On Mon, Oct 28, 2019 at 11:28:08AM -0400, Theodore Y. Ts'o wrote:
> drop:
> jbd_unlock_bh_state(bh);
> __brelse(bh);
> if (drop_reserve) {
> /* no need to reserve log space for this block -bzzz */
> handle->h_buffer_credits++;
> }
> return err;
>
> not_jbd:
> jbd_unlock_bh_state(bh);
> __bforget(bh);
> goto drop;
> ----
>
> And we still have a case we jump to not_jbd, at which point hilarity
> will ensue.
>
> This is cleaned up in the following patch in this sequence, but this
> leaves us in a not-great state if we are ever bisecting.
Proposed fixup: I'll apply the following on top of this commit, and
then fix the merge conflicts in 6/7 so that the end result looks the
same as before.
Jan, any objections? I figure this way it'll save you from resending
the patch series, since the rest of it looks fine to me.
- Ted
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index f2af4afc690a..c7c9a42451c7 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -1541,8 +1541,11 @@ int jbd2_journal_forget (handle_t *handle, struct buffer_head *bh)
jbd_lock_bh_state(bh);
- if (!buffer_jbd(bh))
- goto not_jbd;
+ if (!buffer_jbd(bh)) {
+ jbd_unlock_bh_state(bh);
+ __bforget(bh);
+ return 0;
+ }
jh = bh2jh(bh);
/* Critical error: attempting to delete a bitmap buffer, maybe?
@@ -1671,11 +1674,6 @@ int jbd2_journal_forget (handle_t *handle, struct buffer_head *bh)
handle->h_buffer_credits++;
}
return err;
-
-not_jbd:
- jbd_unlock_bh_state(bh);
- __bforget(bh);
- goto drop;
}
/**
