linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Johannes Thumshirn <jth@kernel.org>
To: Qu Wenruo <wqu@suse.com>
Cc: linux-btrfs@vger.kernel.org,
	Johannes Thumshirn <johannes.thumshirn@wdc.com>
Subject: [RFC 1/2] btrfs: introduce dummy RAID stripes for preallocated data
Date: Wed, 18 Sep 2024 16:08:49 +0200	[thread overview]
Message-ID: <20240918140850.27261-2-jth@kernel.org> (raw)
In-Reply-To: <20240918140850.27261-1-jth@kernel.org>

From: Johannes Thumshirn <johannes.thumshirn@wdc.com>

Introduce BTRFS_RAID_STRIPE_DUMMY_KEY which describes a range in the
RAID stripe-tree that is backed by meta-data but no data. For instance
for preallocated extents.

On read, we can simply ignore the contents of the stripe_extent as
we're not interested in the physical address and device id of these
stripe_extents.

Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
---
 fs/btrfs/raid-stripe-tree.c     | 43 +++++++++++++++++++++++++++++++++
 fs/btrfs/raid-stripe-tree.h     |  2 ++
 include/uapi/linux/btrfs_tree.h |  1 +
 3 files changed, 46 insertions(+)

diff --git a/fs/btrfs/raid-stripe-tree.c b/fs/btrfs/raid-stripe-tree.c
index 075fecd08d87..bbe0689b1d17 100644
--- a/fs/btrfs/raid-stripe-tree.c
+++ b/fs/btrfs/raid-stripe-tree.c
@@ -158,6 +158,40 @@ static int update_raid_extent_item(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
+int btrfs_insert_dummy_raid_extent(struct btrfs_trans_handle *trans,
+				   u64 logical, u64 length)
+{
+	struct btrfs_fs_info *fs_info = trans->fs_info;
+	struct btrfs_key stripe_key;
+	struct btrfs_root *stripe_root = fs_info->stripe_root;
+	struct btrfs_path *path;
+	int ret;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+
+	stripe_key.objectid = logical;
+	stripe_key.type = BTRFS_RAID_STRIPE_DUMMY_KEY;
+	stripe_key.offset = length;
+
+	ret = btrfs_insert_empty_item(trans, stripe_root, path, &stripe_key, 0);
+	if (ret == -EEXIST) {
+		struct extent_buffer *leaf = path->nodes[0];
+		int slot = path->slots[0];
+		struct btrfs_key found_key;
+
+		btrfs_item_key_to_cpu(leaf, &found_key, slot);
+		found_key.objectid = logical;
+		found_key.offset = length;
+		btrfs_mark_buffer_dirty(trans, leaf);
+		ret = 0;
+	}
+	btrfs_free_path(path);
+
+	return ret;
+}
+
 static int btrfs_insert_one_raid_extent(struct btrfs_trans_handle *trans,
 					struct btrfs_io_context *bioc)
 {
@@ -305,6 +339,15 @@ int btrfs_get_raid_extent_offset(struct btrfs_fs_info *fs_info,
 	if (end > found_end)
 		*length -= end - found_end;
 
+	/*
+	 * If we have a BTRFS_RAID_STRIPE_DUMMY_KEY it means we've hit
+	 * an entry for a preallocated extent. Noone will ever check
+	 * the physical, only logical and length, so we're good to
+	 * bail out from here.
+	 */
+	if (found_key.type == BTRFS_RAID_STRIPE_DUMMY_KEY)
+		goto out;
+
 	num_stripes = btrfs_num_raid_stripes(btrfs_item_size(leaf, slot));
 	stripe_extent = btrfs_item_ptr(leaf, slot, struct btrfs_stripe_extent);
 
diff --git a/fs/btrfs/raid-stripe-tree.h b/fs/btrfs/raid-stripe-tree.h
index 1ac1c21aac2f..dbe52789f201 100644
--- a/fs/btrfs/raid-stripe-tree.h
+++ b/fs/btrfs/raid-stripe-tree.h
@@ -27,6 +27,8 @@ int btrfs_get_raid_extent_offset(struct btrfs_fs_info *fs_info,
 				 u32 stripe_index, struct btrfs_io_stripe *stripe);
 int btrfs_insert_raid_extent(struct btrfs_trans_handle *trans,
 			     struct btrfs_ordered_extent *ordered_extent);
+int btrfs_insert_dummy_raid_extent(struct btrfs_trans_handle *trans,
+				   u64 logical, u64 length);
 
 static inline bool btrfs_need_stripe_tree_update(struct btrfs_fs_info *fs_info,
 						 u64 map_type)
diff --git a/include/uapi/linux/btrfs_tree.h b/include/uapi/linux/btrfs_tree.h
index fc29d273845d..76b18013b394 100644
--- a/include/uapi/linux/btrfs_tree.h
+++ b/include/uapi/linux/btrfs_tree.h
@@ -281,6 +281,7 @@
 #define BTRFS_CHUNK_ITEM_KEY	228
 
 #define BTRFS_RAID_STRIPE_KEY	230
+#define BTRFS_RAID_STRIPE_DUMMY_KEY 231
 
 /*
  * Records the overall state of the qgroups.
-- 
2.43.0


  reply	other threads:[~2024-09-18 14:09 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-09-12 14:33 [PATCH] btrfs: scrub: skip PREALLOC extents on RAID stripe-tree Johannes Thumshirn
2024-09-12 21:32 ` Qu Wenruo
2024-09-13  5:42   ` Johannes Thumshirn
2024-09-13  5:47     ` Qu Wenruo
2024-09-18 14:08       ` [RFC 0/2] Add dummy RAID stripe tree entries for PREALLOC data Johannes Thumshirn
2024-09-18 14:08         ` Johannes Thumshirn [this message]
2024-09-18 14:08         ` [RFC 2/2] btrfs: insert dummy RAID stripe extents for preallocated data Johannes Thumshirn
2024-09-18 23:45         ` [RFC 0/2] Add dummy RAID stripe tree entries for PREALLOC data Qu Wenruo
2024-09-19 15:42           ` Johannes Thumshirn
2024-09-19 16:57         ` Gerhard Wiesinger
2024-09-20  9:50           ` Johannes Thumshirn
2024-09-23 15:27         ` Josef Bacik

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=20240918140850.27261-2-jth@kernel.org \
    --to=jth@kernel.org \
    --cc=johannes.thumshirn@wdc.com \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=wqu@suse.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).