Re: [syzbot] [btrfs?] WARNING in btrfs_create_pending_block_groups

From: David Sterba
Date: Tue Feb 13 2024 - 13:03:20 EST


On Tue, Feb 13, 2024 at 05:22:02AM -0800, syzbot wrote:
> syzbot suspects this issue was fixed by commit:
>
> commit a1912f712188291f9d7d434fba155461f1ebef66
> Author: Josef Bacik <josef@xxxxxxxxxxxxxx>
> Date: Wed Nov 22 17:17:55 2023 +0000
>
> btrfs: remove code for inode_cache and recovery mount options

This does not look like a fix, it's an unrelated change, the reported
problem probably depends on timing.

> bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=113ba042180000
> start commit: a4d7d7011219 Merge tag 'spi-fix-v6.4-rc5' of git://git.ker..
> git tree: upstream
> kernel config: https://syzkaller.appspot.com/x/.config?x=7474de833c217bf4
> dashboard link: https://syzkaller.appspot.com/bug?extid=5fd11a1f057a67a03a1b

The log says that there's an attempt do
"balance: start -f -susage=6,vrange=0..9223372036854775809"

which looks suspicious but should be otherwise harmless, the range is
only considered. What is possible problem is that it's for the system
block group:

nfo (device loop0): balance: start -f -susage=6,vrange=0..9223372036854775809
------------[ cut here ]------------
BTRFS: Transaction aborted (error -28)
WARNING: CPU: 0 PID: 5028 at fs/btrfs/block-group.c:2686 btrfs_create_pending_block_groups+0x10b7/0x1220 fs/btrfs/block-group.c:2686
Modules linked in:
CPU: 0 PID: 5028 Comm: syz-executor115 Not tainted 6.6.0-syzkaller-00207-g14ab6d425e80 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/09/2023
RIP: 0010:btrfs_create_pending_block_groups+0x10b7/0x1220 fs/btrfs/block-group.c:2686