FAILED: patch "[PATCH] btrfs: fix log context list corruption after rename whiteout" failed to apply to 4.14-stable tree

PATCH [email protected] 2 participants 3 messages Mar 23, 2020

Subsystems (2)

Stable branch Btrfs file system

lore

Share & tools

lore

[email protected]

author maintainer

2020-03-23 14:43 UTC #0 lore raw

The patch below does not apply to the 4.14-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 236ebc20d9afc5e9ff52f3cf3f365a91583aac10 Mon Sep 17 00:00:00 2001
From: Filipe Manana <[email protected]>
Date: Tue, 10 Mar 2020 12:13:53 +0000
Subject: [PATCH] btrfs: fix log context list corruption after rename whiteout
error

During a rename whiteout, if btrfs_whiteout_for_rename() returns an error we can end up returning from btrfs_rename() with the log context object still in the root's log context list - this happens if 'sync_log' was set to true before we called btrfs_whiteout_for_rename() and it is dangerous because we end up with a corrupt linked list (root->log_ctxs) as the log context object was allocated on the stack.

After btrfs_rename() returns, any task that is running btrfs_sync_log() concurrently can end up crashing because that linked list is traversed by btrfs_sync_log() (through btrfs_remove_all_log_ctxs()). That results in the same issue that commit e6c617102c7e4 ("Btrfs: fix log context list corruption after rename exchange operation") fixed.

Fixes d4682ba03ef618 ("Btrfs: sync log after logging new name")

CC [email protected] # 4.19+

Signed-off-by Filipe Manana <[email protected]>

Signed-off-by David Sterba <[email protected]>

fs/btrfs/inode.c +4 -0

index 27076ebadb36..d267eb5caa7b 100644
@@ -9496,6 +9496,10 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		ret = btrfs_sync_log(trans, BTRFS_I(old_inode)->root, &ctx);
 		if (ret)
 			commit_transaction = true;
+	} else if (sync_log) {
+		mutex_lock(&root->log_mutex);
+		list_del(&ctx.list);
+		mutex_unlock(&root->log_mutex);
 	}
 	if (commit_transaction) {
 		ret = btrfs_commit_transaction(trans);

Sasha Levin

maintainer

2020-03-23 16:45 UTC #1 lore raw

Quoted 31 lines · On Mon, Mar 23, 2020 at 03:43:02PM +0100, [email protected] wrote:

The patch below does not apply to the 4.14-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 236ebc20d9afc5e9ff52f3cf3f365a91583aac10 Mon Sep 17 00:00:00 2001
From: Filipe Manana <[email protected]>
Date: Tue, 10 Mar 2020 12:13:53 +0000
Subject: [PATCH] btrfs: fix log context list corruption after rename whiteout error
During a rename whiteout, if btrfs_whiteout_for_rename() returns an error we can end up returning from btrfs_rename() with the log context object still in the root's log context list - this happens if 'sync_log' was set to true before we called btrfs_whiteout_for_rename() and it is dangerous because we end up with a corrupt linked list (root->log_ctxs) as the log context object was allocated on the stack.
After btrfs_rename() returns, any task that is running btrfs_sync_log() concurrently can end up crashing because that linked list is traversed by btrfs_sync_log() (through btrfs_remove_all_log_ctxs()). That results in the same issue that commit e6c617102c7e4 ("Btrfs: fix log context list corruption after rename exchange operation") fixed.

Quoted 4 lines

Fixes d4682ba03ef618 ("Btrfs: sync log after logging new name")
CC [email protected] # 4.19+
Signed-off-by Filipe Manana <[email protected]>
Signed-off-by David Sterba <[email protected]>

Greg, I'm not sure why you tried this for 4.14 - it's tagged 4.19+ and we don't have d4682ba03ef618 in 4.14 either.

Signature · 2 lines

Thanks,
Sasha

Greg KH

author maintainer

2020-03-23 17:44 UTC #2 lore raw

Quoted 32 lines · Sasha Levin

Quoted 31 lines · On Mon, Mar 23, 2020 at 03:43:02PM +0100, [email protected] wrote:
The patch below does not apply to the 4.14-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
Quoted 1 line
From 236ebc20d9afc5e9ff52f3cf3f365a91583aac10 Mon Sep 17 00:00:00 2001
From: Filipe Manana <[email protected]>
Date: Tue, 10 Mar 2020 12:13:53 +0000
Subject: [PATCH] btrfs: fix log context list corruption after rename whiteout error
During a rename whiteout, if btrfs_whiteout_for_rename() returns an error we can end up returning from btrfs_rename() with the log context object still in the root's log context list - this happens if 'sync_log' was set to true before we called btrfs_whiteout_for_rename() and it is dangerous because we end up with a corrupt linked list (root->log_ctxs) as the log context object was allocated on the stack.
After btrfs_rename() returns, any task that is running btrfs_sync_log() concurrently can end up crashing because that linked list is traversed by btrfs_sync_log() (through btrfs_remove_all_log_ctxs()). That results in the same issue that commit e6c617102c7e4 ("Btrfs: fix log context list corruption after rename exchange operation") fixed.

Quoted 7 lines

Quoted 4 lines
Fixes d4682ba03ef618 ("Btrfs: sync log after logging new name")
CC [email protected] # 4.19+
Signed-off-by Filipe Manana <[email protected]>
Signed-off-by David Sterba <[email protected]>
Greg, I'm not sure why you tried this for 4.14 - it's tagged 4.19+ and we don't have d4682ba03ef618 in 4.14 either.

Ah, I thought this was backported, as "simple" grep shows it showing up in 4.14.110, but that was just a reference to the short git commit id.

I need to fix my scripts to always expand it to the full sha1 to prevent that from catching.

thanks,

greg k-h