btrfs: fix warning when putting transaction with qgroups enabled after abort

Commit Message

Filipe Manana July 14, 2023, 12:12 p.m. UTC

From: Filipe Manana <[e-mail]>

If we have a transaction abort with qgroups enabled we get a warning
triggered when doing the final put on the transaction, like this:

  [161552.678901] ------------[ cut here ]------------
  [161552.681530] WARNING: CPU: 4 PID: 81745 at fs/btrfs/transaction.c:144 btrfs_put_transaction+0x123/0x130 [btrfs]
  [161552.681759] Modules linked in: btrfs blake2b_generic xor (...)
  [161552.681934] CPU: 4 PID: 81745 Comm: btrfs-transacti Tainted: G        W          6.4.0-rc6-btrfs-next-134+ #1
  [161552.681945] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.2-0-gea1b7a073390-prebuilt.qemu.org 04/01/2014
  [161552.681951] RIP: 0010:btrfs_put_transaction+0x123/0x130 [btrfs]
  [161552.682139] Code: bd a0 01 00 (...)
  [161552.682146] RSP: 0018:ffffa168c0527e28 EFLAGS: 00010286
  [161552.682155] RAX: ffff936042caed00 RBX: ffff93604a3eb448 RCX: 0000000000000000
  [161552.682161] RDX: ffff93606421b028 RSI: ffffffff92ff0878 RDI: ffff93606421b010
  [161552.682166] RBP: ffff93606421b000 R08: 0000000000000000 R09: ffffa168c0d07c20
  [161552.682171] R10: 0000000000000000 R11: ffff93608dc52950 R12: ffffa168c0527e70
  [161552.682175] R13: ffff93606421b000 R14: ffff93604a3eb420 R15: ffff93606421b028
  [161552.682181] FS:  0000000000000000(0000) GS:ffff93675fb00000(0000) knlGS:0000000000000000
  [161552.682187] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  [161552.682193] CR2: 0000558ad262b000 CR3: 000000014feda005 CR4: 0000000000370ee0
  [161552.682211] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  [161552.682216] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  [161552.682221] Call Trace:
  [161552.682229]  <TASK>
  [161552.682236]  ? __warn+0x80/0x130
  [161552.682250]  ? btrfs_put_transaction+0x123/0x130 [btrfs]
  [161552.682430]  ? report_bug+0x1f4/0x200
  [161552.682444]  ? handle_bug+0x42/0x70
  [161552.682456]  ? exc_invalid_op+0x14/0x70
  [161552.682467]  ? asm_exc_invalid_op+0x16/0x20
  [161552.682483]  ? btrfs_put_transaction+0x123/0x130 [btrfs]
  [161552.682661]  btrfs_cleanup_transaction+0xe7/0x5e0 [btrfs]
  [161552.682838]  ? _raw_spin_unlock_irqrestore+0x23/0x40
  [161552.682847]  ? try_to_wake_up+0x94/0x5e0
  [161552.682856]  ? __pfx_process_timeout+0x10/0x10
  [161552.682872]  transaction_kthread+0x103/0x1d0 [btrfs]
  [161552.683047]  ? __pfx_transaction_kthread+0x10/0x10 [btrfs]
  [161552.683217]  kthread+0xee/0x120
  [161552.683227]  ? __pfx_kthread+0x10/0x10
  [161552.683237]  ret_from_fork+0x29/0x50
  [161552.683259]  </TASK>
  [161552.683262] ---[ end trace 0000000000000000 ]---

This corresponds to this line of code:

  void btrfs_put_transaction(struct btrfs_transaction *transaction)
  {
      (...)
          WARN_ON(!RB_EMPTY_ROOT(
                          &transaction->delayed_refs.dirty_extent_root));
      (...)
  }

The warning happens because btrfs_qgroup_destroy_extent_records(), called
in the transaction abort path, we free all entries from the rbtree
"dirty_extent_root" with rbtree_postorder_for_each_entry_safe(), but we
don't actually empty the rbtree - it's still pointing to nodes that were
freed.

So set the rbtree's root node to NULL to avoid this warning.

Fixes: 81f7eb00ff5b ("btrfs: destroy qgroup extent records on transaction abort")
Signed-off-by: Filipe Manana <[e-mail]>
---
 fs/btrfs/qgroup.c | 1 +
 1 file changed, 1 insertion(+)

Patch

diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
index da1f84a0eb29..3da859a43c98 100644
--- a/fs/btrfs/qgroup.c
+++ b/fs/btrfs/qgroup.c
@@ -4445,4 +4445,5 @@  void btrfs_qgroup_destroy_extent_records(struct btrfs_transaction *trans)
 		ulist_free(entry->old_roots);
 		kfree(entry);
 	}
+	root->rb_node = NULL;
 }