Merge tag 'for-6.18-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs updates from David Sterba:
 "There are no new features, the changes are in the core code, notably
  tree-log error handling and reporting improvements, and initial
  support for block size > page size.

  Performance improvements:

   - search data checksums in the commit root (previous transaction) to
     avoid locking contention, this improves parallelism of read
     heavy/low write workloads, and also reduces transaction commit
     time; on real and reproducer workload the sync time went from
     minutes to tens of seconds (workload and numbers are in the
     changelog)

  Core:

   - tree-log updates:
      - error handling improvements, transaction aborts
      - add new error state 'O' (printed in status messages) when log
        replay fails and is aborted
      - reduced number of btrfs_path allocations when traversing the
        tree

   - 'block size > page size' support
      - basic implementation with limitations, under experimental build
      - limitations: no direct io, raid56, encoded read (standalone and
        in send ioctl), encoded write
      - preparatory work for compression, removing implicit assumptions
        of page and block sizes
      - compression workspaces are now per-filesystem, we cannot assume
        common block size for work memory among different filesystems

   - tree-checker now verifies INODE_EXTREF item (which is implementing
     hardlinks)

   - tree leaf pretty printer updates, there were missing data from
     items, keys/items

   - move config option CONFIG_BTRFS_REF_VERIFY to CONFIG_BTRFS_DEBUG,
     it's a debugging feature and not needed to be enabled separately

   - more struct btrfs_path auto free updates

   - use ref_tracker API for tracking delayed inodes, enabled by mount
     option 'ref_verify', allowing to better pinpoint leaking references

   - in zoned mode, avoid selecting data relocation zoned for ordinary
     data block groups

   - updated and enhanced error messages

   - lots of cleanups and refactoring"

* tag 'for-6.18-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (113 commits)
  btrfs: use smp_mb__after_atomic() when forcing COW in create_pending_snapshot()
  btrfs: add unlikely annotations to branches leading to transaction abort
  btrfs: add unlikely annotations to branches leading to EIO
  btrfs: add unlikely annotations to branches leading to EUCLEAN
  btrfs: more trivial BTRFS_PATH_AUTO_FREE conversions
  btrfs: zoned: don't fail mount needlessly due to too many active zones
  btrfs: use kmalloc_array() for open-coded arithmetic in kmalloc()
  btrfs: enable experimental bs > ps support
  btrfs: add extra ASSERT()s to catch unaligned bios
  btrfs: fix symbolic link reading when bs > ps
  btrfs: prepare scrub to support bs > ps cases
  btrfs: prepare zlib to support bs > ps cases
  btrfs: prepare lzo to support bs > ps cases
  btrfs: prepare zstd to support bs > ps cases
  btrfs: prepare compression folio alloc/free for bs > ps cases
  btrfs: fix the incorrect max_bytes value for find_lock_delalloc_range()
  btrfs: remove pointless key offset setup in create_pending_snapshot()
  btrfs: annotate btrfs_is_testing() as unlikely and make it return bool
  btrfs: make the rule checking more readable for should_cow_block()
  btrfs: simplify inline extent end calculation at replay_one_extent()
  ...

1 files changed, 5 insertions, 7 deletions

diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c
index 4675bcd5f92e..a4eaef60549e 100644
--- a/fs/btrfs/dev-replace.c
+++ b/fs/btrfs/dev-replace.c
@@ -98,7 +98,7 @@ no_valid_dev_replace_entry_found:
 		 * We don't have a replace item or it's corrupted.  If there is
 		 * a replace target, fail the mount.
 		 */
-		if (btrfs_find_device(fs_info->fs_devices, &args)) {
+		if (unlikely(btrfs_find_device(fs_info->fs_devices, &args))) {
 			btrfs_err(fs_info,
 			"found replace target device without a valid replace item");
 			return -EUCLEAN;
@@ -158,7 +158,7 @@ no_valid_dev_replace_entry_found:
 		 * We don't have an active replace item but if there is a
 		 * replace target, fail the mount.
 		 */
-		if (btrfs_find_device(fs_info->fs_devices, &args)) {
+		if (unlikely(btrfs_find_device(fs_info->fs_devices, &args))) {
 			btrfs_err(fs_info,
 "replace without active item, run 'device scan --forget' on the target device");
 			ret = -EUCLEAN;
@@ -177,8 +177,7 @@ no_valid_dev_replace_entry_found:
 		 * allow 'btrfs dev replace_cancel' if src/tgt device is
 		 * missing
 		 */
-		if (!dev_replace->srcdev &&
-		    !btrfs_test_opt(fs_info, DEGRADED)) {
+		if (unlikely(!dev_replace->srcdev && !btrfs_test_opt(fs_info, DEGRADED))) {
 			ret = -EIO;
 			btrfs_warn(fs_info,
 			   "cannot mount because device replace operation is ongoing and");
@@ -186,8 +185,7 @@ no_valid_dev_replace_entry_found:
 			   "srcdev (devid %llu) is missing, need to run 'btrfs dev scan'?",
 			   src_devid);
 		}
-		if (!dev_replace->tgtdev &&
-		    !btrfs_test_opt(fs_info, DEGRADED)) {
+		if (unlikely(!dev_replace->tgtdev && !btrfs_test_opt(fs_info, DEGRADED))) {
 			ret = -EIO;
 			btrfs_warn(fs_info,
 			   "cannot mount because device replace operation is ongoing and");
@@ -637,7 +635,7 @@ static int btrfs_dev_replace_start(struct btrfs_fs_info *fs_info,
 		break;
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED:
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED:
-		DEBUG_WARN("unexpected STARTED ot SUSPENDED dev-replace state");
+		DEBUG_WARN("unexpected STARTED or SUSPENDED dev-replace state");
 		ret = BTRFS_IOCTL_DEV_REPLACE_RESULT_ALREADY_STARTED;
 		up_write(&dev_replace->rwsem);
 		goto leave;