Merge tag 'xfs-6.8-merge-3' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux

Pull xfs updates from Chandan Babu:
 "New features/functionality:
   - Online repair:
       - Reserve disk space for online repairs
       - Fix misinteraction between the AIL and btree bulkloader because
         of which the bulk load fails to queue a buffer for writeback if
         it happens to be on the AIL list
       - Prevent transaction reservation overflows when reaping blocks
         during online repair
       - Whenever possible, bulkloader now copies multiple records into
         a block
       - Support repairing of
           1. Per-AG free space, inode and refcount btrees
           2. Ondisk inodes
           3. File data and attribute fork mappings
       - Verify the contents of
           1. Inode and data fork of realtime bitmap file
           2. Quota files
   - Introduce MF_MEM_PRE_REMOVE. This will be used to notify tasks
     about a pmem device being removed

  Bug fixes:
   - Fix memory leak of recovered attri intent items
   - Fix UAF during log intent recovery
   - Fix realtime geometry integer overflows
   - Prevent scrub from live locking in xchk_iget
   - Prevent fs shutdown when removing files during low free disk space
   - Prevent transaction reservation overflow when extending an RT
     device
   - Prevent incorrect warning from being printed when extending a
     filesystem
   - Fix an off-by-one error in xreap_agextent_binval
   - Serialize access to perag radix tree during deletion operation
   - Fix perag memory leak during growfs
   - Allow allocation of minlen realtime extent when the maximum sized
     realtime free extent is minlen in size

  Cleanups:
   - Remove duplicate boilerplate code spread across functionality
     associated with different log items
   - Cleanup resblks interfaces
   - Pass defer ops pointer to defer helpers instead of an enum
   - Initialize di_crc in xfs_log_dinode to prevent KMSAN warnings
   - Use static_assert() instead of BUILD_BUG_ON_MSG() to validate size
     of structures and structure member offsets. This is done in order
     to be able to share the code with userspace
   - Move XFS documentation under a new directory specific to XFS
   - Do not invoke deferred ops' ->create_done callback if the deferred
     operation does not have an intent item associated with it
   - Remove duplicate inclusion of header files from scrub/health.c
   - Refactor Realtime code
   - Cleanup attr code"

* tag 'xfs-6.8-merge-3' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (123 commits)
  xfs: use the op name in trace_xlog_intent_recovery_failed
  xfs: fix a use after free in xfs_defer_finish_recovery
  xfs: turn the XFS_DA_OP_REPLACE checks in xfs_attr_shortform_addname into asserts
  xfs: remove xfs_attr_sf_hdr_t
  xfs: remove struct xfs_attr_shortform
  xfs: use xfs_attr_sf_findname in xfs_attr_shortform_getvalue
  xfs: remove xfs_attr_shortform_lookup
  xfs: simplify xfs_attr_sf_findname
  xfs: move the xfs_attr_sf_lookup tracepoint
  xfs: return if_data from xfs_idata_realloc
  xfs: make if_data a void pointer
  xfs: fold xfs_rtallocate_extent into xfs_bmap_rtalloc
  xfs: simplify and optimize the RT allocation fallback cascade
  xfs: reorder the minlen and prod calculations in xfs_bmap_rtalloc
  xfs: remove XFS_RTMIN/XFS_RTMAX
  xfs: remove rt-wrappers from xfs_format.h
  xfs: factor out a xfs_rtalloc_sumlevel helper
  xfs: tidy up xfs_rtallocate_extent_exact
  xfs: merge the calls to xfs_rtallocate_range in xfs_rtallocate_block
  xfs: reflow the tail end of xfs_rtallocate_extent_block
  ...

1 files changed, 17 insertions, 4 deletions

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index a0d9b4ac7d54..4f9b61f4a668 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -678,7 +678,7 @@ static void add_to_kill_fsdax(struct task_struct *tsk, struct page *p,
  */
 static void collect_procs_fsdax(struct page *page,
 		struct address_space *mapping, pgoff_t pgoff,
-		struct list_head *to_kill)
+		struct list_head *to_kill, bool pre_remove)
 {
 	struct vm_area_struct *vma;
 	struct task_struct *tsk;
@@ -686,8 +686,15 @@ static void collect_procs_fsdax(struct page *page,
 	i_mmap_lock_read(mapping);
 	rcu_read_lock();
 	for_each_process(tsk) {
-		struct task_struct *t = task_early_kill(tsk, true);
+		struct task_struct *t = tsk;
 
+		/*
+		 * Search for all tasks while MF_MEM_PRE_REMOVE is set, because
+		 * the current may not be the one accessing the fsdax page.
+		 * Otherwise, search for the current task.
+		 */
+		if (!pre_remove)
+			t = task_early_kill(tsk, true);
 		if (!t)
 			continue;
 		vma_interval_tree_foreach(vma, &mapping->i_mmap, pgoff, pgoff) {
@@ -1793,6 +1800,7 @@ int mf_dax_kill_procs(struct address_space *mapping, pgoff_t index,
 	dax_entry_t cookie;
 	struct page *page;
 	size_t end = index + count;
+	bool pre_remove = mf_flags & MF_MEM_PRE_REMOVE;
 
 	mf_flags |= MF_ACTION_REQUIRED | MF_MUST_KILL;
 
@@ -1804,9 +1812,14 @@ int mf_dax_kill_procs(struct address_space *mapping, pgoff_t index,
 		if (!page)
 			goto unlock;
 
-		SetPageHWPoison(page);
+		if (!pre_remove)
+			SetPageHWPoison(page);
 
-		collect_procs_fsdax(page, mapping, index, &to_kill);
+		/*
+		 * The pre_remove case is revoking access, the memory is still
+		 * good and could theoretically be put back into service.
+		 */
+		collect_procs_fsdax(page, mapping, index, &to_kill, pre_remove);
 		unmap_and_kill(&to_kill, page_to_pfn(page), mapping,
 				index, mf_flags);
 unlock: