Hey Linus,
/* Summary */
This contains changes the changes for files for this cycle:
- Introduce a new reference counting mechanism for files.
As atomic_inc_not_zero() is implemented with a try_cmpxchg() loop it
has O(N^2) behaviour under contention with N concurrent operations and
it is in a hot path in __fget_files_rcu().
The rcuref infrastructures remedies this problem by using an
unconditional increment relying on safe- and dead zones to make this
work and requiring rcu protection for the data structure in question.
This not just scales better it also introduces overflow protection.
However, in contrast to generic rcuref, files require a memory barrier
and thus cannot rely on *_relaxed() atomic operations and also require
to be built on atomic_long_t as having massive amounts of reference
isn't unheard of even if it is just an attack.
This adds a file specific variant instead of making this a generic
library.
This has been tested by various people and it gives consistent
improvement up to 3-5% on workloads with loads of threads.
- Add a fastpath for find_next_zero_bit(). Skip 2-levels searching via
find_next_zero_bit() when there is a free slot in the word that contains
the next fd. This improves pts/blogbench-1.1.0 read by 8% and write by
4% on Intel ICX 160.
- Conditionally clear full_fds_bits since it's very likely that a bit
in full_fds_bits has been cleared during __clear_open_fds(). This
improves pts/blogbench-1.1.0 read up to 13%, and write up to 5% on
Intel ICX 160.
- Get rid of all lookup_*_fdget_rcu() variants. They were used to lookup
files without taking a reference count. That became invalid once files
were switched to SLAB_TYPESAFE_BY_RCU and now we're always taking a
reference count. Switch to an already existing helper and remove the
legacy variants.
- Remove pointless includes of <linux/fdtable.h>.
- Avoid cmpxchg() in close_files() as nobody else has a reference to the
files_struct at that point.
- Move close_range() into fs/file.c and fold __close_range() into it.
- Cleanup calling conventions of alloc_fdtable() and expand_files().
- Merge __{set,clear}_close_on_exec() into one.
- Make __set_open_fd() set cloexec as well instead of doing it
in two separate steps.
/* Testing */
gcc version 14.2.0 (Debian 14.2.0-6)
Debian clang version 16.0.6 (27+b1)
All patches are based on v6.12-rc2 and have been sitting in linux-next.
No build failures or warnings were observed.
/* Conflicts */
Merge conflicts with mainline
=============================
No known conflicts.
Merge conflicts with other trees
================================
No known conflicts.
The following changes since commit 8cf0b93919e13d1e8d4466eb4080a4c4d9d66d7b:
Linux 6.12-rc2 (2024-10-06 15:32:27 -0700)
are available in the Git repository at:
git@xxxxxxxxxxxxxxxxxxx:pub/scm/linux/kernel/git/vfs/vfs tags/vfs-6.13.file
for you to fetch changes up to aab154a442f9ba2a08fc130dbc8d178a33e10345:
selftests: add file SLAB_TYPESAFE_BY_RCU recycling stressor (2024-10-30 09:58:02 +0100)
Please consider pulling these changes from the signed vfs-6.13.file tag.
Thanks!
Christian
----------------------------------------------------------------
vfs-6.13.file
----------------------------------------------------------------
Al Viro (8):
get rid of ...lookup...fdget_rcu() family
remove pointless includes of <linux/fdtable.h>
close_files(): don't bother with xchg()
move close_range(2) into fs/file.c, fold __close_range() into it
alloc_fdtable(): change calling conventions.
file.c: merge __{set,clear}_close_on_exec()
make __set_open_fd() set cloexec state as well
expand_files(): simplify calling conventions
Christian Brauner (6):
fs: protect backing files with rcu
fs: add file_ref
fs: port files to file_ref
Merge patch series "fs: introduce file_ref_t"
Merge branch 'work.fdtable' into vfs.file
selftests: add file SLAB_TYPESAFE_BY_RCU recycling stressor
Yu Ma (3):
fs/file.c: remove sanity_check and add likely/unlikely in alloc_fd()
fs/file.c: conditionally clear full_fds
fs/file.c: add fast path in find_next_fd()
arch/powerpc/platforms/cell/spufs/coredump.c | 4 +-
drivers/gpu/drm/i915/gt/shmem_utils.c | 2 +-
drivers/gpu/drm/vmwgfx/ttm_object.c | 2 +-
fs/eventpoll.c | 2 +-
fs/fcntl.c | 1 -
fs/file.c | 281 +++++++++++----------
fs/file_table.c | 50 +++-
fs/gfs2/glock.c | 12 +-
fs/notify/dnotify/dnotify.c | 5 +-
fs/notify/fanotify/fanotify.c | 1 -
fs/notify/fanotify/fanotify_user.c | 1 -
fs/open.c | 17 --
fs/overlayfs/copy_up.c | 1 -
fs/proc/base.c | 1 -
fs/proc/fd.c | 12 +-
include/linux/fdtable.h | 5 -
include/linux/file.h | 1 +
include/linux/file_ref.h | 177 +++++++++++++
include/linux/fs.h | 10 +-
io_uring/io_uring.c | 1 -
kernel/bpf/bpf_inode_storage.c | 1 -
kernel/bpf/bpf_task_storage.c | 1 -
kernel/bpf/task_iter.c | 6 +-
kernel/bpf/token.c | 1 -
kernel/exit.c | 1 -
kernel/kcmp.c | 4 +-
kernel/module/dups.c | 1 -
kernel/module/kmod.c | 1 -
kernel/umh.c | 1 -
net/handshake/request.c | 1 -
security/apparmor/domain.c | 1 -
tools/testing/selftests/filesystems/.gitignore | 1 +
tools/testing/selftests/filesystems/Makefile | 2 +-
.../testing/selftests/filesystems/file_stressor.c | 194 ++++++++++++++
34 files changed, 576 insertions(+), 226 deletions(-)
create mode 100644 include/linux/file_ref.h
create mode 100644 tools/testing/selftests/filesystems/file_stressor.c
