[GIT PULL] Btrfs updates for 5.1, part 1

From: David Sterba
Date: Mon Mar 04 2019 - 14:20:17 EST

Next message: Logan Gunthorpe: "Re: [PATCH 1/2] PCI: Prevent 64-bit resources from being counted in 32-bit bridge region"
Previous message: Dave Hansen: "Re: [PATCH v4 07/17] x86/split_lock: Enumerate #AC for split lock by MSR IA32_CORE_CAPABILITY"
Next in thread: pr-tracker-bot: "Re: [GIT PULL] Btrfs updates for 5.1, part 1"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hi,

the branch contains usual mix of new features, core changes and fixes;
full list below. I'm planning 2nd pull request, with a few more fixes
that arrived recently but too close to merge window, will send it next
week.

Please pull, thanks.

New features:

- support zstd compression levels

- new ioctl to unregister a device from the module (ie. reverse of device
scan)

- scrub prints a message to log when it's about to start or finish

Core changes:

- qgroups can now skip part of a tree that does not get updated during
relocation, because this does not affect the quota accounting, estimated
speedup in run time is about 20%

- the compression workspace management had to be enhanced due to zstd
requirements

- various enospc fixes, when there's high fragmentation the
over-reservation can cause ENOSPC that might not happen after a flush,
in such cases try to wait if the situation improves

Fixes:

- various ioctls could overwrite previous return value if copy_to_user
fails, fix this so the original error is reported

- more reclaim vs GFP_KERNEL fixes

- other cleanups and refactoring

- fix a (valid) lockdep warning in a test when device replace is
destroying worker threads

- make qgroup async transaction commit more aggressive, this avoids some
'quota limit reached' errors if there are not enough data to trigger
transaction in order to flush

- fix deadlock between snapshot deletion and quotas when backref walking
is called from context that already holds the same locks

- fsync fixes:
- fix fsync after succession of renames of different files
- fix fsync after succession of renames and unlink/rmdir

----------------------------------------------------------------
The following changes since commit 5908e6b738e3357af42c10e1183753c70a0117a9:

Linux 5.0-rc8 (2019-02-24 16:46:45 -0800)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux.git for-5.1-part1-tag

for you to fetch changes up to f65e25e343cfc0e6f4db9a687c4085fad268325d:

btrfs: Remove unnecessary casts in btrfs_read_root_item (2019-02-25 14:19:23 +0100)

----------------------------------------------------------------
Anand Jain (11):
btrfs: merge btrfs_find_device_missing_or_by_path() into parent
btrfs: cleanup btrfs_find_device_by_devspec()
btrfs: refactor btrfs_find_device() take fs_devices as argument
btrfs: refactor btrfs_free_stale_devices() to get return value
btrfs: merge btrfs_find_device and find_device
btrfs: scrub: print messages when started or finished
btrfs: introduce new ioctl to unregister a btrfs device
btrfs: fix comment its device list mutex not volume lock
btrfs: scrub: fix circular locking dependency warning
btrfs: scrub: add scrub_lock lockdep check in scrub_workers_get
btrfs: scrub: convert scrub_workers_refcnt to refcount_t

Anders Roxell (1):
btrfs: let the assertion expression compile in all configs

Dan Carpenter (1):
btrfs: drop the lock on error in btrfs_dev_replace_cancel

Dan Robertson (1):
btrfs: init csum_list before possible free

David Sterba (9):
btrfs: simplify workqueue name when allocating
btrfs: split btrfs_set_lock_blocking_rw to read and write helpers
btrfs: split btrfs_clear_lock_blocking_rw to read and write helpers
btrfs: replace btrfs_set_lock_blocking_rw with appropriate helpers
btrfs: open code now trivial btrfs_set_lock_blocking
btrfs: simplify waiting loop in btrfs_tree_lock
btrfs: merge btrfs_set_lock_blocking_rw with it's caller
btrfs: scrub: add assertions for worker pointers
btrfs: scrub: remove unused nocow worker pointer

Dennis Zhou (12):
btrfs: add helpers for compression type and level
btrfs: rename workspaces_list to workspace_manager
btrfs: manage heuristic workspace as index 0
btrfs: unify compression ops with workspace_manager
btrfs: add helper methods for workspace manager init and cleanup
btrfs: add compression interface in (get/put)_workspace
btrfs: move to function pointers for get/put workspaces
btrfs: plumb level through the compression interface
btrfs: change set_level() to bound the level passed in
btrfs: zstd use the passed through level instead of default
btrfs: make zstd memory requirements monotonic
btrfs: add zstd compression level support

Filipe Manana (14):
Btrfs: do not overwrite scrub error with fault error in scrub ioctl
Btrfs: do not overwrite error return value in scrub progress ioctl
Btrfs: do not overwrite error return value in the get device stats ioctl
Btrfs: setup a nofs context for memory allocation at btrfs_create_tree()
Btrfs: setup a nofs context for memory allocation at __btrfs_set_acl
Btrfs: remove redundant check for swapfiles when reflinking
Btrfs: do not overwrite error return value in the device replace ioctl
Btrfs: do not overwrite error return value in the balance ioctl
Btrfs: move duplicated nodatasum check into common reflink/dedupe helper
Btrfs: fix fsync after succession of renames of different files
Btrfs: fix fsync after succession of renames and unlink/rmdir
Btrfs: remove no longer needed range length checks for deduplication
Btrfs: add missing error handling after doing leaf/node binary search
Btrfs: remove assertion when searching for a key in a node/leaf

Johannes Thumshirn (1):
btrfs: ensure that a DUP or RAID1 block group has exactly two stripes

Josef Bacik (12):
btrfs: make btrfs_destroy_delayed_refs use btrfs_delayed_ref_lock
btrfs: make btrfs_destroy_delayed_refs use btrfs_delete_ref_head
btrfs: call btrfs_create_pending_block_groups unconditionally
btrfs: replace cleaner_delayed_iput_mutex with a waitqueue
btrfs: check if there are free block groups for commit
btrfs: dump block_rsv details when dumping space info
btrfs: don't use global reserve for chunk allocation
btrfs: don't enospc all tickets on flush failure
btrfs: loop in inode_rsv_refill
btrfs: be more explicit about allowed flush states
btrfs: reserve extra space during evict
btrfs: honor path->skip_locking in backref code

Julia Lawall (1):
Btrfs: drop useless LIST_HEAD in merge_reloc_root

Nikolay Borisov (17):
btrfs: Make first argument of btrfs_run_delalloc_range directly an inode
btrfs: Remove unused arguments from btrfs_get_extent_fiemap
btrfs: Refactor btrfs_get_extent_fiemap
btrfs: Remove redundant assignment in btrfs_get_extent_fiemap
btrfs: Remove inode argument from async_cow_submit
btrfs: Remove isize local variable in compress_file_range
btrfs: Use ihold instead of igrab in cow_file_range_async
btrfs: Remove WARN_ON in btrfs_alloc_delalloc_work
btrfs: Document logic regarding inode in async_cow_submit
btrfs: Refactor shrink_delalloc
btrfs: Remove impossible condition from mergable_maps
btrfs: Rename found_type to extent_type in btrfs_get_extent
btrfs: Consolidate retval checking of core btree functions
btrfs: Remove not_found_em label from btrfs_get_extent
btrfs: use WARN_ON in a canonical form btrfs_remove_block_group
btrfs: Remove EXTENT_FIRST_DELALLOC bit
btrfs: Fix grossly misleading argument names in extent io search

Qu Wenruo (10):
btrfs: relocation: Delay reloc tree deletion after merge_reloc_roots
btrfs: qgroup: Refactor btrfs_qgroup_trace_subtree_swap
btrfs: qgroup: Introduce per-root swapped blocks infrastructure
btrfs: qgroup: Use delayed subtree rescan for balance
btrfs: qgroup: Cleanup old subtree swap code
btrfs: qgroup: Remove duplicated trace points for qgroup_rsv_add/release
btrfs: Output ENOSPC debug info in inc_block_group_ro
btrfs: extent_io: Kill the forward declaration of flush_write_bio
btrfs: qgroup: Move reserved data accounting from btrfs_delayed_ref_head to btrfs_qgroup_extent_record
btrfs: qgroup: Make qgroup async transaction commit more aggressive

YueHaibing (2):
btrfs: remove set but not used variable 'num_pages'
btrfs: Remove unnecessary casts in btrfs_read_root_item

fs/btrfs/acl.c | 9 ++
fs/btrfs/async-thread.c | 10 +-
fs/btrfs/backref.c | 22 ++-
fs/btrfs/compression.c | 253 ++++++++++++++---------------
fs/btrfs/compression.h | 52 +++++-
fs/btrfs/ctree.c | 74 ++++-----
fs/btrfs/ctree.h | 61 +++++--
fs/btrfs/delayed-ref.c | 15 +-
fs/btrfs/delayed-ref.h | 11 --
fs/btrfs/dev-replace.c | 9 +-
fs/btrfs/disk-io.c | 39 +++--
fs/btrfs/extent-tree.c | 292 +++++++++++++++++++++------------
fs/btrfs/extent_io.c | 87 +++++-----
fs/btrfs/extent_io.h | 15 +-
fs/btrfs/extent_map.c | 5 +-
fs/btrfs/extent_map.h | 1 +
fs/btrfs/file.c | 3 +-
fs/btrfs/inode.c | 207 +++++++++++++++---------
fs/btrfs/ioctl.c | 60 ++-----
fs/btrfs/locking.c | 108 ++++++-------
fs/btrfs/locking.h | 15 +-
fs/btrfs/lzo.c | 31 +++-
fs/btrfs/qgroup.c | 372 ++++++++++++++++++++++++++++++++-----------
fs/btrfs/qgroup.h | 120 ++++++++++++--
fs/btrfs/ref-verify.c | 4 +-
fs/btrfs/relocation.c | 119 ++++++++++----
fs/btrfs/root-tree.c | 4 +-
fs/btrfs/scrub.c | 49 ++++--
fs/btrfs/super.c | 13 +-
fs/btrfs/transaction.c | 9 +-
fs/btrfs/tree-defrag.c | 2 +-
fs/btrfs/tree-log.c | 282 ++++++++++++++++++++++++++------
fs/btrfs/volumes.c | 202 ++++++++++++-----------
fs/btrfs/volumes.h | 5 +-
fs/btrfs/zlib.c | 45 +++++-
fs/btrfs/zstd.c | 316 ++++++++++++++++++++++++++++++++++--
include/trace/events/btrfs.h | 30 +---
include/uapi/linux/btrfs.h | 2 +
38 files changed, 1993 insertions(+), 960 deletions(-)

Next message: Logan Gunthorpe: "Re: [PATCH 1/2] PCI: Prevent 64-bit resources from being counted in 32-bit bridge region"
Previous message: Dave Hansen: "Re: [PATCH v4 07/17] x86/split_lock: Enumerate #AC for split lock by MSR IA32_CORE_CAPABILITY"
Next in thread: pr-tracker-bot: "Re: [GIT PULL] Btrfs updates for 5.1, part 1"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]