Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-nextgrafted

Pull networking updates from Jakub Kicinski:
 "Core:

   - Add dedicated kmem_cache for typical/small skb->head, avoid having
     to access struct page at kfree time, and improve memory use.

   - Introduce sysctl to set default RPS configuration for new netdevs.

   - Define Netlink protocol specification format which can be used to
     describe messages used by each family and auto-generate parsers.
     Add tools for generating kernel data structures and uAPI headers.

   - Expose all net/core sysctls inside netns.

   - Remove 4s sleep in netpoll if carrier is instantly detected on
     boot.

   - Add configurable limit of MDB entries per port, and port-vlan.

   - Continue populating drop reasons throughout the stack.

   - Retire a handful of legacy Qdiscs and classifiers.

  Protocols:

   - Support IPv4 big TCP (TSO frames larger than 64kB).

   - Add IP_LOCAL_PORT_RANGE socket option, to control local port range
     on socket by socket basis.

   - Track and report in procfs number of MPTCP sockets used.

   - Support mixing IPv4 and IPv6 flows in the in-kernel MPTCP path
     manager.

   - IPv6: don't check net.ipv6.route.max_size and rely on garbage
     collection to free memory (similarly to IPv4).

   - Support Penultimate Segment Pop (PSP) flavor in SRv6 (RFC8986).

   - ICMP: add per-rate limit counters.

   - Add support for user scanning requests in ieee802154.

   - Remove static WEP support.

   - Support minimal Wi-Fi 7 Extremely High Throughput (EHT) rate
     reporting.

   - WiFi 7 EHT channel puncturing support (client & AP).

  BPF:

   - Add a rbtree data structure following the "next-gen data structure"
     precedent set by recently added linked list, that is, by using
     kfunc + kptr instead of adding a new BPF map type.

   - Expose XDP hints via kfuncs with initial support for RX hash and
     timestamp metadata.

   - Add BPF_F_NO_TUNNEL_KEY extension to bpf_skb_set_tunnel_key to
     better support decap on GRE tunnel devices not operating in collect
     metadata.

   - Improve x86 JIT's codegen for PROBE_MEM runtime error checks.

   - Remove the need for trace_printk_lock for bpf_trace_printk and
     bpf_trace_vprintk helpers.

   - Extend libbpf's bpf_tracing.h support for tracing arguments of
     kprobes/uprobes and syscall as a special case.

   - Significantly reduce the search time for module symbols by
     livepatch and BPF.

   - Enable cpumasks to be used as kptrs, which is useful for tracing
     programs tracking which tasks end up running on which CPUs in
     different time intervals.

   - Add support for BPF trampoline on s390x and riscv64.

   - Add capability to export the XDP features supported by the NIC.

   - Add __bpf_kfunc tag for marking kernel functions as kfuncs.

   - Add cgroup.memory=nobpf kernel parameter option to disable BPF
     memory accounting for container environments.

  Netfilter:

   - Remove the CLUSTERIP target. It has been marked as obsolete for
     years, and we still have WARN splats wrt races of the out-of-band
     /proc interface installed by this target.

   - Add 'destroy' commands to nf_tables. They are identical to the
     existing 'delete' commands, but do not return an error if the
     referenced object (set, chain, rule...) did not exist.

  Driver API:

   - Improve cpumask_local_spread() locality to help NICs set the right
     IRQ affinity on AMD platforms.

   - Separate C22 and C45 MDIO bus transactions more clearly.

   - Introduce new DCB table to control DSCP rewrite on egress.

   - Support configuration of Physical Layer Collision Avoidance (PLCA)
     Reconciliation Sublayer (RS) (802.3cg-2019). Modern version of
     shared medium Ethernet.

   - Support for MAC Merge layer (IEEE 802.3-2018 clause 99). Allowing
     preemption of low priority frames by high priority frames.

   - Add support for controlling MACSec offload using netlink SET.

   - Rework devlink instance refcounts to allow registration and
     de-registration under the instance lock. Split the code into
     multiple files, drop some of the unnecessarily granular locks and
     factor out common parts of netlink operation handling.

   - Add TX frame aggregation parameters (for USB drivers).

   - Add a new attr TCA_EXT_WARN_MSG to report TC (offload) warning
     messages with notifications for debug.

   - Allow offloading of UDP NEW connections via act_ct.

   - Add support for per action HW stats in TC.

   - Support hardware miss to TC action (continue processing in SW from
     a specific point in the action chain).

   - Warn if old Wireless Extension user space interface is used with
     modern cfg80211/mac80211 drivers. Do not support Wireless
     Extensions for Wi-Fi 7 devices at all. Everyone should switch to
     using nl80211 interface instead.

   - Improve the CAN bit timing configuration. Use extack to return
     error messages directly to user space, update the SJW handling,
     including the definition of a new default value that will benefit
     CAN-FD controllers, by increasing their oscillator tolerance.

  New hardware / drivers:

   - Ethernet:
      - nVidia BlueField-3 support (control traffic driver)
      - Ethernet support for imx93 SoCs
      - Motorcomm yt8531 gigabit Ethernet PHY
      - onsemi NCN26000 10BASE-T1S PHY (with support for PLCA)
      - Microchip LAN8841 PHY (incl. cable diagnostics and PTP)
      - Amlogic gxl MDIO mux

   - WiFi:
      - RealTek RTL8188EU (rtl8xxxu)
      - Qualcomm Wi-Fi 7 devices (ath12k)

   - CAN:
      - Renesas R-Car V4H

  Drivers:

   - Bluetooth:
      - Set Per Platform Antenna Gain (PPAG) for Intel controllers.

   - Ethernet NICs:
      - Intel (1G, igc):
         - support TSN / Qbv / packet scheduling features of i226 model
      - Intel (100G, ice):
         - use GNSS subsystem instead of TTY
         - multi-buffer XDP support
         - extend support for GPIO pins to E823 devices
      - nVidia/Mellanox:
         - update the shared buffer configuration on PFC commands
         - implement PTP adjphase function for HW offset control
         - TC support for Geneve and GRE with VF tunnel offload
         - more efficient crypto key management method
         - multi-port eswitch support
      - Netronome/Corigine:
         - add DCB IEEE support
         - support IPsec offloading for NFP3800
      - Freescale/NXP (enetc):
         - support XDP_REDIRECT for XDP non-linear buffers
         - improve reconfig, avoid link flap and waiting for idle
         - support MAC Merge layer
      - Other NICs:
         - sfc/ef100: add basic devlink support for ef100
         - ionic: rx_push mode operation (writing descriptors via MMIO)
         - bnxt: use the auxiliary bus abstraction for RDMA
         - r8169: disable ASPM and reset bus in case of tx timeout
         - cpsw: support QSGMII mode for J721e CPSW9G
         - cpts: support pulse-per-second output
         - ngbe: add an mdio bus driver
         - usbnet: optimize usbnet_bh() by avoiding unnecessary queuing
         - r8152: handle devices with FW with NCM support
         - amd-xgbe: support 10Mbps, 2.5GbE speeds and rx-adaptation
         - virtio-net: support multi buffer XDP
         - virtio/vsock: replace virtio_vsock_pkt with sk_buff
         - tsnep: XDP support

   - Ethernet high-speed switches:
      - nVidia/Mellanox (mlxsw):
         - add support for latency TLV (in FW control messages)
      - Microchip (sparx5):
         - separate explicit and implicit traffic forwarding rules, make
           the implicit rules always active
         - add support for egress DSCP rewrite
         - IS0 VCAP support (Ingress Classification)
         - IS2 VCAP filters (protos, L3 addrs, L4 ports, flags, ToS
           etc.)
         - ES2 VCAP support (Egress Access Control)
         - support for Per-Stream Filtering and Policing (802.1Q,
           8.6.5.1)

   - Ethernet embedded switches:
      - Marvell (mv88e6xxx):
         - add MAB (port auth) offload support
         - enable PTP receive for mv88e6390
      - NXP (ocelot):
         - support MAC Merge layer
         - support for the the vsc7512 internal copper phys
      - Microchip:
         - lan9303: convert to PHYLINK
         - lan966x: support TC flower filter statistics
         - lan937x: PTP support for KSZ9563/KSZ8563 and LAN937x
         - lan937x: support Credit Based Shaper configuration
         - ksz9477: support Energy Efficient Ethernet
      - other:
         - qca8k: convert to regmap read/write API, use bulk operations
         - rswitch: Improve TX timestamp accuracy

   - Intel WiFi (iwlwifi):
      - EHT (Wi-Fi 7) rate reporting
      - STEP equalizer support: transfer some STEP (connection to radio
        on platforms with integrated wifi) related parameters from the
        BIOS to the firmware.

   - Qualcomm 802.11ax WiFi (ath11k):
      - IPQ5018 support
      - Fine Timing Measurement (FTM) responder role support
      - channel 177 support

   - MediaTek WiFi (mt76):
      - per-PHY LED support
      - mt7996: EHT (Wi-Fi 7) support
      - Wireless Ethernet Dispatch (WED) reset support
      - switch to using page pool allocator

   - RealTek WiFi (rtw89):
      - support new version of Bluetooth co-existance

   - Mobile:
      - rmnet: support TX aggregation"

* tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1872 commits)
  page_pool: add a comment explaining the fragment counter usage
  net: ethtool: fix __ethtool_dev_mm_supported() implementation
  ethtool: pse-pd: Fix double word in comments
  xsk: add linux/vmalloc.h to xsk.c
  sefltests: netdevsim: wait for devlink instance after netns removal
  selftest: fib_tests: Always cleanup before exit
  net/mlx5e: Align IPsec ASO result memory to be as required by hardware
  net/mlx5e: TC, Set CT miss to the specific ct action instance
  net/mlx5e: Rename CHAIN_TO_REG to MAPPED_OBJ_TO_REG
  net/mlx5: Refactor tc miss handling to a single function
  net/mlx5: Kconfig: Make tc offload depend on tc skb extension
  net/sched: flower: Support hardware miss to tc action
  net/sched: flower: Move filter handle initialization earlier
  net/sched: cls_api: Support hardware miss to tc action
  net/sched: Rename user cookie and act cookie
  sfc: fix builds without CONFIG_RTC_LIB
  sfc: clean up some inconsistent indentings
  net/mlx4_en: Introduce flexible array to silence overflow warning
  net: lan966x: Fix possible deadlock inside PTP
  net/ulp: Remove redundant ->clone() test in inet_clone_ulp().
  ...

1 files changed, 724 insertions, 0 deletions

diff --git a/drivers/md/dm-region-hash.c b/drivers/md/dm-region-hash.c
new file mode 100644
index 000000000..1f760451e
--- /dev/null
+++ b/drivers/md/dm-region-hash.c
@@ -0,0 +1,724 @@
+/*
+ * Copyright (C) 2003 Sistina Software Limited.
+ * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
+ *
+ * This file is released under the GPL.
+ */
+
+#include <linux/dm-dirty-log.h>
+#include <linux/dm-region-hash.h>
+
+#include <linux/ctype.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/vmalloc.h>
+
+#include "dm.h"
+
+#define	DM_MSG_PREFIX	"region hash"
+
+/*-----------------------------------------------------------------
+ * Region hash
+ *
+ * The mirror splits itself up into discrete regions.  Each
+ * region can be in one of three states: clean, dirty,
+ * nosync.  There is no need to put clean regions in the hash.
+ *
+ * In addition to being present in the hash table a region _may_
+ * be present on one of three lists.
+ *
+ *   clean_regions: Regions on this list have no io pending to
+ *   them, they are in sync, we are no longer interested in them,
+ *   they are dull.  dm_rh_update_states() will remove them from the
+ *   hash table.
+ *
+ *   quiesced_regions: These regions have been spun down, ready
+ *   for recovery.  rh_recovery_start() will remove regions from
+ *   this list and hand them to kmirrord, which will schedule the
+ *   recovery io with kcopyd.
+ *
+ *   recovered_regions: Regions that kcopyd has successfully
+ *   recovered.  dm_rh_update_states() will now schedule any delayed
+ *   io, up the recovery_count, and remove the region from the
+ *   hash.
+ *
+ * There are 2 locks:
+ *   A rw spin lock 'hash_lock' protects just the hash table,
+ *   this is never held in write mode from interrupt context,
+ *   which I believe means that we only have to disable irqs when
+ *   doing a write lock.
+ *
+ *   An ordinary spin lock 'region_lock' that protects the three
+ *   lists in the region_hash, with the 'state', 'list' and
+ *   'delayed_bios' fields of the regions.  This is used from irq
+ *   context, so all other uses will have to suspend local irqs.
+ *---------------------------------------------------------------*/
+struct dm_region_hash {
+	uint32_t region_size;
+	unsigned region_shift;
+
+	/* holds persistent region state */
+	struct dm_dirty_log *log;
+
+	/* hash table */
+	rwlock_t hash_lock;
+	unsigned mask;
+	unsigned nr_buckets;
+	unsigned prime;
+	unsigned shift;
+	struct list_head *buckets;
+
+	/*
+	 * If there was a flush failure no regions can be marked clean.
+	 */
+	int flush_failure;
+
+	unsigned max_recovery; /* Max # of regions to recover in parallel */
+
+	spinlock_t region_lock;
+	atomic_t recovery_in_flight;
+	struct list_head clean_regions;
+	struct list_head quiesced_regions;
+	struct list_head recovered_regions;
+	struct list_head failed_recovered_regions;
+	struct semaphore recovery_count;
+
+	mempool_t region_pool;
+
+	void *context;
+	sector_t target_begin;
+
+	/* Callback function to schedule bios writes */
+	void (*dispatch_bios)(void *context, struct bio_list *bios);
+
+	/* Callback function to wakeup callers worker thread. */
+	void (*wakeup_workers)(void *context);
+
+	/* Callback function to wakeup callers recovery waiters. */
+	void (*wakeup_all_recovery_waiters)(void *context);
+};
+
+struct dm_region {
+	struct dm_region_hash *rh;	/* FIXME: can we get rid of this ? */
+	region_t key;
+	int state;
+
+	struct list_head hash_list;
+	struct list_head list;
+
+	atomic_t pending;
+	struct bio_list delayed_bios;
+};
+
+/*
+ * Conversion fns
+ */
+static region_t dm_rh_sector_to_region(struct dm_region_hash *rh, sector_t sector)
+{
+	return sector >> rh->region_shift;
+}
+
+sector_t dm_rh_region_to_sector(struct dm_region_hash *rh, region_t region)
+{
+	return region << rh->region_shift;
+}
+EXPORT_SYMBOL_GPL(dm_rh_region_to_sector);
+
+region_t dm_rh_bio_to_region(struct dm_region_hash *rh, struct bio *bio)
+{
+	return dm_rh_sector_to_region(rh, bio->bi_iter.bi_sector -
+				      rh->target_begin);
+}
+EXPORT_SYMBOL_GPL(dm_rh_bio_to_region);
+
+void *dm_rh_region_context(struct dm_region *reg)
+{
+	return reg->rh->context;
+}
+EXPORT_SYMBOL_GPL(dm_rh_region_context);
+
+region_t dm_rh_get_region_key(struct dm_region *reg)
+{
+	return reg->key;
+}
+EXPORT_SYMBOL_GPL(dm_rh_get_region_key);
+
+sector_t dm_rh_get_region_size(struct dm_region_hash *rh)
+{
+	return rh->region_size;
+}
+EXPORT_SYMBOL_GPL(dm_rh_get_region_size);
+
+/*
+ * FIXME: shall we pass in a structure instead of all these args to
+ * dm_region_hash_create()????
+ */
+#define RH_HASH_MULT 2654435387U
+#define RH_HASH_SHIFT 12
+
+#define MIN_REGIONS 64
+struct dm_region_hash *dm_region_hash_create(
+		void *context, void (*dispatch_bios)(void *context,
+						     struct bio_list *bios),
+		void (*wakeup_workers)(void *context),
+		void (*wakeup_all_recovery_waiters)(void *context),
+		sector_t target_begin, unsigned max_recovery,
+		struct dm_dirty_log *log, uint32_t region_size,
+		region_t nr_regions)
+{
+	struct dm_region_hash *rh;
+	unsigned nr_buckets, max_buckets;
+	size_t i;
+	int ret;
+
+	/*
+	 * Calculate a suitable number of buckets for our hash
+	 * table.
+	 */
+	max_buckets = nr_regions >> 6;
+	for (nr_buckets = 128u; nr_buckets < max_buckets; nr_buckets <<= 1)
+		;
+	nr_buckets >>= 1;
+
+	rh = kzalloc(sizeof(*rh), GFP_KERNEL);
+	if (!rh) {
+		DMERR("unable to allocate region hash memory");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	rh->context = context;
+	rh->dispatch_bios = dispatch_bios;
+	rh->wakeup_workers = wakeup_workers;
+	rh->wakeup_all_recovery_waiters = wakeup_all_recovery_waiters;
+	rh->target_begin = target_begin;
+	rh->max_recovery = max_recovery;
+	rh->log = log;
+	rh->region_size = region_size;
+	rh->region_shift = __ffs(region_size);
+	rwlock_init(&rh->hash_lock);
+	rh->mask = nr_buckets - 1;
+	rh->nr_buckets = nr_buckets;
+
+	rh->shift = RH_HASH_SHIFT;
+	rh->prime = RH_HASH_MULT;
+
+	rh->buckets = vmalloc(array_size(nr_buckets, sizeof(*rh->buckets)));
+	if (!rh->buckets) {
+		DMERR("unable to allocate region hash bucket memory");
+		kfree(rh);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	for (i = 0; i < nr_buckets; i++)
+		INIT_LIST_HEAD(rh->buckets + i);
+
+	spin_lock_init(&rh->region_lock);
+	sema_init(&rh->recovery_count, 0);
+	atomic_set(&rh->recovery_in_flight, 0);
+	INIT_LIST_HEAD(&rh->clean_regions);
+	INIT_LIST_HEAD(&rh->quiesced_regions);
+	INIT_LIST_HEAD(&rh->recovered_regions);
+	INIT_LIST_HEAD(&rh->failed_recovered_regions);
+	rh->flush_failure = 0;
+
+	ret = mempool_init_kmalloc_pool(&rh->region_pool, MIN_REGIONS,
+					sizeof(struct dm_region));
+	if (ret) {
+		vfree(rh->buckets);
+		kfree(rh);
+		rh = ERR_PTR(-ENOMEM);
+	}
+
+	return rh;
+}
+EXPORT_SYMBOL_GPL(dm_region_hash_create);
+
+void dm_region_hash_destroy(struct dm_region_hash *rh)
+{
+	unsigned h;
+	struct dm_region *reg, *nreg;
+
+	BUG_ON(!list_empty(&rh->quiesced_regions));
+	for (h = 0; h < rh->nr_buckets; h++) {
+		list_for_each_entry_safe(reg, nreg, rh->buckets + h,
+					 hash_list) {
+			BUG_ON(atomic_read(&reg->pending));
+			mempool_free(reg, &rh->region_pool);
+		}
+	}
+
+	if (rh->log)
+		dm_dirty_log_destroy(rh->log);
+
+	mempool_exit(&rh->region_pool);
+	vfree(rh->buckets);
+	kfree(rh);
+}
+EXPORT_SYMBOL_GPL(dm_region_hash_destroy);
+
+struct dm_dirty_log *dm_rh_dirty_log(struct dm_region_hash *rh)
+{
+	return rh->log;
+}
+EXPORT_SYMBOL_GPL(dm_rh_dirty_log);
+
+static unsigned rh_hash(struct dm_region_hash *rh, region_t region)
+{
+	return (unsigned) ((region * rh->prime) >> rh->shift) & rh->mask;
+}
+
+static struct dm_region *__rh_lookup(struct dm_region_hash *rh, region_t region)
+{
+	struct dm_region *reg;
+	struct list_head *bucket = rh->buckets + rh_hash(rh, region);
+
+	list_for_each_entry(reg, bucket, hash_list)
+		if (reg->key == region)
+			return reg;
+
+	return NULL;
+}
+
+static void __rh_insert(struct dm_region_hash *rh, struct dm_region *reg)
+{
+	list_add(&reg->hash_list, rh->buckets + rh_hash(rh, reg->key));
+}
+
+static struct dm_region *__rh_alloc(struct dm_region_hash *rh, region_t region)
+{
+	struct dm_region *reg, *nreg;
+
+	nreg = mempool_alloc(&rh->region_pool, GFP_ATOMIC);
+	if (unlikely(!nreg))
+		nreg = kmalloc(sizeof(*nreg), GFP_NOIO | __GFP_NOFAIL);
+
+	nreg->state = rh->log->type->in_sync(rh->log, region, 1) ?
+		      DM_RH_CLEAN : DM_RH_NOSYNC;
+	nreg->rh = rh;
+	nreg->key = region;
+	INIT_LIST_HEAD(&nreg->list);
+	atomic_set(&nreg->pending, 0);
+	bio_list_init(&nreg->delayed_bios);
+
+	write_lock_irq(&rh->hash_lock);
+	reg = __rh_lookup(rh, region);
+	if (reg)
+		/* We lost the race. */
+		mempool_free(nreg, &rh->region_pool);
+	else {
+		__rh_insert(rh, nreg);
+		if (nreg->state == DM_RH_CLEAN) {
+			spin_lock(&rh->region_lock);
+			list_add(&nreg->list, &rh->clean_regions);
+			spin_unlock(&rh->region_lock);
+		}
+
+		reg = nreg;
+	}
+	write_unlock_irq(&rh->hash_lock);
+
+	return reg;
+}
+
+static struct dm_region *__rh_find(struct dm_region_hash *rh, region_t region)
+{
+	struct dm_region *reg;
+
+	reg = __rh_lookup(rh, region);
+	if (!reg) {
+		read_unlock(&rh->hash_lock);
+		reg = __rh_alloc(rh, region);
+		read_lock(&rh->hash_lock);
+	}
+
+	return reg;
+}
+
+int dm_rh_get_state(struct dm_region_hash *rh, region_t region, int may_block)
+{
+	int r;
+	struct dm_region *reg;
+
+	read_lock(&rh->hash_lock);
+	reg = __rh_lookup(rh, region);
+	read_unlock(&rh->hash_lock);
+
+	if (reg)
+		return reg->state;
+
+	/*
+	 * The region wasn't in the hash, so we fall back to the
+	 * dirty log.
+	 */
+	r = rh->log->type->in_sync(rh->log, region, may_block);
+
+	/*
+	 * Any error from the dirty log (eg. -EWOULDBLOCK) gets
+	 * taken as a DM_RH_NOSYNC
+	 */
+	return r == 1 ? DM_RH_CLEAN : DM_RH_NOSYNC;
+}
+EXPORT_SYMBOL_GPL(dm_rh_get_state);
+
+static void complete_resync_work(struct dm_region *reg, int success)
+{
+	struct dm_region_hash *rh = reg->rh;
+
+	rh->log->type->set_region_sync(rh->log, reg->key, success);
+
+	/*
+	 * Dispatch the bios before we call 'wake_up_all'.
+	 * This is important because if we are suspending,
+	 * we want to know that recovery is complete and
+	 * the work queue is flushed.  If we wake_up_all
+	 * before we dispatch_bios (queue bios and call wake()),
+	 * then we risk suspending before the work queue
+	 * has been properly flushed.
+	 */
+	rh->dispatch_bios(rh->context, &reg->delayed_bios);
+	if (atomic_dec_and_test(&rh->recovery_in_flight))
+		rh->wakeup_all_recovery_waiters(rh->context);
+	up(&rh->recovery_count);
+}
+
+/* dm_rh_mark_nosync
+ * @ms
+ * @bio
+ *
+ * The bio was written on some mirror(s) but failed on other mirror(s).
+ * We can successfully endio the bio but should avoid the region being
+ * marked clean by setting the state DM_RH_NOSYNC.
+ *
+ * This function is _not_ safe in interrupt context!
+ */
+void dm_rh_mark_nosync(struct dm_region_hash *rh, struct bio *bio)
+{
+	unsigned long flags;
+	struct dm_dirty_log *log = rh->log;
+	struct dm_region *reg;
+	region_t region = dm_rh_bio_to_region(rh, bio);
+	int recovering = 0;
+
+	if (bio->bi_opf & REQ_PREFLUSH) {
+		rh->flush_failure = 1;
+		return;
+	}
+
+	if (bio_op(bio) == REQ_OP_DISCARD)
+		return;
+
+	/* We must inform the log that the sync count has changed. */
+	log->type->set_region_sync(log, region, 0);
+
+	read_lock(&rh->hash_lock);
+	reg = __rh_find(rh, region);
+	read_unlock(&rh->hash_lock);
+
+	/* region hash entry should exist because write was in-flight */
+	BUG_ON(!reg);
+	BUG_ON(!list_empty(&reg->list));
+
+	spin_lock_irqsave(&rh->region_lock, flags);
+	/*
+	 * Possible cases:
+	 *   1) DM_RH_DIRTY
+	 *   2) DM_RH_NOSYNC: was dirty, other preceding writes failed
+	 *   3) DM_RH_RECOVERING: flushing pending writes
+	 * Either case, the region should have not been connected to list.
+	 */
+	recovering = (reg->state == DM_RH_RECOVERING);
+	reg->state = DM_RH_NOSYNC;
+	BUG_ON(!list_empty(&reg->list));
+	spin_unlock_irqrestore(&rh->region_lock, flags);
+
+	if (recovering)
+		complete_resync_work(reg, 0);
+}
+EXPORT_SYMBOL_GPL(dm_rh_mark_nosync);
+
+void dm_rh_update_states(struct dm_region_hash *rh, int errors_handled)
+{
+	struct dm_region *reg, *next;
+
+	LIST_HEAD(clean);
+	LIST_HEAD(recovered);
+	LIST_HEAD(failed_recovered);
+
+	/*
+	 * Quickly grab the lists.
+	 */
+	write_lock_irq(&rh->hash_lock);
+	spin_lock(&rh->region_lock);
+	if (!list_empty(&rh->clean_regions)) {
+		list_splice_init(&rh->clean_regions, &clean);
+
+		list_for_each_entry(reg, &clean, list)
+			list_del(&reg->hash_list);
+	}
+
+	if (!list_empty(&rh->recovered_regions)) {
+		list_splice_init(&rh->recovered_regions, &recovered);
+
+		list_for_each_entry(reg, &recovered, list)
+			list_del(&reg->hash_list);
+	}
+
+	if (!list_empty(&rh->failed_recovered_regions)) {
+		list_splice_init(&rh->failed_recovered_regions,
+				 &failed_recovered);
+
+		list_for_each_entry(reg, &failed_recovered, list)
+			list_del(&reg->hash_list);
+	}
+
+	spin_unlock(&rh->region_lock);
+	write_unlock_irq(&rh->hash_lock);
+
+	/*
+	 * All the regions on the recovered and clean lists have
+	 * now been pulled out of the system, so no need to do
+	 * any more locking.
+	 */
+	list_for_each_entry_safe(reg, next, &recovered, list) {
+		rh->log->type->clear_region(rh->log, reg->key);
+		complete_resync_work(reg, 1);
+		mempool_free(reg, &rh->region_pool);
+	}
+
+	list_for_each_entry_safe(reg, next, &failed_recovered, list) {
+		complete_resync_work(reg, errors_handled ? 0 : 1);
+		mempool_free(reg, &rh->region_pool);
+	}
+
+	list_for_each_entry_safe(reg, next, &clean, list) {
+		rh->log->type->clear_region(rh->log, reg->key);
+		mempool_free(reg, &rh->region_pool);
+	}
+
+	rh->log->type->flush(rh->log);
+}
+EXPORT_SYMBOL_GPL(dm_rh_update_states);
+
+static void rh_inc(struct dm_region_hash *rh, region_t region)
+{
+	struct dm_region *reg;
+
+	read_lock(&rh->hash_lock);
+	reg = __rh_find(rh, region);
+
+	spin_lock_irq(&rh->region_lock);
+	atomic_inc(&reg->pending);
+
+	if (reg->state == DM_RH_CLEAN) {
+		reg->state = DM_RH_DIRTY;
+		list_del_init(&reg->list);	/* take off the clean list */
+		spin_unlock_irq(&rh->region_lock);
+
+		rh->log->type->mark_region(rh->log, reg->key);
+	} else
+		spin_unlock_irq(&rh->region_lock);
+
+
+	read_unlock(&rh->hash_lock);
+}
+
+void dm_rh_inc_pending(struct dm_region_hash *rh, struct bio_list *bios)
+{
+	struct bio *bio;
+
+	for (bio = bios->head; bio; bio = bio->bi_next) {
+		if (bio->bi_opf & REQ_PREFLUSH || bio_op(bio) == REQ_OP_DISCARD)
+			continue;
+		rh_inc(rh, dm_rh_bio_to_region(rh, bio));
+	}
+}
+EXPORT_SYMBOL_GPL(dm_rh_inc_pending);
+
+void dm_rh_dec(struct dm_region_hash *rh, region_t region)
+{
+	unsigned long flags;
+	struct dm_region *reg;
+	int should_wake = 0;
+
+	read_lock(&rh->hash_lock);
+	reg = __rh_lookup(rh, region);
+	read_unlock(&rh->hash_lock);
+
+	spin_lock_irqsave(&rh->region_lock, flags);
+	if (atomic_dec_and_test(&reg->pending)) {
+		/*
+		 * There is no pending I/O for this region.
+		 * We can move the region to corresponding list for next action.
+		 * At this point, the region is not yet connected to any list.
+		 *
+		 * If the state is DM_RH_NOSYNC, the region should be kept off
+		 * from clean list.
+		 * The hash entry for DM_RH_NOSYNC will remain in memory
+		 * until the region is recovered or the map is reloaded.
+		 */
+
+		/* do nothing for DM_RH_NOSYNC */
+		if (unlikely(rh->flush_failure)) {
+			/*
+			 * If a write flush failed some time ago, we
+			 * don't know whether or not this write made it
+			 * to the disk, so we must resync the device.
+			 */
+			reg->state = DM_RH_NOSYNC;
+		} else if (reg->state == DM_RH_RECOVERING) {
+			list_add_tail(&reg->list, &rh->quiesced_regions);
+		} else if (reg->state == DM_RH_DIRTY) {
+			reg->state = DM_RH_CLEAN;
+			list_add(&reg->list, &rh->clean_regions);
+		}
+		should_wake = 1;
+	}
+	spin_unlock_irqrestore(&rh->region_lock, flags);
+
+	if (should_wake)
+		rh->wakeup_workers(rh->context);
+}
+EXPORT_SYMBOL_GPL(dm_rh_dec);
+
+/*
+ * Starts quiescing a region in preparation for recovery.
+ */
+static int __rh_recovery_prepare(struct dm_region_hash *rh)
+{
+	int r;
+	region_t region;
+	struct dm_region *reg;
+
+	/*
+	 * Ask the dirty log what's next.
+	 */
+	r = rh->log->type->get_resync_work(rh->log, &region);
+	if (r <= 0)
+		return r;
+
+	/*
+	 * Get this region, and start it quiescing by setting the
+	 * recovering flag.
+	 */
+	read_lock(&rh->hash_lock);
+	reg = __rh_find(rh, region);
+	read_unlock(&rh->hash_lock);
+
+	spin_lock_irq(&rh->region_lock);
+	reg->state = DM_RH_RECOVERING;
+
+	/* Already quiesced ? */
+	if (atomic_read(&reg->pending))
+		list_del_init(&reg->list);
+	else
+		list_move(&reg->list, &rh->quiesced_regions);
+
+	spin_unlock_irq(&rh->region_lock);
+
+	return 1;
+}
+
+void dm_rh_recovery_prepare(struct dm_region_hash *rh)
+{
+	/* Extra reference to avoid race with dm_rh_stop_recovery */
+	atomic_inc(&rh->recovery_in_flight);
+
+	while (!down_trylock(&rh->recovery_count)) {
+		atomic_inc(&rh->recovery_in_flight);
+		if (__rh_recovery_prepare(rh) <= 0) {
+			atomic_dec(&rh->recovery_in_flight);
+			up(&rh->recovery_count);
+			break;
+		}
+	}
+
+	/* Drop the extra reference */
+	if (atomic_dec_and_test(&rh->recovery_in_flight))
+		rh->wakeup_all_recovery_waiters(rh->context);
+}
+EXPORT_SYMBOL_GPL(dm_rh_recovery_prepare);
+
+/*
+ * Returns any quiesced regions.
+ */
+struct dm_region *dm_rh_recovery_start(struct dm_region_hash *rh)
+{
+	struct dm_region *reg = NULL;
+
+	spin_lock_irq(&rh->region_lock);
+	if (!list_empty(&rh->quiesced_regions)) {
+		reg = list_entry(rh->quiesced_regions.next,
+				 struct dm_region, list);
+		list_del_init(&reg->list);  /* remove from the quiesced list */
+	}
+	spin_unlock_irq(&rh->region_lock);
+
+	return reg;
+}
+EXPORT_SYMBOL_GPL(dm_rh_recovery_start);
+
+void dm_rh_recovery_end(struct dm_region *reg, int success)
+{
+	struct dm_region_hash *rh = reg->rh;
+
+	spin_lock_irq(&rh->region_lock);
+	if (success)
+		list_add(&reg->list, &reg->rh->recovered_regions);
+	else
+		list_add(&reg->list, &reg->rh->failed_recovered_regions);
+
+	spin_unlock_irq(&rh->region_lock);
+
+	rh->wakeup_workers(rh->context);
+}
+EXPORT_SYMBOL_GPL(dm_rh_recovery_end);
+
+/* Return recovery in flight count. */
+int dm_rh_recovery_in_flight(struct dm_region_hash *rh)
+{
+	return atomic_read(&rh->recovery_in_flight);
+}
+EXPORT_SYMBOL_GPL(dm_rh_recovery_in_flight);
+
+int dm_rh_flush(struct dm_region_hash *rh)
+{
+	return rh->log->type->flush(rh->log);
+}
+EXPORT_SYMBOL_GPL(dm_rh_flush);
+
+void dm_rh_delay(struct dm_region_hash *rh, struct bio *bio)
+{
+	struct dm_region *reg;
+
+	read_lock(&rh->hash_lock);
+	reg = __rh_find(rh, dm_rh_bio_to_region(rh, bio));
+	bio_list_add(&reg->delayed_bios, bio);
+	read_unlock(&rh->hash_lock);
+}
+EXPORT_SYMBOL_GPL(dm_rh_delay);
+
+void dm_rh_stop_recovery(struct dm_region_hash *rh)
+{
+	int i;
+
+	/* wait for any recovering regions */
+	for (i = 0; i < rh->max_recovery; i++)
+		down(&rh->recovery_count);
+}
+EXPORT_SYMBOL_GPL(dm_rh_stop_recovery);
+
+void dm_rh_start_recovery(struct dm_region_hash *rh)
+{
+	int i;
+
+	for (i = 0; i < rh->max_recovery; i++)
+		up(&rh->recovery_count);
+
+	rh->wakeup_workers(rh->context);
+}
+EXPORT_SYMBOL_GPL(dm_rh_start_recovery);
+
+MODULE_DESCRIPTION(DM_NAME " region hash");
+MODULE_AUTHOR("Joe Thornber/Heinz Mauelshagen <dm-devel@redhat.com>");
+MODULE_LICENSE("GPL");