Merge tag 'for-6.14-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs updates from David Sterba:
 "User visible changes, features:

   - rebuilding of the free space tree at mount time is done in more
     transactions, fix potential hangs when the transaction thread is
     blocked due to large amount of block groups

   - more read IO balancing strategies (experimental config), add two
     new ways how to select a device for read if the profiles allow that
     (all RAID1*), the current default selects the device by pid which
     is good on average but less performant for single reader workloads

       - select preferred device for all reads (namely for testing)
       - round-robin, balance reads across devices relevant for the
         requested IO range

   - add encoded write ioctl support to io_uring (read was added in
     6.12), basis for writing send stream using that instead of
     syscalls, non-blocking mode is not yet implemented

   - support FS_IOC_READ_VERITY_METADATA, applications can use the
     metadata to do their own verification

   - pass inode's i_write_hint to bios, for parity with other
     filesystems, ioctls F_GET_RW_HINT/F_SET_RW_HINT

  Core:

   - in zoned mode: allow to directly reclaim a block group by simply
     resetting it, then it can be reused and another block group does
     not need to be allocated

   - super block validation now also does more comprehensive sys array
     validation, adding it to the points where superblock is validated
     (post-read, pre-write)

   - subpage mode fixes:
      - fix double accounting of blocks due to some races
      - improved or fixed error handling in a few cases (compression,
        delalloc)

   - raid stripe tree:
      - fix various cases with extent range splitting or deleting
      - implement hole punching to extent range
      - reduce number of stripe tree lookups during bio submission
      - more self-tests

   - updated self-tests (delayed refs)

   - error handling improvements

   - cleanups, refactoring
      - remove rest of backref caching infrastructure from relocation,
        not needed anymore
      - error message updates
      - remove unnecessary calls when extent buffer was marked dirty
      - unused parameter removal
      - code moved to new files

  Other code changes: add rb_find_add_cached() to the rb-tree API"

* tag 'for-6.14-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (127 commits)
  btrfs: selftests: add a selftest for deleting two out of three extents
  btrfs: selftests: add test for punching a hole into 3 RAID stripe-extents
  btrfs: selftests: add selftest for punching holes into the RAID stripe extents
  btrfs: selftests: test RAID stripe-tree deletion spanning two items
  btrfs: selftests: don't split RAID extents in half
  btrfs: selftests: check for correct return value of failed lookup
  btrfs: don't use btrfs_set_item_key_safe on RAID stripe-extents
  btrfs: implement hole punching for RAID stripe extents
  btrfs: fix deletion of a range spanning parts two RAID stripe extents
  btrfs: fix tail delete of RAID stripe-extents
  btrfs: fix front delete range calculation for RAID stripe extents
  btrfs: assert RAID stripe-extent length is always greater than 0
  btrfs: don't try to delete RAID stripe-extents if we don't need to
  btrfs: selftests: correct RAID stripe-tree feature flag setting
  btrfs: add io_uring interface for encoded writes
  btrfs: remove the unused locked_folio parameter from btrfs_cleanup_ordered_extents()
  btrfs: add extra error messages for delalloc range related errors
  btrfs: subpage: dump the involved bitmap when ASSERT() failed
  btrfs: subpage: fix the bitmap dump of the locked flags
  btrfs: do proper folio cleanup when run_delalloc_nocow() failed
  ...

1 files changed, 41 insertions, 65 deletions

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 14e27473c5bc..36f51c311bb1 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -36,52 +36,7 @@
 #include "ioctl.h"
 #include "file.h"
 #include "super.h"
-
-/*
- * Helper to fault in page and copy.  This should go away and be replaced with
- * calls into generic code.
- */
-static noinline int btrfs_copy_from_user(loff_t pos, size_t write_bytes,
-					 struct folio *folio, struct iov_iter *i)
-{
-	size_t copied = 0;
-	size_t total_copied = 0;
-	int offset = offset_in_page(pos);
-
-	while (write_bytes > 0) {
-		size_t count = min_t(size_t, PAGE_SIZE - offset, write_bytes);
-		/*
-		 * Copy data from userspace to the current page
-		 */
-		copied = copy_folio_from_iter_atomic(folio, offset, count, i);
-
-		/* Flush processor's dcache for this page */
-		flush_dcache_folio(folio);
-
-		/*
-		 * if we get a partial write, we can end up with
-		 * partially up to date page.  These add
-		 * a lot of complexity, so make sure they don't
-		 * happen by forcing this copy to be retried.
-		 *
-		 * The rest of the btrfs_file_write code will fall
-		 * back to page at a time copies after we return 0.
-		 */
-		if (unlikely(copied < count)) {
-			if (!folio_test_uptodate(folio)) {
-				iov_iter_revert(i, copied);
-				copied = 0;
-			}
-			if (!copied)
-				break;
-		}
-
-		write_bytes -= copied;
-		total_copied += copied;
-		offset += copied;
-	}
-	return total_copied;
-}
+#include "print-tree.h"
 
 /*
  * Unlock folio after btrfs_file_write() is done with it.
@@ -106,7 +61,7 @@ static void btrfs_drop_folio(struct btrfs_fs_info *fs_info, struct folio *folio,
 }
 
 /*
- * After btrfs_copy_from_user(), update the following things for delalloc:
+ * After copy_folio_from_iter_atomic(), update the following things for delalloc:
  * - Mark newly dirtied folio as DELALLOC in the io tree.
  *   Used to advise which range is to be written back.
  * - Mark modified folio as Uptodate/Dirty and not needing COW fixup
@@ -224,7 +179,7 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 	if (args->drop_cache)
 		btrfs_drop_extent_map_range(inode, args->start, args->end - 1, false);
 
-	if (args->start >= inode->disk_i_size && !args->replace_extent)
+	if (data_race(args->start >= inode->disk_i_size) && !args->replace_extent)
 		modify_tree = 0;
 
 	update_refs = (btrfs_root_id(root) != BTRFS_TREE_LOG_OBJECTID);
@@ -245,7 +200,11 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 next_slot:
 		leaf = path->nodes[0];
 		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
-			BUG_ON(del_nr > 0);
+			if (WARN_ON(del_nr > 0)) {
+				btrfs_print_leaf(leaf);
+				ret = -EINVAL;
+				break;
+			}
 			ret = btrfs_next_leaf(root, path);
 			if (ret < 0)
 				break;
@@ -321,7 +280,11 @@ next_slot:
 		 *  | -------- extent -------- |
 		 */
 		if (args->start > key.offset && args->end < extent_end) {
-			BUG_ON(del_nr > 0);
+			if (WARN_ON(del_nr > 0)) {
+				btrfs_print_leaf(leaf);
+				ret = -EINVAL;
+				break;
+			}
 			if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 				ret = -EOPNOTSUPP;
 				break;
@@ -351,7 +314,6 @@ next_slot:
 			btrfs_set_file_extent_offset(leaf, fi, extent_offset);
 			btrfs_set_file_extent_num_bytes(leaf, fi,
 							extent_end - args->start);
-			btrfs_mark_buffer_dirty(trans, leaf);
 
 			if (update_refs && disk_bytenr > 0) {
 				struct btrfs_ref ref = {
@@ -397,7 +359,6 @@ next_slot:
 			btrfs_set_file_extent_offset(leaf, fi, extent_offset);
 			btrfs_set_file_extent_num_bytes(leaf, fi,
 							extent_end - args->end);
-			btrfs_mark_buffer_dirty(trans, leaf);
 			if (update_refs && disk_bytenr > 0)
 				args->bytes_found += args->end - key.offset;
 			break;
@@ -409,7 +370,11 @@ next_slot:
 		 *  | -------- extent -------- |
 		 */
 		if (args->start > key.offset && args->end >= extent_end) {
-			BUG_ON(del_nr > 0);
+			if (WARN_ON(del_nr > 0)) {
+				btrfs_print_leaf(leaf);
+				ret = -EINVAL;
+				break;
+			}
 			if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 				ret = -EOPNOTSUPP;
 				break;
@@ -417,7 +382,6 @@ next_slot:
 
 			btrfs_set_file_extent_num_bytes(leaf, fi,
 							args->start - key.offset);
-			btrfs_mark_buffer_dirty(trans, leaf);
 			if (update_refs && disk_bytenr > 0)
 				args->bytes_found += extent_end - args->start;
 			if (args->end == extent_end)
@@ -437,7 +401,11 @@ delete_extent_item:
 				del_slot = path->slots[0];
 				del_nr = 1;
 			} else {
-				BUG_ON(del_slot + del_nr != path->slots[0]);
+				if (WARN_ON(del_slot + del_nr != path->slots[0])) {
+					btrfs_print_leaf(leaf);
+					ret = -EINVAL;
+					break;
+				}
 				del_nr++;
 			}
 
@@ -668,7 +636,6 @@ again:
 							 trans->transid);
 			btrfs_set_file_extent_num_bytes(leaf, fi,
 							end - other_start);
-			btrfs_mark_buffer_dirty(trans, leaf);
 			goto out;
 		}
 	}
@@ -697,7 +664,6 @@ again:
 							other_end - start);
 			btrfs_set_file_extent_offset(leaf, fi,
 						     start - orig_offset);
-			btrfs_mark_buffer_dirty(trans, leaf);
 			goto out;
 		}
 	}
@@ -731,7 +697,6 @@ again:
 		btrfs_set_file_extent_offset(leaf, fi, split - orig_offset);
 		btrfs_set_file_extent_num_bytes(leaf, fi,
 						extent_end - split);
-		btrfs_mark_buffer_dirty(trans, leaf);
 
 		ref.action = BTRFS_ADD_DELAYED_REF;
 		ref.bytenr = bytenr;
@@ -810,7 +775,6 @@ again:
 		btrfs_set_file_extent_type(leaf, fi,
 					   BTRFS_FILE_EXTENT_REG);
 		btrfs_set_file_extent_generation(leaf, fi, trans->transid);
-		btrfs_mark_buffer_dirty(trans, leaf);
 	} else {
 		fi = btrfs_item_ptr(leaf, del_slot - 1,
 			   struct btrfs_file_extent_item);
@@ -819,7 +783,6 @@ again:
 		btrfs_set_file_extent_generation(leaf, fi, trans->transid);
 		btrfs_set_file_extent_num_bytes(leaf, fi,
 						extent_end - key.offset);
-		btrfs_mark_buffer_dirty(trans, leaf);
 
 		ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
 		if (ret < 0) {
@@ -1052,7 +1015,7 @@ int btrfs_check_nocow_lock(struct btrfs_inode *inode, loff_t pos,
 						   &cached_state);
 	}
 	ret = can_nocow_extent(&inode->vfs_inode, lockstart, &num_bytes,
-			       NULL, nowait, false);
+			       NULL, nowait);
 	if (ret <= 0)
 		btrfs_drew_write_unlock(&root->snapshot_lock);
 	else
@@ -1252,7 +1215,23 @@ again:
 			break;
 		}
 
-		copied = btrfs_copy_from_user(pos, write_bytes, folio, i);
+		copied = copy_folio_from_iter_atomic(folio,
+				offset_in_folio(folio, pos), write_bytes, i);
+		flush_dcache_folio(folio);
+
+		/*
+		 * If we get a partial write, we can end up with partially
+		 * uptodate page. Although if sector size < page size we can
+		 * handle it, but if it's not sector aligned it can cause
+		 * a lot of complexity, so make sure they don't happen by
+		 * forcing retry this copy.
+		 */
+		if (unlikely(copied < write_bytes)) {
+			if (!folio_test_uptodate(folio)) {
+				iov_iter_revert(i, copied);
+				copied = 0;
+			}
+		}
 
 		num_sectors = BTRFS_BYTES_TO_BLKS(fs_info, reserve_bytes);
 		dirty_sectors = round_up(copied + sector_offset,
@@ -2029,7 +2008,6 @@ static int fill_holes(struct btrfs_trans_handle *trans,
 		btrfs_set_file_extent_ram_bytes(leaf, fi, num_bytes);
 		btrfs_set_file_extent_offset(leaf, fi, 0);
 		btrfs_set_file_extent_generation(leaf, fi, trans->transid);
-		btrfs_mark_buffer_dirty(trans, leaf);
 		goto out;
 	}
 
@@ -2046,7 +2024,6 @@ static int fill_holes(struct btrfs_trans_handle *trans,
 		btrfs_set_file_extent_ram_bytes(leaf, fi, num_bytes);
 		btrfs_set_file_extent_offset(leaf, fi, 0);
 		btrfs_set_file_extent_generation(leaf, fi, trans->transid);
-		btrfs_mark_buffer_dirty(trans, leaf);
 		goto out;
 	}
 	btrfs_release_path(path);
@@ -2194,7 +2171,6 @@ static int btrfs_insert_replace_extent(struct btrfs_trans_handle *trans,
 	btrfs_set_file_extent_num_bytes(leaf, extent, replace_len);
 	if (extent_info->is_new_extent)
 		btrfs_set_file_extent_generation(leaf, extent, trans->transid);
-	btrfs_mark_buffer_dirty(trans, leaf);
 	btrfs_release_path(path);
 
 	ret = btrfs_inode_set_file_extent_range(inode, extent_info->file_offset,