This reverts commit ffb6cf19e06334062744b7e3493f71e500964f8e.

Users reported regressions due to enabling multi-grained timestamps
unconditionally. As no clear consensus on a solution has come up and the
discussion has gone back to the drawing board revert the infrastructure
changes for. If it isn't code that's here to stay, make it go away.

Message-ID: <20230920-keine-eile-c9755b5825db@brauner>
Acked-by: Jan Kara <jack@suse.cz>
Acked-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>

Pull superblock updates from Christian Brauner:
 "This contains the super rework that was ready for this cycle. The
  first part changes the order of how we open block devices and allocate
  superblocks, contains various cleanups, simplifications, and a new
  mechanism to wait on superblock state changes.

  This unblocks work to ultimately limit the number of writers to a
  block device. Jan has already scheduled follow-up work that will be
  ready for v6.7 and allows us to restrict the number of writers to a
  given block device. That series builds on this work right here.

  The second part contains filesystem freezing updates.

  Overview:

  The generic superblock changes are rougly organized as follows
  (ignoring additional minor cleanups):

   (1) Removal of the bd_super member from struct block_device.

       This was a very odd back pointer to struct super_block with
       unclear rules. For all relevant places we have other means to get
       the same information so just get rid of this.

   (2) Simplify rules for superblock cleanup.

       Roughly, everything that is allocated during fs_context
       initialization and that's stored in fs_context->s_fs_info needs
       to be cleaned up by the fs_context->free() implementation before
       the superblock allocation function has been called successfully.

       After sget_fc() returned fs_context->s_fs_info has been
       transferred to sb->s_fs_info at which point sb->kill_sb() if
       fully responsible for cleanup. Adhering to these rules means that
       cleanup of sb->s_fs_info in fill_super() is to be avoided as it's
       brittle and inconsistent.

       Cleanup shouldn't be duplicated between sb->put_super() as
       sb->put_super() is only called if sb->s_root has been set aka
       when the filesystem has been successfully born (SB_BORN). That
       complexity should be avoided.

       This also means that block devices are to be closed in
       sb->kill_sb() instead of sb->put_super(). More details in the
       lower section.

   (3) Make it possible to lookup or create a superblock before opening
       block devices

       There's a subtle dependency on (2) as some filesystems did rely
       on fill_super() to be called in order to correctly clean up
       sb->s_fs_info. All these filesystems have been fixed.

   (4) Switch most filesystem to follow the same logic as the generic
       mount code now does as outlined in (3).

   (5) Use the superblock as the holder of the block device. We can now
       easily go back from block device to owning superblock.

   (6) Export and extend the generic fs_holder_ops and use them as
       holder ops everywhere and remove the filesystem specific holder
       ops.

   (7) Call from the block layer up into the filesystem layer when the
       block device is removed, allowing to shut down the filesystem
       without risk of deadlocks.

   (8) Get rid of get_super().

       We can now easily go back from the block device to owning
       superblock and can call up from the block layer into the
       filesystem layer when the device is removed. So no need to wade
       through all registered superblock to find the owning superblock
       anymore"

Link: https://lore.kernel.org/lkml/20230824-prall-intakt-95dbffdee4a0@brauner/

* tag 'v6.6-vfs.super' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: (47 commits)
  super: use higher-level helper for {freeze,thaw}
  super: wait until we passed kill super
  super: wait for nascent superblocks
  super: make locking naming consistent
  super: use locking helpers
  fs: simplify invalidate_inodes
  fs: remove get_super
  block: call into the file system for ioctl BLKFLSBUF
  block: call into the file system for bdev_mark_dead
  block: consolidate __invalidate_device and fsync_bdev
  block: drop the "busy inodes on changed media" log message
  dasd: also call __invalidate_device when setting the device offline
  amiflop: don't call fsync_bdev in FDFMTBEG
  floppy: call disk_force_media_change when changing the format
  block: simplify the disk_force_media_change interface
  nbd: call blk_mark_disk_dead in nbd_clear_sock_ioctl
  xfs use fs_holder_ops for the log and RT devices
  xfs: drop s_umount over opening the log and RT devices
  ext4: use fs_holder_ops for the log device
  ext4: drop s_umount over opening the log device
  ...

Pull vfs timestamp updates from Christian Brauner:
 "This adds VFS support for multi-grain timestamps and converts tmpfs,
  xfs, ext4, and btrfs to use them. This carries acks from all relevant
  filesystems.

  The VFS always uses coarse-grained timestamps when updating the ctime
  and mtime after a change. This has the benefit of allowing filesystems
  to optimize away a lot of 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 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 introduces fine-grained timestamps that are used when they are
  actively queried.

  This uses the 31st bit of the ctime tv_nsec field to indicate that
  something has queried the inode for the mtime or ctime. When this flag
  is set, on the next mtime or ctime update, the kernel will fetch a
  fine-grained timestamp instead of the usual coarse-grained one.

  As POSIX generally mandates that when the mtime changes, the ctime
  must also change the kernel always stores normalized ctime values, so
  only the first 30 bits of the tv_nsec field are ever used.

  Filesytems can opt into this behavior by setting the FS_MGTIME flag in
  the fstype. Filesystems that don't set this flag will continue to use
  coarse-grained timestamps.

  Various preparatory changes, fixes and cleanups are included:

   - Fixup all relevant places where POSIX requires updating ctime
     together with mtime. This is a wide-range of places and all
     maintainers provided necessary Acks.

   - Add new accessors for inode->i_ctime directly and change all
     callers to rely on them. Plain accesses to inode->i_ctime are now
     gone and it is accordingly rename to inode->__i_ctime and commented
     as requiring accessors.

   - Extend generic_fillattr() to pass in a request mask mirroring in a
     sense the statx() uapi. This allows callers to pass in a request
     mask to only get a subset of attributes filled in.

   - Rework timestamp updates so it's possible to drop the @now
     parameter the update_time() inode operation and associated helpers.

   - Add inode_update_timestamps() and convert all filesystems to it
     removing a bunch of open-coding"

* tag 'v6.6-vfs.ctime' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: (107 commits)
  btrfs: convert to multigrain timestamps
  ext4: switch to multigrain timestamps
  xfs: switch to multigrain timestamps
  tmpfs: add support for multigrain timestamps
  fs: add infrastructure for multigrain timestamps
  fs: drop the timespec64 argument from update_time
  xfs: have xfs_vn_update_time gets its own timestamp
  fat: make fat_update_time get its own timestamp
  fat: remove i_version handling from fat_update_time
  ubifs: have ubifs_update_time use inode_update_timestamps
  btrfs: have it use inode_update_timestamps
  fs: drop the timespec64 arg from generic_update_time
  fs: pass the request_mask to generic_fillattr
  fs: remove silly warning from current_time
  gfs2: fix timestamp handling on quota inodes
  fs: rename i_ctime field to __i_ctime
  selinux: convert to ctime accessor functions
  security: convert to ctime accessor functions
  apparmor: convert to ctime accessor functions
  sunrpc: convert to ctime accessor functions
  ...

kill_dirty has always been true for a long time, so hard code it and
remove the unused return value.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Message-Id: <20230811100828.1897174-18-hch@lst.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>

The only remaining consumer is new_inode, where it showed up in 2001 as
commit c37fa164f793 ("v2.4.9.9 -> v2.4.9.10") in a historical repo [1]
with a changelog which does not mention it.

Since then the line got only touched up to keep compiling.

While it may have been of benefit back in the day, it is guaranteed to
at best not get in the way in the multicore setting -- as the code
performs *a lot* of work between the prefetch and actual lock acquire,
any contention means the cacheline is already invalid by the time the
routine calls spin_lock().  It adds spurious traffic, for short.

On top of it prefetch is notoriously tricky to use for single-threaded
purposes, making it questionable from the get go.

As such, remove it.

I admit upfront I did not see value in benchmarking this change, but I
can do it if that is deemed appropriate.

Removal from new_inode and of the entire thing are in the same patch as
requested by Linus, so whatever weird looks can be directed at that guy.

Link: https://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git/commit/fs/inode.c?id=c37fa164f793735b32aa3f53154ff1a7659e6442 [1]
Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

The VFS always uses 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 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.

What we need is a way to only use fine-grained timestamps when they are
being actively queried.

POSIX generally mandates that when the the mtime changes, the ctime must
also change. The kernel always stores normalized ctime values, so only
the first 30 bits of the tv_nsec field are ever used.

Use the 31st bit of the ctime tv_nsec field to indicate that something
has queried the inode for the mtime or ctime. When this flag is set,
on the next mtime or ctime update, the kernel will fetch a fine-grained
timestamp instead of the usual coarse-grained one.

Filesytems can opt into this behavior by setting the FS_MGTIME flag in
the fstype. Filesystems that don't set this flag will continue to use
coarse-grained timestamps.

Later patches will convert individual filesystems to use the new
infrastructure.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Message-Id: <20230807-mgctime-v7-9-d1dec143a704@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>

Now that all of the update_time operations are prepared for it, we can
drop the timespec64 argument from the update_time operation. Do that and
remove it from some associated functions like inode_update_time and
inode_needs_update_time.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Message-Id: <20230807-mgctime-v7-8-d1dec143a704@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>

In future patches we're going to change how the ctime is updated
to keep track of when it has been queried. The way that the update_time
operation works (and a lot of its callers) make this difficult, since
they grab a timestamp early and then pass it down to eventually be
copied into the inode.

All of the existing update_time callers pass in the result of
current_time() in some fashion. Drop the "time" parameter from
generic_update_time, and rework it to fetch its own timestamp.

This change means that an update_time could fetch a different timestamp
than was seen in inode_needs_update_time. update_time is only ever
called with one of two flag combinations: Either S_ATIME is set, or
S_MTIME|S_CTIME|S_VERSION are set.

With this change we now treat the flags argument as an indicator that
some value needed to be updated when last checked, rather than an
indication to update specific timestamps.

Rework the logic for updating the timestamps and put it in a new
inode_update_timestamps helper that other update_time routines can use.
S_ATIME is as treated as we always have, but if any of the other three
are set, then we attempt to update all three.

Also, some callers of generic_update_time need to know what timestamps
were actually updated. Change it to return an S_* flag mask to indicate
that and rework the callers to expect it.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Message-Id: <20230807-mgctime-v7-3-d1dec143a704@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>

An inode with no superblock? Unpossible!

Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Message-Id: <20230807-mgctime-v7-1-d1dec143a704@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>

In later patches, we're going to change how the inode's ctime field is
used. Switch to using accessor functions instead of raw accesses of
inode->i_ctime.

Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Message-Id: <20230705190309.579783-23-jlayton@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>