From: Christoph Hellwig <hch@lst.de>
To: linux-xfs@vger.kernel.org
Cc: bfoster@redhat.com
Subject: [PATCH 5/6] xfs: add comments documenting the rebalance algorithm
Date: Thu, 9 Nov 2017 14:26:55 +0100 [thread overview]
Message-ID: <20171109132656.1649-6-hch@lst.de> (raw)
In-Reply-To: <20171109132656.1649-1-hch@lst.de>
Reported-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
fs/xfs/libxfs/xfs_iext_tree.c | 24 ++++++++++++++++++++++++
1 file changed, 24 insertions(+)
diff --git a/fs/xfs/libxfs/xfs_iext_tree.c b/fs/xfs/libxfs/xfs_iext_tree.c
index 3974989b0929..81e0480822d8 100644
--- a/fs/xfs/libxfs/xfs_iext_tree.c
+++ b/fs/xfs/libxfs/xfs_iext_tree.c
@@ -672,6 +672,11 @@ xfs_iext_rebalance_node(
struct xfs_iext_node *node,
int nr_entries)
{
+ /*
+ * If the neighbouring nodes are completely full, or have different
+ * parents, we might never be able to merge our node, and will only
+ * delete it once the number of entries hits zero.
+ */
if (nr_entries == 0)
return node;
@@ -693,6 +698,11 @@ xfs_iext_rebalance_node(
int nr_next = xfs_iext_node_nr_entries(next, 0), i;
if (nr_entries + nr_next <= KEYS_PER_NODE) {
+ /*
+ * Merge the next node into this node so that we don't
+ * have to do an additional update of the keys in the
+ * higher levels.
+ */
for (i = 0; i < nr_next; i++) {
node->keys[nr_entries + i] = next->keys[i];
node->ptrs[nr_entries + i] = next->ptrs[i];
@@ -741,6 +751,11 @@ xfs_iext_remove_node(
return;
if (level < ifp->if_height) {
+ /*
+ * If we aren't at the root yet try to find a neighbour node to
+ * merge with (or delete the node if it is empty), and then
+ * recurse up to the next level.
+ */
level++;
parent = xfs_iext_find_level(ifp, offset, level);
pos = xfs_iext_node_pos(parent, offset);
@@ -755,6 +770,10 @@ xfs_iext_remove_node(
goto again;
}
} else if (nr_entries == 1) {
+ /*
+ * If we are at the root and only one entry is left we can just
+ * free this node and update the root pointer.
+ */
ASSERT(node == ifp->if_u1.if_root);
ifp->if_u1.if_root = node->ptrs[0];
ifp->if_height--;
@@ -789,6 +808,11 @@ xfs_iext_rebalance_leaf(
int nr_next = xfs_iext_leaf_nr_entries(ifp, leaf->next, 0), i;
if (fill + nr_next <= RECS_PER_LEAF) {
+ /*
+ * Merge the next node into this node so that we don't
+ * have to do an additional update of the keys in the
+ * higher levels.
+ */
for (i = 0; i < nr_next; i++)
leaf->recs[fill + i] = leaf->next->recs[i];
--
2.14.2
next prev parent reply other threads:[~2017-11-09 13:27 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-11-09 13:26 incore extent b+tree fixups Christoph Hellwig
2017-11-09 13:26 ` [PATCH 1/6] xfs: fix number of records handling in xfs_iext_split_leaf Christoph Hellwig
2017-11-09 13:26 ` [PATCH 2/6] xfs: add some comments to xfs_iext_insert/xfs_iext_insert_node Christoph Hellwig
2017-11-09 13:26 ` [PATCH 3/6] xfs: remove a superflous assignment in xfs_iext_remove_node Christoph Hellwig
2017-11-09 13:26 ` [PATCH 4/6] xfs: trivial indentation fixup for xfs_iext_remove_node Christoph Hellwig
2017-11-09 13:26 ` Christoph Hellwig [this message]
2017-11-09 13:26 ` [PATCH 6/6] xfs: handle zero entries case in xfs_iext_rebalance_leaf Christoph Hellwig
2017-11-09 14:32 ` incore extent b+tree fixups Brian Foster
2017-11-10 1:11 ` Darrick J. Wong
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=20171109132656.1649-6-hch@lst.de \
--to=hch@lst.de \
--cc=bfoster@redhat.com \
--cc=linux-xfs@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).