exfat: speed up iterate/lookup by fixing start point of traversing cluster chain

When directory iterate and lookup is called, there's a buggy rewinding
of start point for traversing cluster chain to

exfat: speed up iterate/lookup by fixing start point of traversing cluster chain

When directory iterate and lookup is called, there's a buggy rewinding
of start point for traversing cluster chain to the parent directory
entry's first cluster. This caused repeated cluster chain traversing
from the first entry of the parent directory that would show worse
performance if huge amounts of files exist under the parent directory.
Fix not to rewind, make continue from currently referenced cluster and
dir entry.

Tested with 50,000 files under single directory / 256GB sdcard,
with command "time ls -l > /dev/null",
Before : 0m08.69s real 0m00.27s user 0m05.91s system
After : 0m07.01s real 0m00.25s user 0m04.34s system

Signed-off-by: Hyeongseok Kim <hyeongseok@gmail.com>
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: improve write performance when dirsync enabled

Degradation of write speed caused by frequent disk access for cluster
bitmap update on every cluster allocation could be improved by
selective s

exfat: improve write performance when dirsync enabled

Degradation of write speed caused by frequent disk access for cluster
bitmap update on every cluster allocation could be improved by
selective syncing bitmap buffer. Change to flush bitmap buffer only
for the directory related operations.

Signed-off-by: Hyeongseok Kim <hyeongseok@gmail.com>
Acked-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: add support ioctl and FITRIM function

Add FITRIM ioctl to enable discarding unused blocks while mounted.
As current exFAT doesn't have generic ioctl handler, add empty ioctl
function first, a

exfat: add support ioctl and FITRIM function

Add FITRIM ioctl to enable discarding unused blocks while mounted.
As current exFAT doesn't have generic ioctl handler, add empty ioctl
function first, and add FITRIM handler.

Signed-off-by: Hyeongseok Kim <hyeongseok@gmail.com>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Acked-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: introduce bitmap_lock for cluster bitmap access

s_lock which is for protecting concurrent access of file operations is
too huge for cluster bitmap protection, so introduce a new bitmap_lock
t

exfat: introduce bitmap_lock for cluster bitmap access

s_lock which is for protecting concurrent access of file operations is
too huge for cluster bitmap protection, so introduce a new bitmap_lock
to narrow the lock range if only need to access cluster bitmap.

Signed-off-by: Hyeongseok Kim <hyeongseok@gmail.com>
Acked-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: improve performance of exfat_free_cluster when using dirsync mount option

There are stressful update of cluster allocation bitmap when using
dirsync mount option which is doing sync buffer on

exfat: improve performance of exfat_free_cluster when using dirsync mount option

There are stressful update of cluster allocation bitmap when using
dirsync mount option which is doing sync buffer on every cluster bit
clearing. This could result in performance degradation when deleting
big size file.
Fix to update only when the bitmap buffer index is changed would make
less disk access, improving performance especially for truncate operation.

Testing with Samsung 256GB sdcard, mounted with dirsync option
(mount -t exfat /dev/block/mmcblk0p1 /temp/mount -o dirsync)

Remove 4GB file, blktrace result.
[Before] : 39 secs.
Total (blktrace):
Reads Queued: 0, 0KiB Writes Queued: 32775, 16387KiB
Read Dispatches: 0, 0KiB Write Dispatches: 32775, 16387KiB
Reads Requeued: 0 Writes Requeued: 0
Reads Completed: 0, 0KiB Writes Completed: 32775, 16387KiB
Read Merges: 0, 0KiB Write Merges: 0, 0KiB
IO unplugs: 2 Timer unplugs: 0

[After] : 1 sec.
Total (blktrace):
Reads Queued: 0, 0KiB Writes Queued: 13, 6KiB
Read Dispatches: 0, 0KiB Write Dispatches: 13, 6KiB
Reads Requeued: 0 Writes Requeued: 0
Reads Completed: 0, 0KiB Writes Completed: 13, 6KiB
Read Merges: 0, 0KiB Write Merges: 0, 0KiB
IO unplugs: 1 Timer unplugs: 0

Signed-off-by: Hyeongseok Kim <hyeongseok@gmail.com>
Acked-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

fs: make helpers idmap mount aware

Extend some inode methods with an additional user namespace argument. A
filesystem that is aware of idmapped mounts will receive the user
namespace the mount has b

fs: make helpers idmap mount aware

Extend some inode methods with an additional user namespace argument. A
filesystem that is aware of idmapped mounts will receive the user
namespace the mount has been marked with. This can be used for
additional permission checking and also to enable filesystems to
translate between uids and gids if they need to. We have implemented all
relevant helpers in earlier patches.

As requested we simply extend the exisiting inode method instead of
introducing new ones. This is a little more code churn but it's mostly
mechanical and doesnt't leave us with additional inode methods.

Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Howells <dhowells@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: linux-fsdevel@vger.kernel.org
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

show more ...

exfat: remove 'rwoffset' in exfat_inode_info

Remove 'rwoffset' in exfat_inode_info and replace it with the parameter of
exfat_readdir().
Since rwoffset is referenced only by exfat_readdir(), it is n

exfat: remove 'rwoffset' in exfat_inode_info

Remove 'rwoffset' in exfat_inode_info and replace it with the parameter of
exfat_readdir().
Since rwoffset is referenced only by exfat_readdir(), it is not necessary
a exfat_inode_info's member.
Also, change cpos to point to the next of entry-set, and return the index
of dir-entry via dir_entry->entry.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Acked-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: fix misspellings using codespell tool

Sedat reported typos using codespell tool.

$ codespell fs/exfat/*.c | grep -v iput
fs/exfat/namei.c:293: upto ==> up to
fs/exfat/nls.c:14: tabel ==>

exfat: fix misspellings using codespell tool

Sedat reported typos using codespell tool.

$ codespell fs/exfat/*.c | grep -v iput
fs/exfat/namei.c:293: upto ==> up to
fs/exfat/nls.c:14: tabel ==> table

$ codespell fs/exfat/*.h
fs/exfat/exfat_fs.h:133: usally ==> usually

Fix typos found by codespell.

Reported-by: Sedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: fix use of uninitialized spinlock on error path

syzbot reported warning message:

Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x1d6/0x29e lib/dump_stack.c:118
register

exfat: fix use of uninitialized spinlock on error path

syzbot reported warning message:

Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x1d6/0x29e lib/dump_stack.c:118
register_lock_class+0xf06/0x1520 kernel/locking/lockdep.c:893
__lock_acquire+0xfd/0x2ae0 kernel/locking/lockdep.c:4320
lock_acquire+0x148/0x720 kernel/locking/lockdep.c:5029
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
spin_lock include/linux/spinlock.h:354 [inline]
exfat_cache_inval_inode+0x30/0x280 fs/exfat/cache.c:226
exfat_evict_inode+0x124/0x270 fs/exfat/inode.c:660
evict+0x2bb/0x6d0 fs/inode.c:576
exfat_fill_super+0x1e07/0x27d0 fs/exfat/super.c:681
get_tree_bdev+0x3e9/0x5f0 fs/super.c:1342
vfs_get_tree+0x88/0x270 fs/super.c:1547
do_new_mount fs/namespace.c:2875 [inline]
path_mount+0x179d/0x29e0 fs/namespace.c:3192
do_mount fs/namespace.c:3205 [inline]
__do_sys_mount fs/namespace.c:3413 [inline]
__se_sys_mount+0x126/0x180 fs/namespace.c:3390
do_syscall_64+0x31/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9

If exfat_read_root() returns an error, spinlock is used in
exfat_evict_inode() without initialization. This patch combines
exfat_cache_init_inode() with exfat_inode_init_once() to initialize
spinlock by slab constructor.

Fixes: c35b6810c495 ("exfat: add exfat cache")
Cc: stable@vger.kernel.org # v5.7+
Reported-by: syzbot <syzbot+b91107320911a26c9a95@syzkaller.appspotmail.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: retain 'VolumeFlags' properly

MediaFailure and VolumeDirty should be retained if these are set before
mounting.

In '3.1.13.3 Media Failure Field' of exfat specification describe:

If, upon

exfat: retain 'VolumeFlags' properly

MediaFailure and VolumeDirty should be retained if these are set before
mounting.

In '3.1.13.3 Media Failure Field' of exfat specification describe:

If, upon mounting a volume, the value of this field is 1,
implementations which scan the entire volume for media failures and
record all failures as "bad" clusters in the FAT (or otherwise resolve
media failures) may clear the value of this field to 0.

Therefore, We should not clear MediaFailure without scanning volume.

In '8.1 Recommended Write Ordering' of exfat specification describe:

Clear the value of the VolumeDirty field to 0, if its value prior to
the first step was 0.

Therefore, We should not clear VolumeDirty after mounting.
Also rename ERR_MEDIUM to MEDIA_FAILURE.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: add error check when updating dir-entries

Add error check when synchronously updating dir-entries.

Suggested-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Tetsuhiro Kohada <kohada

exfat: add error check when updating dir-entries

Add error check when synchronously updating dir-entries.

Suggested-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: write multiple sectors at once

Write multiple sectors at once when updating dir-entries.
Add exfat_update_bhs() for that. It wait for write completion once
instead of sector by sector.
It's o

exfat: write multiple sectors at once

Write multiple sectors at once when updating dir-entries.
Add exfat_update_bhs() for that. It wait for write completion once
instead of sector by sector.
It's only effective if sync enabled.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: remove EXFAT_SB_DIRTY flag

This flag is set/reset in exfat_put_super()/exfat_sync_fs()
to avoid sync_blockdev().
- exfat_put_super():
Before calling this, the VFS has already called sync_file

exfat: remove EXFAT_SB_DIRTY flag

This flag is set/reset in exfat_put_super()/exfat_sync_fs()
to avoid sync_blockdev().
- exfat_put_super():
Before calling this, the VFS has already called sync_filesystem(),
so sync is never performed here.
- exfat_sync_fs():
After calling this, the VFS calls sync_blockdev(), so, it is meaningless
to check EXFAT_SB_DIRTY or to bypass sync_blockdev() here.

Remove the EXFAT_SB_DIRTY check to ensure synchronization.
And remove the code related to the flag.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: fix overflow issue in exfat_cluster_to_sector()

An overflow issue can occur while calculating sector in
exfat_cluster_to_sector(). It needs to cast clus's type to sector_t
before left shiftin

exfat: fix overflow issue in exfat_cluster_to_sector()

An overflow issue can occur while calculating sector in
exfat_cluster_to_sector(). It needs to cast clus's type to sector_t
before left shifting.

Fixes: 1acf1a564b60 ("exfat: add in-memory and on-disk structures and headers")
Cc: stable@vger.kernel.org # v5.7
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: flush dirty metadata in fsync

generic_file_fsync() exfat used could not guarantee the consistency of
a file because it has flushed not dirty metadata but only dirty data pages
for a file.

In

exfat: flush dirty metadata in fsync

generic_file_fsync() exfat used could not guarantee the consistency of
a file because it has flushed not dirty metadata but only dirty data pages
for a file.

Instead of that, use exfat_file_fsync() for files and directories so that
it guarantees to commit both the metadata and data pages for a file.

Signed-off-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: standardize checksum calculation

To clarify that it is a 16-bit checksum, the parts related to the 16-bit
checksum are renamed and change type to u16.
Furthermore, replace checksum calculatio

exfat: standardize checksum calculation

To clarify that it is a 16-bit checksum, the parts related to the 16-bit
checksum are renamed and change type to u16.
Furthermore, replace checksum calculation in exfat_load_upcase_table()
with exfat_calc_checksum32().

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: add boot region verification

Add Boot-Regions verification specified in exFAT specification.
Note that the checksum type is strongly related to the raw structure,
so the'u32 'type is used to

exfat: add boot region verification

Add Boot-Regions verification specified in exFAT specification.
Note that the checksum type is strongly related to the raw structure,
so the'u32 'type is used to clarify the number of bits.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: redefine PBR as boot_sector

Aggregate PBR related definitions and redefine as "boot_sector" to comply
with the exFAT specification.
And, rename variable names including 'pbr'.

Signed-off-by:

exfat: redefine PBR as boot_sector

Aggregate PBR related definitions and redefine as "boot_sector" to comply
with the exFAT specification.
And, rename variable names including 'pbr'.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: optimize dir-cache

Optimize directory access based on exfat_entry_set_cache.
- Hold bh instead of copied d-entry.
- Modify bh->data directly instead of the copied d-entry.
- Write back the

exfat: optimize dir-cache

Optimize directory access based on exfat_entry_set_cache.
- Hold bh instead of copied d-entry.
- Modify bh->data directly instead of the copied d-entry.
- Write back the retained bh instead of rescanning the d-entry-set.
And
- Remove unused cache related definitions.

Signed-off-by: Tetsuhiro Kohada <kohada.tetsuhiro@dc.mitsubishielectric.co.jp>
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: replace 'time_ms' with 'time_cs'

Replace time_ms with time_cs in the file directory entry structure
and related functions.

The unit of create_time_ms/modify_time_ms in File Directory Entry

exfat: replace 'time_ms' with 'time_cs'

Replace time_ms with time_cs in the file directory entry structure
and related functions.

The unit of create_time_ms/modify_time_ms in File Directory Entry are not
'milli-second', but 'centi-second'.
The exfat specification uses the term '10ms', but instead use 'cs' as in
msdos_fs.h.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: Remove unused functions exfat_high_surrogate() and exfat_low_surrogate()

After applying previous two patches, these functions are not used anymore.

Signed-off-by: Pali Rohár <pali@kernel.org

exfat: Remove unused functions exfat_high_surrogate() and exfat_low_surrogate()

After applying previous two patches, these functions are not used anymore.

Signed-off-by: Pali Rohár <pali@kernel.org>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: Use a more common logging style

Remove the direct use of KERN_<LEVEL> in functions by creating
separate exfat_<level> macros.

Miscellanea:

o Remove several unnecessary terminating newlines

exfat: Use a more common logging style

Remove the direct use of KERN_<LEVEL> in functions by creating
separate exfat_<level> macros.

Miscellanea:

o Remove several unnecessary terminating newlines in formats
o Realign arguments and fit to 80 columns where appropriate

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: truncate atimes to 2s granularity

The timestamp for access_time has double seconds granularity(There is no
10msIncrement field for access_time unlike create/modify_time).
exfat's atimes are r

exfat: truncate atimes to 2s granularity

The timestamp for access_time has double seconds granularity(There is no
10msIncrement field for access_time unlike create/modify_time).
exfat's atimes are restricted to only 2s granularity so after
we set an atime, round it down to the nearest 2s and set the
sub-second component of the timestamp to 0.

Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...

exfat: add in-memory and on-disk structures and headers

This adds in-memory and on-disk structures and headers.

Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Sungjong Seo <sj1

exfat: add in-memory and on-disk structures and headers

This adds in-memory and on-disk structures and headers.

Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Sungjong Seo <sj1557.seo@samsung.com>
Reviewed-by: Pali Rohár <pali.rohar@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

show more ...

exfat: remove 'rwoffset' in exfat_inode_info

Remove 'rwoffset' in exfat_inode_info and replace it with the parameter of
exfat_readdir().
Since rwoffset is referenced only by exfat_re

exfat: remove 'rwoffset' in exfat_inode_info

Remove 'rwoffset' in exfat_inode_info and replace it with the parameter of
exfat_readdir().
Since rwoffset is referenced only by exfat_readdir(), it is not necessary
a exfat_inode_info's member.
Also, change cpos to point to the next of entry-set, and return the index
of dir-entry via dir_entry->entry.

Signed-off-by: Tetsuhiro Kohada <kohada.t2@gmail.com>
Acked-by: Sungjong Seo <sj1557.seo@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>

show more ...