All the mail mirrored from lore.kernel.org
 help / color / mirror / Atom feed

From: Dan Carpenter <dan.carpenter@linaro.org>
To: josef@toxicpanda.com
Cc: linux-btrfs@vger.kernel.org
Subject: [bug report] btrfs: clean up our handling of refs == 0 in snapshot delete
Date: Sun, 28 Apr 2024 15:54:15 +0300	[thread overview]
Message-ID: <96789032-42fb-41c0-b16c-561ac00ca7c3@moroto.mountain> (raw)

Hello Josef Bacik,

Commit a07155f4e6b8 ("btrfs: clean up our handling of refs == 0 in
snapshot delete") from Apr 19, 2024 (linux-next), leads to the
following Smatch static checker warning:

	fs/btrfs/extent-tree.c:5875 walk_up_proc()
	warn: inconsistent returns '&eb->lock'.

fs/btrfs/extent-tree.c
    5766 static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
    5767                                  struct btrfs_root *root,
    5768                                  struct btrfs_path *path,
    5769                                  struct walk_control *wc)
    5770 {
    5771         struct btrfs_fs_info *fs_info = root->fs_info;
    5772         int ret;
    5773         int level = wc->level;
    5774         struct extent_buffer *eb = path->nodes[level];
    5775         u64 parent = 0;
    5776 
    5777         if (wc->stage == UPDATE_BACKREF) {
    5778                 ASSERT(wc->shared_level >= level);
    5779                 if (level < wc->shared_level)
    5780                         goto out;
    5781 
    5782                 ret = find_next_key(path, level + 1, &wc->update_progress);
    5783                 if (ret > 0)
    5784                         wc->update_ref = 0;
    5785 
    5786                 wc->stage = DROP_REFERENCE;
    5787                 wc->shared_level = -1;
    5788                 path->slots[level] = 0;
    5789 
    5790                 /*
    5791                  * check reference count again if the block isn't locked.
    5792                  * we should start walking down the tree again if reference
    5793                  * count is one.
    5794                  */
    5795                 if (!path->locks[level]) {
    5796                         ASSERT(level > 0);
    5797                         btrfs_tree_lock(eb);
                                 ^^^^^^^^^^^^^^^^^^^^
Takes the lock

    5798                         path->locks[level] = BTRFS_WRITE_LOCK;
    5799 
    5800                         ret = btrfs_lookup_extent_info(trans, fs_info,
    5801                                                        eb->start, level, 1,
    5802                                                        &wc->refs[level],
    5803                                                        &wc->flags[level],
    5804                                                        NULL);
    5805                         if (ret < 0) {
    5806                                 btrfs_tree_unlock_rw(eb, path->locks[level]);

These paths drop the lock

    5807                                 path->locks[level] = 0;
    5808                                 return ret;
    5809                         }
    5810                         if (unlikely(wc->refs[level] == 0)) {
    5811                                 btrfs_err(fs_info, "Missing refs.");
    5812                                 return -EUCLEAN;

But this new return doesn't.

    5813                         }
    5814                         if (wc->refs[level] == 1) {
    5815                                 btrfs_tree_unlock_rw(eb, path->locks[level]);
    5816                                 path->locks[level] = 0;
    5817                                 return 1;
    5818                         }
    5819                 }
    5820         }
    5821 
    5822         /* wc->stage == DROP_REFERENCE */

regards,
dan carpenter

                 reply	other threads:[~2024-04-28 12:54 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=96789032-42fb-41c0-b16c-561ac00ca7c3@moroto.mountain \
    --to=dan.carpenter@linaro.org \
    --cc=josef@toxicpanda.com \
    --cc=linux-btrfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.