There is a bug in ext4_ext_shift_path_extents() where if we actually
manage to merge a extent we would skip shifting the next extent. This
will result in in one extent in the extent tree not being properly
shifted.
This is causing failure in various xfstests tests using fsx or fsstress
with collapse range support. It will also cause file system corruption
which looks something like:
e2fsck 1.42.9 (4-Feb-2014)
Pass 1: Checking inodes, blocks, and sizes
Inode 20 has out of order extents
(invalid logical block 3, physical block 492938, len 2)
Clear? yes
...
when running e2fsck.
It's also very easily reproducible just by running fsx without any
parameters. I can usually hit the problem within a minute.
Fix it by increasing ex_start only if we're not merging the extent.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
Currently in ext4_collapse_range() and ext4_punch_hole() we're
discarding preallocation twice. Once before we attempt to do any changes
and second time after we're done with the changes.
While the second call to ext4_discard_preallocations() in
ext4_punch_hole() case is not needed, we need to discard preallocation
right after ext4_ext_remove_space() in collapse range case because in
the case we had to restart a transaction in the middle of removing space
we might have new preallocations created.
Remove unneeded ext4_discard_preallocations() ext4_punch_hole() and move
it to the better place in ext4_collapse_range()
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
We're already calling truncate_pagecache() before we attempt to do any
actual job so there is not need to truncate pagecache once more using
truncate_setsize() after we're finished.
Remove truncate_setsize() and replace it just with i_size_write() note
that we're holding appropriate locks.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Currently in ext4_collapse_range() when calling ext4_es_remove_extent() to
remove status extents we're passing (EXT_MAX_BLOCKS - punch_start - 1)
in order to remove all extents from start of the collapse range to the
end of the file. However this is wrong because we might miss the
possible extent covering the last block of the file.
Fix it by removing the -1.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com>
Currently we're passing -1 as lend argumnet for
filemap_write_and_wait_range() which is wrong since lend is signed type
so it would cause some confusion and we might not write_and_wait for the
entire range we're expecting to write.
Fix it by using LLONG_MAX instead.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
We should be using truncate_pagecache() instead of
truncate_pagecache_range() in the collapse range because we're
truncating page cache from offset to the end of file.
truncate_pagecache() also get rid of the private COWed pages from the
range because we're going to shift the end of the file.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Where are some places where logic guaranties us that extent we are
searching exits, but this may not be true due to on-disk data
corruption. If such corruption happens we must prevent possible
null pointer dereferences.
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Unfortunately, we weren't checking to make sure of this the inode was
extent-based before attempt operate on it. Hilarity ensues.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: Namjae Jeon <namjae.jeon@samsung.com>
This commit tries to fix some byte order issues that is found by sparse
check.
$ make M=fs/ext4 C=2 CF=-D__CHECK_ENDIAN__
...
CHECK fs/ext4/extents.c
fs/ext4/extents.c:5232:41: warning: restricted __le32 degrades to integer
fs/ext4/extents.c:5236:52: warning: bad assignment (-=) to restricted __le32
fs/ext4/extents.c:5258:45: warning: bad assignment (-=) to restricted __le32
fs/ext4/extents.c:5303:28: warning: restricted __le32 degrades to integer
fs/ext4/extents.c:5318:18: warning: incorrect type in assignment (different base types)
fs/ext4/extents.c:5318:18: expected unsigned int [unsigned] [usertype] ex_start
fs/ext4/extents.c:5318:18: got restricted __le32 [usertype] ee_block
fs/ext4/extents.c:5319:24: warning: restricted __le32 degrades to integer
fs/ext4/extents.c:5334:31: warning: incorrect type in assignment (different base types)
...
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Zheng Liu <wenqing.lz@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Currently some file system have IS_SWAPFILE check in their fallocate
implementations and some do not. However we should really prevent any
fallocate operation on swapfile so move the check to vfs and remove the
redundant checks from the file systems fallocate implementations.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Currently in do_fallocate in collapse range case we're checking
whether offset + len is not bigger than i_size. However there is
nothing which would prevent i_size from changing so the check is
pointless. It should be done in the file system itself and the file
system needs to make sure that i_size is not going to change. The
i_size check for the other fallocate modes are also done in the
filesystems.
As it is now we can easily crash the kernel by having two processes
doing truncate and fallocate collapse range at the same time. This
can be reproduced on ext4 and it is theoretically possible on xfs even
though I was not able to trigger it with this simple test.
This commit removes the check from do_fallocate and adds it to the
file system.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Acked-by: Dave Chinner <david@fromorbit.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
All the checks IS_APPEND and IS_IMMUTABLE for the fallocate operation on
the inode are done in vfs. No need to do this again in ext4. Remove it.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When mounting ext4 with data=journal option, xfstest shared/002 and
shared/004 are currently failing as checksum computed for testfile
does not match with the checksum computed in other journal modes.
In case of data=journal mode, a call to filemap_write_and_wait_range
will not flush anything to disk as buffers are not marked dirty in
write_end. In collapse range this call is followed by a call to
truncate_pagecache_range. Due to this, when checksum is computed,
a portion of file is re-read from disk which replace valid data with
NULL bytes and hence the reason for the difference in checksum.
Calling ext4_force_commit before filemap_write_and_wait_range solves
the issue as it will mark the buffers dirty during commit transaction
which can be later synced by a call to filemap_write_and_wait_range.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Ashish Sangwan <a.sangwan@samsung.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Xfstests generic/311 and shared/298 fail when run on a bigalloc file
system. Kernel error messages produced during the tests report that
blocks to be freed are already on the to-be-freed list. When e2fsck
is run at the end of the tests, it typically reports bad i_blocks and
bad free blocks counts.
The bug that causes these failures is located in ext4_ext_rm_leaf().
Code at the end of the function frees a partial cluster if it's not
shared with an extent remaining in the leaf. However, if all the
extents in the leaf have been removed, the code dereferences an
invalid extent pointer (off the front of the leaf) when the check for
sharing is made. This generally has the effect of unconditionally
freeing the partial cluster, which leads to the observed failures
when the partial cluster is shared with the last extent in the next
leaf.
Fix this by attempting to free the cluster only if extents remain in
the leaf. Any remaining partial cluster will be freed if possible
when the next leaf is processed or when leaf removal is complete.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
Currently in ext4_fallocate() and ext4_zero_range() we're testing ret
variable along with new_size. However in ext4_fallocate() we just tested
ret before and in ext4_zero_range() if will always be zero when we get
there so there is no need to test it in both cases.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Introduce new FALLOC_FL_ZERO_RANGE flag for fallocate. This has the same
functionality as xfs ioctl XFS_IOC_ZERO_RANGE.
It can be used to convert a range of file to zeros preferably without
issuing data IO. Blocks should be preallocated for the regions that span
holes in the file, and the entire range is preferable converted to
unwritten extents
This can be also used to preallocate blocks past EOF in the same way as
with fallocate. Flag FALLOC_FL_KEEP_SIZE which should cause the inode
size to remain the same.
Also add appropriate tracepoints.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Move block allocation out of the ext4_fallocate into separate function
called ext4_alloc_file_blocks(). This will allow us to use the same
allocation code for other allocation operations such as zero range which
is commit in the next patch.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Currently in ext4_fallocate we would update inode size, c_time and sync
the file with every partial allocation which is entirely unnecessary. It
is true that if the crash happens in the middle of truncate we might end
up with unchanged i size, or c_time which I do not think is really a
problem - it does not mean file system corruption in any way. Note that
xfs is doing things the same way e.g. update all of the mentioned after
the allocation is done.
This commit moves all the updates after the allocation is done. In
addition we also need to change m_time as not only inode has been change
bot also data regions might have changed (unwritten extents). However
m_time will be only updated when i_size changed.
Also we do not need to be paranoid about changing the c_time only if the
actual allocation have happened, we can change it even if we try to
allocate only to find out that there are already block allocated. It's
not really a big deal and it will save us some additional complexity.
Also use ext4_debug, instead of ext4_warning in #ifdef EXT4FS_DEBUG
section.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>-
--
v3: Do not remove the code to set EXT4_INODE_EOFBLOCKS flag
fs/ext4/extents.c | 96 ++++++++++++++++++++++++-------------------------------
1 file changed, 42 insertions(+), 54 deletions(-)
Commit 9cb00419fa, which enables hole punching for bigalloc file
systems, exposed a bug introduced by commit 6ae06ff51e in an earlier
release. When run on a bigalloc file system, xfstests generic/013, 068,
075, 083, 091, 100, 112, 127, 263, 269, and 270 fail with e2fsck errors
or cause kernel error messages indicating that previously freed blocks
are being freed again.
The latter commit optimizes the selection of the starting extent in
ext4_ext_rm_leaf() when hole punching by beginning with the extent
supplied in the path argument rather than with the last extent in the
leaf node (as is still done when truncating). However, the code in
rm_leaf that initially sets partial_cluster to track cluster sharing on
extent boundaries is only guaranteed to run if rm_leaf starts with the
last node in the leaf. Consequently, partial_cluster is not correctly
initialized when hole punching, and a cluster on the boundary of a
punched region that should be retained may instead be deallocated.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
Code deallocating the extent path referenced by an argument to
ext4_ext_handle_uninitialized_extents was made redundant with identical
code in its one caller, ext4_ext_map_blocks, by commit 3779473246.
Allocating and deallocating the path in the same function also makes
the code clearer.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This patch implements fallocate's FALLOC_FL_COLLAPSE_RANGE for Ext4.
The semantics of this flag are following:
1) It collapses the range lying between offset and length by removing any data
blocks which are present in this range and than updates all the logical
offsets of extents beyond "offset + len" to nullify the hole created by
removing blocks. In short, it does not leave a hole.
2) It should be used exclusively. No other fallocate flag in combination.
3) Offset and length supplied to fallocate should be fs block size aligned
in case of xfs and ext4.
4) Collaspe range does not work beyond i_size.
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Ashish Sangwan <a.sangwan@samsung.com>
Tested-by: Dongsu Park <dongsu.park@profitbricks.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Commit 3779473246 breaks the return of error codes from
ext4_ext_handle_uninitialized_extents() in ext4_ext_map_blocks(). A
portion of the patch assigns that function's signed integer return
value to an unsigned int. Consequently, negatively valued error codes
are lost and can be treated as a bogus allocated block count.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
Commit a115f749c1 (ext4: remove wait for unwritten extent conversion from
ext4_truncate) exposed a bug in ext4_ext_handle_uninitialized_extents().
It can be triggered by xfstest generic/299 when run on a test file
system created without a journal. This test continuously fallocates and
truncates files to which random dio/aio writes are simultaneously
performed by a separate process. The test completes successfully, but
if the test filesystem is mounted with the block_validity option, a
warning message stating that a logical block has been mapped to an
illegal physical block is posted in the kernel log.
The bug occurs when an extent is being converted to the written state
by ext4_end_io_dio() and ext4_ext_handle_uninitialized_extents()
discovers a mapping for an existing uninitialized extent. Although it
sets EXT4_MAP_MAPPED in map->m_flags, it fails to set map->m_pblk to
the discovered physical block number. Because map->m_pblk is not
otherwise initialized or set by this function or its callers, its
uninitialized value is returned to ext4_map_blocks(), where it is
stored as a bogus mapping in the extent status tree.
Since map->m_pblk can accidentally contain illegal values that are
larger than the physical size of the file system, calls to
check_block_validity() in ext4_map_blocks() that are enabled if the
block_validity mount option is used can fail, resulting in the logged
warning message.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org # 3.11+
Commit f5a44db5d2 introduced a regression on filesystems created with
the bigalloc feature (cluster size > blocksize). It causes xfstests
generic/006 and /013 to fail with an unexpected JBD2 failure and
transaction abort that leaves the test file system in a read only state.
Other xfstests run on bigalloc file systems are likely to fail as well.
The cause is the accidental use of a cluster mask where a cluster
offset was needed in ext4_ext_map_blocks().
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
The missing casts can cause the high 64-bits of the physical blocks to
be lost. Set up new macros which allows us to make sure the right
thing happen, even if at some point we end up supporting larger
logical block numbers.
Thanks to the Emese Revfy and the PaX security team for reporting this
issue.
Reported-by: PaX Team <pageexec@freemail.hu>
Reported-by: Emese Revfy <re.emese@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
A corrupted ext4 may have out of order leaf extents, i.e.
extent: lblk 0--1023, len 1024, pblk 9217, flags: LEAF UNINIT
extent: lblk 1000--2047, len 1024, pblk 10241, flags: LEAF UNINIT
^^^^ overlap with previous extent
Reading such extent could hit BUG_ON() in ext4_es_cache_extent().
BUG_ON(end < lblk);
The problem is that __read_extent_tree_block() tries to cache holes as
well but assumes 'lblk' is greater than 'prev' and passes underflowed
length to ext4_es_cache_extent(). Fix it by checking for overlapping
extents in ext4_valid_extent_entries().
I hit this when fuzz testing ext4, and am able to reproduce it by
modifying the on-disk extent by hand.
Also add the check for (ee_block + len - 1) in ext4_valid_extent() to
make sure the value is not overflow.
Ran xfstests on patched ext4 and no regression.
Cc: Lukáš Czerner <lczerner@redhat.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
Commit ec22ba8e ("ext4: disable merging of uninitialized extents")
ensured that if either extent under consideration is uninit, we
decline to merge, and ext4_can_extents_be_merged() returns false.
So there is no need for the caller to then test whether the
extent under consideration is unitialized; if it were, we
wouldn't have gotten that far.
The comments were also inaccurate; ext4_can_extents_be_merged()
no longer XORs the states, it fails if *either* is uninit.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>
Commit ec22ba8e ("ext4: disable merging of uninitialized extents")
ensured that if either extent under consideration is uninit, we
decline to merge, and immediately return.
But right after that test, we test again for an uninit
extent; we can never hit this. So just remove the impossible
test and associated variable.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>
After applied the commit (4a092d73), we have reduced the number of
source files that need to #include ext4_extents.h. But we can do
better.
This commit defines ext4_zeroout_es() in extents.c and move
EXT_MAX_BLOCKS into ext4.h in order not to include ext4_extents.h in
indirect.c and ioctl.c. Meanwhile we just need to include this file in
extent_status.c when ES_AGGRESSIVE_TEST is defined. Otherwise, this
commit removes a duplicated declaration in trace/events/ext4.h.
After applied this patch, we just need to include ext4_extents.h file
in {super,migrate,move_extents,extents}.c, and it is easy for us to
define a new extent disk layout.
Signed-off-by: Zheng Liu <wenqing.lz@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If ext_debugging is enabled and path[depth].p_ext is NULL, len
and lblock are printed non initialized
Signed-off-by: Andi Shyti <andi@etezian.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
reaim workfile.dbase test easily triggers warning in
ext4_da_update_reserve_space():
EXT4-fs warning (device ram0): ext4_da_update_reserve_space:365:
ino 12, allocated 1 with only 0 reserved metadata blocks (releasing 1
blocks with reserved 9 data blocks)
The problem is that (one of) tests creates file and then randomly writes
to it with O_SYNC. That results in writing back pages of the file in
random order so we create extents for written blocks say 0, 2, 4, 6, 8
- this last allocation also allocates new block for extents. Then we
writeout block 1 so we have extents 0-2, 4, 6, 8 and we release
indirect extent block because extents fit in the inode again. Then we
writeout block 10 and we need to allocate indirect extent block again
which triggers the warning because we don't have the reservation
anymore.
Fix the problem by giving back freed metadata blocks resulting from
extent merging into inode's reservation pool.
Signed-off-by: Jan Kara <jack@suse.cz>
Add a new fiemap flag which forces the all of the extents in an inode
to be cached in the extent_status tree. This is critically important
when using AIO to a preallocated file, since if we need to read in
blocks from the extent tree, the io_submit(2) system call becomes
synchronous, and the AIO is no longer "A", which is bad.
In addition, for most files which have an external leaf tree block,
the cost of caching the information in the extent status tree will be
less than caching the entire 4k block in the buffer cache. So it is
generally a win to keep the extent information cached.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When we read in an extent tree leaf block from disk, arrange to have
all of its entries cached. In nearly all cases the in-memory
representation will be more compact than the on-disk representation in
the buffer cache, and it allows us to get the information without
having to traverse the extent tree for successive extents.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>
When we find an invalid extent tree block, report the block number of
the bad block for debugging purposes.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>
Refactor out the code needed to read the extent tree block into a
single read_extent_tree_block() function. In addition to simplifying
the code, it also makes sure that we call the ext4_ext_load_extent
tracepoint whenever we need to read an extent tree block from disk.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>
If there are no items in the extent status tree, ext4_es_lru_add() is
a no-op. So it is not sufficient to call ext4_es_lru_add() before we
try to lookup an entry in the extent status tree. We also need to
call it at the end of ext4_ext_map_blocks(), after items have been
added to the extent status tree.
This could lead to inodes with that have extent status trees but which
are not in the LRU list, which means they won't get considered for
eviction by the es_shrinker.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: Zheng Liu <wenqing.lz@taobao.com>
Cc: stable@vger.kernel.org
During large unlink operations on files with extents, we can use a lot
of CPU time. This adds a cond_resched() call when starting to examine
the next level of a multi-level extent tree. Multi-level extent trees
are rare in the first place, and this should rarely be executed.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
In ext4_ext_map_blocks(), if we have successfully allocated the data
blocks, but then run into trouble inserting the extent into the extent
tree, most likely due to an ENOSPC condition, determine the arguments
to ext4_free_blocks() in a simpler way which is easier to prove to be
correct.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Previously ext4_ext_truncate() was ignoring potential error returns
from ext4_es_remove_extent() and ext4_ext_remove_space(). This can
lead to the on-diks extent tree and the extent status tree cache
getting out of sync, which is particuarlly bad, and can lead to file
system corruption and potential data loss.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
Both hole punch and truncate use ext4_ext_rm_leaf() for removing
blocks. Currently we choose the last extent as the starting
point for removing blocks:
ex = EXT_LAST_EXTENT(eh);
This is OK for truncate but for hole punch we can optimize the extent
selection as the path is already initialized. We could use this
information to select proper starting extent. The code change in this
patch will not affect truncate as for truncate path[depth].p_ext will
always be NULL.
Signed-off-by: Ashish Sangwan <a.sangwan@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Translate the bitfields used in various flags argument to strings to
make the tracepoint output more human-readable.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
No need to pass file pointer when we can directly pass inode pointer.
Signed-off-by: Ashish Sangwan <a.sangwan@samsung.com>
Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
