Commit graph

1,403,262 commits

Author SHA1 Message Date
Long Li
99fc33d16b xfs: clean up xfs_end_ioend() to reuse local variables
Use already initialized local variables 'offset' and 'size' instead
of accessing ioend members directly in xfs_setfilesize() call.

This is just a code cleanup with no functional changes.

Signed-off-by: Long Li <leo.lilong@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2025-01-13 14:56:02 +01:00
Long Li
efebe42d95 xfs: fix mount hang during primary superblock recovery failure
When mounting an image containing a log with sb modifications that require
log replay, the mount process hang all the time and stack as follows:

  [root@localhost ~]# cat /proc/557/stack
  [<0>] xfs_buftarg_wait+0x31/0x70
  [<0>] xfs_buftarg_drain+0x54/0x350
  [<0>] xfs_mountfs+0x66e/0xe80
  [<0>] xfs_fs_fill_super+0x7f1/0xec0
  [<0>] get_tree_bdev_flags+0x186/0x280
  [<0>] get_tree_bdev+0x18/0x30
  [<0>] xfs_fs_get_tree+0x1d/0x30
  [<0>] vfs_get_tree+0x2d/0x110
  [<0>] path_mount+0xb59/0xfc0
  [<0>] do_mount+0x92/0xc0
  [<0>] __x64_sys_mount+0xc2/0x160
  [<0>] x64_sys_call+0x2de4/0x45c0
  [<0>] do_syscall_64+0xa7/0x240
  [<0>] entry_SYSCALL_64_after_hwframe+0x76/0x7e

During log recovery, while updating the in-memory superblock from the
primary SB buffer, if an error is encountered, such as superblock
corruption occurs or some other reasons, we will proceed to out_release
and release the xfs_buf. However, this is insufficient because the
xfs_buf's log item has already been initialized and the xfs_buf is held
by the buffer log item as follows, the xfs_buf will not be released,
causing the mount thread to hang.

  xlog_recover_do_primary_sb_buffer
    xlog_recover_do_reg_buffer
      xlog_recover_validate_buf_type
        xfs_buf_item_init(bp, mp)

The solution is straightforward, we simply need to allow it to be
handled by the normal buffer write process. The filesystem will be
shutdown before the submission of buffer_list in xlog_do_recovery_pass(),
ensuring the correct release of the xfs_buf as follows:

  xlog_do_recovery_pass
    error = xlog_recover_process
      xlog_recover_process_data
        xlog_recover_process_ophdr
          xlog_recovery_process_trans
            ...
              xlog_recover_buf_commit_pass2
                error = xlog_recover_do_primary_sb_buffer
                  //Encounter error and return
                if (error)
                  goto out_writebuf
                ...
              out_writebuf:
                xfs_buf_delwri_queue(bp, buffer_list) //add bp to list
                return  error
            ...
    if (!list_empty(&buffer_list))
      if (error)
        xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR); //shutdown first
      xfs_buf_delwri_submit(&buffer_list); //submit buffers in list
        __xfs_buf_submit
          if (bp->b_mount->m_log && xlog_is_shutdown(bp->b_mount->m_log))
            xfs_buf_ioend_fail(bp)  //release bp correctly

Fixes: 6a18765b54 ("xfs: update the file system geometry after recoverying superblock buffers")
Cc: stable@vger.kernel.org # v6.12
Signed-off-by: Long Li <leo.lilong@huawei.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2025-01-13 14:55:37 +01:00
Christoph Hellwig
471511d6ef xfs: remove the t_magic field in struct xfs_trans
The t_magic field is only ever assigned to, but never read.  Remove it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2025-01-13 14:55:19 +01:00
Christoph Hellwig
415dee1e06 xfs: remove XFS_ILOG_NONCORE
XFS_ILOG_NONCORE is not used in the kernel code or xfsprogs, remove it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2025-01-13 14:55:14 +01:00
Christoph Hellwig
23ebf63925 xfs: mark xfs_dir_isempty static
And return bool instead of a boolean condition as int.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2025-01-13 14:55:06 +01:00
Carlos Maiolino
156d1c389c xfs: reflink on the realtime device [v6.2 05/14]
This patchset enables use of the file data block sharing feature (i.e.
 reflink) on the realtime device.  It follows the same basic sequence as
 the realtime rmap series -- first a few cleanups; then  introduction of
 the new btree format and inode fork format.  Next comes enabling CoW and
 remapping for the rt device; new scrub, repair, and health reporting
 code; and at the end we implement some code to lengthen write requests
 so that rt extents are always CoWed fully.
 
 This has been running on the djcloud for months with no problems.  Enjoy!
 
 Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
 -----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZ2nQ6gAKCRBKO3ySh0YR
 phxFAQDsng0DbICm7bQPkjOxCG8SGMv0DRIABYe7kmk4NJuZkQEA/mHiDSA6CLg+
 bmqS9dfMRMSG/hl+9+Vp3e3CsZgMxQ4=
 =dWPC
 -----END PGP SIGNATURE-----

Merge tag 'realtime-reflink_2024-12-23' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into for-next

xfs: reflink on the realtime device [v6.2 05/14]

This patchset enables use of the file data block sharing feature (i.e.
reflink) on the realtime device.  It follows the same basic sequence as
the realtime rmap series -- first a few cleanups; then  introduction of
the new btree format and inode fork format.  Next comes enabling CoW and
remapping for the rt device; new scrub, repair, and health reporting
code; and at the end we implement some code to lengthen write requests
so that rt extents are always CoWed fully.

This has been running on the djcloud for months with no problems.  Enjoy!

Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
2025-01-13 14:54:52 +01:00
Carlos Maiolino
a938bbe473 xfs: realtime reverse-mapping support [v6.2 04/14]
This is the latest revision of a patchset that adds to XFS kernel
 support for reverse mapping for the realtime device.  This time around
 I've fixed some of the bitrot that I've noticed over the past few
 months, and most notably have converted rtrmapbt to use the metadata
 inode directory feature instead of burning more space in the superblock.
 
 At the beginning of the set are patches to implement storing B+tree
 leaves in an inode root, since the realtime rmapbt is rooted in an
 inode, unlike the regular rmapbt which is rooted in an AG block.
 Prior to this, the only btree that could be rooted in the inode fork
 was the block mapping btree; if all the extent records fit in the
 inode, format would be switched from 'btree' to 'extents'.
 
 The next few patches enhance the reverse mapping routines to handle
 the parts that are specific to rtgroups -- adding the new btree type,
 adding a new log intent item type, and wiring up the metadata directory
 tree entries.
 
 Finally, implement GETFSMAP with the rtrmapbt and scrub functionality
 for the rtrmapbt and rtbitmap and online fsck functionality.
 
 This has been running on the djcloud for months with no problems.  Enjoy!
 
 Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
 -----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZ2nQ6gAKCRBKO3ySh0YR
 puNFAP4j44Z3j+WEMSONKXLgl3Goo7Xahm/teGTzSN42l7lUtQEA5zhKcvDgiNIe
 KXxtJcRmHCqq8qaxqcnM6vlcHok92wM=
 =XjiL
 -----END PGP SIGNATURE-----

Merge tag 'realtime-rmap_2024-12-23' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into for-next

xfs: realtime reverse-mapping support [v6.2 04/14]

This is the latest revision of a patchset that adds to XFS kernel
support for reverse mapping for the realtime device.  This time around
I've fixed some of the bitrot that I've noticed over the past few
months, and most notably have converted rtrmapbt to use the metadata
inode directory feature instead of burning more space in the superblock.

At the beginning of the set are patches to implement storing B+tree
leaves in an inode root, since the realtime rmapbt is rooted in an
inode, unlike the regular rmapbt which is rooted in an AG block.
Prior to this, the only btree that could be rooted in the inode fork
was the block mapping btree; if all the extent records fit in the
inode, format would be switched from 'btree' to 'extents'.

The next few patches enhance the reverse mapping routines to handle
the parts that are specific to rtgroups -- adding the new btree type,
adding a new log intent item type, and wiring up the metadata directory
tree entries.

Finally, implement GETFSMAP with the rtrmapbt and scrub functionality
for the rtrmapbt and rtbitmap and online fsck functionality.

This has been running on the djcloud for months with no problems.  Enjoy!

Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
2025-01-13 14:54:41 +01:00
Carlos Maiolino
8a092f440e xfs: enable in-core block reservation for rt metadata [v6.2 03/14]
In preparation for adding reverse mapping and refcounting to the
 realtime device, enhance the metadir code to reserve free space for
 btree shape changes as delayed allocation blocks.
 
 This enables us to pre-allocate space for the rmap and refcount btrees
 in the same manner as we do for the data device counterparts, which is
 how we avoid ENOSPC failures when space is low but we've already
 committed to a COW operation.
 
 This has been running on the djcloud for months with no problems.  Enjoy!
 
 Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
 -----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZ2nQ6gAKCRBKO3ySh0YR
 pluAAP9u4A74HNTi7Df4C70diZIpXynSnyzAxVirQzda/WlUzAD+OjZEo7VkLgG4
 AJIqspcX632tFBFhBAhf3XzqWlM+zws=
 =hOw3
 -----END PGP SIGNATURE-----

Merge tag 'reserve-rt-metadata-space_2024-12-23' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into for-next

xfs: enable in-core block reservation for rt metadata [v6.2 03/14]

In preparation for adding reverse mapping and refcounting to the
realtime device, enhance the metadir code to reserve free space for
btree shape changes as delayed allocation blocks.

This enables us to pre-allocate space for the rmap and refcount btrees
in the same manner as we do for the data device counterparts, which is
how we avoid ENOSPC failures when space is low but we've already
committed to a COW operation.

This has been running on the djcloud for months with no problems.  Enjoy!

Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
2025-01-13 14:54:33 +01:00
Carlos Maiolino
9a2ce7254c xfs: refactor btrees to support records in inode root [v6.2 02/14]
Amend the btree code to support storing btree rcords in the inode root,
 because the current bmbt code does not support this.
 
 This has been running on the djcloud for months with no problems.  Enjoy!
 
 Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
 -----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZ2nQ6QAKCRBKO3ySh0YR
 piSBAP0dmPfWwKtdIx1AQUmfQ4FeNQbSk5PTKLuXIjyfUug9QwEAu7L616xtrgt6
 6/9CBlIUNp3gJriZzLbCjS72F7hssgQ=
 =Jk+2
 -----END PGP SIGNATURE-----

Merge tag 'btree-ifork-records_2024-12-23' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into for-next

xfs: refactor btrees to support records in inode root [v6.2 02/14]

Amend the btree code to support storing btree rcords in the inode root,
because the current bmbt code does not support this.

This has been running on the djcloud for months with no problems.  Enjoy!

Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
2025-01-13 14:54:23 +01:00
Carlos Maiolino
69bf6cd7f3 xfs: bug fixes for 6.13 [01/14]
Bug fixes for 6.13.
 
 This has been running on the djcloud for months with no problems.  Enjoy!
 
 Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
 -----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZ2nQ6QAKCRBKO3ySh0YR
 pjVrAQD9ty5VSpobCR0HRKImR1zCvjreaTzDty+XXnCrYVka1AEA1NbC53RxAPoy
 bW5GJ0ZEvQhamkx9YuBfiJXoqYw1jgI=
 =ZmFb
 -----END PGP SIGNATURE-----

Merge tag 'xfs-6.13-fixes_2024-12-23' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into for-next

xfs: bug fixes for 6.13 [01/14]

Bug fixes for 6.13.

This has been running on the djcloud for months with no problems.  Enjoy!

Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
2025-01-13 14:54:14 +01:00
David Sterba
311473984c btrfs: async-thread: rename DFT_THRESHOLD to DEFAULT_THRESHOLD
Rename the macro so it's obvious what it means.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:23 +01:00
David Sterba
ef8c0047aa btrfs: remove redundant variables from __process_folios_contig() and lock_delalloc_folios()
Same pattern in both functions, we really only use index, start_index is
redundant.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:23 +01:00
David Sterba
248c4ff393 btrfs: split waiting from read_extent_buffer_pages(), drop parameter wait
There are only 2 WAIT_* values left for wait parameter, we can encode
this to the function name if the waiting functionality is split.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:23 +01:00
David Sterba
db9eef2ea8 btrfs: remove unused define WAIT_PAGE_LOCK for extent io
Last use was in the readahead code that got removed by f26c923860
("btrfs: remove reada infrastructure").

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:23 +01:00
David Sterba
f8e0b8f9c2 btrfs: unwrap folio locking helpers
Another conversion to folio API, use the folio locking directly instead
of back and forth page <-> folio conversions.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:23 +01:00
David Sterba
549a88acbe btrfs: change return type to bool type of check_eb_alignment()
The check function pattern is supposed to return true/false, currently
there's only one error code.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:23 +01:00
David Sterba
a43caf82a1 btrfs: switch grab_extent_buffer() to folios
Use the folio API, remove page_folio/folio_page conversions.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
cc8f51a355 btrfs: rename btrfs_release_extent_buffer_pages() to mention folios
Continue page to folio updates, sync what the function does with it's
name.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
a722c72bef btrfs: open code __free_extent_buffer()
Using the kmem cache freeing directly is clear enough, we don't need to
wrap it.  All the users are in the same file.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
b6160baed3 btrfs: drop one time used local variable in end_bbio_meta_write()
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
075adeeb92 btrfs: make wait_on_extent_buffer_writeback() static inline
The simple helper can be inlined, no need for the separate function.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
011a9a1f24 btrfs: use btrfs_inode in extent_writepage()
As extent_writepage() is internal helper we should use our inode type,
so change it from struct inode.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
06de96faf7 btrfs: rename __get_extent_map() and pass btrfs_inode
The double underscore naming scheme does not apply here, there's only
only get_extent_map(). As the definition is changed also pass the struct
btrfs_inode.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
3a1c46dbc9 btrfs: open code set_page_extent_mapped()
The function set_page_extent_mapped() is now a simple wrapper so use the
folio helper.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
2b41599bff btrfs: rename __unlock_for_delalloc() and drop underscores
Drop the leading underscores in '__unlock_for_delalloc()' and rename it
to 'unlock_delalloc_folio()'. This also ensures naming parity with
'lock_delalloc_folios()'.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
2a1e8378dc btrfs: use SECTOR_SIZE defines in btrfs_issue_discard()
Use the existing define for single sector size.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:22 +01:00
David Sterba
6d67ff1c0b btrfs: remove stray comment about SRCU
The subvol_srcu was removed in c75e839414 ("btrfs: kill the
subvol_srcu") years ago.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
David Sterba
5f14eb12a3 btrfs: drop unused parameter fs_info to btrfs_delete_delayed_insertion_item()
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Jing Xia
2fa07d7a0f btrfs: pass write-hint for buffered IO
Commit 449813515d ("block, fs: Restore the per-bio/request data
lifetime fields") restored write-hint support in btrfs. But that is
applicable only for direct IO. This patch supports passing
write-hint for buffered IO from btrfs file system to block layer
by filling bi_write_hint of struct bio in alloc_new_bio().

There's an ongoing discussion which devices can use that,
https://lore.kernel.org/all/20240910150200.6589-6-joshi.k@samsung.com,
in SCSI there's support using sd_group_number() and
sd_setup_rw32_cmnd().

The hint goes from the application directly to the block device so it's
up to the application to set up everything properly to utilize the
different hint classes.

Link: https://lore.kernel.org/all/20240910150200.6589-6-joshi.k@samsung.com
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Jing Xia <j.xia@samsung.com>
[ Add more context and use case. ]
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
3681dbe0af btrfs: print read policy on module load
Print the read read policy if set as module parameter (with
CONFIG_BTRFS_EXPERIMENTAL).

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
e426286cfa btrfs: configure read policy via module parameter
For testing purposes allow to configure the read policy via module
parameter from the beginning. Available only with CONFIG_BTRFS_EXPERIMENTAL

Examples:

- Set the RAID1 balancing method to round-robin with a custom
  min_contig_read of 4k:
  $ modprobe btrfs read_policy=round-robin:4096

- Set the round-robin balancing method with the default
  min_contiguous_read:
  $ modprobe btrfs read_policy=round-robin

- Set the "devid" balancing method, defaulting to the latest device:
  $ modprobe btrfs read_policy=devid

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
bb4715e967 btrfs: print status of experimental mode when loading module
Commit c9c49e8f157e ("btrfs: split out CONFIG_BTRFS_EXPERIMENTAL from
CONFIG_BTRFS_DEBUG") introduces a way to enable or disable experimental
features, print its status during module load, like:

  Btrfs loaded, experimental=on, debug=on, assert=on, zoned=yes, fsverity=yes

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
c86aae73bd btrfs: add read policy to set a preferred device
Add read policy that will force all reads to go to the given device
(specified by devid) on the RAID1 profiles.

This will be used for testing, e.g. to read from stale device. Users may
find other use cases.

Can be set in sysfs, the value format is "devid:<devid>" to the file

  /sys/fs/btrfs/FSID/read_policy

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
6d7a915495 btrfs: introduce RAID1 round-robin read balancing
Add round-robin read policy that balances reads over available devices
(all RAID1 block group profiles). Switch to the next devices is done
after a number of blocks is read, which is 256K by default and is
configurable in sysfs.

The format is "round-robin:<min-contig-read>" and can be set in file

  /sys/fs/btrfs/FSID/read_policy

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
22fb0d99c9 btrfs: add tracking of read blocks for read policy
Track number of read blocks in the whole filesystem. The counter is
initialized when devices are opened. The counter is increased at
btrfs_submit_dev_bio() if the stats tracking is enabled (depends on the
read policy).  Stats tracking is disabled by default and is enabled
through fs_devices::collect_fs_stats when required.

The code is not under the EXPERIMENTAL define, as stats can be expanded
to include write counts and other performance counters, with the user
interface independent of its internal use.

This is an in-memory-only feature, not related to the dev error stats.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
b6bed20ed3 btrfs: sysfs: handle value associated with read balancing policy
Enable specifying additional configuration values along the RAID1
balancing read policy in a single input string.

Update btrfs_read_policy_to_enum() to parse and handle a value
associated with the policy in the format "policy:value", the value part
if present is converted to 64-bit integer.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
38cae63137 btrfs: sysfs: add btrfs_read_policy_to_enum() helper and refactor read policy store
Introduce btrfs_read_policy_to_enum() helper to simplify the conversion
of a string read policy to its corresponding enum value. This reduces
duplication and improves code clarity in btrfs_read_policy_store().

The parameter is copied locally to allow modification, enabling the
separation of the method and its value. This prepares for the addition of
more functionality in subsequent patches.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:21 +01:00
Anand Jain
83be7f8b9c btrfs: sysfs: refactor output formatting in btrfs_read_policy_show()
Refactor the logic in btrfs_read_policy_show() for easier extension with
more balancing methods.  Streamline the space and bracket handling
around the active policy without altering the functional output.  This
is in preparation to add more methods.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Anand Jain
a5019b7070 btrfs: initialize fs_devices->fs_info earlier in btrfs_init_devices_late()
Currently, fs_devices->fs_info is initialized in btrfs_init_devices_late(),
but this occurs too late for find_live_mirror(), which is invoked by
load_super_root() much earlier than btrfs_init_devices_late().

Fix this by moving the initialization to open_ctree(), before load_super_root().

Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
74973b45a6 btrfs: xattr: remove unnecessary call to btrfs_mark_buffer_dirty()
The call to btrfs_mark_buffer_dirty() at btrfs_setxattr() is not
necessary as we have a path setup for writing with btrfs_search_slot()
having a 'cow' argument set to 1.

This just makes the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove it.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
1ca4e15f41 btrfs: volumes: remove unnecessary calls to btrfs_mark_buffer_dirty()
We have several places explicitly calling btrfs_mark_buffer_dirty() but
that is not necessarily since the target leaf came from a path that was
obtained for a btree search function that modifies the btree, something
like btrfs_insert_empty_item() or anything else that ends up calling
btrfs_search_slot() with a value of 1 for its 'cow' argument.

These just make the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove them.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
c9a4390707 btrfs: uuid-tree: remove unnecessary call to btrfs_mark_buffer_dirty()
The call to btrfs_mark_buffer_dirty() at btrfs_uuid_tree_add() is not
necessary as we have a path setup for writing with btrfs_search_slot()
having a 'cow' argument set to 1 (through btrfs_insert_empty_item()).

This just makes the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove it.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
65733e8d6c btrfs: root-tree: remove unnecessary calls to btrfs_mark_buffer_dirty()
We have several places explicitly calling btrfs_mark_buffer_dirty() but
that is not necessarily since the target leaf came from a path that was
obtained for a btree search function that modifies the btree, something
like btrfs_insert_empty_item() or anything else that ends up calling
btrfs_search_slot() with a value of 1 for its 'cow' argument.

These just make the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove them.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
5a8293a1cc btrfs: relocation: remove unnecessary calls to btrfs_mark_buffer_dirty()
We have several places explicitly calling btrfs_mark_buffer_dirty() but
that is not necessarily since the target leaf came from a path that was
obtained for a btree search function that modifies the btree, something
like btrfs_insert_empty_item() or anything else that ends up calling
btrfs_search_slot() with a value of 1 for its 'cow' argument.

These just make the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove them.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
bdf1660b22 btrfs: raid-stripe-tree: remove unnecessary call to btrfs_mark_buffer_dirty()
The call to btrfs_mark_buffer_dirty() at update_raid_extent_item() is not
necessary as we have a path setup for writing with btrfs_search_slot()
having a 'cow' argument set to 1.

This just makes the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove it.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
d74a36f37e btrfs: qgroup: remove unnecessary calls to btrfs_mark_buffer_dirty()
We have several places explicitly calling btrfs_mark_buffer_dirty() but
that is not necessarily since the target leaf came from a path that was
obtained for a btree search function that modifies the btree, something
like btrfs_insert_empty_item() or anything else that ends up calling
btrfs_search_slot() with a value of 1 for its 'cow' argument.

These just make the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove them.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
bd25bf9dcd btrfs: ioctl: remove unnecessary call to btrfs_mark_buffer_dirty()
The call to btrfs_mark_buffer_dirty() at btrfs_ioctl_default_subvol() is
not necessary as we have a path setup for writing with btrfs_search_slot()
having a 'cow' argument set to 1.

This just makes the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove it.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:20 +01:00
Filipe Manana
212e5f5cb8 btrfs: inode-item: remove unnecessary calls to btrfs_mark_buffer_dirty()
We have several places explicitly calling btrfs_mark_buffer_dirty() but
that is not necessarily since the target leaf came from a path that was
obtained for a btree search function that modifies the btree, something
like btrfs_insert_empty_item() or anything else that ends up calling
btrfs_search_slot() with a value of 1 for its 'cow' argument.

These just make the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove them.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:19 +01:00
Filipe Manana
5c7763312c btrfs: inode: remove unnecessary calls to btrfs_mark_buffer_dirty()
We have several places explicitly calling btrfs_mark_buffer_dirty() but
that is not necessarily since the target leaf came from a path that was
obtained for a btree search function that modifies the btree, something
like btrfs_insert_empty_item() or anything else that ends up calling
btrfs_search_slot() with a value of 1 for its 'cow' argument.

These just make the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove them.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:19 +01:00
Filipe Manana
038d6999ec btrfs: free-space-cache: remove unnecessary calls to btrfs_mark_buffer_dirty()
We have several places explicitly calling btrfs_mark_buffer_dirty() but
that is not necessarily since the target leaf came from a path that was
obtained for a btree search function that modifies the btree, something
like btrfs_insert_empty_item() or anything else that ends up calling
btrfs_search_slot() with a value of 1 for its 'cow' argument.

These just make the code more verbose, confusing and add a little extra
overhead and well as increase the module's text size, so remove them.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
2025-01-13 14:53:19 +01:00