linux.git/fs/stat.c, branch v5.8

Merge tag 'vfs-5.8-merge-1' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux

2020-06-03T02:45:12+00:00

Pull DAX updates part one from Darrick Wong:
 "After many years of LKML-wrangling about how to enable programs to
  query and influence the file data access mode (DAX) when a filesystem
  resides on storage devices such as persistent memory, Ira Weiny has
  emerged with a proposed set of standard behaviors that has not been
  shot down by anyone! We're more or less standardizing on the current
  XFS behavior and adapting ext4 to do the same.

  This is the first of a handful pull requests that will make ext4 and
  XFS present a consistent interface for user programs that care about
  DAX. We add a statx attribute that programs can check to see if DAX is
  enabled on a particular file. Then, we update the DAX documentation to
  spell out the user-visible behaviors that filesystems will guarantee
  (until the next storage industry shakeup). The on-disk inode flag has
  been in XFS for a few years now.

  Summary:

   - Clean up io_is_direct.

   - Add a new statx flag to indicate when file data access is being
     done via DAX (as opposed to the page cache).

   - Update the documentation for how system administrators and
     application programmers can take advantage of the (still
     experimental DAX) feature"

Link: https://lore.kernel.org/lkml/20200505002016.1085071-1-ira.weiny@intel.com/

* tag 'vfs-5.8-merge-1' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
  Documentation/dax: Update Usage section
  fs/stat: Define DAX statx attribute
  fs: Remove unneeded IS_DAX() check in io_is_direct()

Merge tag 'for-5.8/io_uring-2020-06-01' of git://git.kernel.dk/linux-block

2020-06-02T22:42:50+00:00

Pull io_uring updates from Jens Axboe:
 "A relatively quiet round, mostly just fixes and code improvements. In
particular:

   - Make statx just use the generic statx handler, instead of open
     coding it. We don't need that anymore, as we always call it async
     safe (Bijan)

   - Enable closing of the ring itself. Also fixes O_PATH closure (me)

   - Properly name completion members (me)

   - Batch reap of dead file registrations (me)

   - Allow IORING_OP_POLL with double waitqueues (me)

   - Add tee(2) support (Pavel)

   - Remove double off read (Pavel)

   - Fix overflow cancellations (Pavel)

   - Improve CQ timeouts (Pavel)

   - Async defer drain fixes (Pavel)

   - Add support for enabling/disabling notifications on a registered
     eventfd (Stefano)

   - Remove dead state parameter (Xiaoguang)

   - Disable SQPOLL submit on dying ctx (Xiaoguang)

   - Various code cleanups"

* tag 'for-5.8/io_uring-2020-06-01' of git://git.kernel.dk/linux-block: (29 commits)
  io_uring: fix overflowed reqs cancellation
  io_uring: off timeouts based only on completions
  io_uring: move timeouts flushing to a helper
  statx: hide interfaces no longer used by io_uring
  io_uring: call statx directly
  statx: allow system call to be invoked from io_uring
  io_uring: add io_statx structure
  io_uring: get rid of manual punting in io_close
  io_uring: separate DRAIN flushing into a cold path
  io_uring: don't re-read sqe->off in timeout_prep()
  io_uring: simplify io_timeout locking
  io_uring: fix flush req->refs underflow
  io_uring: don't submit sqes when ctx->refs is dying
  io_uring: async task poll trigger cleanup
  io_uring: add tee(2) support
  splice: export do_tee()
  io_uring: don't repeat valid flag list
  io_uring: rename io_file_put()
  io_uring: remove req->needs_fixed_files
  io_uring: cleanup io_poll_remove_one() logic
  ...

statx: hide interfaces no longer used by io_uring

2020-05-26T22:48:06+00:00

The io_uring interfaces have been replaced by do_statx() and are no
longer needed.

Signed-off-by: Bijan Mottahedeh 
Signed-off-by: Jens Axboe

statx: allow system call to be invoked from io_uring

2020-05-26T22:48:06+00:00

This is a prepatory patch to allow io_uring to invoke statx directly.

Signed-off-by: Bijan Mottahedeh 
Signed-off-by: Jens Axboe

statx: add mount_root

2020-05-14T14:44:24+00:00

Determining whether a path or file descriptor refers to a mountpoint (or
more precisely a mount root) is not trivial using current tools.

Add a flag to statx that indicates whether the path or fd refers to the
root of a mount or not.

Cc: linux-api@vger.kernel.org
Cc: linux-man@vger.kernel.org
Reported-by: Lennart Poettering 
Reported-by: J. Bruce Fields 
Signed-off-by: Miklos Szeredi 
Reviewed-by: Christoph Hellwig

statx: add mount ID

2020-05-14T14:44:24+00:00

Systemd is hacking around to get it and it's trivial to add to statx, so...

Cc: linux-api@vger.kernel.org
Cc: linux-man@vger.kernel.org
Signed-off-by: Miklos Szeredi 
Reviewed-by: Christoph Hellwig

statx: don't clear STATX_ATIME on SB_RDONLY

2020-05-14T14:44:24+00:00

IS_NOATIME(inode) is defined as __IS_FLG(inode, SB_RDONLY|SB_NOATIME), so
generic_fillattr() will clear STATX_ATIME from the result_mask if the super
block is marked read only.

This was probably not the intention, so fix to only clear STATX_ATIME if
the fs doesn't support atime at all.

Signed-off-by: Miklos Szeredi 
Acked-by: David Howells 
Reviewed-by: Christoph Hellwig

uapi: deprecate STATX_ALL

2020-05-14T14:44:24+00:00

Constants of the *_ALL type can be actively harmful due to the fact that
developers will usually fail to consider the possible effects of future
changes to the definition.

Deprecate STATX_ALL in the uapi, while no damage has been done yet.

We could keep something like this around in the kernel, but there's
actually no point, since all filesystems should be explicitly checking
flags that they support and not rely on the VFS masking unknown ones out: a
flag could be known to the VFS, yet not known to the filesystem.

Cc: David Howells 
Cc: linux-api@vger.kernel.org
Cc: linux-man@vger.kernel.org
Signed-off-by: Miklos Szeredi 
Reviewed-by: Christoph Hellwig

fs/stat: Define DAX statx attribute

2020-05-04T15:49:39+00:00

In order for users to determine if a file is currently operating in DAX
state (effective DAX).  Define a statx attribute value and set that
attribute if the effective DAX flag is set.

To go along with this we propose the following addition to the statx man
page:

STATX_ATTR_DAX

	The file is in the DAX (cpu direct access) state.  DAX state
	attempts to minimize software cache effects for both I/O and
	memory mappings of this file.  It requires a file system which
	has been configured to support DAX.

	DAX generally assumes all accesses are via cpu load / store
	instructions which can minimize overhead for small accesses, but
	may adversely affect cpu utilization for large transfers.

	File I/O is done directly to/from user-space buffers and memory
	mapped I/O may be performed with direct memory mappings that
	bypass kernel page cache.

	While the DAX property tends to result in data being transferred
	synchronously, it does not give the same guarantees of O_SYNC
	where data and the necessary metadata are transferred together.

	A DAX file may support being mapped with the MAP_SYNC flag,
	which enables a program to use CPU cache flush instructions to
	persist CPU store operations without an explicit fsync(2).  See
	mmap(2) for more information.

Reviewed-by: Dave Chinner 
Reviewed-by: Jan Kara 
Reviewed-by: Darrick J. Wong 
Signed-off-by: Ira Weiny 
Signed-off-by: Darrick J. Wong

fs: make two stat prep helpers available

2020-01-21T00:03:54+00:00

To implement an async stat, we need to provide the flags mapping and
the statx user copy. Make them available internally, through
fs/internal.h.

Signed-off-by: Jens Axboe