Since sit_i.min_mtime's type is unsigned long long, so we should
initialize it with max value of the type ULLONG_MAX instead of
LLONG_MAX.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
1. thread A: commit_inmem_pages submit data into block layer, but
haven't waited it writeback.
2. thread A: commit_inmem_pages update related node.
3. thread B: do checkpoint, flush all nodes to disk.
4. SPOR
Then, atomic file becomes corrupted since nodes is flushed before data.
This patch fixes to treat atomic page as checkpoint guaranteed one,
then in checkpoint, we can make sure all atomic page can be writebacked
with metadata of atomic file.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
i_current_depth is used only for directory inode, but its space is
shared with i_gc_failures field used for regular inode, in order to
avoid affecting i_gc_failures' value, this patch fixes to initialize
the union's fields according to inode type.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
For extreme case:
10 section, op = 10%, no_fggc_threshold = 90%
All section usage: 85% 85% 85% 85% 90% 90% 95% 95% 95% 95%
During foreground GC, if we skip select dirty section whose usage
is larger than no_fggc_threshold, we can only recycle 80% invalid
space from four 85% usage sections and two 90% usage sections,
result in encountering out-of-space issue.
This reverts commit e93b9865251a0503d83fd570e7d5a7c8bc351715 to
fix this issue, besides, we keep the logic that we scan all dirty
section when searching a victim, so that GC can select victim with
least valid blocks.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Trivial fix to spelling mistake in extension list text
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
In order to avoid the below overflow issue, we should have checked the
boundaries in superblock before reaching out to allocation. As Linus suggested,
the right place should be sanity_check_raw_super().
Dr Silvio Cesare of InfoSect reported:
There are integer overflows with using the cp_payload superblock field in the
f2fs filesystem potentially leading to memory corruption.
include/linux/f2fs_fs.h
struct f2fs_super_block {
...
__le32 cp_payload;
fs/f2fs/f2fs.h
typedef u32 block_t; /*
* should not change u32, since it is the on-disk block
* address format, __le32.
*/
...
static inline block_t __cp_payload(struct f2fs_sb_info *sbi)
{
return le32_to_cpu(F2FS_RAW_SUPER(sbi)->cp_payload);
}
fs/f2fs/checkpoint.c
block_t start_blk, orphan_blocks, i, j;
...
start_blk = __start_cp_addr(sbi) + 1 + __cp_payload(sbi);
orphan_blocks = __start_sum_addr(sbi) - 1 - __cp_payload(sbi);
+++ integer overflows
...
unsigned int cp_blks = 1 + __cp_payload(sbi);
...
sbi->ckpt = kzalloc(cp_blks * blk_size, GFP_KERNEL);
+++ integer overflow leading to incorrect heap allocation.
int cp_payload_blks = __cp_payload(sbi);
...
ckpt->cp_pack_start_sum = cpu_to_le32(1 + cp_payload_blks +
orphan_blocks);
+++ sign bug and integer overflow
...
for (i = 1; i < 1 + cp_payload_blks; i++)
+++ integer overflow
...
sbi->max_orphans = (sbi->blocks_per_seg - F2FS_CP_PACKS -
NR_CURSEG_TYPE - __cp_payload(sbi)) *
F2FS_ORPHANS_PER_BLOCK;
+++ integer overflow
Reported-by: Greg KH <greg@kroah.com>
Reported-by: Silvio Cesare <silvio.cesare@gmail.com>
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Volatile file's data will be updated oftenly, so it'd better to place
its data into hot data segment.
In addition, for atomic file, we change to check FI_ATOMIC_FILE instead
of FI_HOT_DATA to make code readability better.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Introduce release_discard_addr() to include common codes for cleanup.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
[Fengguang Wu: declare static function, reported by kbuild test robot]
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
In build_sit_entries(), if valid_blocks in SIT block is smaller than
valid_blocks in journal, for below calculation:
sbi->discard_blks += old_valid_blocks - se->valid_blocks;
There will be two times potential overflow:
- old_valid_blocks - se->valid_blocks will overflow, and be a very
large number.
- sbi->discard_blks += result will overflow again, comes out a correct
result accidently.
Anyway, it should be fixed.
Fixes: d600af236da5 ("f2fs: avoid unneeded loop in build_sit_entries")
Fixes: 1f43e2ad7bff ("f2fs: introduce CP_TRIMMED_FLAG to avoid unneeded discard")
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
RW semphore dio_rwsem in struct f2fs_inode_info is introduced to avoid
race between dio and data gc, but now, it is more wildly used to avoid
foreground operation vs data gc. So rename it to i_gc_rwsem to improve
its readability.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch move mnt_want_write_file after range check,
it's needless to check arguments with it.
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch fix missing clear FI_NO_PREALLOC in some error case
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This is to give a option for user to be able to recover B/foo in the below
case.
mkdir A
sync()
rename(A, B)
creat (B/foo)
fsync (B/foo)
---crash---
Sugessted-by: Velayudhan Pillai <vijay@cs.utexas.edu>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch moves error handling from commit_inmem_pages() into
__commit_inmem_page() for cleanup, no logic change.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Only dir may have F2FS_INLINE_DOTS flag, so there is no need to check
the flag in recover flow.
Signed-off-by: Sheng Yong <shengyong1@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch removes duplicated dquot_initialize in recover_orphan_inode(),
and fix the error handling if dquot_initialize fails.
Signed-off-by: Sheng Yong <shengyong1@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
dquot_initialize() can fail due to any exception inside quota subsystem,
f2fs needs to be aware of it, and return correct return value to caller.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
v4->v5: move data corruption check to __submit_discard_cmd, in order to
control discard io submitted more accurately, besides, increase async
thread wait time if data corruption detected.
This patch stop async thread and umount process to issue discard
if something wrong with f2fs, which is similar to fstrim.
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
In f2fs_ioc_commit_atomic_write, if file is volatile, return -EINVAL to
indicate that commit failure.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
If a file not set type as hot, has dirty pages more than
threshold 64 before starting atomic write, may be lose hot
flag.
v1->v2: move set FI_ATOMIC_FILE flag behind flush dirty pages too,
in case of dirty pages before starting atomic use atomic mode to
write back.
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
`cur' will never be NULL, we should check inmem_pages list instead.
Signed-off-by: Sheng Yong <shengyong1@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Thread GC thread
- f2fs_ioc_start_atomic_write
- get_dirty_pages
- filemap_write_and_wait_range
- f2fs_gc
- do_garbage_collect
- gc_data_segment
- move_data_page
- f2fs_is_atomic_file
- set_page_dirty
- set_inode_flag(, FI_ATOMIC_FILE)
Dirty data page can still be generated by GC in race condition as
above call stack.
This patch adds fi->dio_rwsem[WRITE] in f2fs_ioc_start_atomic_write
to avoid such race.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
In the structure of f2fs_inode, i_extra_size's type is __le16,
so we should keep type consistent when using it.
Fixes: 704956ecf5bc ("f2fs: support inode checksum")
Signed-off-by: Zhikang Zhang <zhangzhikang1@huawei.com>
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
We should check valid_map_mir and block count to ensure
the flushed raw_sit is correct.
Signed-off-by: Zhikang Zhang <zhangzhikang1@huawei.com>
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Correct return value in two cases:
- return EINVAL if end boundary is out-of-range.
- return EIO if fs needs off-line check.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch fixes to show missing encrypt/inline_data flag in
FS_IOC_GETFLAGS like ext4 does.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Now F2FS_FL_USER_VISIBLE and F2FS_FL_USER_MODIFIABLE has included
F2FS_PROJINHERIT_FL, so remove unneeded F2FS_PROJINHERIT_FL when
using visible/modifiable flag macro.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Related to https://lkml.org/lkml/2018/4/8/661
Sometimes, we need to write meta data to new allocated block address,
then we will allocate a zeroed page in inner inode's address space, and
fill partial data in it, and leave other place with zero value which means
some fields are initial status.
There are two inner inodes (meta inode and node inode) setting __GFP_ZERO,
I have just checked them, for both of them, we can avoid using __GFP_ZERO,
and do initialization by ourselves to avoid unneeded/redundant zeroing
from mm.
Cc: <stable@vger.kernel.org>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch modify max_requests to UINT_MAX, to issue
all big range discards in umount.
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
For buffered IO, we don't need to use block plug to cache bio,
for direct IO, generic f2fs_direct_IO has already added block
plug, so let's remove redundant one in .write_iter.
As Yunlei described in his patch:
-f2fs_file_write_iter
-blk_start_plug
-__generic_file_write_iter
...
-do_blockdev_direct_IO
-blk_start_plug
...
-blk_finish_plug
...
-blk_finish_plug
which may conduct performance decrease in our platform
Signed-off-by: Yunlei He <heyunlei@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Since the layout of regular dentry block is different from inline dentry
block, zero_user_segment starting from MAX_INLINE_DATA(dir) is not
correct for regular dentry block, besides, bitmap is already copied and
used, so there is no necessary to zero page at all, so just remove the
zero_user_segment is OK.
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Previously, we use generic FS_*_FL defined by vfs to indicate inode status
for each bit of i_flags, so f2fs's flag status definition is tied to vfs'
one, it will be hard for f2fs to reuse bits f2fs never used to indicate
new status..
In order to solve this issue, we introduce private inode status mapping,
Note, for these bits have already been persisted into disk, we should
never change their definition, for other ones, we can remap them for
later new coming status.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Log the crypto algorithm driver name for each fscrypt encryption mode on
its first use, also showing a friendly name for the mode.
This will help people determine whether the expected implementations are
being used. In some cases we've seen people do benchmarks and reject
using encryption for performance reasons, when in fact they used a much
slower implementation of AES-XTS than was possible on the hardware. It
can make an enormous difference; e.g., AES-XTS on ARM is about 10x
faster with the crypto extensions (AES instructions) than without.
This also makes it more obvious which modes are being used, now that
fscrypt supports multiple combinations of modes.
Example messages (with default modes, on x86_64):
[ 35.492057] fscrypt: AES-256-CTS-CBC using implementation "cts(cbc-aes-aesni)"
[ 35.492171] fscrypt: AES-256-XTS using implementation "xts-aes-aesni"
Note: algorithms can be dynamically added to the crypto API, which can
result in different implementations being used at different times. But
this is rare; for most users, showing the first will be good enough.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
fscrypt currently only supports AES encryption. However, many low-end
mobile devices have older CPUs that don't have AES instructions, e.g.
the ARMv8 Cryptography Extensions. Currently, user data on such devices
is not encrypted at rest because AES is too slow, even when the NEON
bit-sliced implementation of AES is used. Unfortunately, it is
infeasible to encrypt these devices at all when AES is the only option.
Therefore, this patch updates fscrypt to support the Speck block cipher,
which was recently added to the crypto API. The C implementation of
Speck is not especially fast, but Speck can be implemented very
efficiently with general-purpose vector instructions, e.g. ARM NEON.
For example, on an ARMv7 processor, we measured the NEON-accelerated
Speck128/256-XTS at 69 MB/s for both encryption and decryption, while
AES-256-XTS with the NEON bit-sliced implementation was only 22 MB/s
encryption and 19 MB/s decryption.
There are multiple variants of Speck. This patch only adds support for
Speck128/256, which is the variant with a 128-bit block size and 256-bit
key size -- the same as AES-256. This is believed to be the most secure
variant of Speck, and it's only about 6% slower than Speck128/128.
Speck64/128 would be at least 20% faster because it has 20% rounds, and
it can be even faster on CPUs that can't efficiently do the 64-bit
operations needed for Speck128. However, Speck64's 64-bit block size is
not preferred security-wise. ARM NEON also supports the needed 64-bit
operations even on 32-bit CPUs, resulting in Speck128 being fast enough
for our targeted use cases so far.
The chosen modes of operation are XTS for contents and CTS-CBC for
filenames. These are the same modes of operation that fscrypt defaults
to for AES. Note that as with the other fscrypt modes, Speck will not
be used unless userspace chooses to use it. Nor are any of the existing
modes (which are all AES-based) being removed, of course.
We intentionally don't make CONFIG_FS_ENCRYPTION select
CONFIG_CRYPTO_SPECK, so people will have to enable Speck support
themselves if they need it. This is because we shouldn't bloat the
FS_ENCRYPTION dependencies with every new cipher, especially ones that
aren't recommended for most users. Moreover, CRYPTO_SPECK is just the
generic implementation, which won't be fast enough for many users; in
practice, they'll need to enable CRYPTO_SPECK_NEON to get acceptable
performance.
More details about our choice of Speck can be found in our patches that
added Speck to the crypto API, and the follow-on discussion threads.
We're planning a publication that explains the choice in more detail.
But briefly, we can't use ChaCha20 as we previously proposed, since it
would be insecure to use a stream cipher in this context, with potential
IV reuse during writes on f2fs and/or on wear-leveling flash storage.
We also evaluated many other lightweight and/or ARX-based block ciphers
such as Chaskey-LTS, RC5, LEA, CHAM, Threefish, RC6, NOEKEON, SPARX, and
XTEA. However, all had disadvantages vs. Speck, such as insufficient
performance with NEON, much less published cryptanalysis, or an
insufficient security level. Various design choices in Speck make it
perform better with NEON than competing ciphers while still having a
security margin similar to AES, and in the case of Speck128 also the
same available security levels. Unfortunately, Speck does have some
political baggage attached -- it's an NSA designed cipher, and was
rejected from an ISO standard (though for context, as far as I know none
of the above-mentioned alternatives are ISO standards either).
Nevertheless, we believe it is a good solution to the problem from a
technical perspective.
Certain algorithms constructed from ChaCha or the ChaCha permutation,
such as MEM (Masked Even-Mansour) or HPolyC, may also meet our
performance requirements. However, these are new constructions that
need more time to receive the cryptographic review and acceptance needed
to be confident in their security. HPolyC hasn't been published yet,
and we are concerned that MEM makes stronger assumptions about the
underlying permutation than the ChaCha stream cipher does. In contrast,
the XTS mode of operation is relatively well accepted, and Speck has
over 70 cryptanalysis papers. Of course, these ChaCha-based algorithms
can still be added later if they become ready.
The best known attack on Speck128/256 is a differential cryptanalysis
attack on 25 of 34 rounds with 2^253 time complexity and 2^125 chosen
plaintexts, i.e. only marginally faster than brute force. There is no
known attack on the full 34 rounds.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Currently the key derivation function in fscrypt uses the master key
length as the amount of output key material to derive. This works, but
it means we can waste time deriving more key material than is actually
used, e.g. most commonly, deriving 64 bytes for directories which only
take a 32-byte AES-256-CTS-CBC key. It also forces us to validate that
the master key length is a multiple of AES_BLOCK_SIZE, which wouldn't
otherwise be necessary.
Fix it to only derive the needed length key.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Refactor the confusingly-named function 'validate_user_key()' into a new
function 'find_and_derive_key()' which first finds the keyring key, then
does the key derivation. Among other benefits this avoids the strange
behavior we had previously where if key derivation failed for some
reason, then we would fall back to the alternate key prefix. Now, we'll
only fall back to the alternate key prefix if a valid key isn't found.
This patch also improves the warning messages that are logged when the
keyring key's payload is invalid.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Use a common function for fscrypt warning and error messages so that all
the messages are consistently ratelimited, include the "fscrypt:"
prefix, and include the filesystem name if applicable.
Also fix up a few of the log messages to be more descriptive.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
With one exception, the internal key size constants such as
FS_AES_256_XTS_KEY_SIZE are only used for the 'available_modes' array,
where they really only serve to obfuscate what the values are. Also
some of the constants are unused, and the key sizes tend to be in the
names of the algorithms anyway. In the past these values were also
misused, e.g. we used to have FS_AES_256_XTS_KEY_SIZE in places that
technically should have been FS_MAX_KEY_SIZE.
The exception is that FS_AES_128_ECB_KEY_SIZE is used for key
derivation. But it's more appropriate to use
FS_KEY_DERIVATION_NONCE_SIZE for that instead.
Thus, just put the sizes directly in the 'available_modes' array.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
We're passing 'key_type_logon' to request_key(), so the found key is
guaranteed to be of type "logon". Thus, there is no reason to check
later that the key is really a "logon" key.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Now ->max_namelen() is only called to limit the filename length when
adding NUL padding, and only for real filenames -- not symlink targets.
It also didn't give the correct length for symlink targets anyway since
it forgot to subtract 'sizeof(struct fscrypt_symlink_data)'.
Thus, change ->max_namelen from a function to a simple 'unsigned int'
that gives the filesystem's maximum filename length.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
fname_decrypt() is validating that the encrypted filename is nonempty.
However, earlier a stronger precondition was already enforced: the
encrypted filename must be at least 16 (FS_CRYPTO_BLOCK_SIZE) bytes.
Drop the redundant check for an empty filename.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
fname_decrypt() returns an error if the input filename is longer than
the inode's ->max_namelen() as given by the filesystem. But, this
doesn't actually make sense because the filesystem provided the input
filename in the first place, where it was subject to the filesystem's
limits. And fname_decrypt() has no internal limit itself.
Thus, remove this unnecessary check.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
In fscrypt_setup_filename(), remove the unnecessary check for
fscrypt_get_encryption_info() returning EOPNOTSUPP. There's no reason
to handle this error differently from any other. I think there may have
been some confusion because the "notsupp" version of
fscrypt_get_encryption_info() returns EOPNOTSUPP -- but that's not
applicable from inside fs/crypto/.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
fscrypt is clearing the flags on the crypto_skcipher it allocates for
each inode. But, this is unnecessary and may cause problems in the
future because it will even clear flags that are meant to be internal to
the crypto API, e.g. CRYPTO_TFM_NEED_KEY.
Remove the unnecessary flag clearing.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
