In case of removing of xattr there must be XATTR_REPLACE flag and
zero length. We already check XATTR_REPLACE in ntfs_set_ea, so
now we pass XATTR_REPLACE to ntfs_set_ea.
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We can safely move set_cached_acl because it works with NULL acl too.
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Now ntfs3 locks mutex for smaller time.
Theoretically in successful cases those locks aren't needed at all.
But proving the same for error cases is difficult.
So instead of removing them we just move them.
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We need to always call indx_delete_entry after indx_insert_entry
if error occurred.
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Some discussion has been spoken that this deprecated mount options
should be removed before 5.15 lands. This driver is not never seen day
light so it was decided that nls mount option has to be removed. We have
always possibility to add this if needed.

One possible need is example if current ntfs driver will be taken out of
kernel and ntfs3 needs to support mount options what it has.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

There is already a 'u8 mask' defined at the top of the function.
There is no need to define a new one here.

Remove the useless and shadowing new 'mask' variable.
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

'fnd' has been dereferenced several time before, so testing it here is
pointless.
Moreover, all callers of 'indx_find()' already have some error handling
code that makes sure that no NULL 'fnd' is passed.

So, remove the useless test.
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

This allows to wait only when it's requested.
It speeds up creation of hardlinks.
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

xfstest generic/041 works with 3003 hardlinks.
Because of this we raise hardlinks limit to 4000.
There are no drawbacks or regressions.
Theoretically we can raise all the way up to ffff,
but there is no practical use for this.
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Do not try to insert attribute if there is no room in record.
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Currently a failed allocation on sbi->upcase will cause an exit via
the label free_sbi causing a memory leak on object opts. Fix this by
re-ordering the exit paths free_opts and free_sbi so that kfree's occur
in the reverse allocation order.

Addresses-Coverity: ("Resource leak")
Fixes: 27fac777 ("fs/ntfs3: Init spi more in init_fs_context than fill_super")
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Reviewed-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Current ntfs3 rst documentation is broken. I turn table to list table as
this is current Linux documentation quide line. Simple table also did
not quite work in our situation as we need to span rows together.

It still look quite good as text so we did not loss anything. This will
also make diffing quite bit more pleasure.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Right now sb blocksize first get initiliazed in fill_super but in can be
changed in helper function. It makes more sense to that this happened
only in one place.

Because we move this to helper function it makes more sense that
s_maxbytes will also be there. I rather have every sb releted thing in
fill_super, but because there is already sb releted stuff in this
helper. This will have to do for now.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Initializing should be as close as possible when we use it so that
we do not need to scroll up to see what is happening.

Also bdev_get_queue() can never return NULL so we do not need to check
for !rq.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We can survive without this tmp point upcase. So remove it we don't have
so many tmp pointer in this function.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Drop tmp pointer bd_inode because this is only used ones in fill_super.
Also we have so many initializing happening at the beginning that it is
already way too much to follow.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We only use this in two places so we do not really need it. Also
wrapper sb_rdonly() is pretty self explanatory. This will make little
bit easier to read this super long variable list in the beginning of
ntfs_fill_super().
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Use sb instead of sbi->sb in fill_super. We have sb so why not use
it. Also makes code more readable.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Remove some unnecessary variable loading. These look like copy paste
work and they are not used to anything.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

In many places it is not needed to use goto out. We can just return
right away. This will make code little bit more cleaner as we won't
need to check error path.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Remove root drop when we fault out. This can never happened because
when we allocate root we eather fault when no root or success.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Change EINVAL to ENOMEM when d_make_root fails because that is right
errno.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Fix wrong error message $Logfile -> $UpCase. Probably copy paste.

Fixes: 203c2b3a406a ("fs/ntfs3: Add initialization of super block")
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We can make code little bit more readable by using min/max macros.

These were found with Coccinelle.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We can make code little more readable by using kernel macros clamp/max.

This were found with kernel included Coccinelle minmax script.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We do not need this check as this is same thing as
NTFS_MIN_MFT_ZONE > zlen. We already check NTFS_MIN_MFT_ZONE <= zlen and
exit because is too big request. Remove it so code is cleaner.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

If ntfs_refresh_zone() returns error it will be changed to -ENOSPC. It
is not right. Also caller of this functions also check other errors.

Fixes: 78ab59fe ("fs/ntfs3: Rework file operations")
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Remove tabs before spaces from comment as recommended by kernel coding
style. Checkpatch also warn about these.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Remove braces from single statment block as they are not needed. Also
Linux kernel coding style guide recommend this and checkpatch warn about
this.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

For better code readability place constant always right side of the
test. This will also address checkpatch warning.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

No need for plus sign here. So remove it. Checkpatch will also be happy.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We do not have any reason to keep old linear search in. Before this was
used for error path or if table was so big that it cannot be allocated.
Current binary search implementation won't need error path. Remove old
references to linear entry search.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We could try to optimize algorithm to first fill just small table and
after that use bigger table all the way up to ARRAY_SIZE(offs). This
way we can use bigger search array, but not lose benefits with entry
count smaller < ARRAY_SIZE(offs).
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

Current binary search allocates memory for table and fill whole table
before we start actual binary search. This is quite inefficient because
table fill will always be O(n). Also if table is huge we need to
reallocate memory which is costly.

This implementation use just stack memory and always when table is full
we will check if last element is <= and if not start table fill again.
The idea was that it would be same cost as table reallocation.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We have lot of unnecessary headers in these files. Remove them so that
we help compiler a little bit.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

There is lot of headers which we do not need in this file. Delete them
and add what we really need. Here is list which identify why we need
this header.

<linux/kernel.h> // min()
<linux/slab.h> // kzalloc()
<linux/stddef.h> // offsetof()
<linux/string.h> // memcpy(), memset()
<linux/types.h> // u8, size_t, etc.

"debug.h" // PtrOffset()
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

There is no headers. They will be included through ntfs_fs.c, but that
is not right thing to do. Let's include headers what this file need
straight away.

types.h is needed for __le16, u8 etc.
kernel.h is needed for le16_to_cpu()
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We only need linux/types.h for types like u8 etc. So we can remove rest
and help compiler a little bit.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

size_t needs header. Add missing header guards so that compiler will
only include these ones.
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>

We do not have headers at all in this file. We should have them so that
not every .c file needs to include all of the stuff which this file need
for building. This way we can remove some headers from other files and
get better picture what is needed. This can save some compilation time.
And this can help if we sometimes want to separate this one big header.

Also use forward declarations for structs and enums when it not included
straight with include and it is used in function declarations input.
This will prevent possible compiler warning:
  xxx declared inside parameter list will not be visible
  outside of this definition or declaration

Here is list which I made when parsing this. There is not necessarily
all example from this header file, but this just proofs we need it.

<linux/blkdev.h> SECTOR_SHIFT
<linux/buffer_head.h> sb_bread(), put_bh
<linux/cleancache.h> put_page()
<linux/fs.h> struct inode (Just struct ntfs_inode need it)
<linux/highmem.h> kunmap(), kmap()
<linux/kernel.h> cpu_to_leXX() ALIGN
<linux/mm.h> kvfree()
<linux/mutex.h> struct mutex, mutex_(un/try)lock()
<linux/page-flags.h> PageError()
<linux/pagemap.h> read_mapping_page()
<linux/rbtree.h> struct rb_root
<linux/rwsem.h> struct rw_semaphore
<linux/slab.h> krfree(), kzalloc()
<linux/string.h> memset()
<linux/time64.h> struct timespec64
<linux/types.h> uXX, __leXX
<linux/uidgid.h> kuid_t, kgid_t
<asm/div64.h> do_div()
<asm/page.h> PAGE_SIZE

"debug.h" ntfs_err() (Just one entry. Maybe we can drop this)
"ntfs.h" Do you even ask?
Signed-off-by: NKari Argillander <kari.argillander@gmail.com>
Signed-off-by: NKonstantin Komarov <almaz.alexandrovich@paragon-software.com>