Merge tag 'for-6.12-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs updates from David Sterba:
 "This brings mostly refactoring, cleanups, minor performance
  optimizations and usual fixes. The folio API conversions are most
  noticeable.

  There's one less visible change that could have a high impact. The
  extent lock scope for read is reduced, not held for the entire
  operation. In the buffered read case it's left to page or inode lock,
  some direct io read synchronization is still needed.

  This used to prevent deadlocks induced by page faults during direct
  io, so there was a 4K limitation on the requests, e.g. for io_uring.
  In the future this will allow smoother integration with iomap where
  the extent read lock was a major obstacle.

  User visible changes:

   - the FSTRIM ioctl updates the processed range even after an error or
     interruption

   - cleaner thread is woken up in SYNC ioctl instead of waking the
     transaction thread that can take some delay before waking up the
     cleaner, this can speed up cleaning of deleted subvolumes

   - print an error message when opening a device fail, e.g. when it's
     unexpectedly read-only

  Core changes:

   - improved extent map handling in various ways (locking, iteration, ...)

   - new assertions and locking annotations

   - raid-stripe-tree locking fixes

   - use xarray for tracking dirty qgroup extents, switched from rb-tree

   - turn the subpage test to compile-time condition if possible (e.g.
     on x86_64 with 4K pages), this allows to skip a lot of ifs and
     remove dead code

   - more preparatory work for compression in subpage mode

  Cleanups and refactoring

   - folio API conversions, many simple cases where page is passed so
     switch it to folios

   - more subpage code refactoring, update page state bitmap processing

   - introduce auto free for btrfs_path structure, use for the simple
     cases"

* tag 'for-6.12-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (110 commits)
  btrfs: only unlock the to-be-submitted ranges inside a folio
  btrfs: merge btrfs_folio_unlock_writer() into btrfs_folio_end_writer_lock()
  btrfs: BTRFS_PATH_AUTO_FREE in orphan.c
  btrfs: use btrfs_path auto free in zoned.c
  btrfs: DEFINE_FREE for struct btrfs_path
  btrfs: remove btrfs_folio_end_all_writers()
  btrfs: constify more pointer parameters
  btrfs: rework BTRFS_I as macro to preserve parameter const
  btrfs: add and use helper to verify the calling task has locked the inode
  btrfs: always update fstrim_range on failure in FITRIM ioctl
  btrfs: convert copy_inline_to_page() to use folio
  btrfs: convert btrfs_decompress() to take a folio
  btrfs: convert zstd_decompress() to take a folio
  btrfs: convert lzo_decompress() to take a folio
  btrfs: convert zlib_decompress() to take a folio
  btrfs: convert try_release_extent_mapping() to take a folio
  btrfs: convert try_release_extent_state() to take a folio
  btrfs: convert submit_eb_page() to take a folio
  btrfs: convert submit_eb_subpage() to take a folio
  btrfs: convert read_key_bytes() to take a folio
  ...

1 files changed, 25 insertions, 29 deletions

diff --git a/fs/btrfs/bio.c b/fs/btrfs/bio.c
index b4e31ae17cd9..ce13416bc10f 100644
--- a/fs/btrfs/bio.c
+++ b/fs/btrfs/bio.c
@@ -53,7 +53,7 @@ void btrfs_bio_init(struct btrfs_bio *bbio, struct btrfs_fs_info *fs_info,
 
 /*
  * Allocate a btrfs_bio structure.  The btrfs_bio is the main I/O container for
- * btrfs, and is used for all I/O submitted through btrfs_submit_bio.
+ * btrfs, and is used for all I/O submitted through btrfs_submit_bbio().
  *
  * Just like the underlying bio_alloc_bioset it will not fail as it is backed by
  * a mempool.
@@ -120,12 +120,6 @@ static void __btrfs_bio_end_io(struct btrfs_bio *bbio)
 	}
 }
 
-void btrfs_bio_end_io(struct btrfs_bio *bbio, blk_status_t status)
-{
-	bbio->bio.bi_status = status;
-	__btrfs_bio_end_io(bbio);
-}
-
 static void btrfs_orig_write_end_io(struct bio *bio);
 
 static void btrfs_bbio_propagate_error(struct btrfs_bio *bbio,
@@ -147,8 +141,9 @@ static void btrfs_bbio_propagate_error(struct btrfs_bio *bbio,
 	}
 }
 
-static void btrfs_orig_bbio_end_io(struct btrfs_bio *bbio)
+void btrfs_bio_end_io(struct btrfs_bio *bbio, blk_status_t status)
 {
+	bbio->bio.bi_status = status;
 	if (bbio->bio.bi_pool == &btrfs_clone_bioset) {
 		struct btrfs_bio *orig_bbio = bbio->private;
 
@@ -179,7 +174,7 @@ static int prev_repair_mirror(struct btrfs_failed_bio *fbio, int cur_mirror)
 static void btrfs_repair_done(struct btrfs_failed_bio *fbio)
 {
 	if (atomic_dec_and_test(&fbio->repair_count)) {
-		btrfs_orig_bbio_end_io(fbio->bbio);
+		btrfs_bio_end_io(fbio->bbio, fbio->bbio->bio.bi_status);
 		mempool_free(fbio, &btrfs_failed_bio_pool);
 	}
 }
@@ -211,7 +206,7 @@ static void btrfs_end_repair_bio(struct btrfs_bio *repair_bbio,
 			goto done;
 		}
 
-		btrfs_submit_bio(repair_bbio, mirror);
+		btrfs_submit_bbio(repair_bbio, mirror);
 		return;
 	}
 
@@ -280,7 +275,7 @@ static struct btrfs_failed_bio *repair_one_sector(struct btrfs_bio *failed_bbio,
 
 	mirror = next_repair_mirror(fbio, failed_bbio->mirror_num);
 	btrfs_debug(fs_info, "submitting repair read to mirror %d", mirror);
-	btrfs_submit_bio(repair_bbio, mirror);
+	btrfs_submit_bbio(repair_bbio, mirror);
 	return fbio;
 }
 
@@ -326,7 +321,7 @@ static void btrfs_check_read_bio(struct btrfs_bio *bbio, struct btrfs_device *de
 	if (fbio)
 		btrfs_repair_done(fbio);
 	else
-		btrfs_orig_bbio_end_io(bbio);
+		btrfs_bio_end_io(bbio, bbio->bio.bi_status);
 }
 
 static void btrfs_log_dev_io_error(struct bio *bio, struct btrfs_device *dev)
@@ -360,7 +355,7 @@ static void btrfs_end_bio_work(struct work_struct *work)
 	if (is_data_bbio(bbio))
 		btrfs_check_read_bio(bbio, bbio->bio.bi_private);
 	else
-		btrfs_orig_bbio_end_io(bbio);
+		btrfs_bio_end_io(bbio, bbio->bio.bi_status);
 }
 
 static void btrfs_simple_end_io(struct bio *bio)
@@ -380,7 +375,7 @@ static void btrfs_simple_end_io(struct bio *bio)
 	} else {
 		if (bio_op(bio) == REQ_OP_ZONE_APPEND && !bio->bi_status)
 			btrfs_record_physical_zoned(bbio);
-		btrfs_orig_bbio_end_io(bbio);
+		btrfs_bio_end_io(bbio, bbio->bio.bi_status);
 	}
 }
 
@@ -394,7 +389,7 @@ static void btrfs_raid56_end_io(struct bio *bio)
 	if (bio_op(bio) == REQ_OP_READ && is_data_bbio(bbio))
 		btrfs_check_read_bio(bbio, NULL);
 	else
-		btrfs_orig_bbio_end_io(bbio);
+		btrfs_bio_end_io(bbio, bbio->bio.bi_status);
 
 	btrfs_put_bioc(bioc);
 }
@@ -424,7 +419,7 @@ static void btrfs_orig_write_end_io(struct bio *bio)
 	if (bio_op(bio) == REQ_OP_ZONE_APPEND && !bio->bi_status)
 		stripe->physical = bio->bi_iter.bi_sector << SECTOR_SHIFT;
 
-	btrfs_orig_bbio_end_io(bbio);
+	btrfs_bio_end_io(bbio, bbio->bio.bi_status);
 	btrfs_put_bioc(bioc);
 }
 
@@ -502,8 +497,8 @@ static void btrfs_submit_mirrored_bio(struct btrfs_io_context *bioc, int dev_nr)
 	btrfs_submit_dev_bio(bioc->stripes[dev_nr].dev, bio);
 }
 
-static void __btrfs_submit_bio(struct bio *bio, struct btrfs_io_context *bioc,
-			       struct btrfs_io_stripe *smap, int mirror_num)
+static void btrfs_submit_bio(struct bio *bio, struct btrfs_io_context *bioc,
+			     struct btrfs_io_stripe *smap, int mirror_num)
 {
 	if (!bioc) {
 		/* Single mirror read/write fast path. */
@@ -593,7 +588,7 @@ static void run_one_async_done(struct btrfs_work *work, bool do_free)
 
 	/* If an error occurred we just want to clean up the bio and move on. */
 	if (bio->bi_status) {
-		btrfs_orig_bbio_end_io(async->bbio);
+		btrfs_bio_end_io(async->bbio, async->bbio->bio.bi_status);
 		return;
 	}
 
@@ -603,7 +598,7 @@ static void run_one_async_done(struct btrfs_work *work, bool do_free)
 	 * context.  This changes nothing when cgroups aren't in use.
 	 */
 	bio->bi_opf |= REQ_BTRFS_CGROUP_PUNT;
-	__btrfs_submit_bio(bio, async->bioc, &async->smap, async->mirror_num);
+	btrfs_submit_bio(bio, async->bioc, &async->smap, async->mirror_num);
 }
 
 static bool should_async_write(struct btrfs_bio *bbio)
@@ -678,7 +673,10 @@ static bool btrfs_submit_chunk(struct btrfs_bio *bbio, int mirror_num)
 	blk_status_t ret;
 	int error;
 
-	smap.is_scrub = !bbio->inode;
+	if (!bbio->inode || btrfs_is_data_reloc_root(inode->root))
+		smap.rst_search_commit_root = true;
+	else
+		smap.rst_search_commit_root = false;
 
 	btrfs_bio_counter_inc_blocked(fs_info);
 	error = btrfs_map_block(fs_info, btrfs_op(bio), logical, &map_length,
@@ -749,7 +747,7 @@ static bool btrfs_submit_chunk(struct btrfs_bio *bbio, int mirror_num)
 		}
 	}
 
-	__btrfs_submit_bio(bio, bioc, &smap, mirror_num);
+	btrfs_submit_bio(bio, bioc, &smap, mirror_num);
 done:
 	return map_length == length;
 
@@ -765,16 +763,14 @@ fail:
 		ASSERT(bbio->bio.bi_pool == &btrfs_clone_bioset);
 		ASSERT(remaining);
 
-		remaining->bio.bi_status = ret;
-		btrfs_orig_bbio_end_io(remaining);
+		btrfs_bio_end_io(remaining, ret);
 	}
-	bbio->bio.bi_status = ret;
-	btrfs_orig_bbio_end_io(bbio);
+	btrfs_bio_end_io(bbio, ret);
 	/* Do not submit another chunk */
 	return true;
 }
 
-void btrfs_submit_bio(struct btrfs_bio *bbio, int mirror_num)
+void btrfs_submit_bbio(struct btrfs_bio *bbio, int mirror_num)
 {
 	/* If bbio->inode is not populated, its file_offset must be 0. */
 	ASSERT(bbio->inode || bbio->file_offset == 0);
@@ -786,7 +782,7 @@ void btrfs_submit_bio(struct btrfs_bio *bbio, int mirror_num)
 /*
  * Submit a repair write.
  *
- * This bypasses btrfs_submit_bio deliberately, as that writes all copies in a
+ * This bypasses btrfs_submit_bbio() deliberately, as that writes all copies in a
  * RAID setup.  Here we only want to write the one bad copy, so we do the
  * mapping ourselves and submit the bio directly.
  *
@@ -875,7 +871,7 @@ void btrfs_submit_repair_write(struct btrfs_bio *bbio, int mirror_num, bool dev_
 		ASSERT(smap.dev == fs_info->dev_replace.srcdev);
 		smap.dev = fs_info->dev_replace.tgtdev;
 	}
-	__btrfs_submit_bio(&bbio->bio, NULL, &smap, mirror_num);
+	btrfs_submit_bio(&bbio->bio, NULL, &smap, mirror_num);
 	return;
 
 fail: