Without this change, we will write data past the end of the
mmp buf. Valgrind catches this:
==6373== Syscall param write(buf) points to unaddressable byte(s)
==6373== at 0x362260E470: __write_nocancel (in /lib64/libpthread-2.12.2.so)
==6373== by 0x41CF83: raw_write_blk (unix_io.c:255)
==6373== by 0x41D2BC: unix_write_blk64 (unix_io.c:757)
==6373== by 0x41A05D: ext2fs_mmp_write (mmp.c:130)
==6373== by 0x40B0C9: do_set_mmp_value (set_fields.c:806)
==6373== by 0x421B61: really_execute_command (execute_cmd.c:108)
==6373== by 0x421C54: ss_execute_line (execute_cmd.c:234)
==6373== by 0x403743: main (debugfs.c:2339)
==6373== Address 0x63f000 is not stack'd, malloc'd or (recently) free'd
and in my testing it led to silent failures while writing the mmp
block in debugfs:
write(3, "xV4\22PMM\342\325V\274N\0\0\0\0host.name."..., 4096) = -1 EFAULT (Bad address)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Commit 6b56f3d92d introduced the use of HAVE_STAT64 without arranging
that it be defined in configure.in. Previously ext4.h used
HAVE_OPEN64, but apparently there are (broken) platforms that have
open64() but not stat64(). Go figure.
We do need to consistently use a single test for ext2fs_stat(),
ext2fs_fstat(), and struct ext2fs_struct_stat, or we could end up
passing a struct stat64 to a fstat() system call, or some such. I've
elected to use HAVE_FSTAT64 because: (a) it's already defined in the
configure script, and (b) if we ever come across a really broken
platform that defines fstat64() but not stat64(), we can always
emulate stat64() using open64() followed by a fstat64().
This commit fixed a bug whose symptoms were that mke2fs would not work
if given a file > 2GB on 32-bit platforms.
Addresses-Debian-Bug: #647245
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Without this change, we go back to getting group descriptor
"0" each time we go around the "for i" loop. It must properly
advance through the filesystem.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The ext2fs_file_acl_block() and ext2fs_set_file_acl_block() needs to
only check i_file_acl_high if the 64-bit flag is set. This is needed
because otherwise we will run into problems on Hurd systems which
actually use that field for h_i_mode_high.
This involves an ABI change since we need to pass ext2_filsys to these
functions. Fortunately these functions were first included in the
1.42-WIP series, so it's OK for us to change them now. (This is why
we have 1.42-WIP releases. :-)
Addresses-Sourceforge-Bug: #3379227
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Remove unused variables, places where 'return' was used with no value
in a non-void function, missing function declarations, etc. Don't
assume that all systems have quotactl(), and use <sys/quota.h> if it
exists to define the quotactl interfaces.
One of the unused variables also got rid of a non-portable use of
PATH_MAX.
Cc: Aditya Kali <adityakali@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If the number of block groups exceeds 2**32, a bad cast would lead to
a bogus "Not enough space to build proposed filesystem while setting
up superblock" failure.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Surely we should be setting s_clusters_per_group, not
s_blocks_per_group, to EXT2_MAX_CLUSTERS_PER_GROUP here.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The byte swap functions which are defined in ext2fs.h are only needed
by crc32.c, and not by gen_crc32ctable.c. The gen_crc32ctable program
needs to be compiled on the host OS, where ext2fs.h may not be
present. So move the use of the header function to crc32c.c, to avoid
compilation problems.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
ext2fs.h now calls open() so it should include the headers needed
for this system call as well.
Addresses-Red-Hat-Bugzilla: #742147
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The MMP code in libext2fs tries to gate MMP block swab'ing with this
test:
if (fs->super->s_magic == ext2fs_swab16(EXT2_SUPER_MAGIC))
However, EXT2FS_ENABLE_SWAPFS never seems to be defined anywhere (all
possible existed, the field fs->super->s_magic is always in host
byteorder, so the test always fails. So, we can change the #ifdef to
WORDS_BIGENDIAN (which is conditionally defined on BE platforms) and
get rid of the broken if test.
Signed-off-by: Darrick J. Wong <djwong@us.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Fix several minor errors in structure definitions, the byteswap code,
and Makefiles that result from merging the crc32c and initial parts of
the metadata checksumming patchset.
Signed-off-by: Darrick J. Wong <djwong@us.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The functions htole32(), le32toh(), be32toh(), htobe32() aren't
defined in all environments. Use the ext2fs byte swap functions for
portability.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Multi-mount protection is feature that allows mke2fs, e2fsck, and
others to detect if the filesystem is mounted on a remote node (on
SAN disks) and avoid corrupting the filesystem. For e2fsprogs this
means that it checks the MMP block to see if the filesystem is in use,
and marks the filesystem busy while e2fsck is running on the system.
This is useful on SAN disks that are shared between high-availability
servers, or accessible by multiple nodes that aren't in HA pairs. MMP
isn't intended to serve as a primary HA exclusion mechanism, but as a
failsafe to protect against user, software, or hardware errors.
There is no requirement that e2fsck updates the MMP block at regular
intervals, but e2fsck does this occasionally to provide useful
information to the sysadmin in case of a detected conflict.
For the kernel (since Linux 3.0) MMP adds a "heartbeat" mechanism to
periodically write to disk (every few seconds by default) to notify
other nodes that the filesystem is still in use and unsafe to modify.
Originally-by: Kalpak Shah <kalpak@clusterfs.com>
Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Several compiler errors are quieted:
- zero-length gnu_printf format string
- unused variable
- uninitalized variable (though it isn't actually used for anything)
- fixed a bug in ext2fs_stat() if stat64() does not exist
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
This adds new APIs: ext2fs_flush2 and ext2fs_close2 which take an
extra 'int flags' parameter.
This allows us to pass in an EXT2_FLAG_FLUSH_NO_SYNC flag which avoids
fsync'ing the filesystem when closing it. For the case we have in
mind where we are just constructing a throwaway ext2 filesystem in a
file in order to boot a VM, this saves over 5 seconds during the boot
process and avoids many unnecessary disk writes.
Existing code using ext2fs_flush and ext2fs_close remains unaffected
by this change.
Signed-off-by: Richard W.M. Jones <rjones@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The DEFS line in MCONFIG had gotten so long that it exceeded 4k, and
this was starting to cause some tools heartburn. It also made "make
V=1" almost useless, since trying to following the individual commands
run by make was lost in the noise of all of the defines.
So fix this by putting the configure-generated defines in lib/config.h
and the directory pathnames to lib/dirpaths.h.
In addition, clean up some vestigal defines in configure.in and in the
Makefiles to further shorten the cc command lines.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
block_buf and/or inode_buf may not be properly freed on an error
return.
Create a new errout: target to free them as needed in error conditions.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
In inode_open(), if the allocation of &io fails, we go to cleanup
and dereference io to test io->name, which is a bug.
Similarly in undo_open() if allocation of &data fails, we
go to cleanup and dereference data to test data->real.
In the test_open() case we explicitly set retval to the only
possible error return from ext2fs_get_mem(), so remove that
for tidiness.
The other changes just make make earlier returns go through
the error goto for consistency.
In many cases we returned directly from the first error, but
"goto cleanup" etc for every subsequent error. In some
cases this leads to "impossible" tests such as:
if (ptr)
ext2fs_free_mem(&ptr)
on paths where ptr cannot be null because we would have
returned directly earlier, and Coverity flags this.
This isn't really indicative of an error in most cases, but
I think it can be clearer to always exit through the error goto
if it's used later in the function.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Using the /* fallthrough */ comment lets Coverity (and humans)
know that we really do want to fall through in these case statements.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
In the !undo_io_backing_manager case, undo_err_handler_init
will be passed a null data->real, which will be dereferenced.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If !WORDS_BIGENDIAN, it is pointless to test whether buf
is NULL, because it is initialized to NULL and never changed.
This makes Coverity complain, so we can just move all handling
of "buf" under the #ifdef.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
EXT2_LIB_SOFTSUPP_INCOMPAT_* are supposed to be bitmasks
of features which can be opened even though they are
under development. The intent is that these are masked
out of the features list, so that they will be ignored
on open.
However, the code does a logical not vs. a bitwise not:
features &= !EXT2_LIB_SOFTSUPP_INCOMPAT;
which will not have the desired effect...
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Also cleaned up ext2_fs.h, and improved the byte swapping code so the
extra fields in the large inode are properly byte swapped.
Addresses-Debian-Bug: #641838
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The set_fields commands (set_super_value, set_inode_field,
set_block_group) now handle fields which store in split fields on
ext4's on-disk format. For example, the superblock fields
s_blocks_count and s_blocks_count_hi.
The user can either set the low or high part of the field via
"blocks_count_lo" or "blocks_count_hi", or both parts can be set via
"blocks_count".
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reserve EXT4_FEATURE_RO_COMPAT_METADATA_CSUM and
EXT2_FEATURE_COMPAT_EXCLUDE_BITMAP. Also reserve fields in the
superblock and the inode for the checksums. In the block group
descriptor, reserve the exclude bitmap field for the snapshot feature,
and checksums for the inode and block allocation bitmaps.
With this commit, the metadata checksum and exclude bitmap features
should have reserved all of the fields they need in ext4's on-disk
format.
This commit also fixes an a missing byte swap for s_overhead_blocks.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: Darrick J. Wong <djwong@us.ibm.com>
Cc: Amir Goldstein <amir73il@gmail.com>
ext2fs_group_last_block2() already properly calculates
the last block in the last group, so there is no need
to special-case this after the call.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Code to count the number of blocks in the last partial
group is cut and pasted around the e2fsprogs codebase
a few times.
Making this a helper function should improve matters.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If e2fsprogs tools (mke2fs, e2fsck) is run on regular file instead of
on block device, we can use punch hole instead of regular discard
command which would not work on regular file anyway. This gives us
several advantages. First of all when e2fsck is run with '-E discard'
parameter it will punch out all ununsed space from the image, hence
trimming down the file system image. And secondly, when creating an
file system on regular file (with '-E discard' which is default), we
can use punch hole to clear the file content, hence we can skip inode
table initialization, because reads from sparse area returns zeros. This
will result in faster file system creation (without the need to specify
lazy_itable_init) and smaller images.
This commit also fixes some tests that would fail due to mke2fs showing
discard progress, hence the output would differ.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
In many places we are using #ifdef HAVE_OPEN64 to determine if we can
use open64() but that's ugly. This commit creates two new helpers
ext2fs_open_file() for open() and ext2fs_stat() for stat(). Also we need
new typedef ext2fs_struct_stat for struct stat.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Add a slicing-by-8 CRC32c implementation for metadata checksumming.
Adapted from Bob Pearson's kernel patch.
Also added a self-test mechanism so we can verify that the crc32c
implementation is working correctly.
Signed-off-by: Darrick J. Wong <djwong@us.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The condition ((start+num) & ~0xffffffffULL) in bitmap_range2
and generic_bmap_range funcs in get_bitmap64.c was wrong and
inconsistent with the condition (start+num-1 > bmap->real_end)
in generic_bitmap_range funcs in get_bitmap.c.
I got the following error from tune2fs on a 16TB fs:
Illegal block number passed to ext2fs_unmark_block_bitmap #4294967295
for block bitmap for 16TB.img
tune2fs: Invalid argument while reading bitmaps
Fix to condition to ((start+num-1) & ~0xffffffffULL), because
the bit (start+num) is not going to be changed by the funcs.
Signed-off-by: Amir Goldstein <amir73il@users.sf.net>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The dump program relies on fs->frag_size and the
EXT2_FRAGS_PER_BLOCK() macro. Kind of silly for it to do so, but it's
part of the kludgy way the dump program (which was originally written
for the BSD FFS was ported over to support ext2/3.) Given how it
makes assumptions about the ext2/3/4 file system being similar to the
BSD FFS, it's a bit of a miracle it works for ext4 --- or at least
appears to work...
Addresses-Debian-Bug: #636418
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This is needed to support online resizing for > 32-bit file systems
Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This patch adds support for doing quota accounting during full
e2fsck scan if the 'quota' feature was set on the superblock.
If user-visible quota inodes are in use, they will be hidden
and converted to the reserved quota inodes.
Signed-off-by: Aditya Kali <adityakali@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Remove the interpolation because there is a bug in icount which can
cause a core dump if calculated range gets turned into a NaN and then
do an out-of-bounds array access. We could fix this with some more
tests, but the complexity is such that nuking all of the interpolation
code will be faster than fixing the interpolation.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The f_lotsbad regression test was failing on some systems
with:
Restarting e2fsck from the beginning...
Pass 1: Checking inodes, blocks, and sizes
+Illegal block number passed to ext2fs_test_block_bitmap #0 for in-use block map
Pass 2: Checking directory structure
Entry 'termcap' in / (2) has deleted/unused inode 12. Clear? yes
Running with valgrind (./test_script --valgrind f_lotsbad) we
see:
+==31409== Conditional jump or move depends on uninitialised value(s)
+==31409== at 0x42927A: ext2fs_test_generic_bmap (gen_bitmap64.c:378)
among others.
Looking at gen_bitmap64.c:
376: arg >>= bitmap->cluster_bits;
377:
378: if ((arg < bitmap->start) || (arg > bitmap->end)) {
A little more debugging showed that it was actually
bitmap->cluster_bits which was uninitialized, because it never
gets copied over in ext2fs_copy_generic_bmap()
Patch below resolves the issue.
Reported-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If the callback interator modifies a block in the middle of an extent
during a call to the block iterator, causing the extent to be split,
ext2_block_iterate3() will end up calling the callback function twice
for some number of blocks. Fix this.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Use the EXT2_I_SIZE() macro consistently to access the inode size.
The i_size/i_size_high combination is open coded in several places.
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Eric Sandeen