llvm.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
* Re: [PATCH] xfs: fix deadlock between busy flushing and t_busy
       [not found] <20251114152147.66688-1-haoqinhuang7@gmail.com>
@ 2025-11-15 15:50 ` kernel test robot
  0 siblings, 0 replies; only message in thread
From: kernel test robot @ 2025-11-15 15:50 UTC (permalink / raw)
  To: Haoqin Huang, chandan.babu, djwong, linux-xfs
  Cc: llvm, oe-kbuild-all, Haoqin Huang, Rongwei Wang

Hi Haoqin,

kernel test robot noticed the following build errors:

[auto build test ERROR on xfs-linux/for-next]
[also build test ERROR on linus/master v6.18-rc5 next-20251114]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Haoqin-Huang/xfs-fix-deadlock-between-busy-flushing-and-t_busy/20251115-002142
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
patch link:    https://lore.kernel.org/r/20251114152147.66688-1-haoqinhuang7%40gmail.com
patch subject: [PATCH] xfs: fix deadlock between busy flushing and t_busy
config: x86_64-kexec (https://download.01.org/0day-ci/archive/20251115/202511152311.UvZyd6Gi-lkp@intel.com/config)
compiler: clang version 20.1.8 (https://github.com/llvm/llvm-project 87f0227cb60147a26a1eeb4fb06e3b505e9c7261)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251115/202511152311.UvZyd6Gi-lkp@intel.com/reproduce)

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>
| Closes: https://lore.kernel.org/oe-kbuild-all/202511152311.UvZyd6Gi-lkp@intel.com/

All errors (new ones prefixed by >>):

>> fs/xfs/xfs_extent_busy.c:639:45: error: use of undeclared identifier 'pag'
     639 |                 if (!alloc_flags && busy_gen == READ_ONCE(pag->pagb_gen))
         |                                                           ^
>> fs/xfs/xfs_extent_busy.c:639:45: error: use of undeclared identifier 'pag'
>> fs/xfs/xfs_extent_busy.c:639:45: error: use of undeclared identifier 'pag'
>> fs/xfs/xfs_extent_busy.c:639:45: error: use of undeclared identifier 'pag'
>> fs/xfs/xfs_extent_busy.c:639:45: error: use of undeclared identifier 'pag'
>> fs/xfs/xfs_extent_busy.c:639:45: error: use of undeclared identifier 'pag'
>> fs/xfs/xfs_extent_busy.c:639:45: error: use of undeclared identifier 'pag'
   7 errors generated.


vim +/pag +639 fs/xfs/xfs_extent_busy.c

   594	
   595	/*
   596	 * Flush out all busy extents for this group.
   597	 *
   598	 * If the current transaction is holding busy extents, the caller may not want
   599	 * to wait for committed busy extents to resolve. If we are being told just to
   600	 * try a flush or progress has been made since we last skipped a busy extent,
   601	 * return immediately to allow the caller to try again.
   602	 *
   603	 * If we are freeing extents, we might actually be holding the only free extents
   604	 * in the transaction busy list and the log force won't resolve that situation.
   605	 * In this case, we must return -EAGAIN to avoid a deadlock by informing the
   606	 * caller it needs to commit the busy extents it holds before retrying the
   607	 * extent free operation.
   608	 */
   609	int
   610	xfs_extent_busy_flush(
   611		struct xfs_trans	*tp,
   612		struct xfs_group	*xg,
   613		unsigned		busy_gen,
   614		uint32_t		alloc_flags)
   615	{
   616		struct xfs_extent_busy_tree *eb = xg->xg_busy_extents;
   617		DEFINE_WAIT		(wait);
   618		int			error;
   619	
   620		error = xfs_log_force(tp->t_mountp, XFS_LOG_SYNC);
   621		if (error)
   622			return error;
   623	
   624		/* Avoid deadlocks on uncommitted busy extents. */
   625		if (!list_empty(&tp->t_busy)) {
   626			if (alloc_flags & XFS_ALLOC_FLAG_TRYFLUSH)
   627				return 0;
   628	
   629			if (busy_gen != READ_ONCE(eb->eb_gen))
   630				return 0;
   631	
   632			if (alloc_flags & XFS_ALLOC_FLAG_FREEING)
   633				return -EAGAIN;
   634	
   635			/*
   636			 * To avoid deadlocks if alloc_flags without any FLAG set
   637			 * and t_busy is not empty.
   638			 */
 > 639			if (!alloc_flags && busy_gen == READ_ONCE(pag->pagb_gen))
   640				return -EAGAIN;
   641		}
   642	
   643		/* Wait for committed busy extents to resolve. */
   644		do {
   645			prepare_to_wait(&eb->eb_wait, &wait, TASK_KILLABLE);
   646			if  (busy_gen != READ_ONCE(eb->eb_gen))
   647				break;
   648			schedule();
   649		} while (1);
   650	
   651		finish_wait(&eb->eb_wait, &wait);
   652		return 0;
   653	}
   654	

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

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2025-11-15 15:51 UTC | newest]

Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
     [not found] <20251114152147.66688-1-haoqinhuang7@gmail.com>
2025-11-15 15:50 ` [PATCH] xfs: fix deadlock between busy flushing and t_busy kernel test robot

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).