From: Qu Wenruo <wqu@suse.com>
To: linux-btrfs@vger.kernel.org
Subject: [PATCH RFC] btrfs: relocation: Exit balance explicitly if there is no more extents to relocate
Date: Mon, 27 Apr 2020 14:58:14 +0800 [thread overview]
Message-ID: <20200427065814.46952-1-wqu@suse.com> (raw)
There are several reported runaway balance, some of them are even unable
to be canceled with latest kernel.
Current balance only exit if rc::found_extents is 0, but there is a more
obvious condition to terminate relocation: is there any new extents to
relocate.
This patch will make balance to check such condition, as an final safe
net to catch the runaway balance.
Signed-off-by: Qu Wenruo <wqu@suse.com>
---
This is definitely not the right solution, as we should find the root
cause, e.g. where we missed a rc->found_extents--;, but this would provide
a quick test to determine if it's rc::found_extents get un-synced from the
real situation.
And an extra safenet is never a bad thing.
---
fs/btrfs/relocation.c | 16 +++++++++++++++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 03bc7134e8cb..fb59eda27930 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -3978,6 +3978,14 @@ int prepare_to_relocate(struct reloc_control *rc)
return 0;
}
+/*
+ * Return 0 if current balance stage is finished without problem.
+ * For MOVE_DATA_EXTENTS stage, we need to progress to UPDATE_DATA_PTRS stage.
+ *
+ * Return >0 if there is no extents to relocate anymore.
+ *
+ * Return <0 for fatal error.
+ */
static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
{
struct btrfs_fs_info *fs_info = rc->extent_root->fs_info;
@@ -3988,6 +3996,7 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
struct btrfs_extent_item *ei;
u64 flags;
u32 item_size;
+ bool no_more_extent = true;
int ret;
int err = 0;
int progress = 0;
@@ -4033,6 +4042,7 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
break;
rc->extents_found++;
+ no_more_extent = false;
ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
struct btrfs_extent_item);
@@ -4154,7 +4164,9 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
err = ret;
btrfs_free_block_rsv(fs_info, rc->block_rsv);
btrfs_free_path(path);
- return err;
+ if (err)
+ return err;
+ return no_more_extent;
}
static int __insert_orphan_inode(struct btrfs_trans_handle *trans,
@@ -4374,6 +4386,8 @@ int btrfs_relocate_block_group(struct btrfs_fs_info *fs_info, u64 group_start)
if (ret < 0)
err = ret;
+ if (ret > 0)
+ break;
finishes_stage = rc->stage;
/*
* We may have gotten ENOSPC after we already dirtied some
--
2.23.0
reply other threads:[~2020-04-27 6:58 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20200427065814.46952-1-wqu@suse.com \
--to=wqu@suse.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).