Comment # 21 on bug 1020030 from Randy Wright

Reproducing pretty reliably on my SUT with the previously mentioned KOTD.  
Most recently if produced a full call trace:

[   21.155564] Call Trace:
[   21.158047]  [<ffffffff81019b19>] dump_trace+0x59/0x310
[   21.163310]  [<ffffffff81019eba>] show_stack_log_lvl+0xea/0x170
[   21.169270]  [<ffffffff8101ac41>] show_stack+0x21/0x40
[   21.174448]  [<ffffffff81320810>] dump_stack+0x5c/0x7c
[   21.179628]  [<ffffffff8107e061>] warn_slowpath_common+0x81/0xb0
[   21.185708]  [<ffffffffa034526d>] btrfs_qgroup_free_refroot+0x14d/0x180
[btrfs]
[   21.193104]  [<ffffffffa02d6797>] commit_fs_roots.isra.21+0x147/0x180
[btrfs]
[   21.200315]  [<ffffffffa02d8f86>]
btrfs_commit_transaction.part.27+0x3f6/0xa90 [btrfs]
[   21.208290]  [<ffffffff812369ee>] sync_filesystem+0x6e/0xa0
[   21.213905]  [<ffffffff812086e2>] generic_shutdown_super+0x22/0xf0
[   21.220126]  [<ffffffff812087be>] kill_anon_super+0xe/0x20
[   21.225667]  [<ffffffffa02a7983>] btrfs_kill_super+0x13/0x100 [btrfs]
[   21.232152]  [<ffffffff81208089>] deactivate_locked_super+0x39/0x70
[   21.238466]  [<ffffffff812234ab>] cleanup_mnt+0x3b/0x80
[   21.243730]  [<ffffffff8109a112>] task_work_run+0x72/0x90
[   21.249168]  [<ffffffff81078ad9>] exit_to_usermode_loop+0x91/0xc2
[   21.255305]  [<ffffffff81003ae5>] syscall_return_slowpath+0x85/0xa0
[   21.261618]  [<ffffffff81609348>] int_ret_from_sys_call+0x25/0x8f
[   21.271028] DWARF2 unwinder stuck at int_ret_from_sys_call+0x25/0x8f


There is a supportconfig bundle from this SUT attached to a different
confidential bug 1043261.