/*
- * dm-snapshot.c
+ * dm-exception-store.c
*
* Copyright (C) 2001-2002 Sistina Software (UK) Limited.
+ * Copyright (C) 2006 Red Hat GmbH
*
* This file is released under the GPL.
*/
#include "dm.h"
#include "dm-snap.h"
-#include "dm-io.h"
-#include "kcopyd.h"
#include <linux/mm.h>
#include <linux/pagemap.h>
#include <linux/vmalloc.h>
#include <linux/slab.h>
+#include <linux/dm-io.h>
+#include <linux/dm-kcopyd.h>
#define DM_MSG_PREFIX "snapshots"
+#define DM_CHUNK_SIZE_DEFAULT_SECTORS 32 /* 16KB */
/*-----------------------------------------------------------------
* Persistent snapshots, by persistent we mean that the snapshot
atomic_t pending_count;
uint32_t callback_count;
struct commit_callback *callbacks;
+ struct dm_io_client *io_client;
+
+ struct workqueue_struct *metadata_wq;
};
-static inline unsigned int sectors_to_pages(unsigned int sectors)
+static unsigned sectors_to_pages(unsigned sectors)
{
- return sectors / (PAGE_SIZE >> 9);
+ return DIV_ROUND_UP(sectors, PAGE_SIZE >> 9);
}
static int alloc_area(struct pstore *ps)
static void free_area(struct pstore *ps)
{
vfree(ps->area);
+ ps->area = NULL;
+}
+
+struct mdata_req {
+ struct dm_io_region *where;
+ struct dm_io_request *io_req;
+ struct work_struct work;
+ int result;
+};
+
+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);
}
/*
* Read or write a chunk aligned and sized block of data from a device.
*/
-static int chunk_io(struct pstore *ps, uint32_t chunk, int rw)
+static int chunk_io(struct pstore *ps, uint32_t chunk, int rw, int metadata)
{
- struct io_region where;
- unsigned long bits;
+ struct dm_io_region where = {
+ .bdev = ps->snap->cow->bdev,
+ .sector = ps->snap->chunk_size * chunk,
+ .count = ps->snap->chunk_size,
+ };
+ struct dm_io_request io_req = {
+ .bi_rw = rw,
+ .mem.type = DM_IO_VMA,
+ .mem.ptr.vma = ps->area,
+ .client = ps->io_client,
+ .notify.fn = NULL,
+ };
+ struct mdata_req req;
+
+ if (!metadata)
+ return dm_io(&io_req, 1, &where, NULL);
+
+ req.where = &where;
+ req.io_req = &io_req;
- where.bdev = ps->snap->cow->bdev;
- where.sector = ps->snap->chunk_size * chunk;
- where.count = ps->snap->chunk_size;
+ /*
+ * Issue the synchronous I/O from a different thread
+ * to avoid generic_make_request recursion.
+ */
+ INIT_WORK(&req.work, do_metadata);
+ queue_work(ps->metadata_wq, &req.work);
+ flush_workqueue(ps->metadata_wq);
- return dm_io_sync_vm(1, &where, rw, ps->area, &bits);
+ return req.result;
}
/*
/* convert a metadata area index to a chunk index */
chunk = 1 + ((ps->exceptions_per_area + 1) * area);
- r = chunk_io(ps, chunk, rw);
+ r = chunk_io(ps, chunk, rw, 0);
if (r)
return r;
int r;
struct disk_header *dh;
chunk_t chunk_size;
+ int chunk_size_supplied = 1;
- r = chunk_io(ps, 0, READ);
+ /*
+ * Use default chunk size (or hardsect_size, if larger) if none supplied
+ */
+ if (!ps->snap->chunk_size) {
+ ps->snap->chunk_size = max(DM_CHUNK_SIZE_DEFAULT_SECTORS,
+ bdev_hardsect_size(ps->snap->cow->bdev) >> 9);
+ ps->snap->chunk_mask = ps->snap->chunk_size - 1;
+ ps->snap->chunk_shift = ffs(ps->snap->chunk_size) - 1;
+ chunk_size_supplied = 0;
+ }
+
+ ps->io_client = dm_io_client_create(sectors_to_pages(ps->snap->
+ chunk_size));
+ if (IS_ERR(ps->io_client))
+ return PTR_ERR(ps->io_client);
+
+ r = alloc_area(ps);
if (r)
return r;
+ r = chunk_io(ps, 0, READ, 1);
+ if (r)
+ goto bad;
+
dh = (struct disk_header *) ps->area;
if (le32_to_cpu(dh->magic) == 0) {
*new_snapshot = 1;
+ return 0;
+ }
- } else if (le32_to_cpu(dh->magic) == SNAP_MAGIC) {
- *new_snapshot = 0;
- ps->valid = le32_to_cpu(dh->valid);
- ps->version = le32_to_cpu(dh->version);
- chunk_size = le32_to_cpu(dh->chunk_size);
- if (ps->snap->chunk_size != chunk_size) {
- DMWARN("chunk size %llu in device metadata overrides "
- "table chunk size of %llu.",
- (unsigned long long)chunk_size,
- (unsigned long long)ps->snap->chunk_size);
-
- /* We had a bogus chunk_size. Fix stuff up. */
- dm_io_put(sectors_to_pages(ps->snap->chunk_size));
- free_area(ps);
-
- ps->snap->chunk_size = chunk_size;
- ps->snap->chunk_mask = chunk_size - 1;
- ps->snap->chunk_shift = ffs(chunk_size) - 1;
-
- r = alloc_area(ps);
- if (r)
- return r;
-
- r = dm_io_get(sectors_to_pages(chunk_size));
- if (r)
- return r;
- }
- } else {
- DMWARN("Invalid/corrupt snapshot");
+ if (le32_to_cpu(dh->magic) != SNAP_MAGIC) {
+ DMWARN("Invalid or corrupt snapshot");
r = -ENXIO;
+ goto bad;
}
+ *new_snapshot = 0;
+ ps->valid = le32_to_cpu(dh->valid);
+ ps->version = le32_to_cpu(dh->version);
+ chunk_size = le32_to_cpu(dh->chunk_size);
+
+ if (!chunk_size_supplied || ps->snap->chunk_size == chunk_size)
+ return 0;
+
+ DMWARN("chunk size %llu in device metadata overrides "
+ "table chunk size of %llu.",
+ (unsigned long long)chunk_size,
+ (unsigned long long)ps->snap->chunk_size);
+
+ /* We had a bogus chunk_size. Fix stuff up. */
+ free_area(ps);
+
+ ps->snap->chunk_size = chunk_size;
+ ps->snap->chunk_mask = chunk_size - 1;
+ ps->snap->chunk_shift = ffs(chunk_size) - 1;
+
+ r = dm_io_client_resize(sectors_to_pages(ps->snap->chunk_size),
+ ps->io_client);
+ if (r)
+ return r;
+
+ r = alloc_area(ps);
+ return r;
+
+bad:
+ free_area(ps);
return r;
}
dh->version = cpu_to_le32(ps->version);
dh->chunk_size = cpu_to_le32(ps->snap->chunk_size);
- return chunk_io(ps, 0, WRITE);
+ return chunk_io(ps, 0, WRITE, 1);
}
/*
*/
static struct disk_exception *get_exception(struct pstore *ps, uint32_t index)
{
- if (index >= ps->exceptions_per_area)
- return NULL;
+ BUG_ON(index >= ps->exceptions_per_area);
return ((struct disk_exception *) ps->area) + index;
}
-static int read_exception(struct pstore *ps,
- uint32_t index, struct disk_exception *result)
+static void read_exception(struct pstore *ps,
+ uint32_t index, struct disk_exception *result)
{
- struct disk_exception *e;
-
- e = get_exception(ps, index);
- if (!e)
- return -EINVAL;
+ struct disk_exception *e = get_exception(ps, index);
/* copy it */
result->old_chunk = le64_to_cpu(e->old_chunk);
result->new_chunk = le64_to_cpu(e->new_chunk);
-
- return 0;
}
-static int write_exception(struct pstore *ps,
- uint32_t index, struct disk_exception *de)
+static void write_exception(struct pstore *ps,
+ uint32_t index, struct disk_exception *de)
{
- struct disk_exception *e;
-
- e = get_exception(ps, index);
- if (!e)
- return -EINVAL;
+ struct disk_exception *e = get_exception(ps, index);
/* copy it */
e->old_chunk = cpu_to_le64(de->old_chunk);
e->new_chunk = cpu_to_le64(de->new_chunk);
-
- return 0;
}
/*
*full = 1;
for (i = 0; i < ps->exceptions_per_area; i++) {
- r = read_exception(ps, i, &de);
-
- if (r)
- return r;
+ read_exception(ps, i, &de);
/*
* If the new_chunk is pointing at the start of
return 0;
}
-static inline struct pstore *get_info(struct exception_store *store)
+static struct pstore *get_info(struct exception_store *store)
{
return (struct pstore *) store->context;
}
{
struct pstore *ps = get_info(store);
- dm_io_put(sectors_to_pages(ps->snap->chunk_size));
+ destroy_workqueue(ps->metadata_wq);
+ dm_io_client_destroy(ps->io_client);
vfree(ps->callbacks);
free_area(ps);
kfree(ps);
static int persistent_read_metadata(struct exception_store *store)
{
- int r, new_snapshot;
+ int r, uninitialized_var(new_snapshot);
struct pstore *ps = get_info(store);
/*
/*
* Sanity checks.
*/
- if (!ps->valid) {
- DMWARN("snapshot is marked invalid");
- return -EINVAL;
- }
-
if (ps->version != SNAPSHOT_DISK_VERSION) {
DMWARN("unable to handle snapshot disk version %d",
ps->version);
return -EINVAL;
}
+ /*
+ * Metadata are valid, but snapshot is invalidated
+ */
+ if (!ps->valid)
+ return 1;
+
/*
* Read the metadata.
*/
}
static int persistent_prepare(struct exception_store *store,
- struct exception *e)
+ struct dm_snap_exception *e)
{
struct pstore *ps = get_info(store);
uint32_t stride;
}
static void persistent_commit(struct exception_store *store,
- struct exception *e,
+ struct dm_snap_exception *e,
void (*callback) (void *, int success),
void *callback_context)
{
if (r)
ps->valid = 0;
+ /*
+ * Have we completely filled the current area ?
+ */
+ if (ps->current_committed == ps->exceptions_per_area) {
+ ps->current_committed = 0;
+ r = zero_area(ps, ps->current_area + 1);
+ if (r)
+ ps->valid = 0;
+ }
+
for (i = 0; i < ps->callback_count; i++) {
cb = ps->callbacks + i;
cb->callback(cb->context, r == 0 ? 1 : 0);
ps->callback_count = 0;
}
-
- /*
- * Have we completely filled the current area ?
- */
- if (ps->current_committed == ps->exceptions_per_area) {
- ps->current_committed = 0;
- r = zero_area(ps, ps->current_area + 1);
- if (r)
- ps->valid = 0;
- }
}
static void persistent_drop(struct exception_store *store)
DMWARN("write header failed");
}
-int dm_create_persistent(struct exception_store *store, uint32_t chunk_size)
+int dm_create_persistent(struct exception_store *store)
{
- int r;
struct pstore *ps;
- r = dm_io_get(sectors_to_pages(chunk_size));
- if (r)
- return r;
-
/* allocate the pstore */
ps = kmalloc(sizeof(*ps), GFP_KERNEL);
- if (!ps) {
- r = -ENOMEM;
- goto bad;
- }
+ if (!ps)
+ return -ENOMEM;
ps->snap = store->snap;
ps->valid = 1;
ps->version = SNAPSHOT_DISK_VERSION;
+ ps->area = NULL;
ps->next_free = 2; /* skipping the header and first area */
ps->current_committed = 0;
- r = alloc_area(ps);
- if (r)
- goto bad;
-
ps->callback_count = 0;
atomic_set(&ps->pending_count, 0);
ps->callbacks = NULL;
+ ps->metadata_wq = create_singlethread_workqueue("ksnaphd");
+ if (!ps->metadata_wq) {
+ kfree(ps);
+ DMERR("couldn't start header metadata update thread");
+ return -ENOMEM;
+ }
+
store->destroy = persistent_destroy;
store->read_metadata = persistent_read_metadata;
store->prepare_exception = persistent_prepare;
store->context = ps;
return 0;
-
- bad:
- dm_io_put(sectors_to_pages(chunk_size));
- if (ps && ps->area)
- free_area(ps);
- kfree(ps);
- return r;
}
/*-----------------------------------------------------------------
return 0;
}
-static int transient_prepare(struct exception_store *store, struct exception *e)
+static int transient_prepare(struct exception_store *store,
+ struct dm_snap_exception *e)
{
struct transient_c *tc = (struct transient_c *) store->context;
sector_t size = get_dev_size(store->snap->cow->bdev);
}
static void transient_commit(struct exception_store *store,
- struct exception *e,
- void (*callback) (void *, int success),
- void *callback_context)
+ struct dm_snap_exception *e,
+ void (*callback) (void *, int success),
+ void *callback_context)
{
/* Just succeed */
callback(callback_context, 1);
*denominator = get_dev_size(store->snap->cow->bdev);
}
-int dm_create_transient(struct exception_store *store,
- struct dm_snapshot *s, int blocksize)
+int dm_create_transient(struct exception_store *store)
{
struct transient_c *tc;
- memset(store, 0, sizeof(*store));
store->destroy = transient_destroy;
store->read_metadata = transient_read_metadata;
store->prepare_exception = transient_prepare;
store->commit_exception = transient_commit;
+ store->drop_snapshot = NULL;
store->fraction_full = transient_fraction_full;
- store->snap = s;
tc = kmalloc(sizeof(struct transient_c), GFP_KERNEL);
if (!tc)