* 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