1 files changed, 162 insertions, 80 deletions

diff --git a/drivers/md/dm-snap-persistent.c b/drivers/md/dm-snap-persistent.c
index 3ac415675b6c..568d10842b1f 100644
--- a/drivers/md/dm-snap-persistent.c
+++ b/drivers/md/dm-snap-persistent.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0-only
 /*
  * Copyright (C) 2001-2002 Sistina Software (UK) Limited.
  * Copyright (C) 2006-2008 Red Hat GmbH
@@ -7,20 +8,26 @@
 
 #include "dm-exception-store.h"
 
+#include <linux/ctype.h>
 #include <linux/mm.h>
 #include <linux/pagemap.h>
 #include <linux/vmalloc.h>
 #include <linux/export.h>
 #include <linux/slab.h>
 #include <linux/dm-io.h>
+#include <linux/dm-bufio.h>
 
 #define DM_MSG_PREFIX "persistent snapshot"
-#define DM_CHUNK_SIZE_DEFAULT_SECTORS 32	/* 16KB */
+#define DM_CHUNK_SIZE_DEFAULT_SECTORS 32U	/* 16KB */
 
-/*-----------------------------------------------------------------
+#define DM_PREFETCH_CHUNKS		12
+
+/*
+ *---------------------------------------------------------------
  * Persistent snapshots, by persistent we mean that the snapshot
  * will survive a reboot.
- *---------------------------------------------------------------*/
+ *---------------------------------------------------------------
+ */
 
 /*
  * We need to store a record of which parts of the origin have
@@ -88,7 +95,7 @@ struct core_exception {
 };
 
 struct commit_callback {
-	void (*callback)(void *, int success);
+	void (*callback)(void *ref, int success);
 	void *context;
 };
 
@@ -197,16 +204,11 @@ err_area:
 
 static void free_area(struct pstore *ps)
 {
-	if (ps->area)
-		vfree(ps->area);
+	vfree(ps->area);
 	ps->area = NULL;
-
-	if (ps->zero_area)
-		vfree(ps->zero_area);
+	vfree(ps->zero_area);
 	ps->zero_area = NULL;
-
-	if (ps->header_area)
-		vfree(ps->header_area);
+	vfree(ps->header_area);
 	ps->header_area = NULL;
 }
 
@@ -221,13 +223,13 @@ static void do_metadata(struct work_struct *work)
 {
 	struct mdata_req *req = container_of(work, struct mdata_req, work);
 
-	req->result = dm_io(req->io_req, 1, req->where, NULL);
+	req->result = dm_io(req->io_req, 1, req->where, NULL, IOPRIO_DEFAULT);
 }
 
 /*
  * Read or write a chunk aligned and sized block of data from a device.
  */
-static int chunk_io(struct pstore *ps, void *area, chunk_t chunk, int rw,
+static int chunk_io(struct pstore *ps, void *area, chunk_t chunk, blk_opf_t opf,
 		    int metadata)
 {
 	struct dm_io_region where = {
@@ -236,7 +238,7 @@ static int chunk_io(struct pstore *ps, void *area, chunk_t chunk, int rw,
 		.count = ps->store->chunk_size,
 	};
 	struct dm_io_request io_req = {
-		.bi_rw = rw,
+		.bi_opf = opf,
 		.mem.type = DM_IO_VMA,
 		.mem.ptr.vma = area,
 		.client = ps->io_client,
@@ -245,18 +247,19 @@ static int chunk_io(struct pstore *ps, void *area, chunk_t chunk, int rw,
 	struct mdata_req req;
 
 	if (!metadata)
-		return dm_io(&io_req, 1, &where, NULL);
+		return dm_io(&io_req, 1, &where, NULL, IOPRIO_DEFAULT);
 
 	req.where = &where;
 	req.io_req = &io_req;
 
 	/*
 	 * Issue the synchronous I/O from a different thread
-	 * to avoid generic_make_request recursion.
+	 * to avoid submit_bio_noacct recursion.
 	 */
 	INIT_WORK_ONSTACK(&req.work, do_metadata);
 	queue_work(ps->metadata_wq, &req.work);
-	flush_work(&req.work);
+	flush_workqueue(ps->metadata_wq);
+	destroy_work_on_stack(&req.work);
 
 	return req.result;
 }
@@ -269,22 +272,24 @@ static chunk_t area_location(struct pstore *ps, chunk_t area)
 	return NUM_SNAPSHOT_HDR_CHUNKS + ((ps->exceptions_per_area + 1) * area);
 }
 
+static void skip_metadata(struct pstore *ps)
+{
+	uint32_t stride = ps->exceptions_per_area + 1;
+	chunk_t next_free = ps->next_free;
+
+	if (sector_div(next_free, stride) == NUM_SNAPSHOT_HDR_CHUNKS)
+		ps->next_free++;
+}
+
 /*
  * Read or write a metadata area.  Remembering to skip the first
  * chunk which holds the header.
  */
-static int area_io(struct pstore *ps, int rw)
+static int area_io(struct pstore *ps, blk_opf_t opf)
 {
-	int r;
-	chunk_t chunk;
-
-	chunk = area_location(ps, ps->current_area);
+	chunk_t chunk = area_location(ps, ps->current_area);
 
-	r = chunk_io(ps, ps->area, chunk, rw, 0);
-	if (r)
-		return r;
-
-	return 0;
+	return chunk_io(ps, ps->area, chunk, opf, 0);
 }
 
 static void zero_memory_area(struct pstore *ps)
@@ -294,14 +299,15 @@ static void zero_memory_area(struct pstore *ps)
 
 static int zero_disk_area(struct pstore *ps, chunk_t area)
 {
-	return chunk_io(ps, ps->zero_area, area_location(ps, area), WRITE, 0);
+	return chunk_io(ps, ps->zero_area, area_location(ps, area),
+			REQ_OP_WRITE, 0);
 }
 
 static int read_header(struct pstore *ps, int *new_snapshot)
 {
 	int r;
 	struct disk_header *dh;
-	unsigned chunk_size;
+	unsigned int chunk_size;
 	int chunk_size_supplied = 1;
 	char *chunk_err;
 
@@ -314,7 +320,7 @@ static int read_header(struct pstore *ps, int *new_snapshot)
 		    bdev_logical_block_size(dm_snap_cow(ps->store->snap)->
 					    bdev) >> 9);
 		ps->store->chunk_mask = ps->store->chunk_size - 1;
-		ps->store->chunk_shift = ffs(ps->store->chunk_size) - 1;
+		ps->store->chunk_shift = __ffs(ps->store->chunk_size);
 		chunk_size_supplied = 0;
 	}
 
@@ -326,7 +332,7 @@ static int read_header(struct pstore *ps, int *new_snapshot)
 	if (r)
 		return r;
 
-	r = chunk_io(ps, ps->header_area, 0, READ, 1);
+	r = chunk_io(ps, ps->header_area, 0, REQ_OP_READ, 1);
 	if (r)
 		goto bad;
 
@@ -352,8 +358,7 @@ static int read_header(struct pstore *ps, int *new_snapshot)
 		return 0;
 
 	if (chunk_size_supplied)
-		DMWARN("chunk size %u in device metadata overrides "
-		       "table chunk size of %u.",
+		DMWARN("chunk size %u in device metadata overrides table chunk size of %u.",
 		       chunk_size, ps->store->chunk_size);
 
 	/* We had a bogus chunk_size. Fix stuff up. */
@@ -387,23 +392,24 @@ static int write_header(struct pstore *ps)
 	dh->version = cpu_to_le32(ps->version);
 	dh->chunk_size = cpu_to_le32(ps->store->chunk_size);
 
-	return chunk_io(ps, ps->header_area, 0, WRITE, 1);
+	return chunk_io(ps, ps->header_area, 0, REQ_OP_WRITE, 1);
 }
 
 /*
  * Access functions for the disk exceptions, these do the endian conversions.
  */
-static struct disk_exception *get_exception(struct pstore *ps, uint32_t index)
+static struct disk_exception *get_exception(struct pstore *ps, void *ps_area,
+					    uint32_t index)
 {
 	BUG_ON(index >= ps->exceptions_per_area);
 
-	return ((struct disk_exception *) ps->area) + index;
+	return ((struct disk_exception *) ps_area) + index;
 }
 
-static void read_exception(struct pstore *ps,
+static void read_exception(struct pstore *ps, void *ps_area,
 			   uint32_t index, struct core_exception *result)
 {
-	struct disk_exception *de = get_exception(ps, index);
+	struct disk_exception *de = get_exception(ps, ps_area, index);
 
 	/* copy it */
 	result->old_chunk = le64_to_cpu(de->old_chunk);
@@ -413,7 +419,7 @@ static void read_exception(struct pstore *ps,
 static void write_exception(struct pstore *ps,
 			    uint32_t index, struct core_exception *e)
 {
-	struct disk_exception *de = get_exception(ps, index);
+	struct disk_exception *de = get_exception(ps, ps->area, index);
 
 	/* copy it */
 	de->old_chunk = cpu_to_le64(e->old_chunk);
@@ -422,7 +428,7 @@ static void write_exception(struct pstore *ps,
 
 static void clear_exception(struct pstore *ps, uint32_t index)
 {
-	struct disk_exception *de = get_exception(ps, index);
+	struct disk_exception *de = get_exception(ps, ps->area, index);
 
 	/* clear it */
 	de->old_chunk = 0;
@@ -434,7 +440,7 @@ static void clear_exception(struct pstore *ps, uint32_t index)
  * 'full' is filled in to indicate if the area has been
  * filled.
  */
-static int insert_exceptions(struct pstore *ps,
+static int insert_exceptions(struct pstore *ps, void *ps_area,
 			     int (*callback)(void *callback_context,
 					     chunk_t old, chunk_t new),
 			     void *callback_context,
@@ -448,7 +454,7 @@ static int insert_exceptions(struct pstore *ps,
 	*full = 1;
 
 	for (i = 0; i < ps->exceptions_per_area; i++) {
-		read_exception(ps, i, &e);
+		read_exception(ps, ps_area, i, &e);
 
 		/*
 		 * If the new_chunk is pointing at the start of
@@ -485,29 +491,83 @@ static int read_exceptions(struct pstore *ps,
 			   void *callback_context)
 {
 	int r, full = 1;
+	struct dm_bufio_client *client;
+	chunk_t prefetch_area = 0;
+
+	client = dm_bufio_client_create(dm_snap_cow(ps->store->snap)->bdev,
+					ps->store->chunk_size << SECTOR_SHIFT,
+					1, 0, NULL, NULL, 0);
+
+	if (IS_ERR(client))
+		return PTR_ERR(client);
+
+	/*
+	 * Setup for one current buffer + desired readahead buffers.
+	 */
+	dm_bufio_set_minimum_buffers(client, 1 + DM_PREFETCH_CHUNKS);
 
 	/*
 	 * Keeping reading chunks and inserting exceptions until
 	 * we find a partially full area.
 	 */
 	for (ps->current_area = 0; full; ps->current_area++) {
-		r = area_io(ps, READ);
-		if (r)
-			return r;
+		struct dm_buffer *bp;
+		void *area;
+		chunk_t chunk;
+
+		if (unlikely(prefetch_area < ps->current_area))
+			prefetch_area = ps->current_area;
+
+		if (DM_PREFETCH_CHUNKS) {
+			do {
+				chunk_t pf_chunk = area_location(ps, prefetch_area);
+
+				if (unlikely(pf_chunk >= dm_bufio_get_device_size(client)))
+					break;
+				dm_bufio_prefetch(client, pf_chunk, 1);
+				prefetch_area++;
+				if (unlikely(!prefetch_area))
+					break;
+			} while (prefetch_area <= ps->current_area + DM_PREFETCH_CHUNKS);
+		}
 
-		r = insert_exceptions(ps, callback, callback_context, &full);
-		if (r)
-			return r;
+		chunk = area_location(ps, ps->current_area);
+
+		area = dm_bufio_read(client, chunk, &bp);
+		if (IS_ERR(area)) {
+			r = PTR_ERR(area);
+			goto ret_destroy_bufio;
+		}
+
+		r = insert_exceptions(ps, area, callback, callback_context,
+				      &full);
+
+		if (!full)
+			memcpy(ps->area, area, ps->store->chunk_size << SECTOR_SHIFT);
+
+		dm_bufio_release(bp);
+
+		dm_bufio_forget(client, chunk);
+
+		if (unlikely(r))
+			goto ret_destroy_bufio;
 	}
 
 	ps->current_area--;
 
-	return 0;
+	skip_metadata(ps);
+
+	r = 0;
+
+ret_destroy_bufio:
+	dm_bufio_client_destroy(client);
+
+	return r;
 }
 
 static struct pstore *get_info(struct dm_exception_store *store)
 {
-	return (struct pstore *) store->context;
+	return store->context;
 }
 
 static void persistent_usage(struct dm_exception_store *store,
@@ -541,8 +601,7 @@ static void persistent_dtr(struct dm_exception_store *store)
 	free_area(ps);
 
 	/* Allocated in persistent_read_metadata */
-	if (ps->callbacks)
-		vfree(ps->callbacks);
+	kvfree(ps->callbacks);
 
 	kfree(ps);
 }
@@ -552,7 +611,7 @@ static int persistent_read_metadata(struct dm_exception_store *store,
 						    chunk_t old, chunk_t new),
 				    void *callback_context)
 {
-	int r, uninitialized_var(new_snapshot);
+	int r, new_snapshot;
 	struct pstore *ps = get_info(store);
 
 	/*
@@ -567,8 +626,8 @@ static int persistent_read_metadata(struct dm_exception_store *store,
 	 */
 	ps->exceptions_per_area = (ps->store->chunk_size << SECTOR_SHIFT) /
 				  sizeof(struct disk_exception);
-	ps->callbacks = dm_vcalloc(ps->exceptions_per_area,
-				   sizeof(*ps->callbacks));
+	ps->callbacks = kvcalloc(ps->exceptions_per_area,
+				 sizeof(*ps->callbacks), GFP_KERNEL);
 	if (!ps->callbacks)
 		return -ENOMEM;
 
@@ -616,8 +675,6 @@ static int persistent_prepare_exception(struct dm_exception_store *store,
 					struct dm_exception *e)
 {
 	struct pstore *ps = get_info(store);
-	uint32_t stride;
-	chunk_t next_free;
 	sector_t size = get_dev_size(dm_snap_cow(store->snap)->bdev);
 
 	/* Is there enough room ? */
@@ -630,18 +687,16 @@ static int persistent_prepare_exception(struct dm_exception_store *store,
 	 * Move onto the next free pending, making sure to take
 	 * into account the location of the metadata chunks.
 	 */
-	stride = (ps->exceptions_per_area + 1);
-	next_free = ++ps->next_free;
-	if (sector_div(next_free, stride) == 1)
-		ps->next_free++;
+	ps->next_free++;
+	skip_metadata(ps);
 
 	atomic_inc(&ps->pending_count);
 	return 0;
 }
 
 static void persistent_commit_exception(struct dm_exception_store *store,
-					struct dm_exception *e,
-					void (*callback) (void *, int success),
+					struct dm_exception *e, int valid,
+					void (*callback)(void *, int success),
 					void *callback_context)
 {
 	unsigned int i;
@@ -649,6 +704,9 @@ static void persistent_commit_exception(struct dm_exception_store *store,
 	struct core_exception ce;
 	struct commit_callback *cb;
 
+	if (!valid)
+		ps->valid = 0;
+
 	ce.old_chunk = e->old_chunk;
 	ce.new_chunk = e->new_chunk;
 	write_exception(ps, ps->current_committed++, &ce);
@@ -681,7 +739,8 @@ static void persistent_commit_exception(struct dm_exception_store *store,
 	/*
 	 * Commit exceptions to disk.
 	 */
-	if (ps->valid && area_io(ps, WRITE_FLUSH_FUA))
+	if (ps->valid && area_io(ps, REQ_OP_WRITE | REQ_PREFLUSH | REQ_FUA |
+				 REQ_SYNC))
 		ps->valid = 0;
 
 	/*
@@ -721,13 +780,13 @@ static int persistent_prepare_merge(struct dm_exception_store *store,
 			return 0;
 
 		ps->current_area--;
-		r = area_io(ps, READ);
+		r = area_io(ps, REQ_OP_READ);
 		if (r < 0)
 			return r;
 		ps->current_committed = ps->exceptions_per_area;
 	}
 
-	read_exception(ps, ps->current_committed - 1, &ce);
+	read_exception(ps, ps->area, ps->current_committed - 1, &ce);
 	*last_old_chunk = ce.old_chunk;
 	*last_new_chunk = ce.new_chunk;
 
@@ -737,8 +796,8 @@ static int persistent_prepare_merge(struct dm_exception_store *store,
 	 */
 	for (nr_consecutive = 1; nr_consecutive < ps->current_committed;
 	     nr_consecutive++) {
-		read_exception(ps, ps->current_committed - 1 - nr_consecutive,
-			       &ce);
+		read_exception(ps, ps->area,
+			       ps->current_committed - 1 - nr_consecutive, &ce);
 		if (ce.old_chunk != *last_old_chunk - nr_consecutive ||
 		    ce.new_chunk != *last_new_chunk - nr_consecutive)
 			break;
@@ -758,7 +817,7 @@ static int persistent_commit_merge(struct dm_exception_store *store,
 	for (i = 0; i < nr_merged; i++)
 		clear_exception(ps, ps->current_committed - 1 - i);
 
-	r = area_io(ps, WRITE_FLUSH_FUA);
+	r = area_io(ps, REQ_OP_WRITE | REQ_PREFLUSH | REQ_FUA);
 	if (r < 0)
 		return r;
 
@@ -789,10 +848,10 @@ static void persistent_drop_snapshot(struct dm_exception_store *store)
 		DMWARN("write header failed");
 }
 
-static int persistent_ctr(struct dm_exception_store *store,
-			  unsigned argc, char **argv)
+static int persistent_ctr(struct dm_exception_store *store, char *options)
 {
 	struct pstore *ps;
+	int r;
 
 	/* allocate the pstore */
 	ps = kzalloc(sizeof(*ps), GFP_KERNEL);
@@ -814,27 +873,51 @@ static int persistent_ctr(struct dm_exception_store *store,
 
 	ps->metadata_wq = alloc_workqueue("ksnaphd", WQ_MEM_RECLAIM, 0);
 	if (!ps->metadata_wq) {
-		kfree(ps);
 		DMERR("couldn't start header metadata update thread");
-		return -ENOMEM;
+		r = -ENOMEM;
+		goto err_workqueue;
+	}
+
+	if (options) {
+		char overflow = toupper(options[0]);
+
+		if (overflow == 'O')
+			store->userspace_supports_overflow = true;
+		else {
+			DMERR("Unsupported persistent store option: %s", options);
+			r = -EINVAL;
+			goto err_options;
+		}
 	}
 
 	store->context = ps;
 
 	return 0;
+
+err_options:
+	destroy_workqueue(ps->metadata_wq);
+err_workqueue:
+	kfree(ps);
+
+	return r;
 }
 
-static unsigned persistent_status(struct dm_exception_store *store,
+static unsigned int persistent_status(struct dm_exception_store *store,
 				  status_type_t status, char *result,
-				  unsigned maxlen)
+				  unsigned int maxlen)
 {
-	unsigned sz = 0;
+	unsigned int sz = 0;
 
 	switch (status) {
 	case STATUSTYPE_INFO:
 		break;
 	case STATUSTYPE_TABLE:
-		DMEMIT(" P %llu", (unsigned long long)store->chunk_size);
+		DMEMIT(" %s %llu", store->userspace_supports_overflow ? "PO" : "P",
+		       (unsigned long long)store->chunk_size);
+		break;
+	case STATUSTYPE_IMA:
+		*result = '\0';
+		break;
 	}
 
 	return sz;
@@ -882,8 +965,7 @@ int dm_persistent_snapshot_init(void)
 
 	r = dm_exception_store_type_register(&_persistent_compat_type);
 	if (r) {
-		DMERR("Unable to register old-style persistent exception "
-		      "store type");
+		DMERR("Unable to register old-style persistent exception store type");
 		dm_exception_store_type_unregister(&_persistent_type);
 		return r;
 	}