[brauner-vfs:vfs-7.5.super 62/65] fs/super.c:177 super_lock() warn: missing unwind goto?

[brauner-vfs:vfs-7.5.super 62/65] fs/super.c:177 super_lock() warn: missing unwind goto?

[kernel test robot]

Unfiltered warning: this smatch report is sent as-is and has not been
reviewed for false positives, so please treat it with due skepticism.

BCC: lkp@intel.com
CC: oe-kbuild-all@lists.linux.dev
CC: Christian Brauner <christianvanbrauner@gmail.com>
TO: Christian Brauner <brauner@kernel.org>
CC: Josef Bacik <josef@toxicpanda.com>

Hi Christian,

FYI, the error/warning was bisected to this commit, please ignore it if it's irrelevant.

tree:   https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git vfs-7.5.super
head:   2c8a968b488c134d05ccefff890270ac8ab899a8
commit: 3af63add349c1aeaee12fb571e5af95635aa83ab [62/65] super: let super_lock() wait for a superblock being shut down
:::::: branch date: 13 hours ago
:::::: commit date: 13 hours ago
config: nios2-randconfig-r2202-20261024 (https://download.01.org/0day-ci/archive/20261024/202610242131.8KUVdKA5-lkp@intel.com/config)
compiler: nios2-linux-gcc (GCC) 10.5.0
smatch: v0.5.0-9187-g5189e3fb

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Reported-by: Dan Carpenter <error27@gmail.com>
| Closes: https://lore.kernel.org/r/202610242131.8KUVdKA5-lkp@intel.com/

smatch warnings:
fs/super.c:177 super_lock() warn: missing unwind goto?
fs/super.c:177 super_lock() warn: missing unwind goto?

vim +177 fs/super.c

7c4c683849ac80 Christian Brauner 2026-10-23  123  
5d2b0f2107aa50 Christian Brauner 2026-10-23  124  /**
5d2b0f2107aa50 Christian Brauner 2026-10-23  125   * super_lock - wait for superblock to become ready and lock it
5d2b0f2107aa50 Christian Brauner 2026-10-23  126   * @sb: superblock to wait for
5d2b0f2107aa50 Christian Brauner 2026-10-23  127   * @flags: SUPER_LOCK_* flags
5d2b0f2107aa50 Christian Brauner 2026-10-23  128   *
5d2b0f2107aa50 Christian Brauner 2026-10-23  129   * If the superblock has neither passed through vfs_get_tree() or
5d2b0f2107aa50 Christian Brauner 2026-10-23  130   * generic_shutdown_super() yet wait for it to happen. Either superblock
5d2b0f2107aa50 Christian Brauner 2026-10-23  131   * creation will succeed and SB_BORN is set by vfs_get_tree() or we're
5d2b0f2107aa50 Christian Brauner 2026-10-23  132   * woken and we'll see SB_DYING.
5d2b0f2107aa50 Christian Brauner 2026-10-23  133   *
5d2b0f2107aa50 Christian Brauner 2026-10-23  134   * With SUPER_LOCK_EXCL s_umount is taken for writing, otherwise for reading.
3af63add349c1a Christian Brauner 2026-10-23  135   *
3af63add349c1a Christian Brauner 2026-10-23  136   * A superblock which is being shut down is not waited for unless
3af63add349c1a Christian Brauner 2026-10-23  137   * SUPER_LOCK_WAIT_SHUTDOWN is passed.  That is for those who have nothing to
3af63add349c1a Christian Brauner 2026-10-23  138   * do on a superblock which is going away, but who promise their caller that
3af63add349c1a Christian Brauner 2026-10-23  139   * what it had to write has been written, or who must not carry on while the
3af63add349c1a Christian Brauner 2026-10-23  140   * filesystem is still being written out.  The wait is over once
3af63add349c1a Christian Brauner 2026-10-23  141   * generic_shutdown_super() is done with @sb, and -ENOENT is still returned.
3af63add349c1a Christian Brauner 2026-10-23  142   *
3af63add349c1a Christian Brauner 2026-10-23  143   * With SUPER_LOCK_KILLABLE all the waiting can be cut short by a fatal
3af63add349c1a Christian Brauner 2026-10-23  144   * signal.  That is only for callers which give up altogether when they are
3af63add349c1a Christian Brauner 2026-10-23  145   * killed.  If what comes after the wait happens either way, the wait must
3af63add349c1a Christian Brauner 2026-10-23  146   * not be cut short.
5d2b0f2107aa50 Christian Brauner 2026-10-23  147   *
5d2b0f2107aa50 Christian Brauner 2026-10-23  148   * The caller must have acquired a temporary reference on @sb->s_passive.
5d2b0f2107aa50 Christian Brauner 2026-10-23  149   *
5d2b0f2107aa50 Christian Brauner 2026-10-23  150   * Return: 0 if SB_BORN was set and with s_umount held. Without s_umount held,
5d2b0f2107aa50 Christian Brauner 2026-10-23  151   *         -ENOENT if the superblock is being shut down or SB_DYING was set,
5d2b0f2107aa50 Christian Brauner 2026-10-23  152   *         -EINTR or -ERESTARTSYS if the task was killed.
5e87491415217d Christian Brauner 2023-08-18  153   */
5d2b0f2107aa50 Christian Brauner 2026-10-23  154  static __must_check int super_lock(struct super_block *sb,
5d2b0f2107aa50 Christian Brauner 2026-10-23  155  				   enum super_lock_flags_t flags)
5e87491415217d Christian Brauner 2023-08-18  156  {
5d2b0f2107aa50 Christian Brauner 2026-10-23  157  	unsigned int state = TASK_UNINTERRUPTIBLE;
5d2b0f2107aa50 Christian Brauner 2026-10-23  158  	int ret;
5d2b0f2107aa50 Christian Brauner 2026-10-23  159  
5d2b0f2107aa50 Christian Brauner 2026-10-23  160  	if (flags & SUPER_LOCK_KILLABLE)
5d2b0f2107aa50 Christian Brauner 2026-10-23  161  		state = TASK_KILLABLE;
be167baa28da9a Josef Bacik       2026-10-21  162  
5e87491415217d Christian Brauner 2023-08-18  163  	lockdep_assert_not_held(&sb->s_umount);
5e87491415217d Christian Brauner 2023-08-18  164  
a99d496efb788a Josef Bacik       2026-10-21  165  	/*
a99d496efb788a Josef Bacik       2026-10-21  166  	 * The last active reference is dropped with s_umount held, which
a99d496efb788a Josef Bacik       2026-10-21  167  	 * stays held until SB_DYING is set, and there is no getting a new
a99d496efb788a Josef Bacik       2026-10-21  168  	 * one.  All there is to find at the end of that wait is SB_DYING,
a99d496efb788a Josef Bacik       2026-10-21  169  	 * and it can be a long one: the filesystem is written out first.
a99d496efb788a Josef Bacik       2026-10-21  170  	 */
a99d496efb788a Josef Bacik       2026-10-21  171  	if (!atomic_read(&sb->s_active))
3af63add349c1a Christian Brauner 2026-10-23  172  		goto dying;
a99d496efb788a Josef Bacik       2026-10-21  173  
b30850c58b5b9b Christian Brauner 2023-11-27  174  	/* wait until the superblock is ready or dying */
5d2b0f2107aa50 Christian Brauner 2026-10-23  175  	ret = super_wait(sb, SB_BORN | SB_DYING, state);
be167baa28da9a Josef Bacik       2026-10-21  176  	if (ret)
be167baa28da9a Josef Bacik       2026-10-21 @177  		return ret;
b30850c58b5b9b Christian Brauner 2023-11-27  178  
b30850c58b5b9b Christian Brauner 2023-11-27  179  	/* Don't pointlessly acquire s_umount. */
b30850c58b5b9b Christian Brauner 2023-11-27  180  	if (super_flags(sb, SB_DYING))
3af63add349c1a Christian Brauner 2026-10-23  181  		goto dying;
b30850c58b5b9b Christian Brauner 2023-11-27  182  
5d2b0f2107aa50 Christian Brauner 2026-10-23  183  	ret = __super_lock(sb, flags);
be167baa28da9a Josef Bacik       2026-10-21  184  	if (ret)
be167baa28da9a Josef Bacik       2026-10-21  185  		return ret;
5e87491415217d Christian Brauner 2023-08-18  186  
5e87491415217d Christian Brauner 2023-08-18  187  	/*
5e87491415217d Christian Brauner 2023-08-18  188  	 * Has gone through generic_shutdown_super() in the meantime.
5e87491415217d Christian Brauner 2023-08-18  189  	 * @sb->s_root is NULL and @sb->s_active is 0. No one needs to
5e87491415217d Christian Brauner 2023-08-18  190  	 * grab a reference to this. Tell them so.
5e87491415217d Christian Brauner 2023-08-18  191  	 */
f0cd988016f679 Christian Brauner 2023-10-24  192  	if (sb->s_flags & SB_DYING) {
5d2b0f2107aa50 Christian Brauner 2026-10-23  193  		super_unlock(sb, flags);
3af63add349c1a Christian Brauner 2026-10-23  194  		goto dying;
f0cd988016f679 Christian Brauner 2023-10-24  195  	}
5e87491415217d Christian Brauner 2023-08-18  196  
b30850c58b5b9b Christian Brauner 2023-11-27  197  	WARN_ON_ONCE(!(sb->s_flags & SB_BORN));
be167baa28da9a Josef Bacik       2026-10-21  198  	return 0;
be167baa28da9a Josef Bacik       2026-10-21  199  
3af63add349c1a Christian Brauner 2026-10-23  200  dying:
3af63add349c1a Christian Brauner 2026-10-23  201  	if (flags & SUPER_LOCK_WAIT_SHUTDOWN) {
3af63add349c1a Christian Brauner 2026-10-23  202  		ret = super_wait(sb, SB_DYING, state);
3af63add349c1a Christian Brauner 2026-10-23  203  		if (ret)
3af63add349c1a Christian Brauner 2026-10-23  204  			return ret;
c61f54949db7b5 Josef Bacik       2026-10-21  205  	}
3af63add349c1a Christian Brauner 2026-10-23  206  	return -ENOENT;
6174e0c41df7a7 Josef Bacik       2026-10-21  207  }
6174e0c41df7a7 Josef Bacik       2026-10-21  208  

:::::: The code at line 177 was first introduced by commit
:::::: be167baa28da9a1df7fe384c4edeb9fb1c07772d fs: make a mount waiting on another mount killable

:::::: TO: Josef Bacik <josef@toxicpanda.com>
:::::: CC: Christian Brauner <brauner@kernel.org>

--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki