linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Liu Bo <liubo2009@cn.fujitsu.com>
To: <linux-btrfs@vger.kernel.org>
Subject: [PATCH 1/6 v3][RFC] Btrfs: merge adjacent states as much as possible
Date: Wed, 25 Jul 2012 13:58:37 +0800	[thread overview]
Message-ID: <1343195922-31405-2-git-send-email-liubo2009@cn.fujitsu.com> (raw)
In-Reply-To: <1343195922-31405-1-git-send-email-liubo2009@cn.fujitsu.com>

In order to reduce write locks, we do merge_state as much as much as possible.

Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
---
 fs/btrfs/extent_io.c |   47 +++++++++++++++++++++++++++--------------------
 1 files changed, 27 insertions(+), 20 deletions(-)

diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 01c21b6..1858d86 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -275,29 +275,36 @@ static void merge_state(struct extent_io_tree *tree,
 	if (state->state & (EXTENT_IOBITS | EXTENT_BOUNDARY))
 		return;
 
-	other_node = rb_prev(&state->rb_node);
-	if (other_node) {
+	while (1) {
+		other_node = rb_prev(&state->rb_node);
+		if (!other_node)
+			break;
 		other = rb_entry(other_node, struct extent_state, rb_node);
-		if (other->end == state->start - 1 &&
-		    other->state == state->state) {
-			merge_cb(tree, state, other);
-			state->start = other->start;
-			other->tree = NULL;
-			rb_erase(&other->rb_node, &tree->state);
-			free_extent_state(other);
-		}
+		if (other->end != state->start - 1 ||
+		    other->state != state->state)
+			break;
+
+		merge_cb(tree, state, other);
+		state->start = other->start;
+		other->tree = NULL;
+		rb_erase(&other->rb_node, &tree->state);
+		free_extent_state(other);
 	}
-	other_node = rb_next(&state->rb_node);
-	if (other_node) {
+
+	while (1) {
+		other_node = rb_next(&state->rb_node);
+		if (!other_node)
+			break;
 		other = rb_entry(other_node, struct extent_state, rb_node);
-		if (other->start == state->end + 1 &&
-		    other->state == state->state) {
-			merge_cb(tree, state, other);
-			state->end = other->end;
-			other->tree = NULL;
-			rb_erase(&other->rb_node, &tree->state);
-			free_extent_state(other);
-		}
+		if (other->start != state->end + 1 ||
+		    other->state != state->state)
+			break;
+
+		merge_cb(tree, state, other);
+		state->end = other->end;
+		other->tree = NULL;
+		rb_erase(&other->rb_node, &tree->state);
+		free_extent_state(other);
 	}
 }
 
-- 
1.6.5.2


  reply	other threads:[~2012-07-25  5:59 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-07-25  5:58 [PATCH 0/6 v3][RFC] rwlock for extent state Liu Bo
2012-07-25  5:58 ` Liu Bo [this message]
2012-07-25  5:58 ` [PATCH 2/6 v3][RFC] Btrfs: add helper function to test if we can merge state Liu Bo
2012-07-25  5:58 ` [PATCH 3/6 v3][RFC] Btrfs: break clear_state_bit into two parts Liu Bo
2012-07-25  5:58 ` [PATCH 4/6 v3][RFC] Btrfs: apply rwlock for extent state Liu Bo
2012-07-25  5:58 ` [PATCH 5/6 v3][RFC] Btrfs: batch merge state in readpage endio Liu Bo
2012-07-25  5:58 ` [PATCH 6/6 v3][RFC] Btrfs: async helper for state merge Liu Bo

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=1343195922-31405-2-git-send-email-liubo2009@cn.fujitsu.com \
    --to=liubo2009@cn.fujitsu.com \
    --cc=linux-btrfs@vger.kernel.org \
    /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).