This patch introduces new sysfs entries for users to control the policy of
in-place-updates, namely IPU, in f2fs.
Sometimes f2fs suffers from performance degradation due to its out-of-place
update policy that produces many additional node block writes.
If the storage performance is very dependant on the amount of data writes
instead of IO patterns, we'd better drop this out-of-place update policy.
This patch suggests 5 polcies and their triggering conditions as follows.
[sysfs entry name = ipu_policy]
0: F2FS_IPU_FORCE all the time,
1: F2FS_IPU_SSR if SSR mode is activated,
2: F2FS_IPU_UTIL if FS utilization is over threashold,
3: F2FS_IPU_SSR_UTIL if SSR mode is activated and FS utilization is over
threashold,
4: F2FS_IPU_DISABLE disable IPU. (=default option)
[sysfs entry name = min_ipu_util]
This parameter controls the threshold to trigger in-place-updates.
The number indicates percentage of the filesystem utilization, and used by
F2FS_IPU_UTIL and F2FS_IPU_SSR_UTIL policies.
For more details, see need_inplace_update() in segment.h.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch introduces f2fs_io_info to mitigate the complex parameter list.
struct f2fs_io_info {
enum page_type type; /* contains DATA/NODE/META/META_FLUSH */
int rw; /* contains R/RS/W/WS */
int rw_flag; /* contains REQ_META/REQ_PRIO */
}
1. f2fs_write_data_pages
- DATA
- WRITE_SYNC is set when wbc->WB_SYNC_ALL.
2. sync_node_pages
- NODE
- WRITE_SYNC all the time
3. sync_meta_pages
- META
- WRITE_SYNC all the time
- REQ_META | REQ_PRIO all the time
** f2fs_submit_merged_bio() handles META_FLUSH.
4. ra_nat_pages, ra_sit_pages, ra_sum_pages
- META
- READ_SYNC
Cc: Fan Li <fanofcode.li@samsung.com>
Cc: Changman Lee <cm224.lee@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Previously f2fs submits most of write requests using WRITE_SYNC, but f2fs_write_data_pages
submits last write requests by sync_mode flags callers pass.
This causes a performance problem since continuous pages with different sync flags
can't be merged in cfq IO scheduler(thanks yu chao for pointing it out), and synchronous
requests often take more time.
This patch makes the following modifies to DATA writebacks:
1. every page will be written back using the sync mode caller pass.
2. only pages with the same sync mode can be merged in one bio request.
These changes are restricted to DATA pages.Other types of writebacks are modified
To remain synchronous.
In my test with tiotest, f2fs sequence write performance is improved by about 7%-10% ,
and this patch has no obvious impact on other performance tests.
Signed-off-by: Fan Li <fanofcode.li@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
As we know, some of our branch condition will rarely be true. So we could add
'unlikely' to let compiler optimize these code, by this way we could drop
unneeded 'jump' assemble code to improve performance.
change log:
o add *unlikely* as many as possible across the whole source files at once
suggested by Jaegeuk Kim.
Suggested-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch integrates redundant bio operations on read and write IOs.
1. Move bio-related codes to the top of data.c.
2. Replace f2fs_submit_bio with f2fs_submit_merged_bio, which handles read
bios additionally.
3. Introduce __submit_merged_bio to submit the merged bio.
4. Change f2fs_readpage to f2fs_submit_page_bio.
5. Introduce f2fs_submit_page_mbio to integrate previous submit_read_page and
submit_write_page.
Reviewed-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Reviewed-by: Chao Yu <chao2.yu@samsung.com >
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
The recover_orphan_inodes() returns no error all the time, so we don't need to
check its errors.
Signed-off-by: Chao Yu <chao2.yu@samsung.com>
[Jaegeuk Kim: add description]
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch removes the unnecessary condition checks on:
fs/f2fs/gc.c:667 do_garbage_collect() warn: 'sum_page' isn't an ERR_PTR
fs/f2fs/f2fs.h:795 f2fs_put_page() warn: 'page' isn't an ERR_PTR
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Add new inode flags F2FS_INLINE_DATA and FI_INLINE_DATA to indicate
whether the inode has inline data.
Inline data makes use of inode block's data indices region to save small
file. Currently there are 923 data indices in an inode block. Since
inline xattr has made use of the last 50 indices to save its data, there
are 873 indices left which can be used for inline data. When
FI_INLINE_DATA is set, the layout of inode block's indices region is
like below:
+-----------------+
| | Reserved. reserve_new_block() will make use of
| i_addr[0] | i_addr[0] when we need to reserve a new data block
| | to convert inline data into regular one's.
|-----------------|
| | Used by inline data. A file whose size is less than
| i_addr[1~872] | 3488 bytes(~3.4k) and doesn't reserve extra
| | blocks by fallocate() can be saved here.
|-----------------|
| |
| i_addr[873~922] | Reserved for inline xattr
| |
+-----------------+
Signed-off-by: Haicheng Li <haicheng.li@linux.intel.com>
Signed-off-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Weihong Xu <weihong.xu@intel.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Add the function f2fs_reserve_block() to easily reserve new blocks, and
use it to clean up more codes.
Signed-off-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Haicheng Li <haicheng.li@linux.intel.com>
Signed-off-by: Weihong Xu <weihong.xu@intel.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
For better read performance, we add a new function to support for merging
contiguous read as the one for write.
v1-->v2:
o add declarations here as Gu Zheng suggested.
o use new structure f2fs_bio_info introduced by Jaegeuk Kim.
Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Acked-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
The f2fs has three bio types, NODE, DATA, and META, and manages some data
structures per each bio types.
The codes are a little bit messy, thus, this patch introduces a bio array
which groups individual data structures as follows.
struct f2fs_bio_info {
struct bio *bio; /* bios to merge */
sector_t last_block_in_bio; /* last block number */
struct mutex io_mutex; /* mutex for bio */
};
struct f2fs_sb_info {
...
struct f2fs_bio_info write_io[NR_PAGE_TYPE]; /* for write bios */
...
};
The code changes from this new data structure are trivial.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
The f2fs manages an extent cache to search a number of consecutive data blocks
very quickly.
However it conducts unnecessary cache operations if the file is highly
fragmented with no valid extent cache.
In such the case, we don't need to handle the extent cache, but just can disable
the cache facility.
Nevertheless, this patch gives one more chance to enable the extent cache.
For example,
1. create a file
2. write data sequentially which produces a large valid extent cache
3. update some data, resulting in a fragmented extent
4. if the fragmented extent is too small, then drop extent cache
5. close the file
6. open the file again
7. give another chance to make a new extent cache
8. write data sequentially again which creates another big extent cache.
...
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch removes an unnecessary semaphore (i.e., sbi->bio_sem).
There is no reason to use the semaphore when f2fs submits read and write IOs.
Instead, let's use a write mutex and cover the sbi->bio[] by the lock.
Change log from v1:
o split write_mutex suggested by Chao Yu
Chao described,
"All DATA/NODE/META bio buffers in superblock is protected by
'sbi->write_mutex', but each bio buffer area is independent, So we
should split write_mutex to three for DATA/NODE/META."
Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch adds a slab cache entry for small discards.
Each entry consists of:
struct discard_entry {
struct list_head list; /* list head */
block_t blkaddr; /* block address to be discarded */
int len; /* # of consecutive blocks of the discard */
};
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
use genernal method supported by kernel
o changes from v1
If any waiter exists at end io, wake up it.
Signed-off-by: Changman Lee <cm224.lee@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Previously, f2fs_sync_file() waits for all the node blocks to be written.
But, we don't need to do that, but wait only the inode-related node blocks.
This patch adds wait_on_node_pages_writeback() in which waits inode-related
node blocks that are on writeback.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
If you want to remove unnecessary BUG_ONs, you can just turn off F2FS_CHECK_FS
in your kernel config.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch merges some background jobs into this new function.
Signed-off-by: Changman Lee <cm224.lee@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Previously, f2fs postpones reclaiming prefree segments into free segments
as much as possible.
However, if user writes and deletes a bunch of data without any sync or fsync
calls, some flash storages can suffer from garbage collections.
So, this patch adds the reclaiming codes to f2fs_write_node_pages and background
GC thread.
If there are a lot of prefree segments, let's do checkpoint so that f2fs
submits discard commands for the prefree regions to the flash storage.
Signed-off-by: Changman Lee <cm224.lee@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Introduce the unfailed version of kmem_cache_alloc named f2fs_kmem_cache_alloc
to hide the retry routine and make the code a bit cleaner.
v2:
Fix the wrong use of 'retry' tag pointed out by Gao feng.
Use more neat code to remove redundant tag suggested by Haicheng Li.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Previously, do_checkpoint() will call congestion_wait() for waiting the pages
(previous submitted node/meta/data pages) to be written back.
Because congestion_wait() will set a regular period (e.g. HZ / 50 ) for waiting, and
no additional wake up mechanism was introduced if IO ends up before regular period costed.
Yuan Zhong found there is a situation that after the pages have been written back,
but the checkpoint thread still wait for congestion_wait to exit.
So here we store checkpoint task into f2fs_sb when doing checkpoint, it'll wait for IO completes
if there's IO going on, and in the end IO path, wake up checkpoint task when IO ends up.
Thanks to Yuan Zhong's pre work about this problem.
Reported-by: Yuan Zhong <yuan.mark.zhong@samsung.com>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch removes the logic previously introduced to address the starvation
on cp_rwsem.
One potential there-in bug is that we should cover the wait.list with spin_lock,
but the previous code broke this rule.
And, actually current rwsem handles this starvation issue reasonably, so that we
didn't need to do this before neither.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
The fs_locks is used to block other ops(ex, recovery) when doing checkpoint.
And each other operate routine(besides checkpoint) needs to acquire a fs_lock,
there is a terrible problem here, if these are too many concurrency threads acquiring
fs_lock, so that they will block each other and may lead to some performance problem,
but this is not the phenomenon we want to see.
Though there are some optimization patches introduced to enhance the usage of fs_lock,
but the thorough solution is using a *rw_sem* to replace the fs_lock.
Checkpoint routine takes write_sem, and other ops take read_sem, so that we can block
other ops(ex, recovery) when doing checkpoint, and other ops will not disturb each other,
this can avoid the problem described above completely.
Because of the weakness of rw_sem, the above change may introduce a potential problem
that the checkpoint thread might get starved if other threads are intensively locking
the read semaphore for I/O.(Pointed out by Xu Jin)
In order to avoid this, a wait_list is introduced, the appending read semaphore ops
will be dropped into the wait_list if checkpoint thread is waiting for write semaphore,
and will be waked up when checkpoint thread gives up write semaphore.
Thanks to Kim's previous review and test, and will be very glad to see other guys'
performance tests about this patch.
V2:
-fix the potential starvation problem.
-use more suitable func name suggested by Xu Jin.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
[Jaegeuk Kim: adjust minor coding standard]
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
There is a performance problem: when all sbi->fs_lock are holded, then
all the following threads may get the same next_lock value from sbi->next_lock_num
in function mutex_lock_op, and wait for the same lock(fs_lock[next_lock]),
it may cause performance reduce.
So we move the sbi->next_lock_num++ before getting lock, this will average the
following threads if all sbi->fs_lock are holded.
v1-->v2:
Drop the needless spin_lock as Jaegeuk suggested.
Suggested-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Signed-off-by: Yu Chao <chao2.yu@samsung.com>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
0. modified inode structure
--------------------------------------
metadata (e.g., i_mtime, i_ctime, etc)
--------------------------------------
direct pointers [0 ~ 873]
inline xattrs (200 bytes by default)
indirect pointers [0 ~ 4]
--------------------------------------
node footer
--------------------------------------
1. setxattr flow
- read_all_xattrs copies all the xattrs from inline and xattr node block.
- handle xattr entries
- write_all_xattrs copies modified xattrs into inline and xattr node block.
2. getxattr flow
- read_all_xattrs copies all the xattrs from inline and xattr node block.
- check target entries
3. Usage
# mount -t f2fs -o inline_xattr $DEV $MNT
Once mounted with the inline_xattr option, f2fs marks all the newly created
files to reserve an amount of inline xattr space explicitly inside the inode
block. Without the mount option, f2fs will not touch any existing files and
newly created files as well.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch enables the number of direct pointers inside on-disk inode block to
be changed dynamically according to the size of inline xattr space.
The number of direct pointers, ADDRS_PER_INODE, can be changed only if the file
has inline xattr flag.
The number of direct pointers that will be used by inline xattrs is defined as
F2FS_INLINE_XATTR_ADDRS.
Current patch assigns F2FS_INLINE_XATTR_ADDRS to 0 temporarily.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch adds basic inode flags for inline xattrs, F2FS_INLINE_XATTR,
and add a mount option, inline_xattr, which is enabled when xattr is set.
If the mount option is enabled, all the files are marked with the inline_xattrs
flag.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Previously xattr node blocks are stored to the COLD_NODE log, which means that
our roll-forward mechanism doesn't recover the xattr node blocks at all.
Only the direct node blocks in the WARM_NODE log can be recovered.
So, let's resolve the issue simply by conducting checkpoint during fsync when a
file has a modified xattr node block.
This approach is able to degrade the performance, but normally the checkpoint
overhead is shown at the initial fsync call after the xattr entry changes.
Once the checkpoint is done, no additional overhead would be occurred.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch fixes the use of XATTR_NODE_OFFSET.
o The offset should not use several MSB bits which are used by marking node
blocks.
o IS_DNODE should handle XATTR_NODE_OFFSET to avoid potential abnormality
during the fsync call.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch should resolve the following error reported by kbuild test robot.
All error/warnings:
In file included from fs/f2fs/dir.c:13:0:
>> fs/f2fs/f2fs.h:435:17: error: field 's_kobj' has incomplete type
struct kobject s_kobj;
The failure was caused by missing the kobject header file in dir.c.
So, this patch move the header file to the right location, f2fs.h.
CC: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch fixes a deadlock bug that occurs quite often when there are
concurrent write and fsync on a same file.
Following is the simplified call trace when tasks get hung.
fsync thread:
- f2fs_sync_file
...
- f2fs_write_data_pages
...
- update_extent_cache
...
- update_inode
- wait_on_page_writeback
bdi writeback thread
- __writeback_single_inode
- f2fs_write_data_pages
- mutex_lock(sbi->writepages)
The deadlock happens when the fsync thread waits on a inode page that has
been added to the f2fs' cached bio sbi->bio[NODE], and unfortunately,
no one else could be able to submit the cached bio to block layer for
writeback. This is because the fsync thread already hold a sbi->fs_lock and
the sbi->writepages lock, causing the bdi thread being blocked when attempt
to write data pages for the same inode. At the same time, f2fs_gc thread
does not notice the situation and could not help. Even the sync syscall
gets blocked.
To fix it, we could submit the cached bio first before waiting on a inode page
that is being written back.
Signed-off-by: Jin Xu <jinuxstyle@gmail.com>
[Jaegeuk Kim: add more cases to use f2fs_wait_on_page_writeback]
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Add sysfs entries to control the timing parameters for
f2fs gc thread.
Various Sysfs options introduced are:
gc_min_sleep_time: Min Sleep time for GC in ms
gc_max_sleep_time: Max Sleep time for GC in ms
gc_no_gc_sleep_time: Default Sleep time for GC in ms
Cc: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Pankaj Kumar <pankaj.km@samsung.com>
Reviewed-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
[Jaegeuk Kim: fix an umount bug and some minor changes]
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch fixes mishandling of the sbi->n_orphans variable.
If users request lots of f2fs_unlink(), check_orphan_space() could be contended.
In such the case, sbi->n_orphans can be read incorrectly so that f2fs_unlink()
would fall into the wrong state which results in the failure of
add_orphan_inode().
So, let's increment sbi->n_orphans virtually prior to the actual orphan inode
stuffs. After that, let's release sbi->n_orphans by calling release_orphan_inode
or remove_orphan_inode.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
The error is reproducible by:
0. mkfs.f2fs /dev/sdb1 & mount
1. touch test1
2. touch test2
3. mv test1 test2
4. umount
5. dumpt.f2fs -i 4 /dev/sdb1
After this, when we retrieve the inode->i_name of test2 by dump.f2fs, we get
test1 instead of test2.
This is because f2fs didn't update the file name during the f2fs_rename.
So, this patch fixes that.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Introduce help function F2FS_NODE() to simplify the conversion of node_page to
f2fs_node.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Add a help func F2FS_STAT() to get the f2fs_stat_info.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
While calculating CRC for the checkpoint block, we use __u32, but when storing
the crc value to the disk, we use __le32.
Let's fix the inconsistency.
Reported-and-Tested-by: Oded Gabbay <ogabbay@advaoptical.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
If a file is linked, f2fs loose its parent inode number so that fsync calls
for the linked file should do checkpoint all the time.
But, if we can recover its parent inode number after the checkpoint, we can
adjust roll-forward mechanism for the further fsync calls, which is able to
improve the fsync performance significatly.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
It's used only locally and could be static.
Signed-off-by: Haicheng Li <haicheng.li@linux.intel.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
If update_inode is called, we don't need to do write_inode.
So, let's use a *dirty* flag for each inode.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
If new dentry block is allocated and its i_size is updated, we should update
its inode block together in order to sync i_size and its block allocation.
Otherwise, we can loose additional dentry block due to the unconsistent i_size.
Errorneous Scenario
-------------------
In the recovery routine,
- recovery_dentry
| - __f2fs_add_link
| | - get_new_data_page
| | | - i_size_write(new_i_size)
| | | - mark_inode_dirty_sync(dir)
| | - update_parent_metadata
| | | - mark_inode_dirty(dir)
|
- write_checkpoint
- sync_dirty_dir_inodes
- filemap_flush(dentry_blocks)
- f2fs_write_data_page
- skip to write the last dentry block due to index < i_size
In the above flow, new_i_size is not updated to its inode block so that the
last dentry block will be lost accordingly.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Basically an inode manages the number of allocated blocks with inode->i_blocks
which is represented in a unit of sectors, not file system blocks.
But, f2fs has used i_blocks in a unit of file system blocks, and f2fs_getattr
translates it to the number of sectors when fstat is called.
However, previously f2fs_file_inode_operations only has this, so this patch adds
it to all the types of inode_operations.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch adds the support of security labels for f2fs, which will be used
by Linus Security Models (LSMs).
Quote from http://en.wikipedia.org/wiki/Linux_Security_Modules:
"Linux Security Modules (LSM) is a framework that allows the Linux kernel to
support a variety of computer security models while avoiding favoritism toward
any single security implementation. The framework is licensed under the terms of
the GNU General Public License and is standard part of the Linux kernel since
Linux 2.6. AppArmor, SELinux, Smack and TOMOYO Linux are the currently accepted
modules in the official kernel.".
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
It is possible that iput is skipped after iget during the recovery.
In recover_dentry(),
dir = f2fs_iget();
...
if (de && inode->i_ino == le32_to_cpu(de->ino))
goto out;
In this case, this dir is not able to be added in dirty_dir_inode_list.
The actual linking is done only when set_page_dirty() is called.
So let's add this newly got inode into the list explicitly, and put it at the
end of the recovery routine.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Some, counters are needed only for the statistical information
while debugging.
So, those can be controlled using CONFIG_F2FS_STAT_FS,
pushing the usage for few variables under this flag.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
The on-disk block address is defined as __le32, but in-memory block address,
block_t, does as u64.
Let's synchronize them to 32 bits.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch fixes the following deadlock bug during the recovery.
INFO: task mount:1322 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
mount D ffffffff81125870 0 1322 1266 0x00000000
ffff8801207e39d8 0000000000000046 ffff88012ab1dee0 0000000000000046
ffff8801207e3a08 ffff880115903f40 ffff8801207e3fd8 ffff8801207e3fd8
ffff8801207e3fd8 ffff880115903f40 ffff8801207e39d8 ffff88012fc94520
Call Trace:
[<ffffffff81125870>] ? __lock_page+0x70/0x70
[<ffffffff816a92d9>] schedule+0x29/0x70
[<ffffffff816a93af>] io_schedule+0x8f/0xd0
[<ffffffff8112587e>] sleep_on_page+0xe/0x20
[<ffffffff816a649a>] __wait_on_bit_lock+0x5a/0xc0
[<ffffffff81125867>] __lock_page+0x67/0x70
[<ffffffff8106c7b0>] ? autoremove_wake_function+0x40/0x40
[<ffffffff81126857>] find_lock_page+0x67/0x80
[<ffffffff8112698f>] find_or_create_page+0x3f/0xb0
[<ffffffffa03901a8>] ? sync_inode_page+0xa8/0xd0 [f2fs]
[<ffffffffa038fdf7>] get_node_page+0x67/0x180 [f2fs]
[<ffffffffa039818b>] recover_fsync_data+0xacb/0xff0 [f2fs]
[<ffffffff816aaa1e>] ? _raw_spin_unlock+0x3e/0x40
[<ffffffffa0389634>] f2fs_fill_super+0x7d4/0x850 [f2fs]
[<ffffffff81184cf9>] mount_bdev+0x1c9/0x210
[<ffffffffa0388e60>] ? validate_superblock+0x180/0x180 [f2fs]
[<ffffffffa0387635>] f2fs_mount+0x15/0x20 [f2fs]
[<ffffffff81185a13>] mount_fs+0x43/0x1b0
[<ffffffff81145ba0>] ? __alloc_percpu+0x10/0x20
[<ffffffff811a0796>] vfs_kern_mount+0x76/0x120
[<ffffffff811a2cb7>] do_mount+0x237/0xa10
[<ffffffff81140b9b>] ? strndup_user+0x5b/0x80
[<ffffffff811a3520>] SyS_mount+0x90/0xe0
[<ffffffff816b3502>] system_call_fastpath+0x16/0x1b
The bug is triggered when check_index_in_prev_nodes tries to get the direct
node page by calling get_node_page.
At this point, if the direct node page is already locked by get_dnode_of_data,
its caller, we got a deadlock condition.
This patch adds additional condition check for the reuse of locked direct node
pages prior to the get_node_page call.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Majianpeng reported a lockdep splat for f2fs. It turns out mutex_lock_all()
acquires an array of locks (in global/local lock style).
Any such operation is always serialized using cp_mutex, therefore there is no
fs_lock[] lock-order issue; tell lockdep about this using the
mutex_lock_nest_lock() primitive.
Reported-by: majianpeng <majianpeng@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
I found a bug when testing power-off-recovery as follows.
[Bug Scenario]
1. create a file
2. fsync the file
3. reboot w/o any sync
4. try to recover the file
- found its fsync mark
- found its dentry mark
: try to recover its dentry
- get its file name
- get its parent inode number
: here we got zero value
The reason why we get the wrong parent inode number is that we didn't
synchronize the inode page with its newly created inode information perfectly.
Especially, previous f2fs stores fi->i_pino and writes it to the cached
node page in a wrong order, which incurs the zero-valued i_pino during the
recovery.
So, this patch modifies the creation flow to fix the synchronization order of
inode page with its inode.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
During the dentry recovery routine, recover_inode() triggers __f2fs_add_link
with its directory inode.
In the following scenario, a bug is captured.
1. dir = f2fs_iget(pino)
2. __f2fs_add_link(dir, name)
3. iput(dir)
-> f2fs_evict_inode() faces with BUG_ON(atomic_read(fi->dirty_dents))
Kernel BUG at ffffffffa01c0676 [verbose debug info unavailable]
[<ffffffffa01c0676>] f2fs_evict_inode+0x276/0x300 [f2fs]
Call Trace:
[<ffffffff8118ea00>] evict+0xb0/0x1b0
[<ffffffff8118f1c5>] iput+0x105/0x190
[<ffffffffa01d2dac>] recover_fsync_data+0x3bc/0x1070 [f2fs]
[<ffffffff81692e8a>] ? io_schedule+0xaa/0xd0
[<ffffffff81690acb>] ? __wait_on_bit_lock+0x7b/0xc0
[<ffffffff8111a0e7>] ? __lock_page+0x67/0x70
[<ffffffff81165e21>] ? kmem_cache_alloc+0x31/0x140
[<ffffffff8118a502>] ? __d_instantiate+0x92/0xf0
[<ffffffff812a949b>] ? security_d_instantiate+0x1b/0x30
[<ffffffff8118a5b4>] ? d_instantiate+0x54/0x70
This means that we should flush all the dentry pages between iget and iput().
But, during the recovery routine, it is unallowed due to consistency, so we
have to wait the whole recovery process.
And then, write_checkpoint flushes all the dirty dentry blocks, and nicely we
can put the stale dir inodes from the dirty_dir_inode_list.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
In order to avoid build_free_nid lock contention, let's change the order of
function calls as follows.
At first, check whether there is enough free nids.
- If available, just get a free nid with spin_lock without any overhead.
- Otherwise, conduct build_free_nids.
: scan nat pages, journal nat entries, and nat cache entries.
We should consider carefullly not to serve free nids intermediately made by
build_free_nids.
We can get stable free nids only after build_free_nids is done.
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
In the previous version, f2fs uses global locks according to the usage types,
such as directory operations, block allocation, block write, and so on.
Reference the following lock types in f2fs.h.
enum lock_type {
RENAME, /* for renaming operations */
DENTRY_OPS, /* for directory operations */
DATA_WRITE, /* for data write */
DATA_NEW, /* for data allocation */
DATA_TRUNC, /* for data truncate */
NODE_NEW, /* for node allocation */
NODE_TRUNC, /* for node truncate */
NODE_WRITE, /* for node write */
NR_LOCK_TYPE,
};
In that case, we lose the performance under the multi-threading environment,
since every types of operations must be conducted one at a time.
In order to address the problem, let's share the locks globally with a mutex
array regardless of any types.
So, let users grab a mutex and perform their jobs in parallel as much as
possbile.
For this, I propose a new global lock scheme as follows.
0. Data structure
- f2fs_sb_info -> mutex_lock[NR_GLOBAL_LOCKS]
- f2fs_sb_info -> node_write
1. mutex_lock_op(sbi)
- try to get an avaiable lock from the array.
- returns the index of the gottern lock variable.
2. mutex_unlock_op(sbi, index of the lock)
- unlock the given index of the lock.
3. mutex_lock_all(sbi)
- grab all the locks in the array before the checkpoint.
4. mutex_unlock_all(sbi)
- release all the locks in the array after checkpoint.
5. block_operations()
- call mutex_lock_all()
- sync_dirty_dir_inodes()
- grab node_write
- sync_node_pages()
Note that,
the pairs of mutex_lock_op()/mutex_unlock_op() and
mutex_lock_all()/mutex_unlock_all() should be used together.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch removes a bitmap for victim segments selected by foreground GC, and
modifies the other bitmap for victim segments selected by background GC.
1) foreground GC bitmap
: We don't need to manage this, since we just only one previous victim section
number instead of the whole victim history.
The f2fs uses the victim section number in order not to allocate currently
GC'ed section to current active logs.
2) background GC bitmap
: This bitmap is used to avoid selecting victims repeatedly by background GCs.
In addition, the victims are able to be selected by foreground GCs, since
there is no need to read victim blocks during foreground GCs.
By the fact that the foreground GC reclaims segments in a section unit, it'd
be better to manage this bitmap based on the section granularity.
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
When we recover fsync'ed data after power-off-recovery, we should guarantee
that any parent inode number should be correct for each direct inode blocks.
So, let's make the following rules.
- The fsync should do checkpoint to all the inodes that were experienced hard
links.
- So, the only normal files can be recovered by roll-forward.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch removes data_version check flow during the fsync call.
The original purpose for the use of data_version was to avoid writng inode
pages redundantly by the fsync calls repeatedly.
However, when user can modify file meta and then call fsync, we should not
skip fsync procedure.
So, let's remove this condition check and hope that user triggers in right
manner.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
We should handle errors during the recovery flow correctly.
For example, if we get -ENOMEM, we should report a mount failure instead of
conducting the remained mount procedure.
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
In function check_nid_range, there is no need to trigger BUG_ON and make kernel stop.
Instead it could just check and indicate the inode number to be EINVAL.
Update the return path in do_read_inode to use the return from check_nid_range.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
[Jaegeuk: replace BUG_ON with WARN_ON]
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Previously, f2fs reads several node pages ahead when get_dnode_of_data is called
with RDONLY_NODE flag.
And, this flag is set by the following functions.
- get_data_block_ro
- get_lock_data_page
- do_write_data_page
- truncate_blocks
- truncate_hole
However, this readahead mechanism is initially introduced for the use of
get_data_block_ro to enhance the sequential read performance.
So, let's clarify all the cases with the additional modes as follows.
enum {
ALLOC_NODE, /* allocate a new node page if needed */
LOOKUP_NODE, /* look up a node without readahead */
LOOKUP_NODE_RA, /*
* look up a node with readahead called
* by get_datablock_ro.
*/
}
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
Pull f2fs cleanup patches from Al Viro:
f2fs: get rid of fake on-stack dentries
f2fs: switch init_inode_metadata() to passing parent and name separately
f2fs: switch new_inode_page() from dentry to qstr
f2fs: init_dent_inode() should take qstr
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Conflicts:
fs/f2fs/recovery.c
adding compat_ioctl to provide support for backward comptability - 32bit binary
execution on 64bit kernel.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch makes clearer the ambiguous f2fs_gc flow as follows.
1. Remove intermediate checkpoint condition during f2fs_gc
(i.e., should_do_checkpoint() and GC_BLOCKED)
2. Remove unnecessary return values of f2fs_gc because of #1.
(i.e., GC_NODE, GC_OK, etc)
3. Simplify write_checkpoint() because of #2.
4. Clarify the main f2fs_gc flow.
o monitor how many freed sections during one iteration of do_garbage_collect().
o do GC more without checkpoints if we can't get enough free sections.
o do checkpoint once we've got enough free sections through forground GCs.
5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data
log types. See. get_ssr_segement()
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Introduce accessor to get the sections based upon the block type
(node,dents...) and modify the functions : should_do_checkpoint,
has_not_enough_free_secs to use this accessor function to get
the node sections and dent sections.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
1. Background
Previously, if f2fs tries to move data blocks of an *evicting* inode during the
cleaning process, it stops the process incompletely and then restarts the whole
process, since it needs a locked inode to grab victim data pages in its address
space. In order to get a locked inode, iget_locked() by f2fs_iget() is normally
used, but, it waits if the inode is on freeing.
So, here is a deadlock scenario.
1. f2fs_evict_inode() <- inode "A"
2. f2fs_balance_fs()
3. f2fs_gc()
4. gc_data_segment()
5. f2fs_iget() <- inode "A" too!
If step #1 and #5 treat a same inode "A", step #5 would fall into deadlock since
the inode "A" is on freeing. In order to resolve this, f2fs_iget_nowait() which
skips __wait_on_freeing_inode() was introduced in step #5, and stops f2fs_gc()
to complete f2fs_evict_inode().
1. f2fs_evict_inode() <- inode "A"
2. f2fs_balance_fs()
3. f2fs_gc()
4. gc_data_segment()
5. f2fs_iget_nowait() <- inode "A", then stop f2fs_gc() w/ -ENOENT
2. Problem and Solution
In the above scenario, however, f2fs cannot finish f2fs_evict_inode() only if:
o there are not enough free sections, and
o f2fs_gc() tries to move data blocks of the *evicting* inode repeatedly.
So, the final solution is to use f2fs_iget() and remove f2fs_balance_fs() in
f2fs_evict_inode().
The f2fs_evict_inode() actually truncates all the data and node blocks, which
means that it doesn't produce any dirty node pages accordingly.
So, we don't need to do f2fs_balance_fs() in practical.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
In f2fs_inode_info structure, the description for data_version
has a typo mistake. It should be latest instead of lastes.
So, correcting that.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch enhances the checkpoint routine to cope with IO errors.
Basically f2fs detects IO errors from end_io_write, and the errors are able to
be occurred during one of data, node, and meta page writes.
In the previous code, when an IO error is occurred during writes, f2fs sets a
flag, CP_ERROR_FLAG, in the raw ckeckpoint buffer which will be written to disk.
Afterwards, write_checkpoint() will check the flag and remount f2fs as a
read-only (ro) mode.
However, even once f2fs is remounted as a ro mode, dirty checkpoint pages are
freely able to be written to disk by flusher or kswapd in background.
In such a case, after cold reboot, f2fs would restore the checkpoint data having
CP_ERROR_FLAG, resulting in disabling write_checkpoint and remounting f2fs as
a ro mode again.
Therefore, let's prevent any checkpoint page (meta) writes once an IO error is
occurred, and remount f2fs as a ro mode right away at that moment.
Reported-by: Oliver Winker <oliver@oli1170.net>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
for one thing, it doesn't (and shouldn't) use anything else from dentry;
for another, on some call chains the dentry is fake and should
be eliminated completely.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Add __init to functions in init_f2fs_fs for code consistency.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
As the "status" debugfs entry will be maintained for entire F2FS filesystem
irrespective of the number of partitions.
So, we can move the initialization to the init part of the f2fs and destroy will
be done from exit part. After making changes, for individual partition mount -
entry creation code will not be executed.
Signed-off-by: Jianpeng Ma <majianpeng@gmail.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
I'd like to revisit the f2fs_gc flow and rewrite as follows.
1. In practical, the nGC parameter of f2fs_gc is meaningless. So, let's
remove it.
2. Background GC marks victim blocks as dirty one at a time.
3. Foreground GC should do cleaning job until acquiring enough free
sections. Afterwards, it needs to do checkpoint.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch resolves Coverity #753112.
In practical, the existing code flow does not fall into the reported errorneous
path. But, anyway, let's avoid this for future.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Introduced f2fs_msg function to differentiate f2fs specific messages in
the log. And, added few informative prints in the mount path, to convey
proper error in case of mount failure.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch is intended to unify string length declarations and usage.
There are number of calls to strlen which return size_t object.
The size of this object depends on compiler if it will be bigger,
equal or even smaller than an unsigned int
Signed-off-by: Leon Romanovsky <leon@leon.nu>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Previously, f2fs didn't track the parent inode number correctly which is stored
in each f2fs_inode. In the case of the following scenario, a bug can be occured.
Let's suppose there are one directory, "/b", and two files, "/a" and "/b/a".
- pino of "/a" is ROOT_INO.
- pino of "/b/a" is DIR_B_INO.
Then,
# sync
: The inode pages of "/a" and "/b/a" contain the parent inode numbers as
ROOT_INO and DIR_B_INO respectively.
# mv /a /b/a
: The parent inode number of "/a" should be changed to DIR_B_INO, but f2fs
didn't do that. Ref. f2fs_set_link().
In order to fix this clearly, I added i_pino in f2fs_inode_info, and whenever
it needs to be changed like in f2fs_add_link() and f2fs_set_link(), it is
updated temporarily in f2fs_inode_info.
And later, f2fs_write_inode() stores the latest information to the inode pages.
For power-off-recovery, f2fs_sync_file() triggers simply f2fs_write_inode().
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Do cleanup more for better code readability.
- Change the parameter set of f2fs_bio_alloc()
This function should allocate a bio only since it is not something like
f2fs_bio_init(). Instead, the caller should initialize the allocated bio.
- Introduce SECTOR_FROM_BLOCK
This macro translates a block address to its sector address.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
As pointed out by Randy Dunlap, this patch removes all usage of "/**" for comment
blocks. Instead, just use "/*".
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This patch should resolve the bugs reported by the sparse tool.
Initial reports were written by "kbuild test robot" managed by fengguang.wu.
In my local machines, I've tested also by running:
> make C=2 CF="-D__CHECK_ENDIAN__"
Accordingly, I've found lots of warnings and bugs related to the endian
conversion. And I've fixed all at this moment.
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
This adds the following major in-memory structures in f2fs.
- f2fs_sb_info:
contains f2fs-specific information, two special inode pointers for node and
meta address spaces, and orphan inode management.
- f2fs_inode_info:
contains vfs_inode and other fs-specific information.
- f2fs_nm_info:
contains node manager information such as NAT entry cache, free nid list,
and NAT page management.
- f2fs_node_info:
represents a node as node id, inode number, block address, and its version.
- f2fs_sm_info:
contains segment manager information such as SIT entry cache, free segment
map, current active logs, dirty segment management, and segment utilization.
The specific structures are sit_info, free_segmap_info, dirty_seglist_info,
curseg_info.
In addition, add F2FS_SUPER_MAGIC in magic.h.
Signed-off-by: Chul Lee <chur.lee@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
