[PATCH V9 00/11] XFS - Enable per-file/per-directory DAX operations V9

From: ira . weiny
Date: Tue Apr 21 2020 - 15:18:05 EST


From: Ira Weiny <ira.weiny@xxxxxxxxx>

Changes from V8:
Rebase to 5.7-rc2
Change ALWAYS/NEVER bits to be 26/27
Remove XFS_IDONTCACHE -> lift to I_DONTCACHE
use mark_inode_dontcache() in XFS
create xfs_dax_mode enum
use xfs signature styles
Change xfs_inode_enabe_dax() -> xfs_inode_should_enable()
Based on feedback to ext4 series
Fix locking of DCACHE_DONTCACHE
Change flag_inode_dontcache() -> mark_inode_dontcache()
Change xfs_ioctl_setattr_dax_invalidate() -> xfs_ioctl_dax_check_set_cache()
Documentation cleanups
Clean up all commit messages


At LSF/MM'19 [1] [2] we discussed applications that overestimate memory
consumption due to their inability to detect whether the kernel will
instantiate page cache for a file, and cases where a global dax enable via a
mount option is too coarse.

The following patch series enables the use of DAX on individual files and/or
directories on xfs, and lays some groundwork to do so in ext4. It further
enhances the dax mount option to be a tri-state of 'always', 'never', or
'iflag' (default). Furthermore, it maintians '-o dax' to be equivalent to '-o
dax=always'.

The insight at LSF/MM was to separate the per-mount or per-file "physical"
(FS_XFLAG_DAX) capability switch from an "effective" (S_DAX) attribute for the
file.

At LSF/MM we discussed the difficulties of switching the DAX state of a file
with active mappings / page cache. It was thought the races could be avoided
by limiting DAX state flips to 0-length files.

However, this turns out to not be true.[3][5] This is because address space
operations (a_ops) may be in use at any time the inode is referenced.

For this reason direct manipulation of the FS_XFLAG_DAX is allowed but the
operation of the file (S_DAX) is not immediately changed.

Details of when and how DAX state can be changed on a file is included in a
documentation patch.

It should be noted that FS_XFLAG_DAX inheritance is not shown in this patch set
as it was maintained from previous work on XFS. FS_XFLAG_DAX and it's
inheritance will need to be added to other file systems for user control.


[1] https://lwn.net/Articles/787973/
[2] https://lwn.net/Articles/787233/
[3] https://lkml.org/lkml/2019/10/20/96
[4] https://patchwork.kernel.org/patch/11310511/
[5] https://lore.kernel.org/lkml/20200405061945.GA94792@xxxxxxxxxxxxxxxxxxxxxxxxx/


To: linux-kernel@xxxxxxxxxxxxxxx
Cc: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
Cc: "Darrick J. Wong" <darrick.wong@xxxxxxxxxx>
Cc: Dan Williams <dan.j.williams@xxxxxxxxx>
Cc: Dave Chinner <david@xxxxxxxxxxxxx>
Cc: Christoph Hellwig <hch@xxxxxx>
Cc: "Theodore Y. Ts'o" <tytso@xxxxxxx>
Cc: Jan Kara <jack@xxxxxxx>
Cc: linux-ext4@xxxxxxxxxxxxxxx
Cc: linux-xfs@xxxxxxxxxxxxxxx
Cc: linux-fsdevel@xxxxxxxxxxxxxxx

Changes from V7:
Add DCACHE_DONTCACHE
If mount override don't worry about inode caching
Change mount flags to NEVER/ALWAYS
Clean up xfs_inode_enable_dax()
Clarify comments
Update documentation

Changes from V6:
Incorporate feedback on patches
Add ability to change FS_XFLAG_DAX on files at any time.
Add a don't cache flag to the VFS layer
Preserve internal XFS IDONTCACHE behavior for bulkstat
operations

Changes from V5:
* make dax mount option a tri-state
* Reject changes to FS_XFLAG_DAX for regular files
- Allow only on directories
* Update documentation

Changes from V4:
* Open code the aops lock rather than add it to the xfs_ilock()
subsystem (Darrick's comments were obsoleted by this change)
* Fix lkp build suggestions and bugs

Changes from V3:
* Remove global locking... :-D
* put back per inode locking and remove pre-mature optimizations
* Fix issues with Directories having IS_DAX() set
* Fix kernel crash issues reported by Jeff
* Add some clean up patches
* Consolidate diflags to iflags functions
* Update/add documentation
* Reorder/rename patches quite a bit

Changes from V2:

* Move i_dax_sem to be a global percpu_rw_sem rather than per inode
Internal discussions with Dan determined this would be easier,
just as performant, and slightly less overhead that having it
in the SB as suggested by Jan
* Fix locking order in comments and throughout code
* Change "mode" to "state" throughout commits
* Add CONFIG_FS_DAX wrapper to disable inode_[un]lock_state() when not
configured
* Add static branch for which is activated by a device which supports
DAX in XFS
* Change "lock/unlock" to up/down read/write as appropriate
Previous names were over simplified
* Update comments/documentation

* Remove the xfs specific lock to the vfs (global) layer.
* Fix i_dax_sem locking order and comments

* Move 'i_mapped' count from struct inode to struct address_space and
rename it to mmap_count
* Add inode_has_mappings() call

* Fix build issues
* Clean up syntax spacing and minor issues
* Update man page text for STATX_ATTR_DAX
* Add reviewed-by's
* Rebase to 5.6

Rename patch:
from: fs/xfs: Add lock/unlock state to xfs
to: fs/xfs: Add write DAX lock to xfs layer
Add patch:
fs/xfs: Clarify lockdep dependency for xfs_isilocked()
Drop patch:
fs/xfs: Fix truncate up


Ira Weiny (11):
fs/xfs: Remove unnecessary initialization of i_rwsem
fs: Remove unneeded IS_DAX() check in io_is_direct()
fs/stat: Define DAX statx attribute
fs/xfs: Change XFS_MOUNT_DAX to XFS_MOUNT_DAX_ALWAYS
fs/xfs: Make DAX mount option a tri-state
fs/xfs: Create function xfs_inode_should_enable_dax()
fs/xfs: Combine xfs_diflags_to_linux() and xfs_diflags_to_iflags()
fs: Lift XFS_IDONTCACNE to the VFS layer
fs: Introduce DCACHE_DONTCACHE
fs/xfs: Update xfs_ioctl_setattr_dax_invalidate()
Documentation/dax: Update Usage section

Documentation/filesystems/dax.txt | 164 +++++++++++++++++++++++++++++-
drivers/block/loop.c | 6 +-
fs/dcache.c | 4 +
fs/inode.c | 15 +++
fs/stat.c | 3 +
fs/xfs/xfs_icache.c | 8 +-
fs/xfs/xfs_inode.h | 3 +-
fs/xfs/xfs_ioctl.c | 141 ++++---------------------
fs/xfs/xfs_iops.c | 72 ++++++++-----
fs/xfs/xfs_mount.h | 4 +-
fs/xfs/xfs_super.c | 55 ++++++++--
include/linux/dcache.h | 2 +
include/linux/fs.h | 14 +--
include/uapi/linux/stat.h | 1 +
14 files changed, 321 insertions(+), 171 deletions(-)

--
2.25.1