bcache superblock reading / writing v2

bcache superblock reading / writing v2

[Christoph Hellwig]

Hi Coly and Liang,

can you review this series to sort out the bcache superblock reading for
larger page sizes?  I don't have bcache test setup so this is compile
tested only.

Changes since v1:
 - actually assing ->sb_disk
 - fix the cover letter subject line

[PATCH 1/7] bcache: cached_dev_free needs to put the sb page

[Christoph Hellwig]

From: Liang Chen <liangchen.linux@gmail.com>

Same as cache device, the buffer page needs to be put while
freeing cached_dev.  Otherwise a page would be leaked every
time a cached_dev is stopped.

Signed-off-by: Liang Chen <liangchen.linux@gmail.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/super.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index 77e9869345e7..a573ce1d85aa 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1275,6 +1275,9 @@ static void cached_dev_free(struct closure *cl)
 
 	mutex_unlock(&bch_register_lock);
 
+	if (dc->sb_bio.bi_inline_vecs[0].bv_page)
+		put_page(bio_first_page_all(&dc->sb_bio));
+
 	if (!IS_ERR_OR_NULL(dc->bdev))
 		blkdev_put(dc->bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
 
-- 
2.20.1

[PATCH 2/7] bcache: use a separate data structure for the on-disk super block

[Christoph Hellwig]

Split out an on-disk version struct cache_sb with the proper endianess
annotations.  This fixes a fair chunk of sparse warnings, but there are
some left due to the way the checksum is defined.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/super.c   |  6 ++---
 include/uapi/linux/bcache.h | 51 +++++++++++++++++++++++++++++++++++++
 2 files changed, 54 insertions(+), 3 deletions(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index a573ce1d85aa..3045f27e0d67 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -63,14 +63,14 @@ static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
 			      struct page **res)
 {
 	const char *err;
-	struct cache_sb *s;
+	struct cache_sb_disk *s;
 	struct buffer_head *bh = __bread(bdev, 1, SB_SIZE);
 	unsigned int i;
 
 	if (!bh)
 		return "IO error";
 
-	s = (struct cache_sb *) bh->b_data;
+	s = (struct cache_sb_disk *)bh->b_data;
 
 	sb->offset		= le64_to_cpu(s->offset);
 	sb->version		= le64_to_cpu(s->version);
@@ -209,7 +209,7 @@ static void write_bdev_super_endio(struct bio *bio)
 
 static void __write_super(struct cache_sb *sb, struct bio *bio)
 {
-	struct cache_sb *out = page_address(bio_first_page_all(bio));
+	struct cache_sb_disk *out = page_address(bio_first_page_all(bio));
 	unsigned int i;
 
 	bio->bi_iter.bi_sector	= SB_SECTOR;
diff --git a/include/uapi/linux/bcache.h b/include/uapi/linux/bcache.h
index 5d4f58e059fd..1d8b3a9fc080 100644
--- a/include/uapi/linux/bcache.h
+++ b/include/uapi/linux/bcache.h
@@ -156,6 +156,57 @@ static inline struct bkey *bkey_idx(const struct bkey *k, unsigned int nr_keys)
 
 #define BDEV_DATA_START_DEFAULT		16	/* sectors */
 
+struct cache_sb_disk {
+	__le64			csum;
+	__le64			offset;	/* sector where this sb was written */
+	__le64			version;
+
+	__u8			magic[16];
+
+	__u8			uuid[16];
+	union {
+		__u8		set_uuid[16];
+		__le64		set_magic;
+	};
+	__u8			label[SB_LABEL_SIZE];
+
+	__le64			flags;
+	__le64			seq;
+	__le64			pad[8];
+
+	union {
+	struct {
+		/* Cache devices */
+		__le64		nbuckets;	/* device size */
+
+		__le16		block_size;	/* sectors */
+		__le16		bucket_size;	/* sectors */
+
+		__le16		nr_in_set;
+		__le16		nr_this_dev;
+	};
+	struct {
+		/* Backing devices */
+		__le64		data_offset;
+
+		/*
+		 * block_size from the cache device section is still used by
+		 * backing devices, so don't add anything here until we fix
+		 * things to not need it for backing devices anymore
+		 */
+	};
+	};
+
+	__le32			last_mount;	/* time overflow in y2106 */
+
+	__le16			first_bucket;
+	union {
+		__le16		njournal_buckets;
+		__le16		keys;
+	};
+	__le64			d[SB_JOURNAL_BUCKETS];	/* journal buckets */
+};
+
 struct cache_sb {
 	__u64			csum;
 	__u64			offset;	/* sector where this sb was written */
-- 
2.20.1

[PATCH 3/7] bcache: rework error unwinding in register_bcache

[Christoph Hellwig]

Split the successful and error return path, and use one goto label for each
resource to unwind.  This also fixes some small errors like leaking the
module reference count in the reboot case (which seems entirely harmless)
or printing the wrong warning messages for early failures.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/super.c | 75 +++++++++++++++++++++++----------------
 1 file changed, 45 insertions(+), 30 deletions(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index 3045f27e0d67..e8013e1b0a14 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -2375,29 +2375,33 @@ static bool bch_is_open(struct block_device *bdev)
 static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			       const char *buffer, size_t size)
 {
-	ssize_t ret = -EINVAL;
-	const char *err = "cannot allocate memory";
-	char *path = NULL;
-	struct cache_sb *sb = NULL;
+	const char *err;
+	char *path;
+	struct cache_sb *sb;
 	struct block_device *bdev = NULL;
-	struct page *sb_page = NULL;
+	struct page *sb_page;
+	ssize_t ret;
 
+	ret = -EBUSY;
 	if (!try_module_get(THIS_MODULE))
-		return -EBUSY;
+		goto out;
 
 	/* For latest state of bcache_is_reboot */
 	smp_mb();
 	if (bcache_is_reboot)
-		return -EBUSY;
+		goto out_module_put;
 
+	ret = -ENOMEM;
+	err = "cannot allocate memory";
 	path = kstrndup(buffer, size, GFP_KERNEL);
 	if (!path)
-		goto err;
+		goto out_module_put;
 
 	sb = kmalloc(sizeof(struct cache_sb), GFP_KERNEL);
 	if (!sb)
-		goto err;
+		goto out_free_path;
 
+	ret = -EINVAL;
 	err = "failed to open device";
 	bdev = blkdev_get_by_path(strim(path),
 				  FMODE_READ|FMODE_WRITE|FMODE_EXCL,
@@ -2414,57 +2418,68 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			if (!IS_ERR(bdev))
 				bdput(bdev);
 			if (attr == &ksysfs_register_quiet)
-				goto quiet_out;
+				goto done;
 		}
-		goto err;
+		goto out_free_sb;
 	}
 
 	err = "failed to set blocksize";
 	if (set_blocksize(bdev, 4096))
-		goto err_close;
+		goto out_blkdev_put;
 
 	err = read_super(sb, bdev, &sb_page);
 	if (err)
-		goto err_close;
+		goto out_blkdev_put;
 
 	err = "failed to register device";
 	if (SB_IS_BDEV(sb)) {
 		struct cached_dev *dc = kzalloc(sizeof(*dc), GFP_KERNEL);
 
 		if (!dc)
-			goto err_close;
+			goto out_put_sb_page;
 
 		mutex_lock(&bch_register_lock);
 		ret = register_bdev(sb, sb_page, bdev, dc);
 		mutex_unlock(&bch_register_lock);
 		/* blkdev_put() will be called in cached_dev_free() */
-		if (ret < 0)
-			goto err;
+		if (ret < 0) {
+			bdev = NULL;
+			goto out_put_sb_page;
+		}
 	} else {
 		struct cache *ca = kzalloc(sizeof(*ca), GFP_KERNEL);
 
 		if (!ca)
-			goto err_close;
+			goto out_put_sb_page;
 
 		/* blkdev_put() will be called in bch_cache_release() */
-		if (register_cache(sb, sb_page, bdev, ca) != 0)
-			goto err;
+		if (register_cache(sb, sb_page, bdev, ca) != 0) {
+			bdev = NULL;
+			goto out_put_sb_page;
+		}
 	}
-quiet_out:
-	ret = size;
-out:
-	if (sb_page)
-		put_page(sb_page);
+
+	put_page(sb_page);
+done:
 	kfree(sb);
 	kfree(path);
 	module_put(THIS_MODULE);
-	return ret;
-
-err_close:
-	blkdev_put(bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
-err:
+	return size;
+
+out_put_sb_page:
+	put_page(sb_page);
+out_blkdev_put:
+	if (bdev)
+		blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
+out_free_sb:
+	kfree(sb);
+out_free_path:
+	kfree(path);
+out_module_put:
+	module_put(THIS_MODULE);
+out:
 	pr_info("error %s: %s", path, err);
-	goto out;
+	return ret;
 }
 
 
-- 
2.20.1

[PATCH 4/7] bcache: transfer the sb_page reference to register_{bdev,cache}

[Christoph Hellwig]

Avoid an extra reference count roundtrip by transferring the sb_page
ownership to the lower level register helpers.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/super.c | 21 ++++++---------------
 1 file changed, 6 insertions(+), 15 deletions(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index e8013e1b0a14..c11bc0135ae6 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1368,8 +1368,6 @@ static int register_bdev(struct cache_sb *sb, struct page *sb_page,
 
 	bio_init(&dc->sb_bio, dc->sb_bio.bi_inline_vecs, 1);
 	bio_first_bvec_all(&dc->sb_bio)->bv_page = sb_page;
-	get_page(sb_page);
-
 
 	if (cached_dev_init(dc, sb->block_size << 9))
 		goto err;
@@ -2275,7 +2273,6 @@ static int register_cache(struct cache_sb *sb, struct page *sb_page,
 
 	bio_init(&ca->sb_bio, ca->sb_bio.bi_inline_vecs, 1);
 	bio_first_bvec_all(&ca->sb_bio)->bv_page = sb_page;
-	get_page(sb_page);
 
 	if (blk_queue_discard(bdev_get_queue(bdev)))
 		ca->discard = CACHE_DISCARD(&ca->sb);
@@ -2378,7 +2375,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	const char *err;
 	char *path;
 	struct cache_sb *sb;
-	struct block_device *bdev = NULL;
+	struct block_device *bdev;
 	struct page *sb_page;
 	ssize_t ret;
 
@@ -2442,10 +2439,8 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 		ret = register_bdev(sb, sb_page, bdev, dc);
 		mutex_unlock(&bch_register_lock);
 		/* blkdev_put() will be called in cached_dev_free() */
-		if (ret < 0) {
-			bdev = NULL;
-			goto out_put_sb_page;
-		}
+		if (ret < 0)
+			goto out_free_sb;
 	} else {
 		struct cache *ca = kzalloc(sizeof(*ca), GFP_KERNEL);
 
@@ -2453,13 +2448,10 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			goto out_put_sb_page;
 
 		/* blkdev_put() will be called in bch_cache_release() */
-		if (register_cache(sb, sb_page, bdev, ca) != 0) {
-			bdev = NULL;
-			goto out_put_sb_page;
-		}
+		if (register_cache(sb, sb_page, bdev, ca) != 0)
+			goto out_free_sb;
 	}
 
-	put_page(sb_page);
 done:
 	kfree(sb);
 	kfree(path);
@@ -2469,8 +2461,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 out_put_sb_page:
 	put_page(sb_page);
 out_blkdev_put:
-	if (bdev)
-		blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
+	blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
 out_free_sb:
 	kfree(sb);
 out_free_path:
-- 
2.20.1

[PATCH 5/7] bcache: return a pointer to the on-disk sb from read_super

[Christoph Hellwig]

Returning the properly typed actual data structure insteaf of the
containing struct page will save the callers some work going
forward.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/super.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index c11bc0135ae6..0b7620d9f087 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -60,7 +60,7 @@ struct workqueue_struct *bch_journal_wq;
 /* Superblock */
 
 static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
-			      struct page **res)
+			      struct cache_sb_disk **res)
 {
 	const char *err;
 	struct cache_sb_disk *s;
@@ -191,7 +191,7 @@ static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
 	err = NULL;
 
 	get_page(bh->b_page);
-	*res = bh->b_page;
+	*res = s;
 err:
 	put_bh(bh);
 	return err;
@@ -1353,7 +1353,7 @@ static int cached_dev_init(struct cached_dev *dc, unsigned int block_size)
 
 /* Cached device - bcache superblock */
 
-static int register_bdev(struct cache_sb *sb, struct page *sb_page,
+static int register_bdev(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 				 struct block_device *bdev,
 				 struct cached_dev *dc)
 {
@@ -1367,7 +1367,7 @@ static int register_bdev(struct cache_sb *sb, struct page *sb_page,
 	dc->bdev->bd_holder = dc;
 
 	bio_init(&dc->sb_bio, dc->sb_bio.bi_inline_vecs, 1);
-	bio_first_bvec_all(&dc->sb_bio)->bv_page = sb_page;
+	bio_first_bvec_all(&dc->sb_bio)->bv_page = virt_to_page(sb_disk);
 
 	if (cached_dev_init(dc, sb->block_size << 9))
 		goto err;
@@ -2260,7 +2260,7 @@ static int cache_alloc(struct cache *ca)
 	return ret;
 }
 
-static int register_cache(struct cache_sb *sb, struct page *sb_page,
+static int register_cache(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 				struct block_device *bdev, struct cache *ca)
 {
 	const char *err = NULL; /* must be set for any error case */
@@ -2272,7 +2272,7 @@ static int register_cache(struct cache_sb *sb, struct page *sb_page,
 	ca->bdev->bd_holder = ca;
 
 	bio_init(&ca->sb_bio, ca->sb_bio.bi_inline_vecs, 1);
-	bio_first_bvec_all(&ca->sb_bio)->bv_page = sb_page;
+	bio_first_bvec_all(&ca->sb_bio)->bv_page = virt_to_page(sb_disk);
 
 	if (blk_queue_discard(bdev_get_queue(bdev)))
 		ca->discard = CACHE_DISCARD(&ca->sb);
@@ -2375,8 +2375,8 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	const char *err;
 	char *path;
 	struct cache_sb *sb;
+	struct cache_sb_disk *sb_disk;
 	struct block_device *bdev;
-	struct page *sb_page;
 	ssize_t ret;
 
 	ret = -EBUSY;
@@ -2424,7 +2424,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	if (set_blocksize(bdev, 4096))
 		goto out_blkdev_put;
 
-	err = read_super(sb, bdev, &sb_page);
+	err = read_super(sb, bdev, &sb_disk);
 	if (err)
 		goto out_blkdev_put;
 
@@ -2436,7 +2436,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			goto out_put_sb_page;
 
 		mutex_lock(&bch_register_lock);
-		ret = register_bdev(sb, sb_page, bdev, dc);
+		ret = register_bdev(sb, sb_disk, bdev, dc);
 		mutex_unlock(&bch_register_lock);
 		/* blkdev_put() will be called in cached_dev_free() */
 		if (ret < 0)
@@ -2448,7 +2448,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			goto out_put_sb_page;
 
 		/* blkdev_put() will be called in bch_cache_release() */
-		if (register_cache(sb, sb_page, bdev, ca) != 0)
+		if (register_cache(sb, sb_disk, bdev, ca) != 0)
 			goto out_free_sb;
 	}
 
@@ -2459,7 +2459,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	return size;
 
 out_put_sb_page:
-	put_page(sb_page);
+	put_page(virt_to_page(sb_disk));
 out_blkdev_put:
 	blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
 out_free_sb:
-- 
2.20.1

[PATCH 6/7] bcache: store a pointer to the on-disk sb in the cache and cached_dev structures

[Christoph Hellwig]

This allows to properly build the superblock bio including the offset in the
page using the normal bio helpers.  This fixes writing the superblock for
page sizes larger than 4k where the sb write bio would need an offset in the
bio_vec.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/bcache.h |  2 ++
 drivers/md/bcache/super.c  | 34 +++++++++++++++-------------------
 2 files changed, 17 insertions(+), 19 deletions(-)

diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index 9198c1b480d9..adf26a21fcd1 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -301,6 +301,7 @@ struct cached_dev {
 	struct block_device	*bdev;
 
 	struct cache_sb		sb;
+	struct cache_sb_disk	*sb_disk;
 	struct bio		sb_bio;
 	struct bio_vec		sb_bv[1];
 	struct closure		sb_write;
@@ -403,6 +404,7 @@ enum alloc_reserve {
 struct cache {
 	struct cache_set	*set;
 	struct cache_sb		sb;
+	struct cache_sb_disk	*sb_disk;
 	struct bio		sb_bio;
 	struct bio_vec		sb_bv[1];
 
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index 0b7620d9f087..c3bfbc813daf 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -207,15 +207,15 @@ static void write_bdev_super_endio(struct bio *bio)
 	closure_put(&dc->sb_write);
 }
 
-static void __write_super(struct cache_sb *sb, struct bio *bio)
+static void __write_super(struct cache_sb *sb, struct cache_sb_disk *out,
+		struct bio *bio)
 {
-	struct cache_sb_disk *out = page_address(bio_first_page_all(bio));
 	unsigned int i;
 
+	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_META;
 	bio->bi_iter.bi_sector	= SB_SECTOR;
-	bio->bi_iter.bi_size	= SB_SIZE;
-	bio_set_op_attrs(bio, REQ_OP_WRITE, REQ_SYNC|REQ_META);
-	bch_bio_map(bio, NULL);
+	__bio_add_page(bio, virt_to_page(out), SB_SIZE,
+			offset_in_page(out));
 
 	out->offset		= cpu_to_le64(sb->offset);
 	out->version		= cpu_to_le64(sb->version);
@@ -257,14 +257,14 @@ void bch_write_bdev_super(struct cached_dev *dc, struct closure *parent)
 	down(&dc->sb_write_mutex);
 	closure_init(cl, parent);
 
-	bio_reset(bio);
+	bio_init(bio, dc->sb_bv, 1);
 	bio_set_dev(bio, dc->bdev);
 	bio->bi_end_io	= write_bdev_super_endio;
 	bio->bi_private = dc;
 
 	closure_get(cl);
 	/* I/O request sent to backing device */
-	__write_super(&dc->sb, bio);
+	__write_super(&dc->sb, dc->sb_disk, bio);
 
 	closure_return_with_destructor(cl, bch_write_bdev_super_unlock);
 }
@@ -306,13 +306,13 @@ void bcache_write_super(struct cache_set *c)
 
 		SET_CACHE_SYNC(&ca->sb, CACHE_SYNC(&c->sb));
 
-		bio_reset(bio);
+		bio_init(bio, ca->sb_bv, 1);
 		bio_set_dev(bio, ca->bdev);
 		bio->bi_end_io	= write_super_endio;
 		bio->bi_private = ca;
 
 		closure_get(cl);
-		__write_super(&ca->sb, bio);
+		__write_super(&ca->sb, ca->sb_disk, bio);
 	}
 
 	closure_return_with_destructor(cl, bcache_write_super_unlock);
@@ -1275,8 +1275,8 @@ static void cached_dev_free(struct closure *cl)
 
 	mutex_unlock(&bch_register_lock);
 
-	if (dc->sb_bio.bi_inline_vecs[0].bv_page)
-		put_page(bio_first_page_all(&dc->sb_bio));
+	if (dc->sb_disk)
+		put_page(virt_to_page(dc->sb_disk));
 
 	if (!IS_ERR_OR_NULL(dc->bdev))
 		blkdev_put(dc->bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
@@ -1365,9 +1365,7 @@ static int register_bdev(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 	memcpy(&dc->sb, sb, sizeof(struct cache_sb));
 	dc->bdev = bdev;
 	dc->bdev->bd_holder = dc;
-
-	bio_init(&dc->sb_bio, dc->sb_bio.bi_inline_vecs, 1);
-	bio_first_bvec_all(&dc->sb_bio)->bv_page = virt_to_page(sb_disk);
+	dc->sb_disk = sb_disk;
 
 	if (cached_dev_init(dc, sb->block_size << 9))
 		goto err;
@@ -2137,8 +2135,8 @@ void bch_cache_release(struct kobject *kobj)
 	for (i = 0; i < RESERVE_NR; i++)
 		free_fifo(&ca->free[i]);
 
-	if (ca->sb_bio.bi_inline_vecs[0].bv_page)
-		put_page(bio_first_page_all(&ca->sb_bio));
+	if (ca->sb_disk)
+		put_page(virt_to_page(ca->sb_disk));
 
 	if (!IS_ERR_OR_NULL(ca->bdev))
 		blkdev_put(ca->bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
@@ -2270,9 +2268,7 @@ static int register_cache(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 	memcpy(&ca->sb, sb, sizeof(struct cache_sb));
 	ca->bdev = bdev;
 	ca->bdev->bd_holder = ca;
-
-	bio_init(&ca->sb_bio, ca->sb_bio.bi_inline_vecs, 1);
-	bio_first_bvec_all(&ca->sb_bio)->bv_page = virt_to_page(sb_disk);
+	ca->sb_disk = sb_disk;
 
 	if (blk_queue_discard(bdev_get_queue(bdev)))
 		ca->discard = CACHE_DISCARD(&ca->sb);
-- 
2.20.1

[PATCH 7/7] bcache: use read_cache_page_gfp to read the superblock

[Christoph Hellwig]

Avoid a pointless dependency on buffer heads in bcache by simply open
coding reading a single page.  Also add a SB_OFFSET define for the
byte offset of the superblock instead of using magic numbers.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/super.c   | 16 +++++++---------
 include/uapi/linux/bcache.h |  1 +
 2 files changed, 8 insertions(+), 9 deletions(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index c3bfbc813daf..7ebd355f51b1 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -15,7 +15,6 @@
 #include "writeback.h"
 
 #include <linux/blkdev.h>
-#include <linux/buffer_head.h>
 #include <linux/debugfs.h>
 #include <linux/genhd.h>
 #include <linux/idr.h>
@@ -64,13 +63,14 @@ static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
 {
 	const char *err;
 	struct cache_sb_disk *s;
-	struct buffer_head *bh = __bread(bdev, 1, SB_SIZE);
+	struct page *page;
 	unsigned int i;
 
-	if (!bh)
+	page = read_cache_page_gfp(bdev->bd_inode->i_mapping,
+				   SB_OFFSET >> PAGE_SHIFT, GFP_KERNEL);
+	if (IS_ERR(page))
 		return "IO error";
-
-	s = (struct cache_sb_disk *)bh->b_data;
+	s = page_address(page) + offset_in_page(SB_OFFSET);
 
 	sb->offset		= le64_to_cpu(s->offset);
 	sb->version		= le64_to_cpu(s->version);
@@ -188,12 +188,10 @@ static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
 	}
 
 	sb->last_mount = (u32)ktime_get_real_seconds();
-	err = NULL;
-
-	get_page(bh->b_page);
 	*res = s;
+	return NULL;
 err:
-	put_bh(bh);
+	put_page(page);
 	return err;
 }
 
diff --git a/include/uapi/linux/bcache.h b/include/uapi/linux/bcache.h
index 1d8b3a9fc080..9a1965c6c3d0 100644
--- a/include/uapi/linux/bcache.h
+++ b/include/uapi/linux/bcache.h
@@ -148,6 +148,7 @@ static inline struct bkey *bkey_idx(const struct bkey *k, unsigned int nr_keys)
 #define BCACHE_SB_MAX_VERSION		4
 
 #define SB_SECTOR			8
+#define SB_OFFSET			(SB_SECTOR << SECTOR_SHIFT)
 #define SB_SIZE				4096
 #define SB_LABEL_SIZE			32
 #define SB_JOURNAL_BUCKETS		256U
-- 
2.20.1

[PATCH 5/7] bcache: return a pointer to the on-disk sb from read_super

[Christoph Hellwig]

Returning the properly typed actual data structure insteaf of the
containing struct page will save the callers some work going
forward.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/bcache/super.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index c11bc0135ae6..0b7620d9f087 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -60,7 +60,7 @@ struct workqueue_struct *bch_journal_wq;
 /* Superblock */
 
 static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
-			      struct page **res)
+			      struct cache_sb_disk **res)
 {
 	const char *err;
 	struct cache_sb_disk *s;
@@ -191,7 +191,7 @@ static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
 	err = NULL;
 
 	get_page(bh->b_page);
-	*res = bh->b_page;
+	*res = s;
 err:
 	put_bh(bh);
 	return err;
@@ -1353,7 +1353,7 @@ static int cached_dev_init(struct cached_dev *dc, unsigned int block_size)
 
 /* Cached device - bcache superblock */
 
-static int register_bdev(struct cache_sb *sb, struct page *sb_page,
+static int register_bdev(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 				 struct block_device *bdev,
 				 struct cached_dev *dc)
 {
@@ -1367,7 +1367,7 @@ static int register_bdev(struct cache_sb *sb, struct page *sb_page,
 	dc->bdev->bd_holder = dc;
 
 	bio_init(&dc->sb_bio, dc->sb_bio.bi_inline_vecs, 1);
-	bio_first_bvec_all(&dc->sb_bio)->bv_page = sb_page;
+	bio_first_bvec_all(&dc->sb_bio)->bv_page = virt_to_page(sb_disk);
 
 	if (cached_dev_init(dc, sb->block_size << 9))
 		goto err;
@@ -2260,7 +2260,7 @@ static int cache_alloc(struct cache *ca)
 	return ret;
 }
 
-static int register_cache(struct cache_sb *sb, struct page *sb_page,
+static int register_cache(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 				struct block_device *bdev, struct cache *ca)
 {
 	const char *err = NULL; /* must be set for any error case */
@@ -2272,7 +2272,7 @@ static int register_cache(struct cache_sb *sb, struct page *sb_page,
 	ca->bdev->bd_holder = ca;
 
 	bio_init(&ca->sb_bio, ca->sb_bio.bi_inline_vecs, 1);
-	bio_first_bvec_all(&ca->sb_bio)->bv_page = sb_page;
+	bio_first_bvec_all(&ca->sb_bio)->bv_page = virt_to_page(sb_disk);
 
 	if (blk_queue_discard(bdev_get_queue(bdev)))
 		ca->discard = CACHE_DISCARD(&ca->sb);
@@ -2375,8 +2375,8 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	const char *err;
 	char *path;
 	struct cache_sb *sb;
+	struct cache_sb_disk *sb_disk;
 	struct block_device *bdev;
-	struct page *sb_page;
 	ssize_t ret;
 
 	ret = -EBUSY;
@@ -2424,7 +2424,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	if (set_blocksize(bdev, 4096))
 		goto out_blkdev_put;
 
-	err = read_super(sb, bdev, &sb_page);
+	err = read_super(sb, bdev, &sb_disk);
 	if (err)
 		goto out_blkdev_put;
 
@@ -2436,7 +2436,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			goto out_put_sb_page;
 
 		mutex_lock(&bch_register_lock);
-		ret = register_bdev(sb, sb_page, bdev, dc);
+		ret = register_bdev(sb, sb_disk, bdev, dc);
 		mutex_unlock(&bch_register_lock);
 		/* blkdev_put() will be called in cached_dev_free() */
 		if (ret < 0)
@@ -2448,7 +2448,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			goto out_put_sb_page;
 
 		/* blkdev_put() will be called in bch_cache_release() */
-		if (register_cache(sb, sb_page, bdev, ca) != 0)
+		if (register_cache(sb, sb_disk, bdev, ca) != 0)
 			goto out_free_sb;
 	}
 
@@ -2459,7 +2459,7 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	return size;
 
 out_put_sb_page:
-	put_page(sb_page);
+	put_page(virt_to_page(sb_disk));
 out_blkdev_put:
 	blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
 out_free_sb:
-- 
2.20.1