16, &value);
-	if (ret != -EINVAL) {
-		debugf0("received NBSH= 0x%lx\n", value);
-
-		pvt->ctl_error_info.nbsh = (u32) value;
-
-		return count;
-	}
-	return ret;
-}
-
-/* display back what the last NBSH value written */
-static ssize_t amd64_nbsh_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-	u32 value;
-
-	value = pvt->ctl_error_info.nbsh;
-
-	return sprintf(data, "%x\n", value);
+#define EDAC_DCT_ATTR_SHOW(reg)						\
+static ssize_t amd64_##reg##_show(struct mem_ctl_info *mci, char *data)	\
+{									\
+	struct amd64_pvt *pvt = mci->pvt_info;				\
+		return sprintf(data, "0x%016llx\n", (u64)pvt->reg);	\
 }
 
-/* accept and store the NBCFG (MCA NB Configuration) value user desires */
-static ssize_t amd64_nbcfg_store(struct mem_ctl_info *mci,
-					const char *data, size_t count)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-	unsigned long value;
-	int ret = 0;
-
-	ret = strict_strtoul(data, 16, &value);
-	if (ret != -EINVAL) {
-		debugf0("received NBCFG= 0x%lx\n", value);
-
-		pvt->ctl_error_info.nbcfg = (u32) value;
-
-		return count;
-	}
-	return ret;
-}
-
-/* various show routines for the controls of a MCI */
-static ssize_t amd64_nbcfg_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-
-	return sprintf(data, "%x\n", pvt->ctl_error_info.nbcfg);
-}
-
-
-static ssize_t amd64_dhar_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-
-	return sprintf(data, "%x\n", pvt->dhar);
-}
-
-
-static ssize_t amd64_dbam_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-
-	return sprintf(data, "%x\n", pvt->dbam0);
-}
-
-
-static ssize_t amd64_topmem_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-
-	return sprintf(data, "%llx\n", pvt->top_mem);
-}
-
-
-static ssize_t amd64_topmem2_show(struct mem_ctl_info *mci, char *data)
-{
-	struct amd64_pvt *pvt = mci->pvt_info;
-
-	return sprintf(data, "%llx\n", pvt->top_mem2);
-}
+EDAC_DCT_ATTR_SHOW(dhar);
+EDAC_DCT_ATTR_SHOW(dbam0);
+EDAC_DCT_ATTR_SHOW(top_mem);
+EDAC_DCT_ATTR_SHOW(top_mem2);
 
 static ssize_t amd64_hole_show(struct mem_ctl_info *mci, char *data)
 {
@@ -182,38 +31,6 @@ struct mcidev_sysfs_attribute amd64_dbg_attrs[] = {
 
 	{
 		.attr = {
-			.name = "nbea_ctl",
-			.mode = (S_IRUGO | S_IWUSR)
-		},
-		.show = amd64_nbea_show,
-		.store = amd64_nbea_store,
-	},
-	{
-		.attr = {
-			.name = "nbsl_ctl",
-			.mode = (S_IRUGO | S_IWUSR)
-		},
-		.show = amd64_nbsl_show,
-		.store = amd64_nbsl_store,
-	},
-	{
-		.attr = {
-			.name = "nbsh_ctl",
-			.mode = (S_IRUGO | S_IWUSR)
-		},
-		.show = amd64_nbsh_show,
-		.store = amd64_nbsh_store,
-	},
-	{
-		.attr = {
-			.name = "nbcfg_ctl",
-			.mode = (S_IRUGO | S_IWUSR)
-		},
-		.show = amd64_nbcfg_show,
-		.store = amd64_nbcfg_store,
-	},
-	{
-		.attr = {
 			.name = "dhar",
 			.mode = (S_IRUGO)
 		},
@@ -225,7 +42,7 @@ struct mcidev_sysfs_attribute amd64_dbg_attrs[] = {
 			.name = "dbam",
 			.mode = (S_IRUGO)
 		},
-		.show = amd64_dbam_show,
+		.show = amd64_dbam0_show,
 		.store = NULL,
 	},
 	{
@@ -233,7 +50,7 @@ struct mcidev_sysfs_attribute amd64_dbg_attrs[] = {
 			.name = "topmem",
 			.mode = (S_IRUGO)
 		},
-		.show = amd64_topmem_show,
+		.show = amd64_top_mem_show,
 		.store = NULL,
 	},
 	{
@@ -241,7 +58,7 @@ struct mcidev_sysfs_attribute amd64_dbg_attrs[] = {
 			.name = "topmem2",
 			.mode = (S_IRUGO)
 		},
-		.show = amd64_topmem2_show,
+		.show = amd64_top_mem2_show,
 		.store = NULL,
 	},
 	{
diff --git a/drivers/edac/edac_device_sysfs.c b/drivers/edac/edac_device_sysfs.c
index 070968178a2..2941dca91aa 100644
--- a/drivers/edac/edac_device_sysfs.c
+++ b/drivers/edac/edac_device_sysfs.c
@@ -13,6 +13,7 @@
 #include <linux/ctype.h>
 #include <linux/module.h>
 #include <linux/slab.h>
+#include <linux/edac.h>
 
 #include "edac_core.h"
 #include "edac_module.h"
@@ -235,7 +236,7 @@ int edac_device_register_sysfs_main_kobj(struct edac_device_ctl_info *edac_dev)
 	debugf1("%s()\n", __func__);
 
 	/* get the /sys/devices/system/edac reference */
-	edac_class = edac_get_edac_class();
+	edac_class = edac_get_sysfs_class();
 	if (edac_class == NULL) {
 		debugf1("%s() no edac_class error\n", __func__);
 		err = -ENODEV;
@@ -255,7 +256,7 @@ int edac_device_register_sysfs_main_kobj(struct edac_device_ctl_info *edac_dev)
 
 	if (!try_module_get(edac_dev->owner)) {
 		err = -ENODEV;
-		goto err_out;
+		goto err_mod_get;
 	}
 
 	/* register */
@@ -282,6 +283,9 @@ int edac_device_register_sysfs_main_kobj(struct edac_device_ctl_info *edac_dev)
 err_kobj_reg:
 	module_put(edac_dev->owner);
 
+err_mod_get:
+	edac_put_sysfs_class();
+
 err_out:
 	return err;
 }
@@ -290,12 +294,11 @@ err_out:
  * edac_device_unregister_sysfs_main_kobj:
  *	the '..../edac/<name>' kobject
  */
-void edac_device_unregister_sysfs_main_kobj(
-					struct edac_device_ctl_info *edac_dev)
+void edac_device_unregister_sysfs_main_kobj(struct edac_device_ctl_info *dev)
 {
 	debugf0("%s()\n", __func__);
 	debugf4("%s() name of kobject is: %s\n",
-		__func__, kobject_name(&edac_dev->kobj));
+		__func__, kobject_name(&dev->kobj));
 
 	/*
 	 * Unregister the edac device's kobject and
@@ -304,7 +307,8 @@ void edac_device_unregister_sysfs_main_kobj(
 	 *   a) module_put() this module
 	 *   b) 'kfree' the memory
 	 */
-	kobject_put(&edac_dev->kobj);
+	kobject_put(&dev->kobj);
+	edac_put_sysfs_class();
 }
 
 /* edac_dev -> instance information */
diff --git a/drivers/edac/edac_mc_sysfs.c b/drivers/edac/edac_mc_sysfs.c
index 8aad94d10c0..a4135860149 100644
--- a/drivers/edac/edac_mc_sysfs.c
+++ b/drivers/edac/edac_mc_sysfs.c
@@ -11,6 +11,7 @@
 
 #include <linux/ctype.h>
 #include <linux/slab.h>
+#include <linux/edac.h>
 #include <linux/bug.h>
 
 #include "edac_core.h"
@@ -1011,13 +1012,13 @@ void edac_remove_sysfs_mci_device(struct mem_ctl_info *mci)
  */
 int edac_sysfs_setup_mc_kset(void)
 {
-	int err = 0;
+	int err = -EINVAL;
 	struct sysdev_class *edac_class;
 
 	debugf1("%s()\n", __func__);
 
 	/* get the /sys/devices/system/edac class reference */
-	edac_class = edac_get_edac_class();
+	edac_class = edac_get_sysfs_class();
 	if (edac_class == NULL) {
 		debugf1("%s() no edac_class error=%d\n", __func__, err);
 		goto fail_out;
@@ -1028,15 +1029,16 @@ int edac_sysfs_setup_mc_kset(void)
 	if (!mc_kset) {
 		err = -ENOMEM;
 		debugf1("%s() Failed to register '.../edac/mc'\n", __func__);
-		goto fail_out;
+		goto fail_kset;
 	}
 
 	debugf1("%s() Registered '.../edac/mc' kobject\n", __func__);
 
 	return 0;
 
+fail_kset:
+	edac_put_sysfs_class();
 
-	/* error unwind stack */
 fail_out:
 	return err;
 }
@@ -1049,5 +1051,6 @@ fail_out:
 void edac_sysfs_teardown_mc_kset(void)
 {
 	kset_unregister(mc_kset);
+	edac_put_sysfs_class();
 }
 
diff --git a/drivers/edac/edac_mce_amd.c b/drivers/edac/edac_mce_amd.c
deleted file mode 100644
index 9014df6f605..00000000000
--- a/drivers/edac/edac_mce_amd.c
+++ /dev/null
@@ -1,452 +0,0 @@
-#include <linux/module.h>
-#include "edac_mce_amd.h"
-
-static bool report_gart_errors;
-static void (*nb_bus_decoder)(int node_id, struct err_regs *regs);
-
-void amd_report_gart_errors(bool v)
-{
-	report_gart_errors = v;
-}
-EXPORT_SYMBOL_GPL(amd_report_gart_errors);
-
-void amd_register_ecc_decoder(void (*f)(int, struct err_regs *))
-{
-	nb_bus_decoder = f;
-}
-EXPORT_SYMBOL_GPL(amd_register_ecc_decoder);
-
-void amd_unregister_ecc_decoder(void (*f)(int, struct err_regs *))
-{
-	if (nb_bus_decoder) {
-		WARN_ON(nb_bus_decoder != f);
-
-		nb_bus_decoder = NULL;
-	}
-}
-EXPORT_SYMBOL_GPL(amd_unregister_ecc_decoder);
-
-/*
- * string representation for the different MCA reported error types, see F3x48
- * or MSR0000_0411.
- */
-const char *tt_msgs[] = {        /* transaction type */
-	"instruction",
-	"data",
-	"generic",
-	"reserved"
-};
-EXPORT_SYMBOL_GPL(tt_msgs);
-
-const char *ll_msgs[] = {	/* cache level */
-	"L0",
-	"L1",
-	"L2",
-	"L3/generic"
-};
-EXPORT_SYMBOL_GPL(ll_msgs);
-
-const char *rrrr_msgs[] = {
-	"generic",
-	"generic read",
-	"generic write",
-	"data read",
-	"data write",
-	"inst fetch",
-	"prefetch",
-	"evict",
-	"snoop",
-	"reserved RRRR= 9",
-	"reserved RRRR= 10",
-	"reserved RRRR= 11",
-	"reserved RRRR= 12",
-	"reserved RRRR= 13",
-	"reserved RRRR= 14",
-	"reserved RRRR= 15"
-};
-EXPORT_SYMBOL_GPL(rrrr_msgs);
-
-const char *pp_msgs[] = {	/* participating processor */
-	"local node originated (SRC)",
-	"local node responded to request (RES)",
-	"local node observed as 3rd party (OBS)",
-	"generic"
-};
-EXPORT_SYMBOL_GPL(pp_msgs);
-
-const char *to_msgs[] = {
-	"no timeout",
-	"timed out"
-};
-EXPORT_SYMBOL_GPL(to_msgs);
-
-const char *ii_msgs[] = {	/* memory or i/o */
-	"mem access",
-	"reserved",
-	"i/o access",
-	"generic"
-};
-EXPORT_SYMBOL_GPL(ii_msgs);
-
-/*
- * Map the 4 or 5 (family-specific) bits of Extended Error code to the
- * string table.
- */
-const char *ext_msgs[] = {
-	"K8 ECC error",					/* 0_0000b */
-	"CRC error on link",				/* 0_0001b */
-	"Sync error packets on link",			/* 0_0010b */
-	"Master Abort during link operation",		/* 0_0011b */
-	"Target Abort during link operation",		/* 0_0100b */
-	"Invalid GART PTE entry during table walk",	/* 0_0101b */
-	"Unsupported atomic RMW command received",	/* 0_0110b */
-	"WDT error: NB transaction timeout",		/* 0_0111b */
-	"ECC/ChipKill ECC error",			/* 0_1000b */
-	"SVM DEV Error",				/* 0_1001b */
-	"Link Data error",				/* 0_1010b */
-	"Link/L3/Probe Filter Protocol error",		/* 0_1011b */
-	"NB Internal Arrays Parity error",		/* 0_1100b */
-	"DRAM Address/Control Parity error",		/* 0_1101b */
-	"Link Transmission error",			/* 0_1110b */
-	"GART/DEV Table Walk Data error"		/* 0_1111b */
-	"Res 0x100 error",				/* 1_0000b */
-	"Res 0x101 error",				/* 1_0001b */
-	"Res 0x102 error",				/* 1_0010b */
-	"Res 0x103 error",				/* 1_0011b */
-	"Res 0x104 error",				/* 1_0100b */
-	"Res 0x105 error",				/* 1_0101b */
-	"Res 0x106 error",				/* 1_0110b */
-	"Res 0x107 error",				/* 1_0111b */
-	"Res 0x108 error",				/* 1_1000b */
-	"Res 0x109 error",				/* 1_1001b */
-	"Res 0x10A error",				/* 1_1010b */
-	"Res 0x10B error",				/* 1_1011b */
-	"ECC error in L3 Cache Data",			/* 1_1100b */
-	"L3 Cache Tag error",				/* 1_1101b */
-	"L3 Cache LRU Parity error",			/* 1_1110b */
-	"Probe Filter error"				/* 1_1111b */
-};
-EXPORT_SYMBOL_GPL(ext_msgs);
-
-static void amd_decode_dc_mce(u64 mc0_status)
-{
-	u32 ec  = mc0_status & 0xffff;
-	u32 xec = (mc0_status >> 16) & 0xf;
-
-	pr_emerg("Data Cache Error");
-
-	if (xec == 1 && TLB_ERROR(ec))
-		pr_cont(": %s TLB multimatch.\n", LL_MSG(ec));
-	else if (xec == 0) {
-		if (mc0_status & (1ULL << 40))
-			pr_cont(" during Data Scrub.\n");
-		else if (TLB_ERROR(ec))
-			pr_cont(": %s TLB parity error.\n", LL_MSG(ec));
-		else if (MEM_ERROR(ec)) {
-			u8 ll   = ec & 0x3;
-			u8 tt   = (ec >> 2) & 0x3;
-			u8 rrrr = (ec >> 4) & 0xf;
-
-			/* see F10h BKDG (31116), Table 92. */
-			if (ll == 0x1) {
-				if (tt != 0x1)
-					goto wrong_dc_mce;
-
-				pr_cont(": Data/Tag %s error.\n", RRRR_MSG(ec));
-
-			} else if (ll == 0x2 && rrrr == 0x3)
-				pr_cont(" during L1 linefill from L2.\n");
-			else
-				goto wrong_dc_mce;
-		} else if (BUS_ERROR(ec) && boot_cpu_data.x86 == 0xf)
-			pr_cont(" during system linefill.\n");
-		else
-			goto wrong_dc_mce;
-	} else
-		goto wrong_dc_mce;
-
-	return;
-
-wrong_dc_mce:
-	pr_warning("Corrupted DC MCE info?\n");
-}
-
-static void amd_decode_ic_mce(u64 mc1_status)
-{
-	u32 ec  = mc1_status & 0xffff;
-	u32 xec = (mc1_status >> 16) & 0xf;
-
-	pr_emerg("Instruction Cache Error");
-
-	if (xec == 1 && TLB_ERROR(ec))
-		pr_cont(": %s TLB multimatch.\n", LL_MSG(ec));
-	else if (xec == 0) {
-		if (TLB_ERROR(ec))
-			pr_cont(": %s TLB Parity error.\n", LL_MSG(ec));
-		else if (BUS_ERROR(ec)) {
-			if (boot_cpu_data.x86 == 0xf &&
-			    (mc1_status & (1ULL << 58)))
-				pr_cont(" during system linefill.\n");
-			else
-				pr_cont(" during attempted NB data read.\n");
-		} else if (MEM_ERROR(ec)) {
-			u8 ll   = ec & 0x3;
-			u8 rrrr = (ec >> 4) & 0xf;
-
-			if (ll == 0x2)
-				pr_cont(" during a linefill from L2.\n");
-			else if (ll == 0x1) {
-
-				switch (rrrr) {
-				case 0x5:
-					pr_cont(": Parity error during "
-					       "data load.\n");
-					break;
-
-				case 0x7:
-					pr_cont(": Copyback Parity/Victim"
-						" error.\n");
-					break;
-
-				case 0x8:
-					pr_cont(": Tag Snoop error.\n");
-					break;
-
-				default:
-					goto wrong_ic_mce;
-					break;
-				}
-			}
-		} else
-			goto wrong_ic_mce;
-	} else
-		goto wrong_ic_mce;
-
-	return;
-
-wrong_ic_mce:
-	pr_warning("Corrupted IC MCE info?\n");
-}
-
-static void amd_decode_bu_mce(u64 mc2_status)
-{
-	u32 ec = mc2_status & 0xffff;
-	u32 xec = (mc2_status >> 16) & 0xf;
-
-	pr_emerg("Bus Unit Error");
-
-	if (xec == 0x1)
-		pr_cont(" in the write data buffers.\n");
-	else if (xec == 0x3)
-		pr_cont(" in the victim data buffers.\n");
-	else if (xec == 0x2 && MEM_ERROR(ec))
-		pr_cont(": %s error in the L2 cache tags.\n", RRRR_MSG(ec));
-	else if (xec == 0x0) {
-		if (TLB_ERROR(ec))
-			pr_cont(": %s error in a Page Descriptor Cache or "
-				"Guest TLB.\n", TT_MSG(ec));
-		else if (BUS_ERROR(ec))
-			pr_cont(": %s/ECC error in data read from NB: %s.\n",
-				RRRR_MSG(ec), PP_MSG(ec));
-		else if (MEM_ERROR(ec)) {
-			u8 rrrr = (ec >> 4) & 0xf;
-
-			if (rrrr >= 0x7)
-				pr_cont(": %s error during data copyback.\n",
-					RRRR_MSG(ec));
-			else if (rrrr <= 0x1)
-				pr_cont(": %s parity/ECC error during data "
-					"access from L2.\n", RRRR_MSG(ec));
-			else
-				goto wrong_bu_mce;
-		} else
-			goto wrong_bu_mce;
-	} else
-		goto wrong_bu_mce;
-
-	return;
-
-wrong_bu_mce:
-	pr_warning("Corrupted BU MCE info?\n");
-}
-
-static void amd_decode_ls_mce(u64 mc3_status)
-{
-	u32 ec  = mc3_status & 0xffff;
-	u32 xec = (mc3_status >> 16) & 0xf;
-
-	pr_emerg("Load Store Error");
-
-	if (xec == 0x0) {
-		u8 rrrr = (ec >> 4) & 0xf;
-
-		if (!BUS_ERROR(ec) || (rrrr != 0x3 && rrrr != 0x4))
-			goto wrong_ls_mce;
-
-		pr_cont(" during %s.\n", RRRR_MSG(ec));
-	}
-	return;
-
-wrong_ls_mce:
-	pr_warning("Corrupted LS MCE info?\n");
-}
-
-void amd_decode_nb_mce(int node_id, struct err_regs *regs, int handle_errors)
-{
-	u32 ec  = ERROR_CODE(regs->nbsl);
-
-	if (!handle_errors)
-		return;
-
-	/*
-	 * GART TLB error reporting is disabled by default. Bail out early.
-	 */
-	if (TLB_ERROR(ec) && !report_gart_errors)
-		return;
-
-	pr_emerg("Northbridge Error, node %d", node_id);
-
-	/*
-	 * F10h, revD can disable ErrCpu[3:0] so check that first and also the
-	 * value encoding has changed so interpret those differently
-	 */
-	if ((boot_cpu_data.x86 == 0x10) &&
-	    (boot_cpu_data.x86_model > 7)) {
-		if (regs->nbsh & K8_NBSH_ERR_CPU_VAL)
-			pr_cont(", core: %u\n", (u8)(regs->nbsh & 0xf));
-	} else {
-		u8 assoc_cpus = regs->nbsh & 0xf;
-
-		if (assoc_cpus > 0)
-			pr_cont(", core: %d", fls(assoc_cpus) - 1);
-
-		pr_cont("\n");
-	}
-
-	pr_emerg("%s.\n", EXT_ERR_MSG(regs->nbsl));
-
-	if (BUS_ERROR(ec) && nb_bus_decoder)
-		nb_bus_decoder(node_id, regs);
-}
-EXPORT_SYMBOL_GPL(amd_decode_nb_mce);
-
-static void amd_decode_fr_mce(u64 mc5_status)
-{
-	/* we have only one error signature so match all fields at once. */
-	if ((mc5_status & 0xffff) == 0x0f0f)
-		pr_emerg(" FR Error: CPU Watchdog timer expire.\n");
-	else
-		pr_warning("Corrupted FR MCE info?\n");
-}
-
-static inline void amd_decode_err_code(unsigned int ec)
-{
-	if (TLB_ERROR(ec)) {
-		pr_emerg("Transaction: %s, Cache Level %s\n",
-			 TT_MSG(ec), LL_MSG(ec));
-	} else if (MEM_ERROR(ec)) {
-		pr_emerg("Transaction: %s, Type: %s, Cache Level: %s",
-			 RRRR_MSG(ec), TT_MSG(ec), LL_MSG(ec));
-	} else if (BUS_ERROR(ec)) {
-		pr_emerg("Transaction type: %s(%s), %s, Cache Level: %s, "
-			 "Participating Processor: %s\n",
-			  RRRR_MSG(ec), II_MSG(ec), TO_MSG(ec), LL_MSG(ec),
-			  PP_MSG(ec));
-	} else
-		pr_warning("Huh? Unknown MCE error 0x%x\n", ec);
-}
-
-static int amd_decode_mce(struct notifier_block *nb, unsigned long val,
-			   void *data)
-{
-	struct mce *m = (struct mce *)data;
-	struct err_regs regs;
-	int node, ecc;
-
-	pr_emerg("MC%d_STATUS: ", m->bank);
-
-	pr_cont("%sorrected error, other errors lost: %s, "
-		 "CPU context corrupt: %s",
-		 ((m->status & MCI_STATUS_UC) ? "Unc"  : "C"),
-		 ((m->status & MCI_STATUS_OVER) ? "yes"  : "no"),
-		 ((m->status & MCI_STATUS_PCC) ? "yes" : "no"));
-
-	/* do the two bits[14:13] together */
-	ecc = (m->status >> 45) & 0x3;
-	if (ecc)
-		pr_cont(", %sECC Error", ((ecc == 2) ? "C" : "U"));
-
-	pr_cont("\n");
-
-	switch (m->bank) {
-	case 0:
-		amd_decode_dc_mce(m->status);
-		break;
-
-	case 1:
-		amd_decode_ic_mce(m->status);
-		break;
-
-	case 2:
-		amd_decode_bu_mce(m->status);
-		break;
-
-	case 3:
-		amd_decode_ls_mce(m->status);
-		break;
-
-	case 4:
-		regs.nbsl  = (u32) m->status;
-		regs.nbsh  = (u32)(m->status >> 32);
-		regs.nbeal = (u32) m->addr;
-		regs.nbeah = (u32)(m->addr >> 32);
-		node       = amd_get_nb_id(m->extcpu);
-
-		amd_decode_nb_mce(node, &regs, 1);
-		break;
-
-	case 5:
-		amd_decode_fr_mce(m->status);
-		break;
-
-	default:
-		break;
-	}
-
-	amd_decode_err_code(m->status & 0xffff);
-
-	return NOTIFY_STOP;
-}
-
-static struct notifier_block amd_mce_dec_nb = {
-	.notifier_call	= amd_decode_mce,
-};
-
-static int __init mce_amd_init(void)
-{
-	/*
-	 * We can decode MCEs for K8, F10h and F11h CPUs:
-	 */
-	if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD)
-		return 0;
-
-	if (boot_cpu_data.x86 < 0xf || boot_cpu_data.x86 > 0x11)
-		return 0;
-
-	atomic_notifier_chain_register(&x86_mce_decoder_chain, &amd_mce_dec_nb);
-
-	return 0;
-}
-early_initcall(mce_amd_init);
-
-#ifdef MODULE
-static void __exit mce_amd_exit(void)
-{
-	atomic_notifier_chain_unregister(&x86_mce_decoder_chain, &amd_mce_dec_nb);
-}
-
-MODULE_DESCRIPTION("AMD MCE decoder");
-MODULE_ALIAS("edac-mce-amd");
-MODULE_LICENSE("GPL");
-module_exit(mce_amd_exit);
-#endif
diff --git a/drivers/edac/edac_module.c b/drivers/edac/edac_module.c
index 7e1374afd96..be4b075c309 100644
--- a/drivers/edac/edac_module.c
+++ b/drivers/edac/edac_module.c
@@ -27,15 +27,6 @@ EXPORT_SYMBOL_GPL(edac_debug_level);
 struct workqueue_struct *edac_workqueue;
 
 /*
- * sysfs object: /sys/devices/system/edac
- *	need to export to other files in this modules
- */
-static struct sysdev_class edac_class = {
-	.name = "edac",
-};
-static int edac_class_valid;
-
-/*
  * edac_op_state_to_string()
  */
 char *edac_op_state_to_string(int opstate)
@@ -55,60 +46,6 @@ char *edac_op_state_to_string(int opstate)
 }
 
 /*
- * edac_get_edac_class()
- *
- *	return pointer to the edac class of 'edac'
- */
-struct sysdev_class *edac_get_edac_class(void)
-{
-	struct sysdev_class *classptr = NULL;
-
-	if (edac_class_valid)
-		classptr = &edac_class;
-
-	return classptr;
-}
-
-/*
- * edac_register_sysfs_edac_name()
- *
- *	register the 'edac' into /sys/devices/system
- *
- * return:
- *	0  success
- *	!0 error
- */
-static int edac_register_sysfs_edac_name(void)
-{
-	int err;
-
-	/* create the /sys/devices/system/edac directory */
-	err = sysdev_class_register(&edac_class);
-
-	if (err) {
-		debugf1("%s() error=%d\n", __func__, err);
-		return err;
-	}
-
-	edac_class_valid = 1;
-	return 0;
-}
-
-/*
- * sysdev_class_unregister()
- *
- *	unregister the 'edac' from /sys/devices/system
- */
-static void edac_unregister_sysfs_edac_name(void)
-{
-	/* only if currently registered, then unregister it */
-	if (edac_class_valid)
-		sysdev_class_unregister(&edac_class);
-
-	edac_class_valid = 0;
-}
-
-/*
  * edac_workqueue_setup
  *	initialize the edac work queue for polling operations
  */
@@ -154,21 +91,11 @@ static int __init edac_init(void)
 	edac_pci_clear_parity_errors();
 
 	/*
-	 * perform the registration of the /sys/devices/system/edac class object
-	 */
-	if (edac_register_sysfs_edac_name()) {
-		edac_printk(KERN_ERR, EDAC_MC,
-			"Error initializing 'edac' kobject\n");
-		err = -ENODEV;
-		goto error;
-	}
-
-	/*
 	 * now set up the mc_kset under the edac class object
 	 */
 	err = edac_sysfs_setup_mc_kset();
 	if (err)
-		goto sysfs_setup_fail;
+		goto error;
 
 	/* Setup/Initialize the workq for this core */
 	err = edac_workqueue_setup();
@@ -183,9 +110,6 @@ static int __init edac_init(void)
 workq_fail:
 	edac_sysfs_teardown_mc_kset();
 
-sysfs_setup_fail:
-	edac_unregister_sysfs_edac_name();
-
 error:
 	return err;
 }
@@ -201,7 +125,6 @@ static void __exit edac_exit(void)
 	/* tear down the various subsystems */
 	edac_workqueue_teardown();
 	edac_sysfs_teardown_mc_kset();
-	edac_unregister_sysfs_edac_name();
 }
 
 /*
diff --git a/drivers/edac/edac_module.h b/drivers/edac/edac_module.h
index 233d4798c3a..17aabb7b90e 100644
--- a/drivers/edac/edac_module.h
+++ b/drivers/edac/edac_module.h
@@ -42,7 +42,6 @@ extern void edac_device_unregister_sysfs_main_kobj(
 				struct edac_device_ctl_info *edac_dev);
 extern int edac_device_create_sysfs(struct edac_device_ctl_info *edac_dev);
 extern void edac_device_remove_sysfs(struct edac_device_ctl_info *edac_dev);
-extern struct sysdev_class *edac_get_edac_class(void);
 
 /* edac core workqueue: single CPU mode */
 extern struct workqueue_struct *edac_workqueue;
diff --git a/drivers/edac/edac_pci_sysfs.c b/drivers/edac/edac_pci_sysfs.c
index c39697df9cb..023b01cb517 100644
--- a/drivers/edac/edac_pci_sysfs.c
+++ b/drivers/edac/edac_pci_sysfs.c
@@ -7,7 +7,7 @@
  *
  */
 #include <linux/module.h>
-#include <linux/sysdev.h>
+#include <linux/edac.h>
 #include <linux/slab.h>
 #include <linux/ctype.h>
 
@@ -354,7 +354,7 @@ static int edac_pci_main_kobj_setup(void)
 	/* First time, so create the main kobject and its
 	 * controls and atributes
 	 */
-	edac_class = edac_get_edac_class();
+	edac_class = edac_get_sysfs_class();
 	if (edac_class == NULL) {
 		debugf1("%s() no edac_class\n", __func__);
 		err = -ENODEV;
@@ -368,7 +368,7 @@ static int edac_pci_main_kobj_setup(void)
 	if (!try_module_get(THIS_MODULE)) {
 		debugf1("%s() try_module_get() failed\n", __func__);
 		err = -ENODEV;
-		goto decrement_count_fail;
+		goto mod_get_fail;
 	}
 
 	edac_pci_top_main_kobj = kzalloc(sizeof(struct kobject), GFP_KERNEL);
@@ -403,6 +403,9 @@ kobject_init_and_add_fail:
 kzalloc_fail:
 	module_put(THIS_MODULE);
 
+mod_get_fail:
+	edac_put_sysfs_class();
+
 decrement_count_fail:
 	/* if are on this error exit, nothing to tear down */
 	atomic_dec(&edac_pci_sysfs_refcount);
@@ -429,6 +432,7 @@ static void edac_pci_main_kobj_teardown(void)
 			__func__);
 		kobject_put(edac_pci_top_main_kobj);
 	}
+	edac_put_sysfs_class();
 }
 
 /*
diff --git a/drivers/edac/edac_stub.c b/drivers/edac/edac_stub.c
index 20b428aa155..aab970760b7 100644
--- a/drivers/edac/edac_stub.c
+++ b/drivers/edac/edac_stub.c
@@ -3,10 +3,13 @@
  *
  * Author: Dave Jiang <djiang@mvista.com>
  *
- * 2007 (c) MontaVista Software, Inc. This file is licensed under
- * the terms of the GNU General Public License version 2. This program
- * is licensed "as is" without any warranty of any kind, whether express
- * or implied.
+ * 2007 (c) MontaVista Software, Inc.
+ * 2010 (c) Advanced Micro Devices Inc.
+ *	    Borislav Petkov <borislav.petkov@amd.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
  *
  */
 #include <linux/module.h>
@@ -23,6 +26,8 @@ EXPORT_SYMBOL_GPL(edac_handlers);
 int edac_err_assert = 0;
 EXPORT_SYMBOL_GPL(edac_err_assert);
 
+static atomic_t edac_class_valid = ATOMIC_INIT(0);
+
 /*
  * called to determine if there is an EDAC driver interested in
  * knowing an event (such as NMI) occurred
@@ -44,3 +49,41 @@ void edac_atomic_assert_error(void)
 	edac_err_assert++;
 }
 EXPORT_SYMBOL_GPL(edac_atomic_assert_error);
+
+/*
+ * sysfs object: /sys/devices/system/edac
+ *	need to export to other files
+ */
+struct sysdev_class edac_class = {
+	.name = "edac",
+};
+EXPORT_SYMBOL_GPL(edac_class);
+
+/* return pointer to the 'edac' node in sysfs */
+struct sysdev_class *edac_get_sysfs_class(void)
+{
+	int err = 0;
+
+	if (atomic_read(&edac_class_valid))
+		goto out;
+
+	/* create the /sys/devices/system/edac directory */
+	err = sysdev_class_register(&edac_class);
+	if (err) {
+		printk(KERN_ERR "Error registering toplevel EDAC sysfs dir\n");
+		return NULL;
+	}
+
+out:
+	atomic_inc(&edac_class_valid);
+	return &edac_class;
+}
+EXPORT_SYMBOL_GPL(edac_get_sysfs_class);
+
+void edac_put_sysfs_class(void)
+{
+	/* last user unregisters it */
+	if (atomic_dec_and_test(&edac_class_valid))
+		sysdev_class_unregister(&edac_class);
+}
+EXPORT_SYMBOL_GPL(edac_put_sysfs_class);
diff --git a/drivers/edac/mce_amd.c b/drivers/edac/mce_amd.c
new file mode 100644
index 00000000000..c0181093b49
--- /dev/null
+++ b/drivers/edac/mce_amd.c
@@ -0,0 +1,680 @@
+#include <linux/module.h>
+#include <linux/slab.h>
+
+#include "mce_amd.h"
+
+static struct amd_decoder_ops *fam_ops;
+
+static u8 nb_err_cpumask = 0xf;
+
+static bool report_gart_errors;
+static void (*nb_bus_decoder)(int node_id, struct mce *m, u32 nbcfg);
+
+void amd_report_gart_errors(bool v)
+{
+	report_gart_errors = v;
+}
+EXPORT_SYMBOL_GPL(amd_report_gart_errors);
+
+void amd_register_ecc_decoder(void (*f)(int, struct mce *, u32))
+{
+	nb_bus_decoder = f;
+}
+EXPORT_SYMBOL_GPL(amd_register_ecc_decoder);
+
+void amd_unregister_ecc_decoder(void (*f)(int, struct mce *, u32))
+{
+	if (nb_bus_decoder) {
+		WARN_ON(nb_bus_decoder != f);
+
+		nb_bus_decoder = NULL;
+	}
+}
+EXPORT_SYMBOL_GPL(amd_unregister_ecc_decoder);
+
+/*
+ * string representation for the different MCA reported error types, see F3x48
+ * or MSR0000_0411.
+ */
+
+/* transaction type */
+const char *tt_msgs[] = { "INSN", "DATA", "GEN", "RESV" };
+EXPORT_SYMBOL_GPL(tt_msgs);
+
+/* cache level */
+const char *ll_msgs[] = { "RESV", "L1", "L2", "L3/GEN" };
+EXPORT_SYMBOL_GPL(ll_msgs);
+
+/* memory transaction type */
+const char *rrrr_msgs[] = {
+       "GEN", "RD", "WR", "DRD", "DWR", "IRD", "PRF", "EV", "SNP"
+};
+EXPORT_SYMBOL_GPL(rrrr_msgs);
+
+/* participating processor */
+const char *pp_msgs[] = { "SRC", "RES", "OBS", "GEN" };
+EXPORT_SYMBOL_GPL(pp_msgs);
+
+/* request timeout */
+const char *to_msgs[] = { "no timeout",	"timed out" };
+EXPORT_SYMBOL_GPL(to_msgs);
+
+/* memory or i/o */
+const char *ii_msgs[] = { "MEM", "RESV", "IO", "GEN" };
+EXPORT_SYMBOL_GPL(ii_msgs);
+
+static const char *f10h_nb_mce_desc[] = {
+	"HT link data error",
+	"Protocol error (link, L3, probe filter, etc.)",
+	"Parity error in NB-internal arrays",
+	"Link Retry due to IO link transmission error",
+	"L3 ECC data cache error",
+	"ECC error in L3 cache tag",
+	"L3 LRU parity bits error",
+	"ECC Error in the Probe Filter directory"
+};
+
+static bool f12h_dc_mce(u16 ec)
+{
+	bool ret = false;
+
+	if (MEM_ERROR(ec)) {
+		u8 ll = ec & 0x3;
+		ret = true;
+
+		if (ll == LL_L2)
+			pr_cont("during L1 linefill from L2.\n");
+		else if (ll == LL_L1)
+			pr_cont("Data/Tag %s error.\n", RRRR_MSG(ec));
+		else
+			ret = false;
+	}
+	return ret;
+}
+
+static bool f10h_dc_mce(u16 ec)
+{
+	u8 r4  = (ec >> 4) & 0xf;
+	u8 ll  = ec & 0x3;
+
+	if (r4 == R4_GEN && ll == LL_L1) {
+		pr_cont("during data scrub.\n");
+		return true;
+	}
+	return f12h_dc_mce(ec);
+}
+
+static bool k8_dc_mce(u16 ec)
+{
+	if (BUS_ERROR(ec)) {
+		pr_cont("during system linefill.\n");
+		return true;
+	}
+
+	return f10h_dc_mce(ec);
+}
+
+static bool f14h_dc_mce(u16 ec)
+{
+	u8 r4	 = (ec >> 4) & 0xf;
+	u8 ll	 = ec & 0x3;
+	u8 tt	 = (ec >> 2) & 0x3;
+	u8 ii	 = tt;
+	bool ret = true;
+
+	if (MEM_ERROR(ec)) {
+
+		if (tt != TT_DATA || ll != LL_L1)
+			return false;
+
+		switch (r4) {
+		case R4_DRD:
+		case R4_DWR:
+			pr_cont("Data/Tag parity error due to %s.\n",
+				(r4 == R4_DRD ? "load/hw prf" : "store"));
+			break;
+		case R4_EVICT:
+			pr_cont("Copyback parity error on a tag miss.\n");
+			break;
+		case R4_SNOOP:
+			pr_cont("Tag parity error during snoop.\n");
+			break;
+		default:
+			ret = false;
+		}
+	} else if (BUS_ERROR(ec)) {
+
+		if ((ii != II_MEM && ii != II_IO) || ll != LL_LG)
+			return false;
+
+		pr_cont("System read data error on a ");
+
+		switch (r4) {
+		case R4_RD:
+			pr_cont("TLB reload.\n");
+			break;
+		case R4_DWR:
+			pr_cont("store.\n");
+			break;
+		case R4_DRD:
+			pr_cont("load.\n");
+			break;
+		default:
+			ret = false;
+		}
+	} else {
+		ret = false;
+	}
+
+	return ret;
+}
+
+static void amd_decode_dc_mce(struct mce *m)
+{
+	u16 ec = m->status & 0xffff;
+	u8 xec = (m->status >> 16) & 0xf;
+
+	pr_emerg(HW_ERR "Data Cache Error: ");
+
+	/* TLB error signatures are the same across families */
+	if (TLB_ERROR(ec)) {
+		u8 tt = (ec >> 2) & 0x3;
+
+		if (tt == TT_DATA) {
+			pr_cont("%s TLB %s.\n", LL_MSG(ec),
+				(xec ? "multimatch" : "parity error"));
+			return;
+		}
+		else
+			goto wrong_dc_mce;
+	}
+
+	if (!fam_ops->dc_mce(ec))
+		goto wrong_dc_mce;
+
+	return;
+
+wrong_dc_mce:
+	pr_emerg(HW_ERR "Corrupted DC MCE info?\n");
+}
+
+static bool k8_ic_mce(u16 ec)
+{
+	u8 ll	 = ec & 0x3;
+	u8 r4	 = (ec >> 4) & 0xf;
+	bool ret = true;
+
+	if (!MEM_ERROR(ec))
+		return false;
+
+	if (ll == 0x2)
+		pr_cont("during a linefill from L2.\n");
+	else if (ll == 0x1) {
+		switch (r4) {
+		case R4_IRD:
+			pr_cont("Parity error during data load.\n");
+			break;
+
+		case R4_EVICT:
+			pr_cont("Copyback Parity/Victim error.\n");
+			break;
+
+		case R4_SNOOP:
+			pr_cont("Tag Snoop error.\n");
+			break;
+
+		default:
+			ret = false;
+			break;
+		}
+	} else
+		ret = false;
+
+	return ret;
+}
+
+static bool f14h_ic_mce(u16 ec)
+{
+	u8 ll    = ec & 0x3;
+	u8 tt    = (ec >> 2) & 0x3;
+	u8 r4  = (ec >> 4) & 0xf;
+	bool ret = true;
+
+	if (MEM_ERROR(ec)) {
+		if (tt != 0 || ll != 1)
+			ret = false;
+
+		if (r4 == R4_IRD)
+			pr_cont("Data/tag array parity error for a tag hit.\n");
+		else if (r4 == R4_SNOOP)
+			pr_cont("Tag error during snoop/victimization.\n");
+		else
+			ret = false;
+	}
+	return ret;
+}
+
+static void amd_decode_ic_mce(struct mce *m)
+{
+	u16 ec = m->status & 0xffff;
+	u8 xec = (m->status >> 16) & 0xf;
+
+	pr_emerg(HW_ERR "Instruction Cache Error: ");
+
+	if (TLB_ERROR(ec))
+		pr_cont("%s TLB %s.\n", LL_MSG(ec),
+			(xec ? "multimatch" : "parity error"));
+	else if (BUS_ERROR(ec)) {
+		bool k8 = (boot_cpu_data.x86 == 0xf && (m->status & BIT_64(58)));
+
+		pr_cont("during %s.\n", (k8 ? "system linefill" : "NB data read"));
+	} else if (fam_ops->ic_mce(ec))
+		;
+	else
+		pr_emerg(HW_ERR "Corrupted IC MCE info?\n");
+}
+
+static void amd_decode_bu_mce(struct mce *m)
+{
+	u32 ec = m->status & 0xffff;
+	u32 xec = (m->status >> 16) & 0xf;
+
+	pr_emerg(HW_ERR "Bus Unit Error");
+
+	if (xec == 0x1)
+		pr_cont(" in the write data buffers.\n");
+	else if (xec == 0x3)
+		pr_cont(" in the victim data buffers.\n");
+	else if (xec == 0x2 && MEM_ERROR(ec))
+		pr_cont(": %s error in the L2 cache tags.\n", RRRR_MSG(ec));
+	else if (xec == 0x0) {
+		if (TLB_ERROR(ec))
+			pr_cont(": %s error in a Page Descriptor Cache or "
+				"Guest TLB.\n", TT_MSG(ec));
+		else if (BUS_ERROR(ec))
+			pr_cont(": %s/ECC error in data read from NB: %s.\n",
+				RRRR_MSG(ec), PP_MSG(ec));
+		else if (MEM_ERROR(ec)) {
+			u8 rrrr = (ec >> 4) & 0xf;
+
+			if (rrrr >= 0x7)
+				pr_cont(": %s error during data copyback.\n",
+					RRRR_MSG(ec));
+			else if (rrrr <= 0x1)
+				pr_cont(": %s parity/ECC error during data "
+					"access from L2.\n", RRRR_MSG(ec));
+			else
+				goto wrong_bu_mce;
+		} else
+			goto wrong_bu_mce;
+	} else
+		goto wrong_bu_mce;
+
+	return;
+
+wrong_bu_mce:
+	pr_emerg(HW_ERR "Corrupted BU MCE info?\n");
+}
+
+static void amd_decode_ls_mce(struct mce *m)
+{
+	u16 ec = m->status & 0xffff;
+	u8 xec = (m->status >> 16) & 0xf;
+
+	if (boot_cpu_data.x86 == 0x14) {
+		pr_emerg("You shouldn't be seeing an LS MCE on this cpu family,"
+			 " please report on LKML.\n");
+		return;
+	}
+
+	pr_emerg(HW_ERR "Load Store Error");
+
+	if (xec == 0x0) {
+		u8 r4 = (ec >> 4) & 0xf;
+
+		if (!BUS_ERROR(ec) || (r4 != R4_DRD && r4 != R4_DWR))
+			goto wrong_ls_mce;
+
+		pr_cont(" during %s.\n", RRRR_MSG(ec));
+	} else
+		goto wrong_ls_mce;
+
+	return;
+
+wrong_ls_mce:
+	pr_emerg(HW_ERR "Corrupted LS MCE info?\n");
+}
+
+static bool k8_nb_mce(u16 ec, u8 xec)
+{
+	bool ret = true;
+
+	switch (xec) {
+	case 0x1:
+		pr_cont("CRC error detected on HT link.\n");
+		break;
+
+	case 0x5:
+		pr_cont("Invalid GART PTE entry during GART table walk.\n");
+		break;
+
+	case 0x6:
+		pr_cont("Unsupported atomic RMW received from an IO link.\n");
+		break;
+
+	case 0x0:
+	case 0x8:
+		if (boot_cpu_data.x86 == 0x11)
+			return false;
+
+		pr_cont("DRAM ECC error detected on the NB.\n");
+		break;
+
+	case 0xd:
+		pr_cont("Parity error on the DRAM addr/ctl signals.\n");
+		break;
+
+	default:
+		ret = false;
+		break;
+	}
+
+	return ret;
+}
+
+static bool f10h_nb_mce(u16 ec, u8 xec)
+{
+	bool ret = true;
+	u8 offset = 0;
+
+	if (k8_nb_mce(ec, xec))
+		return true;
+
+	switch(xec) {
+	case 0xa ... 0xc:
+		offset = 10;
+		break;
+
+	case 0xe:
+		offset = 11;
+		break;
+
+	case 0xf:
+		if (TLB_ERROR(ec))
+			pr_cont("GART Table Walk data error.\n");
+		else if (BUS_ERROR(ec))
+			pr_cont("DMA Exclusion Vector Table Walk error.\n");
+		else
+			ret = false;
+
+		goto out;
+		break;
+
+	case 0x1c ... 0x1f:
+		offset = 24;
+		break;
+
+	default:
+		ret = false;
+
+		goto out;
+		break;
+	}
+
+	pr_cont("%s.\n", f10h_nb_mce_desc[xec - offset]);
+
+out:
+	return ret;
+}
+
+static bool nb_noop_mce(u16 ec, u8 xec)
+{
+	return false;
+}
+
+void amd_decode_nb_mce(int node_id, struct mce *m, u32 nbcfg)
+{
+	u8 xec   = (m->status >> 16) & 0x1f;
+	u16 ec   = m->status & 0xffff;
+	u32 nbsh = (u32)(m->status >> 32);
+
+	pr_emerg(HW_ERR "Northbridge Error, node %d: ", node_id);
+
+	/*
+	 * F10h, revD can disable ErrCpu[3:0] so check that first and also the
+	 * value encoding has changed so interpret those differently
+	 */
+	if ((boot_cpu_data.x86 == 0x10) &&
+	    (boot_cpu_data.x86_model > 7)) {
+		if (nbsh & K8_NBSH_ERR_CPU_VAL)
+			pr_cont(", core: %u", (u8)(nbsh & nb_err_cpumask));
+	} else {
+		u8 assoc_cpus = nbsh & nb_err_cpumask;
+
+		if (assoc_cpus > 0)
+			pr_cont(", core: %d", fls(assoc_cpus) - 1);
+	}
+
+	switch (xec) {
+	case 0x2:
+		pr_cont("Sync error (sync packets on HT link detected).\n");
+		return;
+
+	case 0x3:
+		pr_cont("HT Master abort.\n");
+		return;
+
+	case 0x4:
+		pr_cont("HT Target abort.\n");
+		return;
+
+	case 0x7:
+		pr_cont("NB Watchdog timeout.\n");
+		return;
+
+	case 0x9:
+		pr_cont("SVM DMA Exclusion Vector error.\n");
+		return;
+
+	default:
+		break;
+	}
+
+	if (!fam_ops->nb_mce(ec, xec))
+		goto wrong_nb_mce;
+
+	if (boot_cpu_data.x86 == 0xf || boot_cpu_data.x86 == 0x10)
+		if ((xec == 0x8 || xec == 0x0) && nb_bus_decoder)
+			nb_bus_decoder(node_id, m, nbcfg);
+
+	return;
+
+wrong_nb_mce:
+	pr_emerg(HW_ERR "Corrupted NB MCE info?\n");
+}
+EXPORT_SYMBOL_GPL(amd_decode_nb_mce);
+
+static void amd_decode_fr_mce(struct mce *m)
+{
+	if (boot_cpu_data.x86 == 0xf ||
+	    boot_cpu_data.x86 == 0x11)
+		goto wrong_fr_mce;
+
+	/* we have only one error signature so match all fields at once. */
+	if ((m->status & 0xffff) == 0x0f0f) {
+		pr_emerg(HW_ERR "FR Error: CPU Watchdog timer expire.\n");
+		return;
+	}
+
+wrong_fr_mce:
+	pr_emerg(HW_ERR "Corrupted FR MCE info?\n");
+}
+
+static inline void amd_decode_err_code(u16 ec)
+{
+	if (TLB_ERROR(ec)) {
+		pr_emerg(HW_ERR "Transaction: %s, Cache Level: %s\n",
+			 TT_MSG(ec), LL_MSG(ec));
+	} else if (MEM_ERROR(ec)) {
+		pr_emerg(HW_ERR "Transaction: %s, Type: %s, Cache Level: %s\n",
+			 RRRR_MSG(ec), TT_MSG(ec), LL_MSG(ec));
+	} else if (BUS_ERROR(ec)) {
+		pr_emerg(HW_ERR "Transaction: %s (%s), %s, Cache Level: %s, "
+			 "Participating Processor: %s\n",
+			  RRRR_MSG(ec), II_MSG(ec), TO_MSG(ec), LL_MSG(ec),
+			  PP_MSG(ec));
+	} else
+		pr_emerg(HW_ERR "Huh? Unknown MCE error 0x%x\n", ec);
+}
+
+/*
+ * Filter out unwanted MCE signatures here.
+ */
+static bool amd_filter_mce(struct mce *m)
+{
+	u8 xec = (m->status >> 16) & 0x1f;
+
+	/*
+	 * NB GART TLB error reporting is disabled by default.
+	 */
+	if (m->bank == 4 && xec == 0x5 && !report_gart_errors)
+		return true;
+
+	return false;
+}
+
+int amd_decode_mce(struct notifier_block *nb, unsigned long val, void *data)
+{
+	struct mce *m = (struct mce *)data;
+	int node, ecc;
+
+	if (amd_filter_mce(m))
+		return NOTIFY_STOP;
+
+	pr_emerg(HW_ERR "MC%d_STATUS: ", m->bank);
+
+	pr_cont("%sorrected error, other errors lost: %s, "
+		 "CPU context corrupt: %s",
+		 ((m->status & MCI_STATUS_UC) ? "Unc"  : "C"),
+		 ((m->status & MCI_STATUS_OVER) ? "yes"  : "no"),
+		 ((m->status & MCI_STATUS_PCC) ? "yes" : "no"));
+
+	/* do the two bits[14:13] together */
+	ecc = (m->status >> 45) & 0x3;
+	if (ecc)
+		pr_cont(", %sECC Error", ((ecc == 2) ? "C" : "U"));
+
+	pr_cont("\n");
+
+	switch (m->bank) {
+	case 0:
+		amd_decode_dc_mce(m);
+		break;
+
+	case 1:
+		amd_decode_ic_mce(m);
+		break;
+
+	case 2:
+		amd_decode_bu_mce(m);
+		break;
+
+	case 3:
+		amd_decode_ls_mce(m);
+		break;
+
+	case 4:
+		node = amd_get_nb_id(m->extcpu);
+		amd_decode_nb_mce(node, m, 0);
+		break;
+
+	case 5:
+		amd_decode_fr_mce(m);
+		break;
+
+	default:
+		break;
+	}
+
+	amd_decode_err_code(m->status & 0xffff);
+
+	return NOTIFY_STOP;
+}
+EXPORT_SYMBOL_GPL(amd_decode_mce);
+
+static struct notifier_block amd_mce_dec_nb = {
+	.notifier_call	= amd_decode_mce,
+};
+
+static int __init mce_amd_init(void)
+{
+	if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD)
+		return 0;
+
+	if ((boot_cpu_data.x86 < 0xf || boot_cpu_data.x86 > 0x12) &&
+	    (boot_cpu_data.x86 != 0x14 || boot_cpu_data.x86_model > 0xf))
+		return 0;
+
+	fam_ops = kzalloc(sizeof(struct amd_decoder_ops), GFP_KERNEL);
+	if (!fam_ops)
+		return -ENOMEM;
+
+	switch (boot_cpu_data.x86) {
+	case 0xf:
+		fam_ops->dc_mce = k8_dc_mce;
+		fam_ops->ic_mce = k8_ic_mce;
+		fam_ops->nb_mce = k8_nb_mce;
+		break;
+
+	case 0x10:
+		fam_ops->dc_mce = f10h_dc_mce;
+		fam_ops->ic_mce = k8_ic_mce;
+		fam_ops->nb_mce = f10h_nb_mce;
+		break;
+
+	case 0x11:
+		fam_ops->dc_mce = k8_dc_mce;
+		fam_ops->ic_mce = k8_ic_mce;
+		fam_ops->nb_mce = f10h_nb_mce;
+		break;
+
+	case 0x12:
+		fam_ops->dc_mce = f12h_dc_mce;
+		fam_ops->ic_mce = k8_ic_mce;
+		fam_ops->nb_mce = nb_noop_mce;
+		break;
+
+	case 0x14:
+		nb_err_cpumask  = 0x3;
+		fam_ops->dc_mce = f14h_dc_mce;
+		fam_ops->ic_mce = f14h_ic_mce;
+		fam_ops->nb_mce = nb_noop_mce;
+		break;
+
+	default:
+		printk(KERN_WARNING "Huh? What family is that: %d?!\n",
+				    boot_cpu_data.x86);
+		kfree(fam_ops);
+		return -EINVAL;
+	}
+
+	pr_info("MCE: In-kernel MCE decoding enabled.\n");
+
+	atomic_notifier_chain_register(&x86_mce_decoder_chain, &amd_mce_dec_nb);
+
+	return 0;
+}
+early_initcall(mce_amd_init);
+
+#ifdef MODULE
+static void __exit mce_amd_exit(void)
+{
+	atomic_notifier_chain_unregister(&x86_mce_decoder_chain, &amd_mce_dec_nb);
+	kfree(fam_ops);
+}
+
+MODULE_DESCRIPTION("AMD MCE decoder");
+MODULE_ALIAS("edac-mce-amd");
+MODULE_LICENSE("GPL");
+module_exit(mce_amd_exit);
+#endif
diff --git a/drivers/edac/edac_mce_amd.h b/drivers/edac/mce_amd.h
index df23ee065f7..35f6e0e3b29 100644
--- a/drivers/edac/edac_mce_amd.h
+++ b/drivers/edac/mce_amd.h
@@ -1,11 +1,14 @@
 #ifndef _EDAC_MCE_AMD_H
 #define _EDAC_MCE_AMD_H
 
+#include <linux/notifier.h>
+
 #include <asm/mce.h>
 
+#define BIT_64(n)			(U64_C(1) << (n))
+
 #define ERROR_CODE(x)			((x) & 0xffff)
 #define EXT_ERROR_CODE(x)		(((x) >> 16) & 0x1f)
-#define EXT_ERR_MSG(x)			ext_msgs[EXT_ERROR_CODE(x)]
 
 #define LOW_SYNDROME(x)			(((x) >> 15) & 0xff)
 #define HIGH_SYNDROME(x)		(((x) >> 24) & 0xff)
@@ -20,13 +23,14 @@
 #define II_MSG(x)			ii_msgs[II(x)]
 #define LL(x)				(((x) >> 0) & 0x3)
 #define LL_MSG(x)			ll_msgs[LL(x)]
-#define RRRR(x)				(((x) >> 4) & 0xf)
-#define RRRR_MSG(x)			rrrr_msgs[RRRR(x)]
 #define TO(x)				(((x) >> 8) & 0x1)
 #define TO_MSG(x)			to_msgs[TO(x)]
 #define PP(x)				(((x) >> 9) & 0x3)
 #define PP_MSG(x)			pp_msgs[PP(x)]
 
+#define RRRR(x)				(((x) >> 4) & 0xf)
+#define RRRR_MSG(x)			((RRRR(x) < 9) ?  rrrr_msgs[RRRR(x)] : "Wrong R4!")
+
 #define K8_NBSH				0x4C
 
 #define K8_NBSH_VALID_BIT		BIT(31)
@@ -41,13 +45,45 @@
 #define K8_NBSH_UECC			BIT(13)
 #define K8_NBSH_ERR_SCRUBER		BIT(8)
 
+enum tt_ids {
+	TT_INSTR = 0,
+	TT_DATA,
+	TT_GEN,
+	TT_RESV,
+};
+
+enum ll_ids {
+	LL_RESV = 0,
+	LL_L1,
+	LL_L2,
+	LL_LG,
+};
+
+enum ii_ids {
+	II_MEM = 0,
+	II_RESV,
+	II_IO,
+	II_GEN,
+};
+
+enum rrrr_ids {
+	R4_GEN	= 0,
+	R4_RD,
+	R4_WR,
+	R4_DRD,
+	R4_DWR,
+	R4_IRD,
+	R4_PREF,
+	R4_EVICT,
+	R4_SNOOP,
+};
+
 extern const char *tt_msgs[];
 extern const char *ll_msgs[];
 extern const char *rrrr_msgs[];
 extern const char *pp_msgs[];
 extern const char *to_msgs[];
 extern const char *ii_msgs[];
-extern const char *ext_msgs[];
 
 /*
  * relevant NB regs
@@ -60,10 +96,19 @@ struct err_regs {
 	u32 nbeal;
 };
 
+/*
+ * per-family decoder ops
+ */
+struct amd_decoder_ops {
+	bool (*dc_mce)(u16);
+	bool (*ic_mce)(u16);
+	bool (*nb_mce)(u16, u8);
+};
 
 void amd_report_gart_errors(bool);
-void amd_register_ecc_decoder(void (*f)(int, struct err_regs *));
-void amd_unregister_ecc_decoder(void (*f)(int, struct err_regs *));
-void amd_decode_nb_mce(int, struct err_regs *, int);
+void amd_register_ecc_decoder(void (*f)(int, struct mce *, u32));
+void amd_unregister_ecc_decoder(void (*f)(int, struct mce *, u32));
+void amd_decode_nb_mce(int, struct mce *, u32);
+int amd_decode_mce(struct notifier_block *nb, unsigned long val, void *data);
 
 #endif /* _EDAC_MCE_AMD_H */
diff --git a/drivers/edac/mce_amd_inj.c b/drivers/edac/mce_amd_inj.c
new file mode 100644
index 00000000000..8d0688f36d4
--- /dev/null
+++ b/drivers/edac/mce_amd_inj.c
@@ -0,0 +1,171 @@
+/*
+ * A simple MCE injection facility for testing the MCE decoding code. This
+ * driver should be built as module so that it can be loaded on production
+ * kernels for testing purposes.
+ *
+ * This file may be distributed under the terms of the GNU General Public
+ * License version 2.
+ *
+ * Copyright (c) 2010:  Borislav Petkov <borislav.petkov@amd.com>
+ *			Advanced Micro Devices Inc.
+ */
+
+#include <linux/kobject.h>
+#include <linux/sysdev.h>
+#include <linux/edac.h>
+#include <asm/mce.h>
+
+#include "mce_amd.h"
+
+struct edac_mce_attr {
+	struct attribute attr;
+	ssize_t (*show) (struct kobject *kobj, struct edac_mce_attr *attr, char *buf);
+	ssize_t (*store)(struct kobject *kobj, struct edac_mce_attr *attr,
+			 const char *buf, size_t count);
+};
+
+#define EDAC_MCE_ATTR(_name, _mode, _show, _store)			\
+static struct edac_mce_attr mce_attr_##_name = __ATTR(_name, _mode, _show, _store)
+
+static struct kobject *mce_kobj;
+
+/*
+ * Collect all the MCi_XXX settings
+ */
+static struct mce i_mce;
+
+#define MCE_INJECT_STORE(reg)						\
+static ssize_t edac_inject_##reg##_store(struct kobject *kobj,		\
+					 struct edac_mce_attr *attr,	\
+					 const char *data, size_t count)\
+{									\
+	int ret = 0;							\
+	unsigned long value;						\
+									\
+	ret = strict_strtoul(data, 16, &value);				\
+	if (ret < 0)							\
+		printk(KERN_ERR "Error writing MCE " #reg " field.\n");	\
+									\
+	i_mce.reg = value;						\
+									\
+	return count;							\
+}
+
+MCE_INJECT_STORE(status);
+MCE_INJECT_STORE(misc);
+MCE_INJECT_STORE(addr);
+
+#define MCE_INJECT_SHOW(reg)						\
+static ssize_t edac_inject_##reg##_show(struct kobject *kobj,		\
+					struct edac_mce_attr *attr,	\
+					char *buf)			\
+{									\
+	return sprintf(buf, "0x%016llx\n", i_mce.reg);			\
+}
+
+MCE_INJECT_SHOW(status);
+MCE_INJECT_SHOW(misc);
+MCE_INJECT_SHOW(addr);
+
+EDAC_MCE_ATTR(status, 0644, edac_inject_status_show, edac_inject_status_store);
+EDAC_MCE_ATTR(misc, 0644, edac_inject_misc_show, edac_inject_misc_store);
+EDAC_MCE_ATTR(addr, 0644, edac_inject_addr_show, edac_inject_addr_store);
+
+/*
+ * This denotes into which bank we're injecting and triggers
+ * the injection, at the same time.
+ */
+static ssize_t edac_inject_bank_store(struct kobject *kobj,
+				      struct edac_mce_attr *attr,
+				      const char *data, size_t count)
+{
+	int ret = 0;
+	unsigned long value;
+
+	ret = strict_strtoul(data, 10, &value);
+	if (ret < 0) {
+		printk(KERN_ERR "Invalid bank value!\n");
+		return -EINVAL;
+	}
+
+	if (value > 5) {
+		printk(KERN_ERR "Non-existant MCE bank: %lu\n", value);
+		return -EINVAL;
+	}
+
+	i_mce.bank = value;
+
+	amd_decode_mce(NULL, 0, &i_mce);
+
+	return count;
+}
+
+static ssize_t edac_inject_bank_show(struct kobject *kobj,
+				     struct edac_mce_attr *attr, char *buf)
+{
+	return sprintf(buf, "%d\n", i_mce.bank);
+}
+
+EDAC_MCE_ATTR(bank, 0644, edac_inject_bank_show, edac_inject_bank_store);
+
+static struct edac_mce_attr *sysfs_attrs[] = { &mce_attr_status, &mce_attr_misc,
+					       &mce_attr_addr, &mce_attr_bank
+};
+
+static int __init edac_init_mce_inject(void)
+{
+	struct sysdev_class *edac_class = NULL;
+	int i, err = 0;
+
+	edac_class = edac_get_sysfs_class();
+	if (!edac_class)
+		return -EINVAL;
+
+	mce_kobj = kobject_create_and_add("mce", &edac_class->kset.kobj);
+	if (!mce_kobj) {
+		printk(KERN_ERR "Error creating a mce kset.\n");
+		err = -ENOMEM;
+		goto err_mce_kobj;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(sysfs_attrs); i++) {
+		err = sysfs_create_file(mce_kobj, &sysfs_attrs[i]->attr);
+		if (err) {
+			printk(KERN_ERR "Error creating %s in sysfs.\n",
+					sysfs_attrs[i]->attr.name);
+			goto err_sysfs_create;
+		}
+	}
+	return 0;
+
+err_sysfs_create:
+	while (i-- >= 0)
+		sysfs_remove_file(mce_kobj, &sysfs_attrs[i]->attr);
+
+	kobject_del(mce_kobj);
+
+err_mce_kobj:
+	edac_put_sysfs_class();
+
+	return err;
+}
+
+static void __exit edac_exit_mce_inject(void)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(sysfs_attrs); i++)
+		sysfs_remove_file(mce_kobj, &sysfs_attrs[i]->attr);
+
+	kobject_del(mce_kobj);
+
+	edac_put_sysfs_class();
+}
+
+module_init(edac_init_mce_inject);
+module_exit(edac_exit_mce_inject);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Borislav Petkov <borislav.petkov@amd.com>");
+MODULE_AUTHOR("AMD Inc.");
+MODULE_DESCRIPTION("MCE injection facility for testing MCE decoding");
diff --git a/drivers/firmware/Kconfig b/drivers/firmware/Kconfig
index 280c9b5ad9e..88a3ae6cd02 100644
--- a/drivers/firmware/Kconfig
+++ b/drivers/firmware/Kconfig
@@ -125,7 +125,7 @@ config ISCSI_IBFT_FIND
 config ISCSI_IBFT
 	tristate "iSCSI Boot Firmware Table Attributes module"
 	select ISCSI_BOOT_SYSFS
-	depends on ISCSI_IBFT_FIND && SCSI
+	depends on ISCSI_IBFT_FIND && SCSI && SCSI_LOWLEVEL
 	default	n
 	help
 	  This option enables support for detection and exposing of iSCSI
diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c
index c37ef64d146..cb3ccf3ed22 100644
--- a/drivers/idle/intel_idle.c
+++ b/drivers/idle/intel_idle.c
@@ -59,18 +59,11 @@
 #include <linux/hrtimer.h>	/* ktime_get_real() */
 #include <trace/events/power.h>
 #include <linux/sched.h>
+#include <asm/mwait.h>
 
 #define INTEL_IDLE_VERSION "0.4"
 #define PREFIX "intel_idle: "
 
-#define MWAIT_SUBSTATE_MASK	(0xf)
-#define MWAIT_CSTATE_MASK	(0xf)
-#define MWAIT_SUBSTATE_SIZE	(4)
-#define MWAIT_MAX_NUM_CSTATES	8
-#define CPUID_MWAIT_LEAF (5)
-#define CPUID5_ECX_EXTENSIONS_SUPPORTED (0x1)
-#define CPUID5_ECX_INTERRUPT_BREAK	(0x2)
-
 static struct cpuidle_driver intel_idle_driver = {
 	.name = "intel_idle",
 	.owner = THIS_MODULE,
diff --git a/drivers/input/evdev.c b/drivers/input/evdev.c
index 9ddafc30f43..af9ee313c10 100644
--- a/drivers/input/evdev.c
+++ b/drivers/input/evdev.c
@@ -28,7 +28,7 @@ struct evdev {
 	int minor;
 	struct input_handle handle;
 	wait_queue_head_t wait;
-	struct evdev_client *grab;
+	struct evdev_client __rcu *grab;
 	struct list_head client_list;
 	spinlock_t client_lock; /* protects client_list */
 	struct mutex mutex;
diff --git a/drivers/input/misc/hp_sdc_rtc.c b/drivers/input/misc/hp_sdc_rtc.c
index c1906647905..7e2c12a5b83 100644
--- a/drivers/input/misc/hp_sdc_rtc.c
+++ b/drivers/input/misc/hp_sdc_rtc.c
@@ -104,7 +104,7 @@ static int hp_sdc_rtc_do_read_bbrtc (struct rtc_time *rtctm)
 	t.endidx =		91;
 	t.seq =			tseq;
 	t.act.semaphore =	&tsem;
-	init_MUTEX_LOCKED(&tsem);
+	sema_init(&tsem, 0);
 	
 	if (hp_sdc_enqueue_transaction(&t)) return -1;
 	
@@ -698,7 +698,7 @@ static int __init hp_sdc_rtc_init(void)
 		return -ENODEV;
 #endif
 
-	init_MUTEX(&i8042tregs);
+	sema_init(&i8042tregs, 1);
 
 	if ((ret = hp_sdc_request_timer_irq(&hp_sdc_rtc_isr)))
 		return ret;
diff --git a/drivers/input/serio/hil_mlc.c b/drivers/input/serio/hil_mlc.c
index c92f4edfee7..e5624d8f170 100644
--- a/drivers/input/serio/hil_mlc.c
+++ b/drivers/input/serio/hil_mlc.c
@@ -915,15 +915,15 @@ int hil_mlc_register(hil_mlc *mlc)
 	mlc->ostarted = 0;
 
 	rwlock_init(&mlc->lock);
-	init_MUTEX(&mlc->osem);
+	sema_init(&mlc->osem, 1);
 
-	init_MUTEX(&mlc->isem);
+	sema_init(&mlc->isem, 1);
 	mlc->icount = -1;
 	mlc->imatch = 0;
 
 	mlc->opercnt = 0;
 
-	init_MUTEX_LOCKED(&(mlc->csem));
+	sema_init(&(mlc->csem), 0);
 
 	hil_mlc_clear_di_scratch(mlc);
 	hil_mlc_clear_di_map(mlc, 0);
diff --git a/drivers/input/serio/hp_sdc.c b/drivers/input/serio/hp_sdc.c
index bcc2d30ec24..8c0b51c3142 100644
--- a/drivers/input/serio/hp_sdc.c
+++ b/drivers/input/serio/hp_sdc.c
@@ -905,7 +905,7 @@ static int __init hp_sdc_init(void)
 	ts_sync[1]	= 0x0f;
 	ts_sync[2] = ts_sync[3]	= ts_sync[4] = ts_sync[5] = 0;
 	t_sync.act.semaphore = &s_sync;
-	init_MUTEX_LOCKED(&s_sync);
+	sema_init(&s_sync, 0);
 	hp_sdc_enqueue_transaction(&t_sync);
 	down(&s_sync); /* Wait for t_sync to complete */
 
@@ -1039,7 +1039,7 @@ static int __init hp_sdc_register(void)
 		return hp_sdc.dev_err;
 	}
 
-	init_MUTEX_LOCKED(&tq_init_sem);
+	sema_init(&tq_init_sem, 0);
 
 	tq_init.actidx		= 0;
 	tq_init.idx		= 1;
diff --git a/drivers/isdn/act2000/act2000.h b/drivers/isdn/act2000/act2000.h
index d4c50512a1f..88c9423500d 100644
--- a/drivers/isdn/act2000/act2000.h
+++ b/drivers/isdn/act2000/act2000.h
@@ -141,9 +141,9 @@ typedef struct irq_data_isa {
 	__u8            rcvhdr[8];
 } irq_data_isa;
 
-typedef union irq_data {
+typedef union act2000_irq_data {
 	irq_data_isa isa;
-} irq_data;
+} act2000_irq_data;
 
 /*
  * Per card driver data
@@ -176,7 +176,7 @@ typedef struct act2000_card {
 	char   *status_buf_read;
 	char   *status_buf_write;
 	char   *status_buf_end;
-	irq_data idat;			/* Data used for IRQ handler        */
+	act2000_irq_data idat;		/* Data used for IRQ handler        */
 	isdn_if interface;		/* Interface to upper layer         */
 	char regname[35];		/* Name used for request_region     */
 } act2000_card;
diff --git a/drivers/isdn/hisax/config.c b/drivers/isdn/hisax/config.c
index 6f9afcd5ca4..b133378d4dc 100644
--- a/drivers/isdn/hisax/config.c
+++ b/drivers/isdn/hisax/config.c
@@ -801,6 +801,16 @@ static void closecard(int cardnr)
 	ll_unload(csta);
 }
 
+static irqreturn_t card_irq(int intno, void *dev_id)
+{
+	struct IsdnCardState *cs = dev_id;
+	irqreturn_t ret = cs->irq_func(intno, cs);
+
+	if (ret == IRQ_HANDLED)
+		cs->irq_cnt++;
+	return ret;
+}
+
 static int init_card(struct IsdnCardState *cs)
 {
 	int 	irq_cnt, cnt = 3, ret;
@@ -809,10 +819,10 @@ static int init_card(struct IsdnCardState *cs)
 		ret = cs->cardmsg(cs, CARD_INIT, NULL);
 		return(ret);
 	}
-	irq_cnt = kstat_irqs(cs->irq);
+	irq_cnt = cs->irq_cnt = 0;
 	printk(KERN_INFO "%s: IRQ %d count %d\n", CardType[cs->typ],
 	       cs->irq, irq_cnt);
-	if (request_irq(cs->irq, cs->irq_func, cs->irq_flags, "HiSax", cs)) {
+	if (request_irq(cs->irq, card_irq, cs->irq_flags, "HiSax", cs)) {
 		printk(KERN_WARNING "HiSax: couldn't get interrupt %d\n",
 		       cs->irq);
 		return 1;
@@ -822,8 +832,8 @@ static int init_card(struct IsdnCardState *cs)
 		/* Timeout 10ms */
 		msleep(10);
 		printk(KERN_INFO "%s: IRQ %d count %d\n",
-		       CardType[cs->typ], cs->irq, kstat_irqs(cs->irq));
-		if (kstat_irqs(cs->irq) == irq_cnt) {
+		       CardType[cs->typ], cs->irq, cs->irq_cnt);
+		if (cs->irq_cnt == irq_cnt) {
 			printk(KERN_WARNING
 			       "%s: IRQ(%d) getting no interrupts during init %d\n",
 			       CardType[cs->typ], cs->irq, 4 - cnt);
diff --git a/drivers/isdn/hisax/hisax.h b/drivers/isdn/hisax/hisax.h
index 832a87855ff..32ab3924aa7 100644
--- a/drivers/isdn/hisax/hisax.h
+++ b/drivers/isdn/hisax/hisax.h
@@ -959,6 +959,7 @@ struct IsdnCardState {
 	u_long		event;
 	struct work_struct tqueue;
 	struct timer_list dbusytimer;
+	unsigned int	irq_cnt;
 #ifdef ERROR_STATISTIC
 	int		err_crc;
 	int		err_tx;
diff --git a/drivers/macintosh/adb.c b/drivers/macintosh/adb.c
index 1c4ee6e7793..bf64e49d996 100644
--- a/drivers/macintosh/adb.c
+++ b/drivers/macintosh/adb.c
@@ -83,7 +83,7 @@ static struct adb_driver *adb_controller;
 BLOCKING_NOTIFIER_HEAD(adb_client_list);
 static int adb_got_sleep;
 static int adb_inited;
-static DECLARE_MUTEX(adb_probe_mutex);
+static DEFINE_SEMAPHORE(adb_probe_mutex);
 static int sleepy_trackpad;
 static int autopoll_devs;
 int __adb_probe_sync;
diff --git a/drivers/mfd/twl4030-irq.c b/drivers/mfd/twl4030-irq.c
index 097f24d8bce..b9fda7018ce 100644
--- a/drivers/mfd/twl4030-irq.c
+++ b/drivers/mfd/twl4030-irq.c
@@ -78,7 +78,7 @@ struct sih {
 	u8	irq_lines;		/* number of supported irq lines */
 
 	/* SIR ignored -- set interrupt, for testing only */
-	struct irq_data {
+	struct sih_irq_data {
 		u8	isr_offset;
 		u8	imr_offset;
 	} mask[2];
@@ -810,7 +810,7 @@ int twl4030_init_irq(int irq_num, unsigned irq_base, unsigned irq_end)
 	twl4030_irq_chip = dummy_irq_chip;
 	twl4030_irq_chip.name = "twl4030";
 
-	twl4030_sih_irq_chip.ack = dummy_irq_chip.ack;
+	twl4030_sih_irq_chip.irq_ack = dummy_irq_chip.irq_ack;
 
 	for (i = irq_base; i < irq_end; i++) {
 		set_irq_chip_and_handler(i, &twl4030_irq_chip,
diff --git a/drivers/net/3c527.c b/drivers/net/3c527.c
index 70705d1306b..eca55c52bdf 100644
--- a/drivers/net/3c527.c
+++ b/drivers/net/3c527.c
@@ -522,7 +522,7 @@ static int __init mc32_probe1(struct net_device *dev, int slot)
 	lp->tx_len 		= lp->exec_box->data[9];   /* Transmit list count */
 	lp->rx_len 		= lp->exec_box->data[11];  /* Receive list count */
 
-	init_MUTEX_LOCKED(&lp->cmd_mutex);
+	sema_init(&lp->cmd_mutex, 0);
 	init_completion(&lp->execution_cmd);
 	init_completion(&lp->xceiver_cmd);
 
diff --git a/drivers/net/hamradio/6pack.c b/drivers/net/hamradio/6pack.c
index 4b52c767ad0..3e5d0b6b651 100644
--- a/drivers/net/hamradio/6pack.c
+++ b/drivers/net/hamradio/6pack.c
@@ -608,7 +608,7 @@ static int sixpack_open(struct tty_struct *tty)
 
 	spin_lock_init(&sp->lock);
 	atomic_set(&sp->refcnt, 1);
-	init_MUTEX_LOCKED(&sp->dead_sem);
+	sema_init(&sp->dead_sem, 0);
 
 	/* !!! length of the buffers. MTU is IP MTU, not PACLEN!  */
 
diff --git a/drivers/net/hamradio/mkiss.c b/drivers/net/hamradio/mkiss.c
index 66e88bd59ca..4c628393c8b 100644
--- a/drivers/net/hamradio/mkiss.c
+++ b/drivers/net/hamradio/mkiss.c
@@ -747,7 +747,7 @@ static int mkiss_open(struct tty_struct *tty)
 
 	spin_lock_init(&ax->buflock);
 	atomic_set(&ax->refcnt, 1);
-	init_MUTEX_LOCKED(&ax->dead_sem);
+	sema_init(&ax->dead_sem, 0);
 
 	ax->tty = tty;
 	tty->disc_data = ax;
diff --git a/drivers/net/irda/sir_dev.c b/drivers/net/irda/sir_dev.c
index 1b051dab7b2..51d74447f8f 100644
--- a/drivers/net/irda/sir_dev.c
+++ b/drivers/net/irda/sir_dev.c
@@ -909,7 +909,7 @@ struct sir_dev * sirdev_get_instance(const struct sir_driver *drv, const char *n
 	dev->tx_skb = NULL;
 
 	spin_lock_init(&dev->tx_lock);
-	init_MUTEX(&dev->fsm.sem);
+	sema_init(&dev->fsm.sem, 1);
 
 	dev->drv = drv;
 	dev->netdev = ndev;
diff --git a/drivers/net/ppp_async.c b/drivers/net/ppp_async.c
index af50a530dae..78d70a6481b 100644
--- a/drivers/net/ppp_async.c
+++ b/drivers/net/ppp_async.c
@@ -184,7 +184,7 @@ ppp_asynctty_open(struct tty_struct *tty)
 	tasklet_init(&ap->tsk, ppp_async_process, (unsigned long) ap);
 
 	atomic_set(&ap->refcnt, 1);
-	init_MUTEX_LOCKED(&ap->dead_sem);
+	sema_init(&ap->dead_sem, 0);
 
 	ap->chan.private = ap;
 	ap->chan.ops = &async_ops;
diff --git a/drivers/net/wan/cosa.c b/drivers/net/wan/cosa.c
index 04c6cd4333f..10bafd59f9c 100644
--- a/drivers/net/wan/cosa.c
+++ b/drivers/net/wan/cosa.c
@@ -575,7 +575,7 @@ static int cosa_probe(int base, int irq, int dma)
 
 		/* Initialize the chardev data structures */
 		mutex_init(&chan->rlock);
-		init_MUTEX(&chan->wsem);
+		sema_init(&chan->wsem, 1);
 
 		/* Register the network interface */
 		if (!(chan->netdev = alloc_hdlcdev(chan))) {
diff --git a/drivers/parport/share.c b/drivers/parport/share.c
index dffa5d4fb29..a2d9d1e5926 100644
--- a/drivers/parport/share.c
+++ b/drivers/parport/share.c
@@ -306,7 +306,7 @@ struct parport *parport_register_port(unsigned long base, int irq, int dma,
 	spin_lock_init(&tmp->pardevice_lock);
 	tmp->ieee1284.mode = IEEE1284_MODE_COMPAT;
 	tmp->ieee1284.phase = IEEE1284_PH_FWD_IDLE;
-	init_MUTEX_LOCKED (&tmp->ieee1284.irq); /* actually a semaphore at 0 */
+	sema_init(&tmp->ieee1284.irq, 0);
 	tmp->spintime = parport_default_spintime;
 	atomic_set (&tmp->ref_count, 1);
 	INIT_LIST_HEAD(&tmp->full_list);
diff --git a/drivers/pci/dmar.c b/drivers/pci/dmar.c
index 0a19708074c..3de3a436a43 100644
--- a/drivers/pci/dmar.c
+++ b/drivers/pci/dmar.c
@@ -1221,9 +1221,9 @@ const char *dmar_get_fault_reason(u8 fault_reason, int *fault_type)
 	}
 }
 
-void dmar_msi_unmask(unsigned int irq)
+void dmar_msi_unmask(struct irq_data *data)
 {
-	struct intel_iommu *iommu = get_irq_data(irq);
+	struct intel_iommu *iommu = irq_data_get_irq_data(data);
 	unsigned long flag;
 
 	/* unmask it */
@@ -1234,10 +1234,10 @@ void dmar_msi_unmask(unsigned int irq)
 	spin_unlock_irqrestore(&iommu->register_lock, flag);
 }
 
-void dmar_msi_mask(unsigned int irq)
+void dmar_msi_mask(struct irq_data *data)
 {
 	unsigned long flag;
-	struct intel_iommu *iommu = get_irq_data(irq);
+	struct intel_iommu *iommu = irq_data_get_irq_data(data);
 
 	/* mask it */
 	spin_lock_irqsave(&iommu->register_lock, flag);
diff --git a/drivers/pci/htirq.c b/drivers/pci/htirq.c
index 98abf8b9129..834842aa5bb 100644
--- a/drivers/pci/htirq.c
+++ b/drivers/pci/htirq.c
@@ -57,28 +57,22 @@ void fetch_ht_irq_msg(unsigned int irq, struct ht_irq_msg *msg)
 	*msg = cfg->msg;
 }
 
-void mask_ht_irq(unsigned int irq)
+void mask_ht_irq(struct irq_data *data)
 {
-	struct ht_irq_cfg *cfg;
-	struct ht_irq_msg msg;
-
-	cfg = get_irq_data(irq);
+	struct ht_irq_cfg *cfg = irq_data_get_irq_data(data);
+	struct ht_irq_msg msg = cfg->msg;
 
-	msg = cfg->msg;
 	msg.address_lo |= 1;
-	write_ht_irq_msg(irq, &msg);
+	write_ht_irq_msg(data->irq, &msg);
 }
 
-void unmask_ht_irq(unsigned int irq)
+void unmask_ht_irq(struct irq_data *data)
 {
-	struct ht_irq_cfg *cfg;
-	struct ht_irq_msg msg;
-
-	cfg = get_irq_data(irq);
+	struct ht_irq_cfg *cfg = irq_data_get_irq_data(data);
+	struct ht_irq_msg msg = cfg->msg;
 
-	msg = cfg->msg;
 	msg.address_lo &= ~1;
-	write_ht_irq_msg(irq, &msg);
+	write_ht_irq_msg(data->irq, &msg);
 }
 
 /**
diff --git a/drivers/pci/intr_remapping.c b/drivers/pci/intr_remapping.c
index fd1d2867cdc..ec87cd66f3e 100644
--- a/drivers/pci/intr_remapping.c
+++ b/drivers/pci/intr_remapping.c
@@ -46,109 +46,24 @@ static __init int setup_intremap(char *str)
 }
 early_param("intremap", setup_intremap);
 
-struct irq_2_iommu {
-	struct intel_iommu *iommu;
-	u16 irte_index;
-	u16 sub_handle;
-	u8  irte_mask;
-};
-
-#ifdef CONFIG_GENERIC_HARDIRQS
-static struct irq_2_iommu *get_one_free_irq_2_iommu(int node)
-{
-	struct irq_2_iommu *iommu;
-
-	iommu = kzalloc_node(sizeof(*iommu), GFP_ATOMIC, node);
-	printk(KERN_DEBUG "alloc irq_2_iommu on node %d\n", node);
-
-	return iommu;
-}
-
-static struct irq_2_iommu *irq_2_iommu(unsigned int irq)
-{
-	struct irq_desc *desc;
-
-	desc = irq_to_desc(irq);
-
-	if (WARN_ON_ONCE(!desc))
-		return NULL;
-
-	return desc->irq_2_iommu;
-}
-
-static struct irq_2_iommu *irq_2_iommu_alloc(unsigned int irq)
-{
-	struct irq_desc *desc;
-	struct irq_2_iommu *irq_iommu;
-
-	desc = irq_to_desc(irq);
-	if (!desc) {
-		printk(KERN_INFO "can not get irq_desc for %d\n", irq);
-		return NULL;
-	}
-
-	irq_iommu = desc->irq_2_iommu;
-
-	if (!irq_iommu)
-		desc->irq_2_iommu = get_one_free_irq_2_iommu(irq_node(irq));
-
-	return desc->irq_2_iommu;
-}
-
-#else /* !CONFIG_SPARSE_IRQ */
-
-static struct irq_2_iommu irq_2_iommuX[NR_IRQS];
-
-static struct irq_2_iommu *irq_2_iommu(unsigned int irq)
-{
-	if (irq < nr_irqs)
-		return &irq_2_iommuX[irq];
-
-	return NULL;
-}
-static struct irq_2_iommu *irq_2_iommu_alloc(unsigned int irq)
-{
-	return irq_2_iommu(irq);
-}
-#endif
-
 static DEFINE_SPINLOCK(irq_2_ir_lock);
 
-static struct irq_2_iommu *valid_irq_2_iommu(unsigned int irq)
-{
-	struct irq_2_iommu *irq_iommu;
-
-	irq_iommu = irq_2_iommu(irq);
-
-	if (!irq_iommu)
-		return NULL;
-
-	if (!irq_iommu->iommu)
-		return NULL;
-
-	return irq_iommu;
-}
-
-int irq_remapped(int irq)
+static struct irq_2_iommu *irq_2_iommu(unsigned int irq)
 {
-	return valid_irq_2_iommu(irq) != NULL;
+	struct irq_cfg *cfg = get_irq_chip_data(irq);
+	return cfg ? &cfg->irq_2_iommu : NULL;
 }
 
 int get_irte(int irq, struct irte *entry)
 {
-	int index;
-	struct irq_2_iommu *irq_iommu;
+	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
 	unsigned long flags;
+	int index;
 
-	if (!entry)
+	if (!entry || !irq_iommu)
 		return -1;
 
 	spin_lock_irqsave(&irq_2_ir_lock, flags);
-	irq_iommu = valid_irq_2_iommu(irq);
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-		return -1;
-	}
 
 	index = irq_iommu->irte_index + irq_iommu->sub_handle;
 	*entry = *(irq_iommu->iommu->ir_table->base + index);
@@ -160,20 +75,14 @@ int get_irte(int irq, struct irte *entry)
 int alloc_irte(struct intel_iommu *iommu, int irq, u16 count)
 {
 	struct ir_table *table = iommu->ir_table;
-	struct irq_2_iommu *irq_iommu;
+	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
 	u16 index, start_index;
 	unsigned int mask = 0;
 	unsigned long flags;
 	int i;
 
-	if (!count)
-		return -1;
-
-#ifndef CONFIG_SPARSE_IRQ
-	/* protect irq_2_iommu_alloc later */
-	if (irq >= nr_irqs)
+	if (!count || !irq_iommu)
 		return -1;
-#endif
 
 	/*
 	 * start the IRTE search from index 0.
@@ -214,13 +123,6 @@ int alloc_irte(struct intel_iommu *iommu, int irq, u16 count)
 	for (i = index; i < index + count; i++)
 		table->base[i].present = 1;
 
-	irq_iommu = irq_2_iommu_alloc(irq);
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-		printk(KERN_ERR "can't allocate irq_2_iommu\n");
-		return -1;
-	}
-
 	irq_iommu->iommu = iommu;
 	irq_iommu->irte_index =  index;
 	irq_iommu->sub_handle = 0;
@@ -244,17 +146,14 @@ static int qi_flush_iec(struct intel_iommu *iommu, int index, int mask)
 
 int map_irq_to_irte_handle(int irq, u16 *sub_handle)
 {
-	int index;
-	struct irq_2_iommu *irq_iommu;
+	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
 	unsigned long flags;
+	int index;
 
-	spin_lock_irqsave(&irq_2_ir_lock, flags);
-	irq_iommu = valid_irq_2_iommu(irq);
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
+	if (!irq_iommu)
 		return -1;
-	}
 
+	spin_lock_irqsave(&irq_2_ir_lock, flags);
 	*sub_handle = irq_iommu->sub_handle;
 	index = irq_iommu->irte_index;
 	spin_unlock_irqrestore(&irq_2_ir_lock, flags);
@@ -263,18 +162,13 @@ int map_irq_to_irte_handle(int irq, u16 *sub_handle)
 
 int set_irte_irq(int irq, struct intel_iommu *iommu, u16 index, u16 subhandle)
 {
-	struct irq_2_iommu *irq_iommu;
+	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
 	unsigned long flags;
 
-	spin_lock_irqsave(&irq_2_ir_lock, flags);
-
-	irq_iommu = irq_2_iommu_alloc(irq);
-
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-		printk(KERN_ERR "can't allocate irq_2_iommu\n");
+	if (!irq_iommu)
 		return -1;
-	}
+
+	spin_lock_irqsave(&irq_2_ir_lock, flags);
 
 	irq_iommu->iommu = iommu;
 	irq_iommu->irte_index = index;
@@ -286,43 +180,18 @@ int set_irte_irq(int irq, struct intel_iommu *iommu, u16 index, u16 subhandle)
 	return 0;
 }
 
-int clear_irte_irq(int irq, struct intel_iommu *iommu, u16 index)
-{
-	struct irq_2_iommu *irq_iommu;
-	unsigned long flags;
-
-	spin_lock_irqsave(&irq_2_ir_lock, flags);
-	irq_iommu = valid_irq_2_iommu(irq);
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-		return -1;
-	}
-
-	irq_iommu->iommu = NULL;
-	irq_iommu->irte_index = 0;
-	irq_iommu->sub_handle = 0;
-	irq_2_iommu(irq)->irte_mask = 0;
-
-	spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-
-	return 0;
-}
-
 int modify_irte(int irq, struct irte *irte_modified)
 {
-	int rc;
-	int index;
-	struct irte *irte;
+	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
 	struct intel_iommu *iommu;
-	struct irq_2_iommu *irq_iommu;
 	unsigned long flags;
+	struct irte *irte;
+	int rc, index;
 
-	spin_lock_irqsave(&irq_2_ir_lock, flags);
-	irq_iommu = valid_irq_2_iommu(irq);
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
+	if (!irq_iommu)
 		return -1;
-	}
+
+	spin_lock_irqsave(&irq_2_ir_lock, flags);
 
 	iommu = irq_iommu->iommu;
 
@@ -339,31 +208,6 @@ int modify_irte(int irq, struct irte *irte_modified)
 	return rc;
 }
 
-int flush_irte(int irq)
-{
-	int rc;
-	int index;
-	struct intel_iommu *iommu;
-	struct irq_2_iommu *irq_iommu;
-	unsigned long flags;
-
-	spin_lock_irqsave(&irq_2_ir_lock, flags);
-	irq_iommu = valid_irq_2_iommu(irq);
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-		return -1;
-	}
-
-	iommu = irq_iommu->iommu;
-
-	index = irq_iommu->irte_index + irq_iommu->sub_handle;
-
-	rc = qi_flush_iec(iommu, index, irq_iommu->irte_mask);
-	spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-
-	return rc;
-}
-
 struct intel_iommu *map_hpet_to_ir(u8 hpet_id)
 {
 	int i;
@@ -420,16 +264,14 @@ static int clear_entries(struct irq_2_iommu *irq_iommu)
 
 int free_irte(int irq)
 {
-	int rc = 0;
-	struct irq_2_iommu *irq_iommu;
+	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
 	unsigned long flags;
+	int rc;
 
-	spin_lock_irqsave(&irq_2_ir_lock, flags);
-	irq_iommu = valid_irq_2_iommu(irq);
-	if (!irq_iommu) {
-		spin_unlock_irqrestore(&irq_2_ir_lock, flags);
+	if (!irq_iommu)
 		return -1;
-	}
+
+	spin_lock_irqsave(&irq_2_ir_lock, flags);
 
 	rc = clear_entries(irq_iommu);
 
diff --git a/drivers/pci/msi.c b/drivers/pci/msi.c
index 69b7be33b3a..5fcf5aec680 100644
--- a/drivers/pci/msi.c
+++ b/drivers/pci/msi.c
@@ -170,33 +170,31 @@ static void msix_mask_irq(struct msi_desc *desc, u32 flag)
 	desc->masked = __msix_mask_irq(desc, flag);
 }
 
-static void msi_set_mask_bit(unsigned irq, u32 flag)
+static void msi_set_mask_bit(struct irq_data *data, u32 flag)
 {
-	struct msi_desc *desc = get_irq_msi(irq);
+	struct msi_desc *desc = irq_data_get_msi(data);
 
 	if (desc->msi_attrib.is_msix) {
 		msix_mask_irq(desc, flag);
 		readl(desc->mask_base);		/* Flush write to device */
 	} else {
-		unsigned offset = irq - desc->dev->irq;
+		unsigned offset = data->irq - desc->dev->irq;
 		msi_mask_irq(desc, 1 << offset, flag << offset);
 	}
 }
 
-void mask_msi_irq(unsigned int irq)
+void mask_msi_irq(struct irq_data *data)
 {
-	msi_set_mask_bit(irq, 1);
+	msi_set_mask_bit(data, 1);
 }
 
-void unmask_msi_irq(unsigned int irq)
+void unmask_msi_irq(struct irq_data *data)
 {
-	msi_set_mask_bit(irq, 0);
+	msi_set_mask_bit(data, 0);
 }
 
-void read_msi_msg_desc(struct irq_desc *desc, struct msi_msg *msg)
+void __read_msi_msg(struct msi_desc *entry, struct msi_msg *msg)
 {
-	struct msi_desc *entry = get_irq_desc_msi(desc);
-
 	BUG_ON(entry->dev->current_state != PCI_D0);
 
 	if (entry->msi_attrib.is_msix) {
@@ -227,15 +225,13 @@ void read_msi_msg_desc(struct irq_desc *desc, struct msi_msg *msg)
 
 void read_msi_msg(unsigned int irq, struct msi_msg *msg)
 {
-	struct irq_desc *desc = irq_to_desc(irq);
+	struct msi_desc *entry = get_irq_msi(irq);
 
-	read_msi_msg_desc(desc, msg);
+	__read_msi_msg(entry, msg);
 }
 
-void get_cached_msi_msg_desc(struct irq_desc *desc, struct msi_msg *msg)
+void __get_cached_msi_msg(struct msi_desc *entry, struct msi_msg *msg)
 {
-	struct msi_desc *entry = get_irq_desc_msi(desc);
-
 	/* Assert that the cache is valid, assuming that
 	 * valid messages are not all-zeroes. */
 	BUG_ON(!(entry->msg.address_hi | entry->msg.address_lo |
@@ -246,15 +242,13 @@ void get_cached_msi_msg_desc(struct irq_desc *desc, struct msi_msg *msg)
 
 void get_cached_msi_msg(unsigned int irq, struct msi_msg *msg)
 {
-	struct irq_desc *desc = irq_to_desc(irq);
+	struct msi_desc *entry = get_irq_msi(irq);
 
-	get_cached_msi_msg_desc(desc, msg);
+	__get_cached_msi_msg(entry, msg);
 }
 
-void write_msi_msg_desc(struct irq_desc *desc, struct msi_msg *msg)
+void __write_msi_msg(struct msi_desc *entry, struct msi_msg *msg)
 {
-	struct msi_desc *entry = get_irq_desc_msi(desc);
-
 	if (entry->dev->current_state != PCI_D0) {
 		/* Don't touch the hardware now */
 	} else if (entry->msi_attrib.is_msix) {
@@ -292,9 +286,9 @@ void write_msi_msg_desc(struct irq_desc *desc, struct msi_msg *msg)
 
 void write_msi_msg(unsigned int irq, struct msi_msg *msg)
 {
-	struct irq_desc *desc = irq_to_desc(irq);
+	struct msi_desc *entry = get_irq_msi(irq);
 
-	write_msi_msg_desc(desc, msg);
+	__write_msi_msg(entry, msg);
 }
 
 static void free_msi_irqs(struct pci_dev *dev)
diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 7c8008225ee..17927b1f933 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -127,7 +127,10 @@ static void handle_tx(struct vhost_net *net)
 	size_t len, total_len = 0;
 	int err, wmem;
 	size_t hdr_size;
-	struct socket *sock = rcu_dereference(vq->private_data);
+	struct socket *sock;
+
+	sock = rcu_dereference_check(vq->private_data,
+				     lockdep_is_held(&vq->mutex));
 	if (!sock)
 		return;
 
@@ -582,7 +585,10 @@ static void vhost_net_disable_vq(struct vhost_net *n,
 static void vhost_net_enable_vq(struct vhost_net *n,
 				struct vhost_virtqueue *vq)
 {
-	struct socket *sock = vq->private_data;
+	struct socket *sock;
+
+	sock = rcu_dereference_protected(vq->private_data,
+					 lockdep_is_held(&vq->mutex));
 	if (!sock)
 		return;
 	if (vq == n->vqs + VHOST_NET_VQ_TX) {
@@ -598,7 +604,8 @@ static struct socket *vhost_net_stop_vq(struct vhost_net *n,
 	struct socket *sock;
 
 	mutex_lock(&vq->mutex);
-	sock = vq->private_data;
+	sock = rcu_dereference_protected(vq->private_data,
+					 lockdep_is_held(&vq->mutex));
 	vhost_net_disable_vq(n, vq);
 	rcu_assign_pointer(vq->private_data, NULL);
 	mutex_unlock(&vq->mutex);
@@ -736,7 +743,8 @@ static long vhost_net_set_backend(struct vhost_net *n, unsigned index, int fd)
 	}
 
 	/* start polling new socket */
-	oldsock = vq->private_data;
+	oldsock = rcu_dereference_protected(vq->private_data,
+					    lockdep_is_held(&vq->mutex));
 	if (sock != oldsock) {
                 vhost_net_disable_vq(n, vq);
                 rcu_assign_pointer(vq->private_data, sock);
diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index dd3d6f7406f..8b5a1b33d0f 100644
--- a/drivers/vhost/vhost.c
+++ b/drivers/vhost/vhost.c
@@ -320,7 +320,7 @@ long vhost_dev_reset_owner(struct vhost_dev *dev)
 	vhost_dev_cleanup(dev);
 
 	memory->nregions = 0;
-	dev->memory = memory;
+	RCU_INIT_POINTER(dev->memory, memory);
 	return 0;
 }
 
@@ -352,8 +352,9 @@ void vhost_dev_cleanup(struct vhost_dev *dev)
 		fput(dev->log_file);
 	dev->log_file = NULL;
 	/* No one will access memory at this point */
-	kfree(dev->memory);
-	dev->memory = NULL;
+	kfree(rcu_dereference_protected(dev->memory,
+					lockdep_is_held(&dev->mutex)));
+	RCU_INIT_POINTER(dev->memory, NULL);
 	if (dev->mm)
 		mmput(dev->mm);
 	dev->mm = NULL;
@@ -440,14 +441,22 @@ static int vq_access_ok(unsigned int num,
 /* Caller should have device mutex but not vq mutex */
 int vhost_log_access_ok(struct vhost_dev *dev)
 {
-	return memory_access_ok(dev, dev->memory, 1);
+	struct vhost_memory *mp;
+
+	mp = rcu_dereference_protected(dev->memory,
+				       lockdep_is_held(&dev->mutex));
+	return memory_access_ok(dev, mp, 1);
 }
 
 /* Verify access for write logging. */
 /* Caller should have vq mutex and device mutex */
 static int vq_log_access_ok(struct vhost_virtqueue *vq, void __user *log_base)
 {
-	return vq_memory_access_ok(log_base, vq->dev->memory,
+	struct vhost_memory *mp;
+
+	mp = rcu_dereference_protected(vq->dev->memory,
+				       lockdep_is_held(&vq->mutex));
+	return vq_memory_access_ok(log_base, mp,
 			    vhost_has_feature(vq->dev, VHOST_F_LOG_ALL)) &&
 		(!vq->log_used || log_access_ok(log_base, vq->log_addr,
 					sizeof *vq->used +
@@ -487,7 +496,8 @@ static long vhost_set_memory(struct vhost_dev *d, struct vhost_memory __user *m)
 		kfree(newmem);
 		return -EFAULT;
 	}
-	oldmem = d->memory;
+	oldmem = rcu_dereference_protected(d->memory,
+					   lockdep_is_held(&d->mutex));
 	rcu_assign_pointer(d->memory, newmem);
 	synchronize_rcu();
 	kfree(oldmem);
diff --git a/drivers/vhost/vhost.h b/drivers/vhost/vhost.h
index afd77295971..af3c11ded5f 100644
--- a/drivers/vhost/vhost.h
+++ b/drivers/vhost/vhost.h
@@ -106,7 +106,7 @@ struct vhost_virtqueue {
 	 * vhost_work execution acts instead of rcu_read_lock() and the end of
 	 * vhost_work execution acts instead of rcu_read_lock().
 	 * Writers use virtqueue mutex. */
-	void *private_data;
+	void __rcu *private_data;
 	/* Log write descriptors */
 	void __user *log_base;
 	struct vhost_log log[VHOST_NET_MAX_SG];
@@ -116,7 +116,7 @@ struct vhost_dev {
 	/* Readers use RCU to access memory table pointer
 	 * log base pointer and features.
 	 * Writers use mutex below.*/
-	struct vhost_memory *memory;
+	struct vhost_memory __rcu *memory;
 	struct mm_struct *mm;
 	struct mutex mutex;
 	unsigned acked_features;
@@ -173,7 +173,11 @@ enum {
 
 static inline int vhost_has_feature(struct vhost_dev *dev, int bit)
 {
-	unsigned acked_features = rcu_dereference(dev->acked_features);
+	unsigned acked_features;
+
+	acked_features =
+		rcu_dereference_index_check(dev->acked_features,
+					    lockdep_is_held(&dev->mutex));
 	return acked_features & (1 << bit);
 }
 
diff --git a/drivers/xen/events.c b/drivers/xen/events.c
index 13365ba3521..7d24b0d94ed 100644
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -338,30 +338,29 @@ static void unmask_evtchn(int port)
 
 static int find_unbound_irq(void)
 {
-	int irq;
-	struct irq_desc *desc;
+	struct irq_data *data;
+	int irq, res;
 
 	for (irq = 0; irq < nr_irqs; irq++) {
-		desc = irq_to_desc(irq);
+		data = irq_get_irq_data(irq);
 		/* only 0->15 have init'd desc; handle irq > 16 */
-		if (desc == NULL)
+		if (!data)
 			break;
-		if (desc->chip == &no_irq_chip)
+		if (data->chip == &no_irq_chip)
 			break;
-		if (desc->chip != &xen_dynamic_chip)
+		if (data->chip != &xen_dynamic_chip)
 			continue;
 		if (irq_info[irq].type == IRQT_UNBOUND)
-			break;
+			return irq;
 	}
 
 	if (irq == nr_irqs)
 		panic("No available IRQ to bind to: increase nr_irqs!\n");
 
-	desc = irq_to_desc_alloc_node(irq, 0);
-	if (WARN_ON(desc == NULL))
-		return -1;
+	res = irq_alloc_desc_at(irq, 0);
 
-	dynamic_irq_init_keep_chip_data(irq);
+	if (WARN_ON(res != irq))
+		return -1;
 
 	return irq;
 }
@@ -495,7 +494,7 @@ static void unbind_from_irq(unsigned int irq)
 	if (irq_info[irq].type != IRQT_UNBOUND) {
 		irq_info[irq] = mk_unbound_info();
 
-		dynamic_irq_cleanup(irq);
+		irq_free_desc(irq);
 	}
 
 	spin_unlock(&irq_mapping_update_lock);