All of lore.kernel.org
 help / color / mirror / Atom feed
From: Qu Wenruo <quwenruo@cn.fujitsu.com>
To: <linux-btrfs@vger.kernel.org>
Cc: Wang Xiaoguang <wangxg.fnst@cn.fujitsu.com>
Subject: [PATCH v3 15/16] btrfs: dedup: Add support for adding hash for on-disk backend
Date: Thu, 7 Jan 2016 09:08:16 +0800	[thread overview]
Message-ID: <1452128897-5433-16-git-send-email-quwenruo@cn.fujitsu.com> (raw)
In-Reply-To: <1452128897-5433-1-git-send-email-quwenruo@cn.fujitsu.com>

Now on-disk backend can add hash now.

Signed-off-by: Wang Xiaoguang <wangxg.fnst@cn.fujitsu.com>
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>

---
v3:
  Fix a bug in writing extent buffer with wrong offset.
  Add handler for same bytenr different hashes case.
---
 fs/btrfs/dedup.c | 84 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 84 insertions(+)

diff --git a/fs/btrfs/dedup.c b/fs/btrfs/dedup.c
index f448e27..3e09fd8 100644
--- a/fs/btrfs/dedup.c
+++ b/fs/btrfs/dedup.c
@@ -246,6 +246,8 @@ out:
 	return ret;
 }
 
+static int ondisk_search_hash(struct btrfs_dedup_info *dedup_info, u8 *hash,
+			      u64 *bytenr_ret, u64 *num_bytes_ret);
 static void inmem_destroy(struct btrfs_fs_info *fs_info);
 int btrfs_dedup_cleanup(struct btrfs_fs_info *fs_info)
 {
@@ -392,6 +394,86 @@ out:
 	return 0;
 }
 
+static int ondisk_search_bytenr(struct btrfs_trans_handle *trans,
+				struct btrfs_dedup_info *dedup_info,
+				struct btrfs_path *path, u64 bytenr,
+				int prepare_del);
+static int ondisk_add(struct btrfs_trans_handle *trans,
+		      struct btrfs_dedup_info *dedup_info,
+		      struct btrfs_dedup_hash *hash)
+{
+	struct btrfs_path *path;
+	struct btrfs_root *dedup_root = dedup_info->dedup_root;
+	struct btrfs_key key;
+	struct btrfs_dedup_hash_item *hash_item;
+	u64 bytenr;
+	u64 num_bytes;
+	int hash_len = btrfs_dedup_sizes[dedup_info->hash_type];
+	int ret;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+
+	mutex_lock(&dedup_info->ondisk_lock);
+
+	/* Search for duplicated bytenr, refer inmem_add() for reason */
+	ret = ondisk_search_bytenr(NULL, dedup_info, path, hash->bytenr, 0);
+	if (ret < 0)
+		goto out;
+	if (ret > 0) {
+		ret = 0;
+		goto out;
+	}
+	btrfs_release_path(path);
+
+	ret = ondisk_search_hash(dedup_info, hash->hash, &bytenr, &num_bytes);
+	if (ret < 0)
+		goto out;
+	/* Same hash found, don't re-add to save dedup tree space */
+	if (ret > 0) {
+		ret = 0;
+		goto out;
+	}
+
+	/* Insert hash->bytenr item */
+	memcpy(&key.objectid, hash->hash + hash_len - 8, 8);
+	key.type = BTRFS_DEDUP_HASH_ITEM_KEY;
+	key.offset = hash->bytenr;
+
+	ret = btrfs_insert_empty_item(trans, dedup_root, path, &key,
+			sizeof(*hash_item) + hash_len);
+	WARN_ON(ret == -EEXIST);
+	if (ret < 0)
+		goto out;
+	hash_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
+				   struct btrfs_dedup_hash_item);
+	btrfs_set_dedup_hash_len(path->nodes[0], hash_item, hash->num_bytes);
+	write_extent_buffer(path->nodes[0], hash->hash,
+			    (unsigned long)(hash_item + 1), hash_len);
+	btrfs_mark_buffer_dirty(path->nodes[0]);
+	btrfs_release_path(path);
+
+	/* Then bytenr->hash item */
+	key.objectid = hash->bytenr;
+	key.type = BTRFS_DEDUP_BYTENR_ITEM_KEY;
+	memcpy(&key.offset, hash->hash + hash_len - 8, 8);
+
+	ret = btrfs_insert_empty_item(trans, dedup_root, path, &key, hash_len);
+	WARN_ON(ret == -EEXIST);
+	if (ret < 0)
+		goto out;
+	write_extent_buffer(path->nodes[0], hash->hash,
+			btrfs_item_ptr_offset(path->nodes[0], path->slots[0]),
+			hash_len);
+	btrfs_mark_buffer_dirty(path->nodes[0]);
+
+out:
+	mutex_unlock(&dedup_info->ondisk_lock);
+	btrfs_free_path(path);
+	return ret;
+}
+
 int btrfs_dedup_add(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		    struct btrfs_dedup_hash *hash)
 {
@@ -406,6 +488,8 @@ int btrfs_dedup_add(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 
 	if (dedup_info->backend == BTRFS_DEDUP_BACKEND_INMEMORY)
 		return inmem_add(dedup_info, hash);
+	if (dedup_info->backend == BTRFS_DEDUP_BACKEND_ONDISK)
+		return ondisk_add(trans, dedup_info, hash);
 	return -EINVAL;
 }
 
-- 
2.6.4




  parent reply	other threads:[~2016-01-07  1:08 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-01-07  1:08 [PATCH v3 00/14][For 4.6] Btrfs: Add inband (write time) de-duplication framework Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 01/16] btrfs: dedup: Introduce dedup framework and its header Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 02/16] btrfs: dedup: Introduce function to initialize dedup info Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 03/16] btrfs: dedup: Introduce function to add hash into in-memory tree Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 04/16] btrfs: dedup: Introduce function to remove hash from " Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 05/16] btrfs: delayed-ref: Add support for atomic increasing extent ref Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 06/16] btrfs: delayed_ref: Add support for handle dedup hash Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 07/16] btrfs: dedup: Introduce function to search for an existing hash Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 08/16] btrfs: dedup: Implement btrfs_dedup_calc_hash interface Qu Wenruo
2016-01-07 13:21   ` kbuild test robot
2016-01-07  1:08 ` [PATCH v3 09/16] btrfs: ordered-extent: Add support for dedup Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 10/16] btrfs: dedup: Inband in-memory only de-duplication implement Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 11/16] btrfs: dedup: Add basic tree structure for on-disk dedup method Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 12/16] btrfs: dedup: Introduce interfaces to resume and cleanup dedup info Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 13/16] btrfs: dedup: Add support for on-disk hash search Qu Wenruo
2016-01-07  1:08 ` [PATCH v3 14/16] btrfs: dedup: Add support to delete hash for on-disk backend Qu Wenruo
2016-01-07  1:08 ` Qu Wenruo [this message]
2016-01-07  1:08 ` [PATCH v3 16/16] btrfs: dedup: Add ioctl for inband deduplication Qu Wenruo

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1452128897-5433-16-git-send-email-quwenruo@cn.fujitsu.com \
    --to=quwenruo@cn.fujitsu.com \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=wangxg.fnst@cn.fujitsu.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.