]> Git Repo - linux.git/log
linux.git
3 months agoMerge tag 'for-6.13-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Linus Torvalds [Tue, 19 Nov 2024 00:37:41 +0000 (16:37 -0800)]
Merge tag 'for-6.13-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs updates from David Sterba:
 "Changes outside of btrfs: add io_uring command flag to track a dying
  task (the rest will go via the block git tree).

  User visible changes:

   - wire encoded read (ioctl) to io_uring commands, this can be used on
     itself, in the future this will allow 'send' to be asynchronous. As
     a consequence, the encoded read ioctl can also work in non-blocking
     mode

   - new ioctl to wait for cleaned subvolumes, no need to use the
     generic and root-only SEARCH_TREE ioctl, will be used by "btrfs
     subvol sync"

   - recognize different paths/symlinks for the same devices and don't
     report them during rescanning, this can be observed with LVM or DM

   - seeding device use case change, the sprout device (the one
     capturing new writes) will not clear the read-only status of the
     super block; this prevents accumulating space from deleted
     snapshots

  Performance improvements:

   - reduce lock contention when traversing extent buffers

   - reduce extent tree lock contention when searching for inline
     backref

   - switch from rb-trees to xarray for delayed ref tracking,
     improvements due to better cache locality, branching factors and
     more compact data structures

   - enable extent map shrinker again (prevent memory exhaustion under
     some types of IO load), reworked to run in a single worker thread
     (there used to be problems causing long stalls under memory
     pressure)

  Core changes:

   - raid-stripe-tree feature updates:
       - make device replace and scrub work
       - implement partial deletion of stripe extents
       - new selftests

   - split the config option BTRFS_DEBUG and add EXPERIMENTAL for
     features that are experimental or with known problems so we don't
     misuse debugging config for that

   - subpage mode updates (sector < page):
       - update compression implementations
       - update writepage, writeback

   - continued folio API conversions:
       - buffered writes

   - make buffered write copy one page at a time, preparatory work for
     future integration with large folios, may cause performance drop

   - proper locking of root item regarding starting send

   - error handling improvements

   - code cleanups and refactoring:
       - dead code removal
       - unused parameter reduction
       - lockdep assertions"

* tag 'for-6.13-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (119 commits)
  btrfs: send: check for read-only send root under critical section
  btrfs: send: check for dead send root under critical section
  btrfs: remove check for NULL fs_info at btrfs_folio_end_lock_bitmap()
  btrfs: fix warning on PTR_ERR() against NULL device at btrfs_control_ioctl()
  btrfs: fix a typo in btrfs_use_zone_append
  btrfs: avoid superfluous calls to free_extent_map() in btrfs_encoded_read()
  btrfs: simplify logic to decrement snapshot counter at btrfs_mksnapshot()
  btrfs: remove hole from struct btrfs_delayed_node
  btrfs: update stale comment for struct btrfs_delayed_ref_node::add_list
  btrfs: add new ioctl to wait for cleaned subvolumes
  btrfs: simplify range tracking in cow_file_range()
  btrfs: remove conditional path allocation in btrfs_read_locked_inode()
  btrfs: push cleanup into btrfs_read_locked_inode()
  io_uring/cmd: let cmds to know about dying task
  btrfs: add struct io_btrfs_cmd as type for io_uring_cmd_to_pdu()
  btrfs: add io_uring command for encoded reads (ENCODED_READ ioctl)
  btrfs: move priv off stack in btrfs_encoded_read_regular_fill_pages()
  btrfs: don't sleep in btrfs_encoded_read() if IOCB_NOWAIT is set
  btrfs: change btrfs_encoded_read() so that reading of extent is done by caller
  btrfs: remove pointless iocb::ki_pos addition in btrfs_encoded_read()
  ...

3 months agoMerge tag 'ext4_for_linus-6.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Tue, 19 Nov 2024 00:32:58 +0000 (16:32 -0800)]
Merge tag 'ext4_for_linus-6.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4

Pull ext4 updates from Ted Ts'o:
 "A lot of miscellaneous ext4 bug fixes and cleanups this cycle, most
  notably in the journaling code, bufered I/O, and compiler warning
  cleanups"

* tag 'ext4_for_linus-6.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (33 commits)
  jbd2: Fix comment describing journal_init_common()
  ext4: prevent an infinite loop in the lazyinit thread
  ext4: use struct_size() to improve ext4_htree_store_dirent()
  ext4: annotate struct fname with __counted_by()
  jbd2: avoid dozens of -Wflex-array-member-not-at-end warnings
  ext4: use str_yes_no() helper function
  ext4: prevent delalloc to nodelalloc on remount
  jbd2: make b_frozen_data allocation always succeed
  ext4: cleanup variable name in ext4_fc_del()
  ext4: use string choices helpers
  jbd2: remove the 'success' parameter from the jbd2_do_replay() function
  jbd2: remove useless 'block_error' variable
  jbd2: factor out jbd2_do_replay()
  jbd2: refactor JBD2_COMMIT_BLOCK process in do_one_pass()
  jbd2: unified release of buffer_head in do_one_pass()
  jbd2: remove redundant judgments for check v1 checksum
  ext4: use ERR_CAST to return an error-valued pointer
  mm: zero range of eof folio exposed by inode size extension
  ext4: partial zero eof block on unaligned inode size extension
  ext4: disambiguate the return value of ext4_dio_write_end_io()
  ...

3 months agoMerge tag 'pull-statx' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Mon, 18 Nov 2024 22:54:10 +0000 (14:54 -0800)]
Merge tag 'pull-statx' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

Pull statx updates from Al Viro:
 "Sanitize struct filename and lookup flags handling in statx and
  friends"

* tag 'pull-statx' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  libfs: kill empty_dir_getattr()
  fs: Simplify getattr interface function checking AT_GETATTR_NOSEC flag
  fs/stat.c: switch to CLASS(fd_raw)
  kill getname_statx_lookup_flags()
  io_statx_prep(): use getname_uflags()

3 months agoMerge tag 'pull-ufs' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Mon, 18 Nov 2024 20:58:23 +0000 (12:58 -0800)]
Merge tag 'pull-ufs' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

Pull ufs updates from Al Viro:
 "ufs cleanups, fixes and folio conversion"

* tag 'pull-ufs' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  ufs: ufs_sb_private_info: remove unused s_{2,3}apb fields
  ufs: Convert ufs_change_blocknr() to take a folio
  ufs: Pass a folio to ufs_new_fragments()
  ufs: Convert ufs_inode_getfrag() to take a folio
  ufs: Convert ufs_extend_tail() to take a folio
  ufs: Convert ufs_inode_getblock() to take a folio
  ufs: take the handling of free block counters into a helper
  clean ufs_trunc_direct() up a bit...
  ufs: get rid of ubh_{ubhcpymem,memcpyubh}()
  ufs_inode_getfrag(): remove junk comment
  ufs_free_fragments(): fix the braino in sanity check
  ufs_clusteracct(): switch to passing fragment number
  ufs: untangle ubh_...block...(), part 3
  ufs: untangle ubh_...block...(), part 2
  ufs: untangle ubh_...block...() macros, part 1
  ufs: fix ufs_read_cylinder() failure handling
  ufs: missing ->splice_write()
  ufs: fix handling of delete_entry and set_link failures

3 months agoMerge tag 'pull-xattr' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Mon, 18 Nov 2024 20:44:25 +0000 (12:44 -0800)]
Merge tag 'pull-xattr' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

Pull xattr updates from Al Viro:
 "Sanitize xattr and io_uring interactions with it, add *xattrat()
  syscalls, sanitize struct filename handling in there"

* tag 'pull-xattr' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  xattr: remove redundant check on variable err
  fs/xattr: add *at family syscalls
  new helpers: file_removexattr(), filename_removexattr()
  new helpers: file_listxattr(), filename_listxattr()
  replace do_getxattr() with saner helpers.
  replace do_setxattr() with saner helpers.
  new helper: import_xattr_name()
  fs: rename struct xattr_ctx to kernel_xattr_ctx
  xattr: switch to CLASS(fd)
  io_[gs]etxattr_prep(): just use getname()
  io_uring: IORING_OP_F[GS]ETXATTR is fine with REQ_F_FIXED_FILE
  getname_maybe_null() - the third variant of pathname copy-in
  teach filename_lookup() to treat NULL filename as ""

3 months agoMerge tag 'pull-fd' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Mon, 18 Nov 2024 20:24:06 +0000 (12:24 -0800)]
Merge tag 'pull-fd' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

Pull 'struct fd' class updates from Al Viro:
 "The bulk of struct fd memory safety stuff

  Making sure that struct fd instances are destroyed in the same scope
  where they'd been created, getting rid of reassignments and passing
  them by reference, converting to CLASS(fd{,_pos,_raw}).

  We are getting very close to having the memory safety of that stuff
  trivial to verify"

* tag 'pull-fd' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (28 commits)
  deal with the last remaing boolean uses of fd_file()
  css_set_fork(): switch to CLASS(fd_raw, ...)
  memcg_write_event_control(): switch to CLASS(fd)
  assorted variants of irqfd setup: convert to CLASS(fd)
  do_pollfd(): convert to CLASS(fd)
  convert do_select()
  convert vfs_dedupe_file_range().
  convert cifs_ioctl_copychunk()
  convert media_request_get_by_fd()
  convert spu_run(2)
  switch spufs_calls_{get,put}() to CLASS() use
  convert cachestat(2)
  convert do_preadv()/do_pwritev()
  fdget(), more trivial conversions
  fdget(), trivial conversions
  privcmd_ioeventfd_assign(): don't open-code eventfd_ctx_fdget()
  o2hb_region_dev_store(): avoid goto around fdget()/fdput()
  introduce "fd_pos" class, convert fdget_pos() users to it.
  fdget_raw() users: switch to CLASS(fd_raw)
  convert vmsplice() to CLASS(fd)
  ...

3 months agoMerge tag 'vfs-6.13.ecryptfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 19:44:06 +0000 (11:44 -0800)]
Merge tag 'vfs-6.13.ecryptfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull ecryptfs updates from Christian Brauner:
 "The folio project is about to remove page->index. This contains the
  work required for ecryptfs"

* tag 'vfs-6.13.ecryptfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  ecryptfs: Pass the folio index to crypt_extent()
  ecryptfs: Convert lower_offset_for_page() to take a folio
  ecryptfs: Convert ecryptfs_decrypt_page() to take a folio
  ecryptfs: Convert ecryptfs_encrypt_page() to take a folio
  ecryptfs: Convert ecryptfs_write_lower_page_segment() to take a folio
  ecryptfs: Convert ecryptfs_write() to use a folio
  ecryptfs: Convert ecryptfs_read_lower_page_segment() to take a folio
  ecryptfs: Convert ecryptfs_copy_up_encrypted_with_header() to take a folio
  ecryptfs: Use a folio throughout ecryptfs_read_folio()
  ecryptfs: Convert ecryptfs_writepage() to ecryptfs_writepages()

3 months agoMerge tag 'vfs-6.13.untorn.writes' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 18 Nov 2024 19:30:09 +0000 (11:30 -0800)]
Merge tag 'vfs-6.13.untorn.writes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs untorn write support from Christian Brauner:
 "An atomic write is a write issed with torn-write protection. This
  means for a power failure or any hardware failure all or none of the
  data from the write will be stored, never a mix of old and new data.

  This work is already supported for block devices. If a block device is
  opened with O_DIRECT and the block device supports atomic write, then
  FMODE_CAN_ATOMIC_WRITE is added to the file of the opened block
  device.

  This contains the work to expand atomic write support to filesystems,
  specifically ext4 and XFS. Currently, only support for writing exactly
  one filesystem block atomically is added.

  Since it's now possible to have filesystem block size > page size for
  XFS, it's possible to write 4K+ blocks atomically on x86"

* tag 'vfs-6.13.untorn.writes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  iomap: drop an obsolete comment in iomap_dio_bio_iter
  ext4: Do not fallback to buffered-io for DIO atomic write
  ext4: Support setting FMODE_CAN_ATOMIC_WRITE
  ext4: Check for atomic writes support in write iter
  ext4: Add statx support for atomic writes
  xfs: Support setting FMODE_CAN_ATOMIC_WRITE
  xfs: Validate atomic writes
  xfs: Support atomic write for statx
  fs: iomap: Atomic write support
  fs: Export generic_atomic_write_valid()
  block: Add bdev atomic write limits helpers
  fs/block: Check for IOCB_DIRECT in generic_atomic_write_valid()
  block/fs: Pass an iocb to generic_atomic_write_valid()

3 months agoMerge tag 'vfs-6.13.tmpfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 19:05:26 +0000 (11:05 -0800)]
Merge tag 'vfs-6.13.tmpfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull tmpfs case folding updates from Christian Brauner:
 "This adds case-insensitive support for tmpfs.

  The work contained in here adds support for case-insensitive file
  names lookups in tmpfs. The main difference from other casefold
  filesystems is that tmpfs has no information on disk, just on RAM, so
  we can't use mkfs to create a case-insensitive tmpfs. For this
  implementation, there's a mount option for casefolding. The rest of
  the patchset follows a similar approach as ext4 and f2fs.

  The use case for this feature is similar to the use case for ext4, to
  better support compatibility layers (like Wine), particularly in
  combination with sandboxing/container tools (like Flatpak).

  Those containerization tools can share a subset of the host filesystem
  with an application. In the container, the root directory and any
  parent directories required for a shared directory are on tmpfs, with
  the shared directories bind-mounted into the container's view of the
  filesystem.

  If the host filesystem is using case-insensitive directories, then the
  application can do lookups inside those directories in a
  case-insensitive way, without this needing to be implemented in
  user-space. However, if the host is only sharing a subset of a
  case-insensitive directory with the application, then the parent
  directories of the mount point will be part of the container's root
  tmpfs. When the application tries to do case-insensitive lookups of
  those parent directories on a case-sensitive tmpfs, the lookup will
  fail"

* tag 'vfs-6.13.tmpfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  tmpfs: Initialize sysfs during tmpfs init
  tmpfs: Fix type for sysfs' casefold attribute
  libfs: Fix kernel-doc warning in generic_ci_validate_strict_name
  docs: tmpfs: Add casefold options
  tmpfs: Expose filesystem features via sysfs
  tmpfs: Add flag FS_CASEFOLD_FL support for tmpfs dirs
  tmpfs: Add casefold lookup support
  libfs: Export generic_ci_ dentry functions
  unicode: Recreate utf8_parse_version()
  unicode: Export latest available UTF-8 version number
  ext4: Use generic_ci_validate_strict_name helper
  libfs: Create the helper function generic_ci_validate_strict_name()

3 months agoMerge tag 'vfs-6.13.usercopy' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 18:50:09 +0000 (10:50 -0800)]
Merge tag 'vfs-6.13.usercopy' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull copy_struct_to_user helper from Christian Brauner:
 "This adds a copy_struct_to_user() helper which is a companion helper
  to the already widely used copy_struct_from_user().

  It copies a struct from kernel space to userspace, in a way that
  guarantees backwards-compatibility for struct syscall arguments as
  long as future struct extensions are made such that all new fields are
  appended to the old struct, and zeroed-out new fields have the same
  meaning as the old struct.

  The first user is sched_getattr() system call but the new extensible
  pidfs ioctl will be ported to it as well"

* tag 'vfs-6.13.usercopy' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  sched_getattr: port to copy_struct_to_user
  uaccess: add copy_struct_to_user helper

3 months agoMerge tag 'vfs-6.13.pidfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 18:47:46 +0000 (10:47 -0800)]
Merge tag 'vfs-6.13.pidfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull pidfs update from Christian Brauner:
 "This adds a new ioctl to retrieve information about a pidfd.

  A common pattern when using pidfds is having to get information about
  the process, which currently requires /proc being mounted, resolving
  the fd to a pid, and then do manual string parsing of /proc/N/status
  and friends. This needs to be reimplemented over and over in all
  userspace projects (e.g.: it has been reimplemented in systemd, dbus,
  dbus-daemon, polkit so far), and requires additional care in checking
  that the fd is still valid after having parsed the data, to avoid
  races.

  Having a programmatic API that can be used directly removes all these
  requirements, including having /proc mounted.

  As discussed at LPC24, add an ioctl with an extensible struct so that
  more parameters can be added later if needed. Start with returning
  pid/tgid/ppid and some creds unconditionally, and cgroupid optionally"

* tag 'vfs-6.13.pidfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  pidfd: add ioctl to retrieve pid info

3 months agoMerge tag 'vfs-6.13.ovl' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 18:45:06 +0000 (10:45 -0800)]
Merge tag 'vfs-6.13.ovl' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull overlayfs updates from Christian Brauner:
 "Make overlayfs support specifying layers through file descriptors.

  Currently overlayfs only allows specifying layers through path names.
  This is inconvenient for users that want to assemble an overlayfs
  mount purely based on file descriptors:

  This enables user to specify both:

    fsconfig(fd_overlay, FSCONFIG_SET_FD, "upperdir+", NULL, fd_upper);
    fsconfig(fd_overlay, FSCONFIG_SET_FD, "workdir+",  NULL, fd_work);
    fsconfig(fd_overlay, FSCONFIG_SET_FD, "lowerdir+", NULL, fd_lower1);
    fsconfig(fd_overlay, FSCONFIG_SET_FD, "lowerdir+", NULL, fd_lower2);

  in addition to:

    fsconfig(fd_overlay, FSCONFIG_SET_STRING, "upperdir+", "/upper",  0);
    fsconfig(fd_overlay, FSCONFIG_SET_STRING, "workdir+",  "/work",   0);
    fsconfig(fd_overlay, FSCONFIG_SET_STRING, "lowerdir+", "/lower1", 0);
    fsconfig(fd_overlay, FSCONFIG_SET_STRING, "lowerdir+", "/lower2", 0);

  There's also a large set of new overlayfs selftests to test new
  features and some older properties"

* tag 'vfs-6.13.ovl' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  selftests: add test for specifying 500 lower layers
  selftests: add overlayfs fd mounting selftests
  selftests: use shared header
  Documentation,ovl: document new file descriptor based layers
  ovl: specify layers via file descriptors
  fs: add helper to use mount option as path or fd

3 months agoMerge tag 'vfs-6.13.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 18:30:29 +0000 (10:30 -0800)]
Merge tag 'vfs-6.13.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs file updates from Christian Brauner:
 "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"

* tag 'vfs-6.13.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  selftests: add file SLAB_TYPESAFE_BY_RCU recycling stressor
  fs: port files to file_ref
  fs: add file_ref
  expand_files(): simplify calling conventions
  make __set_open_fd() set cloexec state as well
  fs: protect backing files with rcu
  file.c: merge __{set,clear}_close_on_exec()
  alloc_fdtable(): change calling conventions.
  fs/file.c: add fast path in find_next_fd()
  fs/file.c: conditionally clear full_fds
  fs/file.c: remove sanity_check and add likely/unlikely in alloc_fd()
  move close_range(2) into fs/file.c, fold __close_range() into it
  close_files(): don't bother with xchg()
  remove pointless includes of <linux/fdtable.h>
  get rid of ...lookup...fdget_rcu() family

3 months agoMerge tag 'vfs-6.13.netfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 18:26:49 +0000 (10:26 -0800)]
Merge tag 'vfs-6.13.netfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull netfs updates from Christian Brauner:
 "Various fixes for the netfs library and related infrastructure:

  cachefiles:

   - Fix a dentry leak in cachefiles_open_file()

   - Fix incorrect length return value in
     cachefiles_ondemand_fd_write_iter()

   - Fix missing pos updates in cachefiles_ondemand_fd_write_iter()

   - Clean up in cachefiles_commit_tmpfile()

   - Fix NULL pointer dereference in object->file

   - Add a memory barrier for FSCACHE_VOLUME_CREATING

  netfs:

   - Remove call to folio_index()

   - Fix a few minor bugs in netfs_page_mkwrite()

   - Remove unnecessary references to pages"

* tag 'vfs-6.13.netfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  netfs/fscache: Add a memory barrier for FSCACHE_VOLUME_CREATING
  cachefiles: Fix NULL pointer dereference in object->file
  cachefiles: Clean up in cachefiles_commit_tmpfile()
  cachefiles: Fix missing pos updates in cachefiles_ondemand_fd_write_iter()
  cachefiles: Fix incorrect length return value in cachefiles_ondemand_fd_write_iter()
  netfs: Remove unnecessary references to pages
  netfs: Fix a few minor bugs in netfs_page_mkwrite()
  netfs: Remove call to folio_index()

3 months agoMerge tag 'vfs-6.13.pagecache' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 17:54:32 +0000 (09:54 -0800)]
Merge tag 'vfs-6.13.pagecache' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs pagecache updates from Christian Brauner:
 "Cleanup filesystem page flag usage: This continues the work to make
  the mappedtodisk/owner_2 flag available to filesystems which don't use
  buffer heads. Further patches remove uses of Private2. This brings us
  very close to being rid of it entirely"

* tag 'vfs-6.13.pagecache' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  migrate: Remove references to Private2
  ceph: Remove call to PagePrivate2()
  btrfs: Switch from using the private_2 flag to owner_2
  mm: Remove PageMappedToDisk
  nilfs2: Convert nilfs_copy_buffer() to use folios
  fs: Move clearing of mappedtodisk to buffer.c

3 months agoMerge tag 'vfs-6.13.rust.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 17:51:32 +0000 (09:51 -0800)]
Merge tag 'vfs-6.13.rust.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs rust file abstractions from Christian Brauner:
 "This contains the file abstractions needed by the Rust implementation
  of the Binder driver and other parts of the kernel.

  Let's treat this as a first attempt at getting something working but I
  do expect the actual interfaces to change significantly over time.
  Simply because we are still figuring out what actually works. But
  there's no point in further theorizing. Let's see how it holds up with
  actual users"

* tag 'vfs-6.13.rust.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  rust: task: adjust safety comments in Task methods
  rust: add seqfile abstraction
  rust: file: add abstraction for `poll_table`
  rust: file: add `Kuid` wrapper
  rust: file: add `FileDescriptorReservation`
  rust: security: add abstraction for secctx
  rust: cred: add Rust abstraction for `struct cred`
  rust: file: add Rust abstraction for `struct file`
  rust: task: add `Task::current_raw`
  rust: types: add `NotThreadSafe`

3 months agoMerge tag 'vfs-6.13.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 17:35:30 +0000 (09:35 -0800)]
Merge tag 'vfs-6.13.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull misc vfs updates from Christian Brauner:
 "Features:

   - Fixup and improve NLM and kNFSD file lock callbacks

     Last year both GFS2 and OCFS2 had some work done to make their
     locking more robust when exported over NFS. Unfortunately, part of
     that work caused both NLM (for NFS v3 exports) and kNFSD (for
     NFSv4.1+ exports) to no longer send lock notifications to clients

     This in itself is not a huge problem because most NFS clients will
     still poll the server in order to acquire a conflicted lock

     It's important for NLM and kNFSD that they do not block their
     kernel threads inside filesystem's file_lock implementations
     because that can produce deadlocks. We used to make sure of this by
     only trusting that posix_lock_file() can correctly handle blocking
     lock calls asynchronously, so the lock managers would only setup
     their file_lock requests for async callbacks if the filesystem did
     not define its own lock() file operation

     However, when GFS2 and OCFS2 grew the capability to correctly
     handle blocking lock requests asynchronously, they started
     signalling this behavior with EXPORT_OP_ASYNC_LOCK, and the check
     for also trusting posix_lock_file() was inadvertently dropped, so
     now most filesystems no longer produce lock notifications when
     exported over NFS

     Fix this by using an fop_flag which greatly simplifies the problem
     and grooms the way for future uses by both filesystems and lock
     managers alike

   - Add a sysctl to delete the dentry when a file is removed instead of
     making it a negative dentry

     Commit 681ce8623567 ("vfs: Delete the associated dentry when
     deleting a file") introduced an unconditional deletion of the
     associated dentry when a file is removed. However, this led to
     performance regressions in specific benchmarks, such as
     ilebench.sum_operations/s, prompting a revert in commit
     4a4be1ad3a6e ("Revert "vfs: Delete the associated dentry when
     deleting a file""). This reintroduces the concept conditionally
     through a sysctl

   - Expand the statmount() system call:

       * Report the filesystem subtype in a new fs_subtype field to
         e.g., report fuse filesystem subtypes

       * Report the superblock source in a new sb_source field

       * Add a new way to return filesystem specific mount options in an
         option array that returns filesystem specific mount options
         separated by zero bytes and unescaped. This allows caller's to
         retrieve filesystem specific mount options and immediately pass
         them to e.g., fsconfig() without having to unescape or split
         them

       * Report security (LSM) specific mount options in a separate
         security option array. We don't lump them together with
         filesystem specific mount options as security mount options are
         generic and most users aren't interested in them

         The format is the same as for the filesystem specific mount
         option array

   - Support relative paths in fsconfig()'s FSCONFIG_SET_STRING command

   - Optimize acl_permission_check() to avoid costly {g,u}id ownership
     checks if possible

   - Use smp_mb__after_spinlock() to avoid full smp_mb() in evict()

   - Add synchronous wakeup support for ep_poll_callback.

     Currently, epoll only uses wake_up() to wake up task. But sometimes
     there are epoll users which want to use the synchronous wakeup flag
     to give a hint to the scheduler, e.g., the Android binder driver.
     So add a wake_up_sync() define, and use wake_up_sync() when sync is
     true in ep_poll_callback()

  Fixes:

   - Fix kernel documentation for inode_insert5() and iget5_locked()

   - Annotate racy epoll check on file->f_ep

   - Make F_DUPFD_QUERY associative

   - Avoid filename buffer overrun in initramfs

   - Don't let statmount() return empty strings

   - Add a cond_resched() to dump_user_range() to avoid hogging the CPU

   - Don't query the device logical blocksize multiple times for hfsplus

   - Make filemap_read() check that the offset is positive or zero

  Cleanups:

   - Various typo fixes

   - Cleanup wbc_attach_fdatawrite_inode()

   - Add __releases annotation to wbc_attach_and_unlock_inode()

   - Add hugetlbfs tracepoints

   - Fix various vfs kernel doc parameters

   - Remove obsolete TODO comment from io_cancel()

   - Convert wbc_account_cgroup_owner() to take a folio

   - Fix comments for BANDWITH_INTERVAL and wb_domain_writeout_add()

   - Reorder struct posix_acl to save 8 bytes

   - Annotate struct posix_acl with __counted_by()

   - Replace one-element array with flexible array member in freevxfs

   - Use idiomatic atomic64_inc_return() in alloc_mnt_ns()"

* tag 'vfs-6.13.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: (35 commits)
  statmount: retrieve security mount options
  vfs: make evict() use smp_mb__after_spinlock instead of smp_mb
  statmount: add flag to retrieve unescaped options
  fs: add the ability for statmount() to report the sb_source
  writeback: wbc_attach_fdatawrite_inode out of line
  writeback: add a __releases annoation to wbc_attach_and_unlock_inode
  fs: add the ability for statmount() to report the fs_subtype
  fs: don't let statmount return empty strings
  fs:aio: Remove TODO comment suggesting hash or array usage in io_cancel()
  hfsplus: don't query the device logical block size multiple times
  freevxfs: Replace one-element array with flexible array member
  fs: optimize acl_permission_check()
  initramfs: avoid filename buffer overrun
  fs/writeback: convert wbc_account_cgroup_owner to take a folio
  acl: Annotate struct posix_acl with __counted_by()
  acl: Realign struct posix_acl to save 8 bytes
  epoll: Add synchronous wakeup support for ep_poll_callback
  coredump: add cond_resched() to dump_user_range
  mm/page-writeback.c: Fix comment of wb_domain_writeout_add()
  mm/page-writeback.c: Update comment for BANDWIDTH_INTERVAL
  ...

3 months agoMerge tag 'vfs-6.13.mount.api' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 17:33:34 +0000 (09:33 -0800)]
Merge tag 'vfs-6.13.mount.api' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs mount api conversions from Christian Brauner:
 "Convert adfs, affs, befs, hfs, hfsplus, jfs, and hpfs to the new mount
  api"

* tag 'vfs-6.13.mount.api' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  efs: fix the efs new mount api implementation
  ubifs: Convert ubifs to use the new mount API
  hpfs: convert hpfs to use the new mount api
  jfs: convert jfs to use the new mount api
  hfsplus: convert hfsplus to use the new mount api
  hfs: convert hfs to use the new mount api
  befs: convert befs to use the new mount api
  affs: convert affs to use the new mount api
  adfs: convert adfs to use the new mount api

3 months agoMerge tag 'vfs-6.13.mgtime' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Mon, 18 Nov 2024 17:15:39 +0000 (09:15 -0800)]
Merge tag 'vfs-6.13.mgtime' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs multigrain timestamps from Christian Brauner:
 "This is another try at implementing multigrain timestamps. This time
  with significant help from the timekeeping maintainers to reduce the
  performance impact.

  Thomas provided a base branch that contains the required timekeeping
  interfaces for the VFS. It serves as the base for the multi-grain
  timestamp work:

   - Multigrain timestamps allow the kernel to use fine-grained
     timestamps when an inode's attributes is being actively observed
     via ->getattr(). With this support, it's possible for a file to get
     a fine-grained timestamp, and another modified after it to get a
     coarse-grained stamp that is earlier than the fine-grained time. If
     this happens then the files can appear to have been modified in
     reverse order, which breaks VFS ordering guarantees.

     To prevent this, a floor value is maintained for multigrain
     timestamps. Whenever a fine-grained timestamp is handed out, record
     it, and when later coarse-grained stamps are handed out, ensure
     they are not earlier than that value. If the coarse-grained
     timestamp is earlier than the fine-grained floor, return the floor
     value instead.

     The timekeeper changes add a static singleton atomic64_t into
     timekeeper.c that is used to keep track of the latest fine-grained
     time ever handed out. This is tracked as a monotonic ktime_t value
     to ensure that it isn't affected by clock jumps. Because it is
     updated at different times than the rest of the timekeeper object,
     the floor value is managed independently of the timekeeper via a
     cmpxchg() operation, and sits on its own cacheline.

     Two new public timekeeper interfaces are added:

      (1) ktime_get_coarse_real_ts64_mg() fills a timespec64 with the
          later of the coarse-grained clock and the floor time

      (2) ktime_get_real_ts64_mg() gets the fine-grained clock value,
          and tries to swap it into the floor. A timespec64 is filled
          with the result.

   - The VFS has always used coarse-grained timestamps when updating the
     ctime and mtime after a change. This has the benefit of allowing
     filesystems to optimize away a lot metadata updates, down to around
     1 per jiffy, even when a file is under heavy writes.

     Unfortunately, this has always been an issue when we're exporting
     via NFSv3, which relies on timestamps to validate caches. A lot of
     changes can happen in a jiffy, so timestamps aren't sufficient to
     help the client decide when to invalidate the cache. Even with
     NFSv4, a lot of exported filesystems don't properly support a
     change attribute and are subject to the same problems with
     timestamp granularity. Other applications have similar issues with
     timestamps (e.g backup applications).

     If we were to always use fine-grained timestamps, that would
     improve the situation, but that becomes rather expensive, as the
     underlying filesystem would have to log a lot more metadata
     updates.

     This adds a way to only use fine-grained timestamps when they are
     being actively queried. Use the (unused) top bit in
     inode->i_ctime_nsec as a flag that indicates whether the current
     timestamps have been queried via stat() or the like. When it's set,
     we allow the kernel to use a fine-grained timestamp iff it's
     necessary to make the ctime show a different value.

     This solves the problem of being able to distinguish the timestamp
     between updates, but introduces a new problem: it's now possible
     for a file being changed to get a fine-grained timestamp. A file
     that is altered just a bit later can then get a coarse-grained one
     that appears older than the earlier fine-grained time. This
     violates timestamp ordering guarantees.

     This is where the earlier mentioned timkeeping interfaces help. A
     global monotonic atomic64_t value is kept that acts as a timestamp
     floor. When we go to stamp a file, we first get the latter of the
     current floor value and the current coarse-grained time. If the
     inode ctime hasn't been queried then we just attempt to stamp it
     with that value.

     If it has been queried, then first see whether the current coarse
     time is later than the existing ctime. If it is, then we accept
     that value. If it isn't, then we get a fine-grained time and try to
     swap that into the global floor. Whether that succeeds or fails, we
     take the resulting floor time, convert it to realtime and try to
     swap that into the ctime.

     We take the result of the ctime swap whether it succeeds or fails,
     since either is just as valid.

     Filesystems can opt into this by setting the FS_MGTIME fstype flag.
     Others should be unaffected (other than being subject to the same
     floor value as multigrain filesystems)"

* tag 'vfs-6.13.mgtime' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  fs: reduce pointer chasing in is_mgtime() test
  tmpfs: add support for multigrain timestamps
  btrfs: convert to multigrain timestamps
  ext4: switch to multigrain timestamps
  xfs: switch to multigrain timestamps
  Documentation: add a new file documenting multigrain timestamps
  fs: add percpu counters for significant multigrain timestamp events
  fs: tracepoints around multigrain timestamp events
  fs: handle delegated timestamps in setattr_copy_mgtime
  timekeeping: Add percpu counter for tracking floor swap events
  timekeeping: Add interfaces for handling timestamps with a floor value
  fs: have setattr_copy handle multigrain timestamps appropriately
  fs: add infrastructure for multigrain timestamps

3 months agoLinux 6.12 v6.12
Linus Torvalds [Sun, 17 Nov 2024 22:15:08 +0000 (14:15 -0800)]
Linux 6.12

3 months agoMerge tag 'x86_urgent_for_v6.12' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 17 Nov 2024 17:35:51 +0000 (09:35 -0800)]
Merge tag 'x86_urgent_for_v6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86 fixes from Borislav Petkov:

 - Make sure a kdump kernel with CONFIG_IMA_KEXEC enabled and booted on
   an AMD SME enabled hardware properly decrypts the ima_kexec buffer
   information passed to it from the previous kernel

 - Fix building the kernel with Clang where a non-TLS definition of the
   stack protector guard cookie leads to bogus code generation

 - Clear a wrongly advertised virtualized VMLOAD/VMSAVE feature flag on
   some Zen4 client systems as those insns are not supported on client

* tag 'x86_urgent_for_v6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mm: Fix a kdump kernel failure on SME system when CONFIG_IMA_KEXEC=y
  x86/stackprotector: Work around strict Clang TLS symbol requirements
  x86/CPU/AMD: Clear virtualized VMLOAD/VMSAVE on Zen4 client

4 months agoMerge tag 'mm-hotfixes-stable-2024-11-16-15-33' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Sun, 17 Nov 2024 00:00:38 +0000 (16:00 -0800)]
Merge tag 'mm-hotfixes-stable-2024-11-16-15-33' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

Pull hotfixes from Andrew Morton:
 "10 hotfixes, 7 of which are cc:stable. All singletons, please see the
  changelogs for details"

* tag 'mm-hotfixes-stable-2024-11-16-15-33' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
  mm: revert "mm: shmem: fix data-race in shmem_getattr()"
  ocfs2: uncache inode which has failed entering the group
  mm: fix NULL pointer dereference in alloc_pages_bulk_noprof
  mm, doc: update read_ahead_kb for MADV_HUGEPAGE
  fs/proc/task_mmu: prevent integer overflow in pagemap_scan_get_args()
  sched/task_stack: fix object_is_on_stack() for KASAN tagged pointers
  crash, powerpc: default to CRASH_DUMP=n on PPC_BOOK3S_32
  mm/mremap: fix address wraparound in move_page_tables()
  tools/mm: fix compile error
  mm, swap: fix allocation and scanning race with swapoff

4 months agomm: revert "mm: shmem: fix data-race in shmem_getattr()"
Andrew Morton [Sat, 16 Nov 2024 00:57:24 +0000 (16:57 -0800)]
mm: revert "mm: shmem: fix data-race in shmem_getattr()"

Revert d949d1d14fa2 ("mm: shmem: fix data-race in shmem_getattr()") as
suggested by Chuck [1].  It is causing deadlocks when accessing tmpfs over
NFS.

As Hugh commented, "added just to silence a syzbot sanitizer splat: added
where there has never been any practical problem".

Link: https://lkml.kernel.org/r/[email protected]
Fixes: d949d1d14fa2 ("mm: shmem: fix data-race in shmem_getattr()")
Acked-by: Hugh Dickins <[email protected]>
Cc: Chuck Lever <[email protected]>
Cc: Jeongjun Park <[email protected]>
Cc: Yu Zhao <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux
Linus Torvalds [Sat, 16 Nov 2024 23:14:39 +0000 (15:14 -0800)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux

Pull ARM fixes from Russell King:

 - Fix kernel mapping for XIP kernels

 - Fix SMP support for XIP kernels

 - Fix complication corner case with CFI

 - Fix a typo in nommu code

 - Fix cacheflush syscall when PAN is enabled on LPAE platforms

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux:
  ARM: fix cacheflush with PAN
  ARM: 9435/1: ARM/nommu: Fix typo "absence"
  ARM: 9434/1: cfi: Fix compilation corner case
  ARM: 9420/1: smp: Fix SMP for xip kernels
  ARM: 9419/1: mm: Fix kernel memory mapping for xip kernels

4 months agoMerge tag 'drm-fixes-2024-11-17' of https://gitlab.freedesktop.org/drm/kernel
Linus Torvalds [Sat, 16 Nov 2024 23:09:14 +0000 (15:09 -0800)]
Merge tag 'drm-fixes-2024-11-17' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fix from Dave Airlie:
 "Alex sent on a last minute revert for a amdgpu/swsmu regression:

   - revert patch to fix swsmu regression"

* tag 'drm-fixes-2024-11-17' of https://gitlab.freedesktop.org/drm/kernel:
  Revert "drm/amd/pm: correct the workload setting"

4 months agoMerge tag 'amd-drm-fixes-6.12-2024-11-16' of https://gitlab.freedesktop.org/agd5f...
Dave Airlie [Sat, 16 Nov 2024 22:12:44 +0000 (08:12 +1000)]
Merge tag 'amd-drm-fixes-6.12-2024-11-16' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.12-2024-11-16:

amdgpu:
- Revert a swsmu patch to fix a regression

Signed-off-by: Dave Airlie <[email protected]>
From: Alex Deucher <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agoMerge tag 'trace-ringbuffer-v6.12-rc7-2' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 16 Nov 2024 16:12:43 +0000 (08:12 -0800)]
Merge tag 'trace-ringbuffer-v6.12-rc7-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace

Pull ring buffer fixes from Steven Rostedt:

 - Revert: "ring-buffer: Do not have boot mapped buffers hook to CPU
   hotplug"

   A crash that happened on cpu hotplug was actually caused by the
   incorrect ref counting that was fixed by commit 2cf9733891a4
   ("ring-buffer: Fix refcount setting of boot mapped buffers"). The
   removal of calling cpu hotplug callbacks on memory mapped buffers was
   not an issue even though the tests at the time pointed toward it. But
   in fact, there's a check in that code that tests to see if the
   buffers are already allocated or not, and will not allocate them
   again if they are. Not calling the cpu hotplug callbacks ended up not
   initializing the non boot CPU buffers.

   Simply remove that change.

 - Clear all CPU buffers when starting tracing in a boot mapped buffer

   To properly process events from a previous boot, the address space
   needs to be accounted for due to KASLR and the events in the buffer
   are updated accordingly when read. This also requires that when the
   buffer has tracing enabled again in the current boot that the buffers
   are reset so that events from the previous boot do not interact with
   the events of the current boot and cause confusing due to not having
   the proper meta data.

   It was found that if a CPU is taken offline, that its per CPU buffer
   is not reset when tracing starts. This allows for events to be from
   both the previous boot and the current boot to be in the buffer at
   the same time. Clear all CPU buffers when tracing is started in a
   boot mapped buffer.

* tag 'trace-ringbuffer-v6.12-rc7-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  tracing/ring-buffer: Clear all memory mapped CPU ring buffers on first recording
  Revert: "ring-buffer: Do not have boot mapped buffers hook to CPU hotplug"

4 months agoRevert "drm/amd/pm: correct the workload setting"
Alex Deucher [Sat, 16 Nov 2024 14:22:14 +0000 (09:22 -0500)]
Revert "drm/amd/pm: correct the workload setting"

This reverts commit 74e1006430a5377228e49310f6d915628609929e.

This causes a regression in the workload selection.
A more extensive fix is being worked on.
For now, revert.

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/3618
Fixes: 74e1006430a5 ("drm/amd/pm: correct the workload setting")
Signed-off-by: Alex Deucher <[email protected]>
4 months agoMerge tag 'riscv-for-linus-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 15 Nov 2024 19:44:32 +0000 (11:44 -0800)]
Merge tag 'riscv-for-linus-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux

Pull RISC-V fix from Palmer Dabbelt:

 - A fix for the CPU perf driver that avoids leaking CPU ID references
   on systems without snapshot support.

* tag 'riscv-for-linus-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
  drivers: perf: Fix wrong put_cpu() placement

4 months agoMerge tag 'drm-fixes-2024-11-16' of https://gitlab.freedesktop.org/drm/kernel
Linus Torvalds [Fri, 15 Nov 2024 18:53:42 +0000 (10:53 -0800)]
Merge tag 'drm-fixes-2024-11-16' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fixes from Dave Airlie:
 "Final week of fixes, lots of small amdgpu fixes, some i915 and xe
  fixes, the nouveau changes fix a recent regression and some laptop
  panel black screens, then a couple of other misc ones.

  It's probably a little busier than I'd like, but each fix seems fine.

  amdgpu:
   - PSR fix
   - Panel replay fixes
   - DML fix
   - vblank power fix
   - Fix video caps
   - SMU 14.0 fix
   - GPUVM fix
   - MES 12 fix
   - APU carve out fix
   - DC vbios fix
   - NBIO fix

  i915:
   - Don't load GSC on ARL-H and ARL-U if too old FW
   - Avoid potential OOPS in enabling/disabling TV output

  xe:
   - Fix unlock on exec ioctl error path
   - Fix hibernation on LNL due to ggtt getting lost
   - Fix missing runtime PM in OA release

  bridge:
   - tc358768: Fix DSI command tx

  nouveau:
   - Fix GSP AUX error handling
   - dp: Handle retires for AUX CH transfers with GSP
   - fw: Sync DMA after setup

  panthor:
   - Fix partial BO mappings to GPU

  rockchip:
   - vop: Avoid null-ptr deref in plane-state check

  vmwgfx:
   - Avoid null-ptr deref in surface creation"

* tag 'drm-fixes-2024-11-16' of https://gitlab.freedesktop.org/drm/kernel: (27 commits)
  drm/bridge: tc358768: Fix DSI command tx
  drm/vmwgfx: avoid null_ptr_deref in vmw_framebuffer_surface_create_handle
  nouveau/dp: handle retries for AUX CH transfers with GSP.
  nouveau: handle EBUSY and EAGAIN for GSP aux errors.
  nouveau: fw: sync dma after setup is called.
  drm/xe/oa: Fix "Missing outer runtime PM protection" warning
  drm/xe: handle flat ccs during hibernation on igpu
  drm/xe: improve hibernation on igpu
  drm/xe: Restore system memory GGTT mappings
  drm/xe: Ensure all locks released in exec IOCTL
  drm/panthor: Fix handling of partial GPU mapping of BOs
  drm/amd: Fix initialization mistake for NBIO 7.7.0
  Revert "drm/amd/display: parse umc_info or vram_info based on ASIC"
  drm/amd/display: Fix failure to read vram info due to static BP_RESULT
  drm/amdgpu: enable GTT fallback handling for dGPUs only
  drm/i915: Grab intel_display from the encoder to avoid potential oopsies
  drm/i915/gsc: ARL-H and ARL-U need a newer GSC FW.
  drm/amdgpu/mes12: correct kiq unmap latency
  drm/amdgpu: fix check in gmc_v9_0_get_vm_pte()
  drm/amd/pm: print pp_dpm_mclk in ascending order on SMU v14.0.0
  ...

4 months agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Linus Torvalds [Fri, 15 Nov 2024 18:48:28 +0000 (10:48 -0800)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma

Pull rdma fixes from Jason Gunthorpe:

 - Revert a change to the VLAN logic, this broke previously working ROCE
   configurations

 - Fix a memory leak on error unwinding in bnxt_re

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  Revert "RDMA/core: Fix ENODEV error for iWARP test over vlan"
  RDMA/bnxt_re: Remove some dead code
  RDMA/bnxt_re: Fix some error handling paths in bnxt_re_probe()

4 months agoMerge tag 'drm-xe-fixes-2024-11-14' of https://gitlab.freedesktop.org/drm/xe/kernel...
Dave Airlie [Fri, 15 Nov 2024 18:31:09 +0000 (04:31 +1000)]
Merge tag 'drm-xe-fixes-2024-11-14' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-fixes

Driver Changes:
- Fix unlock on exec ioctl error path (Matthew Brost)
- Fix hibernation on LNL due to ggtt getting lost
  (Matthew Brost / Matthew Auld)
- Fix missing runtime PM in OA release (Ashutosh)

Signed-off-by: Dave Airlie <[email protected]>
From: Lucas De Marchi <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/5ntcf2ssmmvo5dsf2mdcee4guwwmpbm3xrlufgt2pdfmznzjo3@62ygo3bxkock
4 months agoMerge tag 'pmdomain-v6.12-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 15 Nov 2024 18:20:17 +0000 (10:20 -0800)]
Merge tag 'pmdomain-v6.12-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/linux-pm

Pull pmdomain fixes from Ulf Hansson:
 "pmdomain core:
   - Add GENPD_FLAG_DEV_NAME_FW flag to generate unique names

  pmdomain providers:
   - arm: Use FLAG_DEV_NAME_FW to ensure unique names
   - imx93-blk-ctrl: Fix the remove path

  arm_scmi/qcom-cpucp:
   - Report duplicate OPPs as firmware bugs for arm_scmi
   - Skip OPP duplicates for arm_scmi
   - Mark the qcom-cpucp mailbox irq with IRQF_NO_SUSPEND flag"

* tag 'pmdomain-v6.12-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/linux-pm:
  mailbox: qcom-cpucp: Mark the irq with IRQF_NO_SUSPEND flag
  firmware: arm_scmi: Report duplicate opps as firmware bugs
  firmware: arm_scmi: Skip opp duplicates
  pmdomain: imx93-blk-ctrl: correct remove path
  pmdomain: arm: Use FLAG_DEV_NAME_FW to ensure unique names
  pmdomain: core: Add GENPD_FLAG_DEV_NAME_FW flag

4 months agoMerge tag 'mmc-v6.12-rc3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc
Linus Torvalds [Fri, 15 Nov 2024 18:16:12 +0000 (10:16 -0800)]
Merge tag 'mmc-v6.12-rc3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc

Pull MMC host fixes from Ulf Hansson:

 - dw_mmc: Revert fix for IDMAC operation with pages bigger than 4K

 - sunxi-mmc: Fix A100 compatible description

* tag 'mmc-v6.12-rc3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
  Revert "mmc: dw_mmc: Fix IDMAC operation with pages bigger than 4K"
  mmc: sunxi-mmc: Fix A100 compatible description

4 months agoMerge tag 'sound-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Linus Torvalds [Fri, 15 Nov 2024 18:09:38 +0000 (10:09 -0800)]
Merge tag 'sound-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound

Pull sound fixes from Takashi Iwai:
 "A few last-minute fixes. All changes are device-specific small fixes
  that should be pretty safe to apply"

* tag 'sound-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: hda/realtek - update set GPIO3 to default for Thinkpad with ALC1318
  ALSA: hda/realtek: fix mute/micmute LEDs for a HP EliteBook 645 G10
  ALSA: hda/realtek - Fixed Clevo platform headset Mic issue
  ALSA: usb-audio: Fix Yamaha P-125 Quirk Entry
  ASoC: max9768: Fix event generation for playback mute
  ASoC: intel: sof_sdw: add quirk for Dell SKU
  ASoC: audio-graph-card2: Purge absent supplies for device tree nodes

4 months agoMerge tag 'v6.12-p5' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Fri, 15 Nov 2024 18:04:39 +0000 (10:04 -0800)]
Merge tag 'v6.12-p5' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6

Pull crypto fix from Herbert Xu:
 "Fix a regression in the MIPS CRC32C code"

* tag 'v6.12-p5' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: mips/crc32 - fix the CRC32C implementation

4 months agoMerge tag 'sched_ext-for-6.12-rc7-fixes-2' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Fri, 15 Nov 2024 17:59:51 +0000 (09:59 -0800)]
Merge tag 'sched_ext-for-6.12-rc7-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext

Pull sched_ext fix from Tejun Heo:
 "One more fix for v6.12-rc7

  ops.cpu_acquire() was being invoked with the wrong kfunc mask allowing
  the operation to call kfuncs which shouldn't be allowed. Fix it by
  using SCX_KF_REST instead, which is trivial and low risk"

* tag 'sched_ext-for-6.12-rc7-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:
  sched_ext: ops.cpu_acquire() should be called with SCX_KF_REST

4 months agoMerge tag 'for-6.12-rc7-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave...
Linus Torvalds [Fri, 15 Nov 2024 17:45:32 +0000 (09:45 -0800)]
Merge tag 'for-6.12-rc7-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fix from David Sterba:
 "One more fix that seems urgent and good to have in 6.12 final.

  It could potentially lead to unexpected transaction aborts, due to
  wrong comparison and order of processing of delayed refs"

* tag 'for-6.12-rc7-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: fix incorrect comparison for delayed refs

4 months agoocfs2: uncache inode which has failed entering the group
Dmitry Antipov [Thu, 14 Nov 2024 04:38:44 +0000 (07:38 +0300)]
ocfs2: uncache inode which has failed entering the group

Syzbot has reported the following BUG:

kernel BUG at fs/ocfs2/uptodate.c:509!
...
Call Trace:
 <TASK>
 ? __die_body+0x5f/0xb0
 ? die+0x9e/0xc0
 ? do_trap+0x15a/0x3a0
 ? ocfs2_set_new_buffer_uptodate+0x145/0x160
 ? do_error_trap+0x1dc/0x2c0
 ? ocfs2_set_new_buffer_uptodate+0x145/0x160
 ? __pfx_do_error_trap+0x10/0x10
 ? handle_invalid_op+0x34/0x40
 ? ocfs2_set_new_buffer_uptodate+0x145/0x160
 ? exc_invalid_op+0x38/0x50
 ? asm_exc_invalid_op+0x1a/0x20
 ? ocfs2_set_new_buffer_uptodate+0x2e/0x160
 ? ocfs2_set_new_buffer_uptodate+0x144/0x160
 ? ocfs2_set_new_buffer_uptodate+0x145/0x160
 ocfs2_group_add+0x39f/0x15a0
 ? __pfx_ocfs2_group_add+0x10/0x10
 ? __pfx_lock_acquire+0x10/0x10
 ? mnt_get_write_access+0x68/0x2b0
 ? __pfx_lock_release+0x10/0x10
 ? rcu_read_lock_any_held+0xb7/0x160
 ? __pfx_rcu_read_lock_any_held+0x10/0x10
 ? smack_log+0x123/0x540
 ? mnt_get_write_access+0x68/0x2b0
 ? mnt_get_write_access+0x68/0x2b0
 ? mnt_get_write_access+0x226/0x2b0
 ocfs2_ioctl+0x65e/0x7d0
 ? __pfx_ocfs2_ioctl+0x10/0x10
 ? smack_file_ioctl+0x29e/0x3a0
 ? __pfx_smack_file_ioctl+0x10/0x10
 ? lockdep_hardirqs_on_prepare+0x43d/0x780
 ? __pfx_lockdep_hardirqs_on_prepare+0x10/0x10
 ? __pfx_ocfs2_ioctl+0x10/0x10
 __se_sys_ioctl+0xfb/0x170
 do_syscall_64+0xf3/0x230
 entry_SYSCALL_64_after_hwframe+0x77/0x7f
...
 </TASK>

When 'ioctl(OCFS2_IOC_GROUP_ADD, ...)' has failed for the particular
inode in 'ocfs2_verify_group_and_input()', corresponding buffer head
remains cached and subsequent call to the same 'ioctl()' for the same
inode issues the BUG() in 'ocfs2_set_new_buffer_uptodate()' (trying
to cache the same buffer head of that inode). Fix this by uncaching
the buffer head with 'ocfs2_remove_from_cache()' on error path in
'ocfs2_group_add()'.

Link: https://lkml.kernel.org/r/[email protected]
Fixes: 7909f2bf8353 ("[PATCH 2/2] ocfs2: Implement group add for online resize")
Signed-off-by: Dmitry Antipov <[email protected]>
Reported-by: [email protected]
Closes: https://syzkaller.appspot.com/bug?extid=453873f1588c2d75b447
Reviewed-by: Joseph Qi <[email protected]>
Cc: Dmitry Antipov <[email protected]>
Cc: Joel Becker <[email protected]>
Cc: Mark Fasheh <[email protected]>
Cc: Junxiao Bi <[email protected]>
Cc: Changwei Ge <[email protected]>
Cc: Jun Piao <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agomm: fix NULL pointer dereference in alloc_pages_bulk_noprof
Jinjiang Tu [Wed, 13 Nov 2024 08:32:35 +0000 (16:32 +0800)]
mm: fix NULL pointer dereference in alloc_pages_bulk_noprof

We triggered a NULL pointer dereference for ac.preferred_zoneref->zone in
alloc_pages_bulk_noprof() when the task is migrated between cpusets.

When cpuset is enabled, in prepare_alloc_pages(), ac->nodemask may be
&current->mems_allowed.  when first_zones_zonelist() is called to find
preferred_zoneref, the ac->nodemask may be modified concurrently if the
task is migrated between different cpusets.  Assuming we have 2 NUMA Node,
when traversing Node1 in ac->zonelist, the nodemask is 2, and when
traversing Node2 in ac->zonelist, the nodemask is 1.  As a result, the
ac->preferred_zoneref points to NULL zone.

In alloc_pages_bulk_noprof(), for_each_zone_zonelist_nodemask() finds a
allowable zone and calls zonelist_node_idx(ac.preferred_zoneref), leading
to NULL pointer dereference.

__alloc_pages_noprof() fixes this issue by checking NULL pointer in commit
ea57485af8f4 ("mm, page_alloc: fix check for NULL preferred_zone") and
commit df76cee6bbeb ("mm, page_alloc: remove redundant checks from alloc
fastpath").

To fix it, check NULL pointer for preferred_zoneref->zone.

Link: https://lkml.kernel.org/r/[email protected]
Fixes: 387ba26fb1cb ("mm/page_alloc: add a bulk page allocator")
Signed-off-by: Jinjiang Tu <[email protected]>
Reviewed-by: Vlastimil Babka <[email protected]>
Cc: Alexander Lobakin <[email protected]>
Cc: David Hildenbrand <[email protected]>
Cc: Kefeng Wang <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Nanyong Sun <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agomm, doc: update read_ahead_kb for MADV_HUGEPAGE
Yafang Shao [Wed, 13 Nov 2024 15:07:11 +0000 (23:07 +0800)]
mm, doc: update read_ahead_kb for MADV_HUGEPAGE

MADV_HUGEPAGE is a new addition to readahead with behavior distinct from
normal pages.  To prevent confusion, we should update the documentation
accordingly.

Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Yafang Shao <[email protected]>
Cc: David Hildenbrand <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agofs/proc/task_mmu: prevent integer overflow in pagemap_scan_get_args()
Dan Carpenter [Thu, 14 Nov 2024 08:59:32 +0000 (11:59 +0300)]
fs/proc/task_mmu: prevent integer overflow in pagemap_scan_get_args()

The "arg->vec_len" variable is a u64 that comes from the user at the start
of the function.  The "arg->vec_len * sizeof(struct page_region))"
multiplication can lead to integer wrapping.  Use size_mul() to avoid
that.

Also the size_add/mul() functions work on unsigned long so for 32bit
systems we need to ensure that "arg->vec_len" fits in an unsigned long.

Link: https://lkml.kernel.org/r/[email protected]
Fixes: 52526ca7fdb9 ("fs/proc/task_mmu: implement IOCTL to get and optionally clear info about PTEs")
Signed-off-by: Dan Carpenter <[email protected]>
Cc: Andrei Vagin <[email protected]>
Cc: Andrii Nakryiko <[email protected]>
Cc: Arnd Bergmann <[email protected]>
Cc: David Hildenbrand <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: Michał Mirosław <[email protected]>
Cc: Muhammad Usama Anjum <[email protected]>
Cc: Oscar Salvador <[email protected]>
Cc: Peter Xu <[email protected]>
Cc: Ryan Roberts <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agosched/task_stack: fix object_is_on_stack() for KASAN tagged pointers
Qun-Wei Lin [Wed, 13 Nov 2024 04:25:43 +0000 (12:25 +0800)]
sched/task_stack: fix object_is_on_stack() for KASAN tagged pointers

When CONFIG_KASAN_SW_TAGS and CONFIG_KASAN_STACK are enabled, the
object_is_on_stack() function may produce incorrect results due to the
presence of tags in the obj pointer, while the stack pointer does not have
tags.  This discrepancy can lead to incorrect stack object detection and
subsequently trigger warnings if CONFIG_DEBUG_OBJECTS is also enabled.

Example of the warning:

ODEBUG: object 3eff800082ea7bb0 is NOT on stack ffff800082ea0000, but annotated.
------------[ cut here ]------------
WARNING: CPU: 0 PID: 1 at lib/debugobjects.c:557 __debug_object_init+0x330/0x364
Modules linked in:
CPU: 0 UID: 0 PID: 1 Comm: swapper/0 Not tainted 6.12.0-rc5 #4
Hardware name: linux,dummy-virt (DT)
pstate: 600000c5 (nZCv daIF -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : __debug_object_init+0x330/0x364
lr : __debug_object_init+0x330/0x364
sp : ffff800082ea7b40
x29: ffff800082ea7b40 x28: 98ff0000c0164518 x27: 98ff0000c0164534
x26: ffff800082d93ec8 x25: 0000000000000001 x24: 1cff0000c00172a0
x23: 0000000000000000 x22: ffff800082d93ed0 x21: ffff800081a24418
x20: 3eff800082ea7bb0 x19: efff800000000000 x18: 0000000000000000
x17: 00000000000000ff x16: 0000000000000047 x15: 206b63617473206e
x14: 0000000000000018 x13: ffff800082ea7780 x12: 0ffff800082ea78e
x11: 0ffff800082ea790 x10: 0ffff800082ea79d x9 : 34d77febe173e800
x8 : 34d77febe173e800 x7 : 0000000000000001 x6 : 0000000000000001
x5 : feff800082ea74b8 x4 : ffff800082870a90 x3 : ffff80008018d3c4
x2 : 0000000000000001 x1 : ffff800082858810 x0 : 0000000000000050
Call trace:
 __debug_object_init+0x330/0x364
 debug_object_init_on_stack+0x30/0x3c
 schedule_hrtimeout_range_clock+0xac/0x26c
 schedule_hrtimeout+0x1c/0x30
 wait_task_inactive+0x1d4/0x25c
 kthread_bind_mask+0x28/0x98
 init_rescuer+0x1e8/0x280
 workqueue_init+0x1a0/0x3cc
 kernel_init_freeable+0x118/0x200
 kernel_init+0x28/0x1f0
 ret_from_fork+0x10/0x20
---[ end trace 0000000000000000 ]---
ODEBUG: object 3eff800082ea7bb0 is NOT on stack ffff800082ea0000, but annotated.
------------[ cut here ]------------

Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Qun-Wei Lin <[email protected]>
Cc: Andrew Yang <[email protected]>
Cc: AngeloGioacchino Del Regno <[email protected]>
Cc: Casper Li <[email protected]>
Cc: Catalin Marinas <[email protected]>
Cc: Chinwen Chang <[email protected]>
Cc: Kent Overstreet <[email protected]>
Cc: Matthias Brugger <[email protected]>
Cc: Pasha Tatashin <[email protected]>
Cc: Shakeel Butt <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agocrash, powerpc: default to CRASH_DUMP=n on PPC_BOOK3S_32
Dave Vasilevsky [Tue, 17 Sep 2024 16:37:20 +0000 (12:37 -0400)]
crash, powerpc: default to CRASH_DUMP=n on PPC_BOOK3S_32

Fixes boot failures on 6.9 on PPC_BOOK3S_32 machines using Open Firmware.
On these machines, the kernel refuses to boot from non-zero
PHYSICAL_START, which occurs when CRASH_DUMP is on.

Since most PPC_BOOK3S_32 machines boot via Open Firmware, it should
default to off for them.  Users booting via some other mechanism can still
turn it on explicitly.

Does not change the default on any other architectures for the
time being.

Link: https://lkml.kernel.org/r/[email protected]
Fixes: 75bc255a7444 ("crash: clean up kdump related config items")
Signed-off-by: Dave Vasilevsky <[email protected]>
Reported-by: Reimar Döffinger <[email protected]>
Closes: https://lists.debian.org/debian-powerpc/2024/07/msg00001.html
Acked-by: Michael Ellerman <[email protected]> [powerpc]
Acked-by: Baoquan He <[email protected]>
Cc: "Eric W. Biederman" <[email protected]>
Cc: John Paul Adrian Glaubitz <[email protected]>
Cc: Reimar Döffinger <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agomm/mremap: fix address wraparound in move_page_tables()
Jann Horn [Mon, 11 Nov 2024 19:34:30 +0000 (20:34 +0100)]
mm/mremap: fix address wraparound in move_page_tables()

On 32-bit platforms, it is possible for the expression `len + old_addr <
old_end` to be false-positive if `len + old_addr` wraps around.
`old_addr` is the cursor in the old range up to which page table entries
have been moved; so if the operation succeeded, `old_addr` is the *end* of
the old region, and adding `len` to it can wrap.

The overflow causes mremap() to mistakenly believe that PTEs have been
copied; the consequence is that mremap() bails out, but doesn't move the
PTEs back before the new VMA is unmapped, causing anonymous pages in the
region to be lost.  So basically if userspace tries to mremap() a
private-anon region and hits this bug, mremap() will return an error and
the private-anon region's contents appear to have been zeroed.

The idea of this check is that `old_end - len` is the original start
address, and writing the check that way also makes it easier to read; so
fix the check by rearranging the comparison accordingly.

(An alternate fix would be to refactor this function by introducing an
"orig_old_start" variable or such.)

Tested in a VM with a 32-bit X86 kernel; without the patch:

```
user@horn:~/big_mremap$ cat test.c
#define _GNU_SOURCE
#include <stdlib.h>
#include <stdio.h>
#include <err.h>
#include <sys/mman.h>

#define ADDR1 ((void*)0x60000000)
#define ADDR2 ((void*)0x10000000)
#define SIZE          0x50000000uL

int main(void) {
  unsigned char *p1 = mmap(ADDR1, SIZE, PROT_READ|PROT_WRITE,
      MAP_ANONYMOUS|MAP_PRIVATE|MAP_FIXED_NOREPLACE, -1, 0);
  if (p1 == MAP_FAILED)
    err(1, "mmap 1");
  unsigned char *p2 = mmap(ADDR2, SIZE, PROT_NONE,
      MAP_ANONYMOUS|MAP_PRIVATE|MAP_FIXED_NOREPLACE, -1, 0);
  if (p2 == MAP_FAILED)
    err(1, "mmap 2");
  *p1 = 0x41;
  printf("first char is 0x%02hhx\n", *p1);
  unsigned char *p3 = mremap(p1, SIZE, SIZE,
      MREMAP_MAYMOVE|MREMAP_FIXED, p2);
  if (p3 == MAP_FAILED) {
    printf("mremap() failed; first char is 0x%02hhx\n", *p1);
  } else {
    printf("mremap() succeeded; first char is 0x%02hhx\n", *p3);
  }
}
user@horn:~/big_mremap$ gcc -static -o test test.c
user@horn:~/big_mremap$ setarch -R ./test
first char is 0x41
mremap() failed; first char is 0x00
```

With the patch:

```
user@horn:~/big_mremap$ setarch -R ./test
first char is 0x41
mremap() succeeded; first char is 0x41
```

Link: https://lkml.kernel.org/r/[email protected]
Fixes: af8ca1c14906 ("mm/mremap: optimize the start addresses in move_page_tables()")
Signed-off-by: Jann Horn <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Reviewed-by: Lorenzo Stoakes <[email protected]>
Acked-by: Qi Zheng <[email protected]>
Reviewed-by: Liam R. Howlett <[email protected]>
Cc: Joel Fernandes (Google) <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agotools/mm: fix compile error
Motiejus JakÅ`tys [Tue, 12 Nov 2024 17:16:55 +0000 (19:16 +0200)]
tools/mm: fix compile error

Add a missing semicolon.

Link: https://lkml.kernel.org/r/[email protected]
Fixes: ece5897e5a10 ("tools/mm: -Werror fixes in page-types/slabinfo")
Signed-off-by: Motiejus JakÅ`tys <[email protected]>
Closes: https://github.com/NixOS/nixpkgs/issues/355369
Reviewed-by: SeongJae Park <[email protected]>
Reviewed-by: Vishal Moola (Oracle) <[email protected]>
Acked-by: Oleksandr Natalenko <[email protected]>
Cc: Wladislav Wiebe <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agomm, swap: fix allocation and scanning race with swapoff
Kairui Song [Tue, 12 Nov 2024 08:34:14 +0000 (16:34 +0800)]
mm, swap: fix allocation and scanning race with swapoff

There are two flags used to synchronize allocation and scanning with
swapoff: SWP_WRITEOK and SWP_SCANNING.

SWP_WRITEOK: Swapoff will first unset this flag, at this point any further
swap allocation or scanning on this device should just abort so no more
new entries will be referencing this device.  Swapoff will then unuse all
existing swap entries.

SWP_SCANNING: This flag is set when device is being scanned.  Swapoff will
wait for all scanner to stop before the final release of the swap device
structures to avoid UAF.  Note this flag is the highest used bit of
si->flags so it could be added up arithmetically, if there are multiple
scanner.

commit 5f843a9a3a1e ("mm: swap: separate SSD allocation from
scan_swap_map_slots()") ignored SWP_SCANNING and SWP_WRITEOK flags while
separating cluster allocation path from the old allocation path.  Add the
flags back to fix swapoff race.  The race is hard to trigger as si->lock
prevents most parallel operations, but si->lock could be dropped for
reclaim or discard.  This issue is found during code review.

This commit fixes this problem.  For SWP_SCANNING, Just like before, set
the flag before scan and remove it afterwards.

For SWP_WRITEOK, there are several places where si->lock could be dropped,
it will be error-prone and make the code hard to follow if we try to cover
these places one by one.  So just do one check before the real allocation,
which is also very similar like before.  With new cluster allocator it may
waste a bit of time iterating the clusters but won't take long, and
swapoff is not performance sensitive.

Link: https://lkml.kernel.org/r/[email protected]
Fixes: 5f843a9a3a1e ("mm: swap: separate SSD allocation from scan_swap_map_slots()")
Reported-by: "Huang, Ying" <[email protected]>
Closes: https://lore.kernel.org/linux-mm/[email protected]/
Signed-off-by: Kairui Song <[email protected]>
Cc: Barry Song <[email protected]>
Cc: Chris Li <[email protected]>
Cc: Hugh Dickins <[email protected]>
Cc: Kalesh Singh <[email protected]>
Cc: Ryan Roberts <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
4 months agoMerge tag 'amd-drm-fixes-6.12-2024-11-14' of https://gitlab.freedesktop.org/agd5f...
Dave Airlie [Thu, 14 Nov 2024 20:48:49 +0000 (06:48 +1000)]
Merge tag 'amd-drm-fixes-6.12-2024-11-14' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.12-2024-11-14:

amdgpu:
- PSR fix
- Panel replay fixes
- DML fix
- vblank power fix
- Fix video caps
- SMU 14.0 fix
- GPUVM fix
- MES 12 fix
- APU carve out fix
- DC vbios fix
- NBIO fix

Signed-off-by: Dave Airlie <[email protected]>
From: Alex Deucher <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agoMerge tag 'drm-misc-fixes-2024-11-14' of https://gitlab.freedesktop.org/drm/misc...
Dave Airlie [Thu, 14 Nov 2024 20:38:32 +0000 (06:38 +1000)]
Merge tag 'drm-misc-fixes-2024-11-14' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-fixes

Short summary of fixes pull:

bridge:
- tc358768: Fix DSI command tx

nouveau:
- Fix GSP AUX error handling
- dp: Handle retires for AUX CH transfers with GSP
- fw: Sync DMA after setup

panthor:
- Fix partial BO mappings to GPU

rockchip:
- vop: Avoid null-ptr deref in plane-state check

vmwgfx:
- Avoid null-ptr deref in surface creation

Signed-off-by: Dave Airlie <[email protected]>
From: Thomas Zimmermann <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/20241114142256.GA86810@2a02-2454-fd5e-fd00-4ce-489-4b34-bd1a.dyn6.pyur.net
4 months agoMerge tag 'drm-intel-fixes-2024-11-14' of https://gitlab.freedesktop.org/drm/i915...
Dave Airlie [Thu, 14 Nov 2024 20:18:34 +0000 (06:18 +1000)]
Merge tag 'drm-intel-fixes-2024-11-14' of https://gitlab.freedesktop.org/drm/i915/kernel into drm-fixes

- Don't load GSC on ARL-H and ARL-U if too old FW
- Avoid potential OOPS in enabling/disabling TV output

Signed-off-by: Dave Airlie <[email protected]>
From: Joonas Lahtinen <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agosched_ext: ops.cpu_acquire() should be called with SCX_KF_REST
Tejun Heo [Thu, 14 Nov 2024 18:50:58 +0000 (08:50 -1000)]
sched_ext: ops.cpu_acquire() should be called with SCX_KF_REST

ops.cpu_acquire() is currently called with 0 kf_maks which is interpreted as
SCX_KF_UNLOCKED which allows all unlocked kfuncs, but ops.cpu_acquire() is
called from balance_one() under the rq lock and should only be allowed call
kfuncs that are safe under the rq lock. Update it to use SCX_KF_REST.

Signed-off-by: Tejun Heo <[email protected]>
Cc: David Vernet <[email protected]>
Cc: Zhao Mengmeng <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Fixes: 245254f7081d ("sched_ext: Implement sched_ext_ops.cpu_acquire/release()")
4 months agoMerge tag 'net-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Thu, 14 Nov 2024 18:05:33 +0000 (10:05 -0800)]
Merge tag 'net-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

Pull networking fixes from Paolo Abeni:
 "Including fixes from bluetooth.

  Quite calm week. No new regression under investigation.

  Current release - regressions:

   - eth: revert "igb: Disable threaded IRQ for igb_msix_other"

  Current release - new code bugs:

   - bluetooth: btintel: direct exception event to bluetooth stack

  Previous releases - regressions:

   - core: fix data-races around sk->sk_forward_alloc

   - netlink: terminate outstanding dump on socket close

   - mptcp: error out earlier on disconnect

   - vsock: fix accept_queue memory leak

   - phylink: ensure PHY momentary link-fails are handled

   - eth: mlx5:
      - fix null-ptr-deref in add rule err flow
      - lock FTE when checking if active

   - eth: dwmac-mediatek: fix inverted handling of mediatek,mac-wol

  Previous releases - always broken:

   - sched: fix u32's systematic failure to free IDR entries for hnodes.

   - sctp: fix possible UAF in sctp_v6_available()

   - eth: bonding: add ns target multicast address to slave device

   - eth: mlx5: fix msix vectors to respect platform limit

   - eth: icssg-prueth: fix 1 PPS sync"

* tag 'net-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (38 commits)
  net: sched: u32: Add test case for systematic hnode IDR leaks
  selftests: bonding: add ns multicast group testing
  bonding: add ns target multicast address to slave device
  net: ti: icssg-prueth: Fix 1 PPS sync
  stmmac: dwmac-intel-plat: fix call balance of tx_clk handling routines
  net: Make copy_safe_from_sockptr() match documentation
  net: stmmac: dwmac-mediatek: Fix inverted handling of mediatek,mac-wol
  ipmr: Fix access to mfc_cache_list without lock held
  samples: pktgen: correct dev to DEV
  net: phylink: ensure PHY momentary link-fails are handled
  mptcp: pm: use _rcu variant under rcu_read_lock
  mptcp: hold pm lock when deleting entry
  mptcp: update local address flags when setting it
  net: sched: cls_u32: Fix u32's systematic failure to free IDR entries for hnodes.
  MAINTAINERS: Re-add cancelled Renesas driver sections
  Revert "igb: Disable threaded IRQ for igb_msix_other"
  Bluetooth: btintel: Direct exception event to bluetooth stack
  Bluetooth: hci_core: Fix calling mgmt_device_connected
  virtio/vsock: Improve MSG_ZEROCOPY error handling
  vsock: Fix sk_error_queue memory leak
  ...

4 months agoMerge tag 'bcachefs-2024-11-13' of git://evilpiepirate.org/bcachefs
Linus Torvalds [Thu, 14 Nov 2024 18:00:23 +0000 (10:00 -0800)]
Merge tag 'bcachefs-2024-11-13' of git://evilpiepirate.org/bcachefs

Pull bcachefs fixes from Kent Overstreet:
 "This fixes one minor regression from the btree cache fixes (in the
  scan_for_btree_nodes repair path) - and the shutdown path fix is the
  big one here, in terms of bugs closed:

   - Assorted tiny syzbot fixes

   - Shutdown path fix: "bch2_btree_write_buffer_flush_going_ro()"

     The shutdown path wasn't flushing the btree write buffer, leading
     to shutting down while we still had operations in flight. This
     fixes a whole slew of syzbot bugs, and undoubtedly other strange
     heisenbugs.

* tag 'bcachefs-2024-11-13' of git://evilpiepirate.org/bcachefs:
  bcachefs: Fix assertion pop in bch2_ptr_swab()
  bcachefs: Fix journal_entry_dev_usage_to_text() overrun
  bcachefs: Allow for unknown key types in backpointers fsck
  bcachefs: Fix assertion pop in topology repair
  bcachefs: Fix hidden btree errors when reading roots
  bcachefs: Fix validate_bset() repair path
  bcachefs: Fix missing validation for bch_backpointer.level
  bcachefs: Fix bch_member.btree_bitmap_shift validation
  bcachefs: bch2_btree_write_buffer_flush_going_ro()

4 months agotracing/ring-buffer: Clear all memory mapped CPU ring buffers on first recording
Steven Rostedt [Thu, 14 Nov 2024 16:28:25 +0000 (11:28 -0500)]
tracing/ring-buffer: Clear all memory mapped CPU ring buffers on first recording

The events of a memory mapped ring buffer from the previous boot should
not be mixed in with events from the current boot. There's meta data that
is used to handle KASLR so that function names can be shown properly.

Also, since the timestamps of the previous boot have no meaning to the
timestamps of the current boot, having them intermingled in a buffer can
also cause confusion because there could possibly be events in the future.

When a trace is activated the meta data is reset so that the pointers of
are now processed for the new address space. The trace buffers are reset
when tracing starts for the first time. The problem here is that the reset
only happens on online CPUs. If a CPU is offline, it does not get reset.

To demonstrate the issue, a previous boot had tracing enabled in the boot
mapped ring buffer on reboot. On the following boot, tracing has not been
started yet so the function trace from the previous boot is still visible.

 # trace-cmd show -B boot_mapped -c 3 | tail
          <idle>-0       [003] d.h2.   156.462395: __rcu_read_lock <-cpu_emergency_disable_virtualization
          <idle>-0       [003] d.h2.   156.462396: vmx_emergency_disable_virtualization_cpu <-cpu_emergency_disable_virtualization
          <idle>-0       [003] d.h2.   156.462396: __rcu_read_unlock <-__sysvec_reboot
          <idle>-0       [003] d.h2.   156.462397: stop_this_cpu <-__sysvec_reboot
          <idle>-0       [003] d.h2.   156.462397: set_cpu_online <-stop_this_cpu
          <idle>-0       [003] d.h2.   156.462397: disable_local_APIC <-stop_this_cpu
          <idle>-0       [003] d.h2.   156.462398: clear_local_APIC <-disable_local_APIC
          <idle>-0       [003] d.h2.   156.462574: mcheck_cpu_clear <-stop_this_cpu
          <idle>-0       [003] d.h2.   156.462575: mce_intel_feature_clear <-stop_this_cpu
          <idle>-0       [003] d.h2.   156.462575: lmce_supported <-mce_intel_feature_clear

Now, if CPU 3 is taken offline, and tracing is started on the memory
mapped ring buffer, the events from the previous boot in the CPU 3 ring
buffer is not reset. Now those events are using the meta data from the
current boot and produces just hex values.

 # echo 0 > /sys/devices/system/cpu/cpu3/online
 # trace-cmd start -B boot_mapped -p function
 # trace-cmd show -B boot_mapped -c 3 | tail
          <idle>-0       [003] d.h2.   156.462395: 0xffffffff9a1e3194 <-0xffffffff9a0f655e
          <idle>-0       [003] d.h2.   156.462396: 0xffffffff9a0a1d24 <-0xffffffff9a0f656f
          <idle>-0       [003] d.h2.   156.462396: 0xffffffff9a1e6bc4 <-0xffffffff9a0f7323
          <idle>-0       [003] d.h2.   156.462397: 0xffffffff9a0d12b4 <-0xffffffff9a0f732a
          <idle>-0       [003] d.h2.   156.462397: 0xffffffff9a1458d4 <-0xffffffff9a0d12e2
          <idle>-0       [003] d.h2.   156.462397: 0xffffffff9a0faed4 <-0xffffffff9a0d12e7
          <idle>-0       [003] d.h2.   156.462398: 0xffffffff9a0faaf4 <-0xffffffff9a0faef2
          <idle>-0       [003] d.h2.   156.462574: 0xffffffff9a0e3444 <-0xffffffff9a0d12ef
          <idle>-0       [003] d.h2.   156.462575: 0xffffffff9a0e4964 <-0xffffffff9a0d12ef
          <idle>-0       [003] d.h2.   156.462575: 0xffffffff9a0e3fb0 <-0xffffffff9a0e496f

Reset all CPUs when starting a boot mapped ring buffer for the first time,
and not just the online CPUs.

Fixes: 7a1d1e4b9639f ("tracing/ring-buffer: Add last_boot_info file to boot instance")
Signed-off-by: Steven Rostedt (Google) <[email protected]>
4 months agostatmount: retrieve security mount options
Christian Brauner [Thu, 14 Nov 2024 15:31:27 +0000 (16:31 +0100)]
statmount: retrieve security mount options

Add the ability to retrieve security mount options. Keep them separate
from filesystem specific mount options so it's easy to tell them apart.
Also allow to retrieve them separate from other mount options as most of
the time users won't be interested in security specific mount options.

Link: https://lore.kernel.org/r/20241114-radtour-ofenrohr-ff34b567b40a@brauner
Reviewed-by: Jeff Layton <[email protected]>
Signed-off-by: Christian Brauner <[email protected]>
4 months agoMerge tag 'asoc-fix-v6.12-rc7' of https://git.kernel.org/pub/scm/linux/kernel/git...
Takashi Iwai [Thu, 14 Nov 2024 15:40:15 +0000 (16:40 +0100)]
Merge tag 'asoc-fix-v6.12-rc7' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: Fixes for v6.12

Some last updates for v6.12, one quirk plus a couple of fixes.  One is a
minor fix for a relatively obscure driver and the other is a relatively
important fix for boot hangs with some audio graph based cards.

4 months agobtrfs: fix incorrect comparison for delayed refs
Josef Bacik [Wed, 13 Nov 2024 16:05:13 +0000 (11:05 -0500)]
btrfs: fix incorrect comparison for delayed refs

When I reworked delayed ref comparison in cf4f04325b2b ("btrfs: move
->parent and ->ref_root into btrfs_delayed_ref_node"), I made a mistake
and returned -1 for the case where ref1->ref_root was > than
ref2->ref_root.  This is a subtle bug that can result in improper
delayed ref running order, which can result in transaction aborts.

Fixes: cf4f04325b2b ("btrfs: move ->parent and ->ref_root into btrfs_delayed_ref_node")
CC: [email protected] # 6.10+
Reviewed-by: Filipe Manana <[email protected]>
Reviewed-by: Qu Wenruo <[email protected]>
Signed-off-by: Josef Bacik <[email protected]>
Reviewed-by: David Sterba <[email protected]>
Signed-off-by: David Sterba <[email protected]>
4 months agoRevert: "ring-buffer: Do not have boot mapped buffers hook to CPU hotplug"
Steven Rostedt [Thu, 14 Nov 2024 04:08:39 +0000 (23:08 -0500)]
Revert: "ring-buffer: Do not have boot mapped buffers hook to CPU hotplug"

A crash happened when testing cpu hotplug with respect to the memory
mapped ring buffers. It was assumed that the hot plug code was adding a
per CPU buffer that was already created that caused the crash. The real
problem was due to ref counting and was fixed by commit 2cf9733891a4
("ring-buffer: Fix refcount setting of boot mapped buffers").

When a per CPU buffer is created, it will not be created again even with
CPU hotplug, so the fix to not use CPU hotplug was a red herring. In fact,
it caused only the boot CPU buffer to be created, leaving the other CPU
per CPU buffers disabled.

Revert that change as it was not the culprit of the fix it was intended to
be.

Cc: Masami Hiramatsu <[email protected]>
Cc: Mathieu Desnoyers <[email protected]>
Link: https://lore.kernel.org/[email protected]
Fixes: 912da2c384d5 ("ring-buffer: Do not have boot mapped buffers hook to CPU hotplug")
Signed-off-by: Steven Rostedt (Google) <[email protected]>
4 months agonet: sched: u32: Add test case for systematic hnode IDR leaks
Alexandre Ferrieux [Wed, 13 Nov 2024 10:04:28 +0000 (11:04 +0100)]
net: sched: u32: Add test case for systematic hnode IDR leaks

Add a tdc test case to exercise the just-fixed systematic leak of
IDR entries in u32 hnode disposal. Given the IDR in question is
confined to the range [1..0x7FF], it is sufficient to create/delete
the same filter 2048 times to fill it up and get a nonzero exit
status from "tc filter add".

Signed-off-by: Alexandre Ferrieux <[email protected]>
Acked-by: Jamal Hadi Salim <[email protected]>
Reviewed-by: Victor Nogueira <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>
4 months agodrm/bridge: tc358768: Fix DSI command tx
Francesco Dolcini [Thu, 26 Sep 2024 14:12:46 +0000 (16:12 +0200)]
drm/bridge: tc358768: Fix DSI command tx

Wait for the command transmission to be completed in the DSI transfer
function polling for the dc_start bit to go back to idle state after the
transmission is started.

This is documented in the datasheet and failures to do so lead to
commands corruption.

Fixes: ff1ca6397b1d ("drm/bridge: Add tc358768 driver")
Cc: [email protected]
Signed-off-by: Francesco Dolcini <[email protected]>
Reviewed-by: Neil Armstrong <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Neil Armstrong <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agoMerge branch 'bonding-fix-ns-targets-not-work-on-hardware-nic'
Paolo Abeni [Thu, 14 Nov 2024 10:16:30 +0000 (11:16 +0100)]
Merge branch 'bonding-fix-ns-targets-not-work-on-hardware-nic'

Hangbin Liu says:

====================
bonding: fix ns targets not work on hardware NIC

The first patch fixed ns targets not work on hardware NIC when bonding
set arp_validate.

The second patch add a related selftest for bonding.

v4: Thanks Nikolay for the comments:
    use bond_slave_ns_maddrs_{add/del} with clear name
    fix comments typos
    remove _slave_set_ns_maddrs underscore directly
    update bond_option_arp_validate_set() change logic
v3: use ndisc_mc_map to convert the mcast mac address (Jay Vosburgh)
v2: only add/del mcast group on backup slaves when arp_validate is set (Jay Vosburgh)
    arp_validate doesn't support 3ad, tlb, alb. So let's only do it on ab mode.
====================

Link: https://patch.msgid.link/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>
4 months agoselftests: bonding: add ns multicast group testing
Hangbin Liu [Mon, 11 Nov 2024 10:16:50 +0000 (10:16 +0000)]
selftests: bonding: add ns multicast group testing

Add a test to make sure the backup slaves join correct multicast group
when arp_validate enabled and ns_ip6_target is set. Here is the result:

TEST: arp_validate (active-backup ns_ip6_target arp_validate 0)     [ OK ]
TEST: arp_validate (join mcast group)                               [ OK ]
TEST: arp_validate (active-backup ns_ip6_target arp_validate 1)     [ OK ]
TEST: arp_validate (join mcast group)                               [ OK ]
TEST: arp_validate (active-backup ns_ip6_target arp_validate 2)     [ OK ]
TEST: arp_validate (join mcast group)                               [ OK ]
TEST: arp_validate (active-backup ns_ip6_target arp_validate 3)     [ OK ]
TEST: arp_validate (join mcast group)                               [ OK ]
TEST: arp_validate (active-backup ns_ip6_target arp_validate 4)     [ OK ]
TEST: arp_validate (join mcast group)                               [ OK ]
TEST: arp_validate (active-backup ns_ip6_target arp_validate 5)     [ OK ]
TEST: arp_validate (join mcast group)                               [ OK ]
TEST: arp_validate (active-backup ns_ip6_target arp_validate 6)     [ OK ]
TEST: arp_validate (join mcast group)                               [ OK ]

Signed-off-by: Hangbin Liu <[email protected]>
Reviewed-by: Nikolay Aleksandrov <[email protected]>
Signed-off-by: Paolo Abeni <[email protected]>
4 months agobonding: add ns target multicast address to slave device
Hangbin Liu [Mon, 11 Nov 2024 10:16:49 +0000 (10:16 +0000)]
bonding: add ns target multicast address to slave device

Commit 4598380f9c54 ("bonding: fix ns validation on backup slaves")
tried to resolve the issue where backup slaves couldn't be brought up when
receiving IPv6 Neighbor Solicitation (NS) messages. However, this fix only
worked for drivers that receive all multicast messages, such as the veth
interface.

For standard drivers, the NS multicast message is silently dropped because
the slave device is not a member of the NS target multicast group.

To address this, we need to make the slave device join the NS target
multicast group, ensuring it can receive these IPv6 NS messages to validate
the slave’s status properly.

There are three policies before joining the multicast group:
1. All settings must be under active-backup mode (alb and tlb do not support
   arp_validate), with backup slaves and slaves supporting multicast.
2. We can add or remove multicast groups when arp_validate changes.
3. Other operations, such as enslaving, releasing, or setting NS targets,
   need to be guarded by arp_validate.

Fixes: 4e24be018eb9 ("bonding: add new parameter ns_targets")
Signed-off-by: Hangbin Liu <[email protected]>
Reviewed-by: Nikolay Aleksandrov <[email protected]>
Signed-off-by: Paolo Abeni <[email protected]>
4 months agofs: reduce pointer chasing in is_mgtime() test
Jeff Layton [Wed, 13 Nov 2024 14:17:51 +0000 (09:17 -0500)]
fs: reduce pointer chasing in is_mgtime() test

The is_mgtime test checks whether the FS_MGTIME flag is set in the
fstype. To get there from the inode though, we have to dereference 3
pointers.

Add a new IOP_MGTIME flag, and have inode_init_always() set that flag
when the fstype flag is set. Then, make is_mgtime test for IOP_MGTIME
instead.

Signed-off-by: Jeff Layton <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Jan Kara <[email protected]>
Signed-off-by: Christian Brauner <[email protected]>
4 months agovfs: make evict() use smp_mb__after_spinlock instead of smp_mb
Mateusz Guzik [Wed, 13 Nov 2024 15:51:03 +0000 (16:51 +0100)]
vfs: make evict() use smp_mb__after_spinlock instead of smp_mb

It literally directly follows a spin_lock() call.

This whacks an explicit barrier on x86-64.

Signed-off-by: Mateusz Guzik <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Jan Kara <[email protected]>
Signed-off-by: Christian Brauner <[email protected]>
4 months agonet: ti: icssg-prueth: Fix 1 PPS sync
Meghana Malladi [Mon, 11 Nov 2024 09:58:42 +0000 (15:28 +0530)]
net: ti: icssg-prueth: Fix 1 PPS sync

The first PPS latch time needs to be calculated by the driver
(in rounded off seconds) and configured as the start time
offset for the cycle. After synchronizing two PTP clocks
running as master/slave, missing this would cause master
and slave to start immediately with some milliseconds
drift which causes the PPS signal to never synchronize with
the PTP master.

Fixes: 186734c15886 ("net: ti: icssg-prueth: add packet timestamping and ptp support")
Signed-off-by: Meghana Malladi <[email protected]>
Reviewed-by: Vadim Fedorenko <[email protected]>
Reviewed-by: MD Danish Anwar <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>
4 months agodrm/vmwgfx: avoid null_ptr_deref in vmw_framebuffer_surface_create_handle
Chen Ridong [Tue, 29 Oct 2024 08:34:29 +0000 (08:34 +0000)]
drm/vmwgfx: avoid null_ptr_deref in vmw_framebuffer_surface_create_handle

The 'vmw_user_object_buffer' function may return NULL with incorrect
inputs. To avoid possible null pointer dereference, add a check whether
the 'bo' is NULL in the vmw_framebuffer_surface_create_handle.

Fixes: d6667f0ddf46 ("drm/vmwgfx: Fix handling of dumb buffers")
Signed-off-by: Chen Ridong <[email protected]>
Signed-off-by: Zack Rusin <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agostmmac: dwmac-intel-plat: fix call balance of tx_clk handling routines
Vitalii Mordan [Fri, 8 Nov 2024 17:33:34 +0000 (20:33 +0300)]
stmmac: dwmac-intel-plat: fix call balance of tx_clk handling routines

If the clock dwmac->tx_clk was not enabled in intel_eth_plat_probe,
it should not be disabled in any path.

Conversely, if it was enabled in intel_eth_plat_probe, it must be disabled
in all error paths to ensure proper cleanup.

Found by Linux Verification Center (linuxtesting.org) with Klever.

Fixes: 9efc9b2b04c7 ("net: stmmac: Add dwmac-intel-plat for GBE driver")
Signed-off-by: Vitalii Mordan <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agonet: Make copy_safe_from_sockptr() match documentation
Michal Luczaj [Sun, 10 Nov 2024 23:17:34 +0000 (00:17 +0100)]
net: Make copy_safe_from_sockptr() match documentation

copy_safe_from_sockptr()
  return copy_from_sockptr()
    return copy_from_sockptr_offset()
      return copy_from_user()

copy_from_user() does not return an error on fault. Instead, it returns a
number of bytes that were not copied. Have it handled.

Patch has a side effect: it un-breaks garbage input handling of
nfc_llcp_setsockopt() and mISDN's data_sock_setsockopt().

Fixes: 6309863b31dd ("net: add copy_safe_from_sockptr() helper")
Signed-off-by: Michal Luczaj <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agonet: stmmac: dwmac-mediatek: Fix inverted handling of mediatek,mac-wol
Nícolas F. R. A. Prado [Sat, 9 Nov 2024 15:16:32 +0000 (10:16 -0500)]
net: stmmac: dwmac-mediatek: Fix inverted handling of mediatek,mac-wol

The mediatek,mac-wol property is being handled backwards to what is
described in the binding: it currently enables PHY WOL when the property
is present and vice versa. Invert the driver logic so it matches the
binding description.

Fixes: fd1d62d80ebc ("net: stmmac: replace the use_phy_wol field with a flag")
Signed-off-by: Nícolas F. R. A. Prado <[email protected]>
Link: https://patch.msgid.link/20241109-mediatek-mac-wol-noninverted-v2-1-0e264e213878@collabora.com
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agoipmr: Fix access to mfc_cache_list without lock held
Breno Leitao [Fri, 8 Nov 2024 14:08:36 +0000 (06:08 -0800)]
ipmr: Fix access to mfc_cache_list without lock held

Accessing `mr_table->mfc_cache_list` is protected by an RCU lock. In the
following code flow, the RCU read lock is not held, causing the
following error when `RCU_PROVE` is not held. The same problem might
show up in the IPv6 code path.

6.12.0-rc5-kbuilder-01145-gbac17284bdcb #33 Tainted: G            E    N
-----------------------------
net/ipv4/ipmr_base.c:313 RCU-list traversed in non-reader section!!

rcu_scheduler_active = 2, debug_locks = 1
   2 locks held by RetransmitAggre/3519:
    #0: ffff88816188c6c0 (nlk_cb_mutex-ROUTE){+.+.}-{3:3}, at: __netlink_dump_start+0x8a/0x290
    #1: ffffffff83fcf7a8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_dumpit+0x6b/0x90

stack backtrace:
    lockdep_rcu_suspicious
    mr_table_dump
    ipmr_rtm_dumproute
    rtnl_dump_all
    rtnl_dumpit
    netlink_dump
    __netlink_dump_start
    rtnetlink_rcv_msg
    netlink_rcv_skb
    netlink_unicast
    netlink_sendmsg

This is not a problem per see, since the RTNL lock is held here, so, it
is safe to iterate in the list without the RCU read lock, as suggested
by Eric.

To alleviate the concern, modify the code to use
list_for_each_entry_rcu() with the RTNL-held argument.

The annotation will raise an error only if RTNL or RCU read lock are
missing during iteration, signaling a legitimate problem, otherwise it
will avoid this false positive.

This will solve the IPv6 case as well, since ip6mr_rtm_dumproute() calls
this function as well.

Signed-off-by: Breno Leitao <[email protected]>
Reviewed-by: David Ahern <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agosamples: pktgen: correct dev to DEV
Wei Fang [Tue, 12 Nov 2024 03:03:47 +0000 (11:03 +0800)]
samples: pktgen: correct dev to DEV

In the pktgen_sample01_simple.sh script, the device variable is uppercase
'DEV' instead of lowercase 'dev'. Because of this typo, the script cannot
enable UDP tx checksum.

Fixes: 460a9aa23de6 ("samples: pktgen: add UDP tx checksum support")
Signed-off-by: Wei Fang <[email protected]>
Reviewed-by: Simon Horman <[email protected]>
Acked-by: Jesper Dangaard Brouer <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agonet: phylink: ensure PHY momentary link-fails are handled
Russell King (Oracle) [Tue, 12 Nov 2024 16:20:00 +0000 (16:20 +0000)]
net: phylink: ensure PHY momentary link-fails are handled

Normally, phylib won't notify changes in quick succession. However, as
a result of commit 3e43b903da04 ("net: phy: Immediately call
adjust_link if only tx_lpi_enabled changes") this is no longer true -
it is now possible that phy_link_down() and phy_link_up() will both
complete before phylink's resolver has run, which means it'll miss that
pl->phy_state.link momentarily became false.

Rename "mac_link_dropped" to be more generic "link_failed" since it will
cover more than the MAC/PCS end of the link failing, and arrange to set
this in phylink_phy_change() if we notice that the PHY reports that the
link is down.

This will ensure that we capture an EEE reconfiguration event.

Fixes: 3e43b903da04 ("net: phy: Immediately call adjust_link if only tx_lpi_enabled changes")
Signed-off-by: Russell King (Oracle) <[email protected]>
Reviewed-by: Oleksij Rempel <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agoMerge branch 'mptcp-pm-a-few-more-fixes'
Jakub Kicinski [Thu, 14 Nov 2024 02:51:09 +0000 (18:51 -0800)]
Merge branch 'mptcp-pm-a-few-more-fixes'

Matthieu Baerts says:

====================
mptcp: pm: a few more fixes

Three small fixes related to the MPTCP path-manager:

- Patch 1: correctly reflect the backup flag to the corresponding local
  address entry of the userspace path-manager. A fix for v5.19.

- Patch 2: hold the PM lock when deleting an entry from the local
  addresses of the userspace path-manager to avoid messing up with this
  list. A fix for v5.19.

- Patch 3: use _rcu variant to iterate the in-kernel path-manager's
  local addresses list, when under rcu_read_lock(). A fix for v5.17.
====================

Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agomptcp: pm: use _rcu variant under rcu_read_lock
Matthieu Baerts (NGI0) [Tue, 12 Nov 2024 19:18:35 +0000 (20:18 +0100)]
mptcp: pm: use _rcu variant under rcu_read_lock

In mptcp_pm_create_subflow_or_signal_addr(), rcu_read_(un)lock() are
used as expected to iterate over the list of local addresses, but
list_for_each_entry() was used instead of list_for_each_entry_rcu() in
__lookup_addr(). It is important to use this variant which adds the
required READ_ONCE() (and diagnostic checks if enabled).

Because __lookup_addr() is also used in mptcp_pm_nl_set_flags() where it
is called under the pernet->lock and not rcu_read_lock(), an extra
condition is then passed to help the diagnostic checks making sure
either the associated spin lock or the RCU lock is held.

Fixes: 86e39e04482b ("mptcp: keep track of local endpoint still available for each msk")
Cc: [email protected]
Reviewed-by: Geliang Tang <[email protected]>
Signed-off-by: Matthieu Baerts (NGI0) <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agomptcp: hold pm lock when deleting entry
Geliang Tang [Tue, 12 Nov 2024 19:18:34 +0000 (20:18 +0100)]
mptcp: hold pm lock when deleting entry

When traversing userspace_pm_local_addr_list and deleting an entry from
it in mptcp_pm_nl_remove_doit(), msk->pm.lock should be held.

This patch holds this lock before mptcp_userspace_pm_lookup_addr_by_id()
and releases it after list_move() in mptcp_pm_nl_remove_doit().

Fixes: d9a4594edabf ("mptcp: netlink: Add MPTCP_PM_CMD_REMOVE")
Cc: [email protected]
Signed-off-by: Geliang Tang <[email protected]>
Reviewed-by: Matthieu Baerts (NGI0) <[email protected]>
Signed-off-by: Matthieu Baerts (NGI0) <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agomptcp: update local address flags when setting it
Geliang Tang [Tue, 12 Nov 2024 19:18:33 +0000 (20:18 +0100)]
mptcp: update local address flags when setting it

Just like in-kernel pm, when userspace pm does set_flags, it needs to send
out MP_PRIO signal, and also modify the flags of the corresponding address
entry in the local address list. This patch implements the missing logic.

Traverse all address entries on userspace_pm_local_addr_list to find the
local address entry, if bkup is true, set the flags of this entry with
FLAG_BACKUP, otherwise, clear FLAG_BACKUP.

Fixes: 892f396c8e68 ("mptcp: netlink: issue MP_PRIO signals from userspace PMs")
Cc: [email protected]
Signed-off-by: Geliang Tang <[email protected]>
Reviewed-by: Matthieu Baerts (NGI0) <[email protected]>
Signed-off-by: Matthieu Baerts (NGI0) <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
4 months agonouveau/dp: handle retries for AUX CH transfers with GSP.
Dave Airlie [Mon, 11 Nov 2024 03:41:25 +0000 (13:41 +1000)]
nouveau/dp: handle retries for AUX CH transfers with GSP.

eb284f4b3781 drm/nouveau/dp: Honor GSP link training retry timeouts

tried to fix a problem with panel retires, however it appears
the auxch also needs the same treatment, so add the same retry
wrapper around it.

This fixes some eDP panels after a suspend/resume cycle.

Fixes: eb284f4b3781 ("drm/nouveau/dp: Honor GSP link training retry timeouts")
Cc: [email protected]
Reviewed-by: Lyude Paul <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agonouveau: handle EBUSY and EAGAIN for GSP aux errors.
Dave Airlie [Mon, 11 Nov 2024 03:41:24 +0000 (13:41 +1000)]
nouveau: handle EBUSY and EAGAIN for GSP aux errors.

The upper layer transfer functions expect EBUSY as a return
for when retries should be done.

Fix the AUX error translation, but also check for both errors
in a few places.

Fixes: eb284f4b3781 ("drm/nouveau/dp: Honor GSP link training retry timeouts")
Cc: [email protected]
Reviewed-by: Lyude Paul <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agonouveau: fw: sync dma after setup is called.
Dave Airlie [Tue, 12 Nov 2024 19:57:03 +0000 (05:57 +1000)]
nouveau: fw: sync dma after setup is called.

When this code moved to non-coherent allocator the sync was put too
early for some firmwares which called the setup function, move the
sync down after the setup function.

Reported-by: Diogo Ivo <[email protected]>
Tested-by: Diogo Ivo <[email protected]>
Reviewed-by: Lyude Paul <[email protected]>
Fixes: 9b340aeb26d5 ("nouveau/firmware: use dma non-coherent allocator")
Cc: [email protected]
Signed-off-by: Dave Airlie <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
4 months agoMerge tag 'pm-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Linus Torvalds [Wed, 13 Nov 2024 21:32:51 +0000 (13:32 -0800)]
Merge tag 'pm-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

Pull power management fix from Rafael Wysocki:
 "Fix a locking issue in the asymmetric CPU capacity setup code in the
  intel_pstate driver that may lead to a deadlock if CPU online/offline
  runs in parallel with the code in question, which is unlikely but not
  impossible (Rafael Wysocki)"

* tag 'pm-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  cpufreq: intel_pstate: Rearrange locking in hybrid_init_cpu_capacity_scaling()

4 months agoMerge tag 'tpmdd-next-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 13 Nov 2024 21:28:58 +0000 (13:28 -0800)]
Merge tag 'tpmdd-next-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/jarkko/linux-tpmdd

Pull tpm fixes from Jarkko Sakkinen:
 "Two bug fixes for TPM bus encryption (the remaining reported issues in
  the feature)"

* tag 'tpmdd-next-6.12-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/jarkko/linux-tpmdd:
  tpm: Disable TPM on tpm2_create_primary() failure
  tpm: Opt-in in disable PCR integrity protection

4 months agodrm/xe/oa: Fix "Missing outer runtime PM protection" warning
Ashutosh Dixit [Sat, 9 Nov 2024 03:20:03 +0000 (19:20 -0800)]
drm/xe/oa: Fix "Missing outer runtime PM protection" warning

Fix the following drm_WARN:

[953.586396] xe 0000:00:02.0: [drm] Missing outer runtime PM protection
...
<4> [953.587090]  ? xe_pm_runtime_get_noresume+0x8d/0xa0 [xe]
<4> [953.587208]  guc_exec_queue_add_msg+0x28/0x130 [xe]
<4> [953.587319]  guc_exec_queue_fini+0x3a/0x40 [xe]
<4> [953.587425]  xe_exec_queue_destroy+0xb3/0xf0 [xe]
<4> [953.587515]  xe_oa_release+0x9c/0xc0 [xe]

Suggested-by: John Harrison <[email protected]>
Suggested-by: Matthew Brost <[email protected]>
Fixes: e936f885f1e9 ("drm/xe/oa/uapi: Expose OA stream fd")
Cc: [email protected]
Signed-off-by: Ashutosh Dixit <[email protected]>
Reviewed-by: Matthew Brost <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
(cherry picked from commit b107c63d2953907908fd0cafb0e543b3c3167b75)
Signed-off-by: Lucas De Marchi <[email protected]>
4 months agotpm: Disable TPM on tpm2_create_primary() failure
Jarkko Sakkinen [Wed, 13 Nov 2024 18:35:39 +0000 (20:35 +0200)]
tpm: Disable TPM on tpm2_create_primary() failure

The earlier bug fix misplaced the error-label when dealing with the
tpm2_create_primary() return value, which the original completely ignored.

Cc: [email protected]
Reported-by: Christoph Anton Mitterer <[email protected]>
Closes: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1087331
Fixes: cc7d8594342a ("tpm: Rollback tpm2_load_null()")
Signed-off-by: Jarkko Sakkinen <[email protected]>
4 months agotpm: Opt-in in disable PCR integrity protection
Jarkko Sakkinen [Wed, 13 Nov 2024 05:54:12 +0000 (07:54 +0200)]
tpm: Opt-in in disable PCR integrity protection

The initial HMAC session feature added TPM bus encryption and/or integrity
protection to various in-kernel TPM operations. This can cause performance
bottlenecks with IMA, as it heavily utilizes PCR extend operations.

In order to mitigate this performance issue, introduce a kernel
command-line parameter to the TPM driver for disabling the integrity
protection for PCR extend operations (i.e. TPM2_PCR_Extend).

Cc: James Bottomley <[email protected]>
Link: https://lore.kernel.org/linux-integrity/[email protected]/
Fixes: 6519fea6fd37 ("tpm: add hmac checks to tpm2_pcr_extend()")
Tested-by: Mimi Zohar <[email protected]>
Co-developed-by: Roberto Sassu <[email protected]>
Signed-off-by: Roberto Sassu <[email protected]>
Co-developed-by: Mimi Zohar <[email protected]>
Signed-off-by: Mimi Zohar <[email protected]>
Signed-off-by: Jarkko Sakkinen <[email protected]>
4 months agojbd2: Fix comment describing journal_init_common()
Daniel Martín Gómez [Thu, 7 Nov 2024 14:45:38 +0000 (15:45 +0100)]
jbd2: Fix comment describing journal_init_common()

The code indicates that journal_init_common() fills the journal_t object
it returns while the comment incorrectly states that only a few fields are
initialised.  Also, the comment claims that journal structures could be
created from scratch which isn't possible as journal_init_common() calls
journal_load_superblock() which loads and checks journal superblock from
disk.

Signed-off-by: Daniel Martín Gómez <[email protected]>
Reviewed-by: Jan Kara <[email protected]>
Reviewed-by: Zhang Yi <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Theodore Ts'o <[email protected]>
4 months agoext4: prevent an infinite loop in the lazyinit thread
Mathieu Othacehe [Wed, 6 Nov 2024 13:47:41 +0000 (14:47 +0100)]
ext4: prevent an infinite loop in the lazyinit thread

Use ktime_get_ns instead of ktime_get_real_ns when computing the lr_timeout
not to be affected by system time jumps.

Use a boolean instead of the MAX_JIFFY_OFFSET value to determine whether
the next_wakeup value has been set. Comparing elr->lr_next_sched to
MAX_JIFFY_OFFSET can cause the lazyinit thread to loop indefinitely.

Co-developed-by: Lukas Skupinski <[email protected]>
Signed-off-by: Lukas Skupinski <[email protected]>
Signed-off-by: Mathieu Othacehe <[email protected]>
Reviewed-by: Jan Kara <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Theodore Ts'o <[email protected]>
4 months agoext4: use struct_size() to improve ext4_htree_store_dirent()
Thorsten Blum [Tue, 5 Nov 2024 10:33:54 +0000 (11:33 +0100)]
ext4: use struct_size() to improve ext4_htree_store_dirent()

Inline and use struct_size() to calculate the number of bytes to
allocate for new_fn and remove the local variable len.

Reviewed-by: Jan Kara <[email protected]>
Signed-off-by: Thorsten Blum <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Theodore Ts'o <[email protected]>
4 months agoext4: annotate struct fname with __counted_by()
Thorsten Blum [Tue, 5 Nov 2024 10:18:14 +0000 (11:18 +0100)]
ext4: annotate struct fname with __counted_by()

Add the __counted_by compiler attribute to the flexible array member
name to improve access bounds-checking via CONFIG_UBSAN_BOUNDS and
CONFIG_FORTIFY_SOURCE.

Reviewed-by: Jan Kara <[email protected]>
Signed-off-by: Thorsten Blum <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Theodore Ts'o <[email protected]>
4 months agojbd2: avoid dozens of -Wflex-array-member-not-at-end warnings
Gustavo A. R. Silva [Fri, 1 Nov 2024 20:45:23 +0000 (14:45 -0600)]
jbd2: avoid dozens of -Wflex-array-member-not-at-end warnings

-Wflex-array-member-not-at-end was introduced in GCC-14, and we
are getting ready to enable it, globally.

Use the `DEFINE_RAW_FLEX()` helper for an on-stack definition of
a flexible structure (`struct shash_desc`) where the size of the
flexible-array member (`__ctx`) is known at compile-time, and
refactor the rest of the code, accordingly.

So, with this, fix 77 of the following warnings:

include/linux/jbd2.h:1800:35: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]

Signed-off-by: Gustavo A. R. Silva <[email protected]>
Reviewed-by: Jan Kara <[email protected]>
Link: https://patch.msgid.link/ZyU94w0IALVhc9Jy@kspp
Signed-off-by: Theodore Ts'o <[email protected]>
4 months agoext4: use str_yes_no() helper function
Thorsten Blum [Mon, 21 Oct 2024 10:00:57 +0000 (12:00 +0200)]
ext4: use str_yes_no() helper function

Remove hard-coded strings by using the str_yes_no() helper function.

Signed-off-by: Thorsten Blum <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Theodore Ts'o <[email protected]>
4 months agoMerge tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Linus Torvalds [Wed, 13 Nov 2024 17:14:19 +0000 (09:14 -0800)]
Merge tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf

Pull bpf fixes from Daniel Borkmann:

 - Fix a mismatching RCU unlock flavor in bpf_out_neigh_v6 (Jiawei Ye)

 - Fix BPF sockmap with kTLS to reject vsock and unix sockets upon kTLS
   context retrieval (Zijian Zhang)

 - Fix BPF bits iterator selftest for s390x (Hou Tao)

* tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
  bpf: Fix mismatched RCU unlock flavour in bpf_out_neigh_v6
  bpf: Add sk_is_inet and IS_ICSK check in tls_sw_has_ctx_tx/rx
  selftests/bpf: Use -4095 as the bad address for bits iterator

4 months agoMerge tag 'loongarch-fixes-6.12-2' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 13 Nov 2024 17:09:00 +0000 (09:09 -0800)]
Merge tag 'loongarch-fixes-6.12-2' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson

Pull LoongArch fixes from Huacai Chen:

 - fix possible CPUs setup logical-physical CPU mapping, in order to
   avoid CPU hotplug issue

 - fix some KASAN bugs

 - fix AP booting issue in VM mode

 - some trivial cleanups

* tag 'loongarch-fixes-6.12-2' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:
  LoongArch: Fix AP booting issue in VM mode
  LoongArch: Add WriteCombine shadow mapping in KASAN
  LoongArch: Disable KASAN if PGDIR_SIZE is too large for cpu_vabits
  LoongArch: Make KASAN work with 5-level page-tables
  LoongArch: Define a default value for VM_DATA_DEFAULT_FLAGS
  LoongArch: Fix early_numa_add_cpu() usage for FDT systems
  LoongArch: For all possible CPUs setup logical-physical CPU mapping

4 months agoMerge tag 'mm-hotfixes-stable-2024-11-12-16-39' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Wed, 13 Nov 2024 16:58:11 +0000 (08:58 -0800)]
Merge tag 'mm-hotfixes-stable-2024-11-12-16-39' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

Pull misc fixes from Andrew Morton:
 "10 hotfixes, 7 of which are cc:stable. 7 are MM, 3 are not. All
  singletons"

* tag 'mm-hotfixes-stable-2024-11-12-16-39' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
  mm: swapfile: fix cluster reclaim work crash on rotational devices
  selftests: hugetlb_dio: fixup check for initial conditions to skip in the start
  mm/thp: fix deferred split queue not partially_mapped: fix
  mm/gup: avoid an unnecessary allocation call for FOLL_LONGTERM cases
  nommu: pass NULL argument to vma_iter_prealloc()
  ocfs2: fix UBSAN warning in ocfs2_verify_volume()
  nilfs2: fix null-ptr-deref in block_dirty_buffer tracepoint
  nilfs2: fix null-ptr-deref in block_touch_buffer tracepoint
  mm: page_alloc: move mlocked flag clearance into free_pages_prepare()
  mm: count zeromap read and set for swapout and swapin

4 months agolibfs: kill empty_dir_getattr()
Al Viro [Sun, 3 Nov 2024 20:20:15 +0000 (15:20 -0500)]
libfs: kill empty_dir_getattr()

It's used only to initialize ->getattr in one inode_operations instance
(empty_dir_inode_operations) and its behaviour had always been equivalent
to what we get with NULL ->getattr.

Just remove that initializer, along with empty_dir_getattr() itself.
While we are at it, the same instance has ->permission initialized to
generic_permission, which is what NULL ->permission ends up doing.
Again, no point keeping it.

Reviewed-by: Christian Brauner <[email protected]>
Signed-off-by: Al Viro <[email protected]>
4 months agofs: Simplify getattr interface function checking AT_GETATTR_NOSEC flag
Stefan Berger [Fri, 1 Nov 2024 19:37:03 +0000 (15:37 -0400)]
fs: Simplify getattr interface function checking AT_GETATTR_NOSEC flag

Commit 8a924db2d7b5 ("fs: Pass AT_GETATTR_NOSEC flag to getattr interface
function")' introduced the AT_GETATTR_NOSEC flag to ensure that the
call paths only call vfs_getattr_nosec if it is set instead of vfs_getattr.
Now, simplify the getattr interface functions of filesystems where the flag
AT_GETATTR_NOSEC is checked.

There is only a single caller of inode_operations getattr function and it
is located in fs/stat.c in vfs_getattr_nosec. The caller there is the only
one from which the AT_GETATTR_NOSEC flag is passed from.

Two filesystems are checking this flag in .getattr and the flag is always
passed to them unconditionally from only vfs_getattr_nosec:

- ecryptfs:  Simplify by always calling vfs_getattr_nosec in
             ecryptfs_getattr. From there the flag is passed to no other
             function and this function is not called otherwise.

- overlayfs: Simplify by always calling vfs_getattr_nosec in
             ovl_getattr. From there the flag is passed to no other
             function and this function is not called otherwise.

The query_flags in vfs_getattr_nosec will mask-out AT_GETATTR_NOSEC from
any caller using AT_STATX_SYNC_TYPE as mask so that the flag is not
important inside this function. Also, since no filesystem is checking the
flag anymore, remove the flag entirely now, including the BUG_ON check that
never triggered.

The net change of the changes here combined with the original commit is
that ecryptfs and overlayfs do not call vfs_getattr but only
vfs_getattr_nosec.

Fixes: 8a924db2d7b5 ("fs: Pass AT_GETATTR_NOSEC flag to getattr interface function")
Reported-by: Al Viro <[email protected]>
Closes: https://lore.kernel.org/linux-fsdevel/20241101011724.GN1350452@ZenIV/T/#u
Cc: Tyler Hicks <[email protected]>
Cc: [email protected]
Cc: Miklos Szeredi <[email protected]>
Cc: Amir Goldstein <[email protected]>
Cc: [email protected]
Cc: Christian Brauner <[email protected]>
Cc: [email protected]
Reviewed-by: Christian Brauner <[email protected]>
Signed-off-by: Stefan Berger <[email protected]>
Signed-off-by: Al Viro <[email protected]>
4 months agofs/stat.c: switch to CLASS(fd_raw)
Al Viro [Sun, 20 Oct 2024 00:48:30 +0000 (20:48 -0400)]
fs/stat.c: switch to CLASS(fd_raw)

... and use fd_empty() consistently

Reviewed-by: Christian Brauner <[email protected]>
Signed-off-by: Al Viro <[email protected]>
4 months agokill getname_statx_lookup_flags()
Al Viro [Fri, 20 Sep 2024 04:38:59 +0000 (00:38 -0400)]
kill getname_statx_lookup_flags()

LOOKUP_EMPTY is ignored by the only remaining user, and without
that 'getname_' prefix makes no sense.

Remove LOOKUP_EMPTY part, rename to statx_lookup_flags() and make
static.  It most likely is _not_ statx() specific, either, but
that's the next step.

Reviewed-by: Christian Brauner <[email protected]>
Signed-off-by: Al Viro <[email protected]>
4 months agoio_statx_prep(): use getname_uflags()
Al Viro [Fri, 20 Sep 2024 04:26:57 +0000 (00:26 -0400)]
io_statx_prep(): use getname_uflags()

the only thing in flags getname_flags() ever cares about is
LOOKUP_EMPTY; anything else is none of its damn business.

Reviewed-by: Jens Axboe <[email protected]>
Reviewed-by: Christian Brauner <[email protected]>
Signed-off-by: Al Viro <[email protected]>
4 months agostatmount: add flag to retrieve unescaped options
Miklos Szeredi [Tue, 12 Nov 2024 10:10:04 +0000 (11:10 +0100)]
statmount: add flag to retrieve unescaped options

Filesystem options can be retrieved with STATMOUNT_MNT_OPTS, which
returns a string of comma separated options, where some characters are
escaped using the \OOO notation.

Add a new flag, STATMOUNT_OPT_ARRAY, which instead returns the raw
option values separated with '\0' charaters.

Since escaped charaters are rare, this inteface is preferable for
non-libmount users which likley don't want to deal with option
de-escaping.

Example code:

if (st->mask & STATMOUNT_OPT_ARRAY) {
const char *opt = st->str + st->opt_array;

for (unsigned int i = 0; i < st->opt_num; i++) {
printf("opt_array[%i]: <%s>\n", i, opt);
opt += strlen(opt) + 1;
}
}

Example ouput:

(1) mnt_opts: <lowerdir+=/l\054w\054r,lowerdir+=/l\054w\054r1,upperdir=/upp\054r,workdir=/w\054rk,redirect_dir=nofollow,uuid=null>

(2) opt_array[0]: <lowerdir+=/l,w,r>
    opt_array[1]: <lowerdir+=/l,w,r1>
    opt_array[2]: <upperdir=/upp,r>
    opt_array[3]: <workdir=/w,rk>
    opt_array[4]: <redirect_dir=nofollow>
    opt_array[5]: <uuid=null>

Signed-off-by: Miklos Szeredi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Acked-by: Jeff Layton <[email protected]>
[brauner: tweak variable naming and parsing add example output]
Signed-off-by: Christian Brauner <[email protected]>
This page took 0.149806 seconds and 4 git commands to generate.