This commit adds support for converting QCOW2 image created previously
with e2image into raw image. The QCOW2 image is detected automatically,
so there is not new option. Just use following command:
e2image -r image.qcow image.raw
No that this tool is aimed to quickly convert qcow2 image created with
e2image into raw image. In order to improve speed we are doing some
assumption I believe might not be true for regular qcow2 images. So it
was not tested with regular QCOW2 images and it might not work with
them. The intention of this tool is only convert images previously
created by e2image.
Note that there is nothing special with QCOW2 images created by e2images
and it can be used with tools like qemu-img, or qemu-nbd without any
problems.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
This commit adds support for exporting filesystem into QCOW2 image
format. Like sparse format this saves space, by writing only necessary
(metadata blocks) into image. Unlike sparse image, QCOW2 image is NOT
sparse, hence does not change its size by copying with not-sparse-aware
tools.
New options '-Q' has been added to tell the e2image to use QCOW2 as an
output image format. QCOW2 supports encryption and compression, however
e2image so far does no support such features, however you can still
scramble filenames with '-s' option.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
This patch adds support for specifying 'reserved_ratio' (percent blocks
reserved for super user, same as '-m' command line option) in mke2fs.conf.
It adds profile_get_double function in profile.c that allows reading
floating point values from profile files.
Signed-off-by: Aditya Kali <adityakali@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
/boot/a: 0 extents found
works properly, but
Filesystem type is: ef53
Filesystem cylinder groups is approximately 61
File size of a is 0 (0 blocks, blocksize 1024)
ext logical physical expected length flags
a: 1 extent found
yields 1 extent when it should be 0.
Fix this up by special-casing no extents returned in verbose
mode; skip printing the header for the columns too, since there
are no columns to print.
Also, in nonverbose mode we can set fm_extent_count to 0
so that FIEMAP will just query the extent count without gathering
details; clarify this with a comment.
Addresses-RedHat-Bugzilla: 653234
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Fix a few typos in manpages.
Reported-by: Branislav Náter <bnater@redhat.com>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Check to see if the device supports discard before starting the
progress bar, and then printing an error about inappropriate ioctl for
device (when creating a file system image to a file, for example).
Also, add a function signature in the ext2_io.h header file for
io_channel_discard() and fix an extra, uneeded argument in mke2fs's
call to that function.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This adds the superblock fields needed so that dumpe2fs works and the
code points and renames the superblock fields from describing
fragments to clusters.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
For some time now we are doing initial discard of the device prior to
filesystem creation. However, there is no feedback for the user and
hence on some devices with slow TRIM implementation it may appear that
mke2fs is stuck.
This commit introduce new function mke2fs_discard_device(), which is a
wrapper for io_channel_discard(). The discard is done in chunks of
2GB, which seems reasonably well for both slow and fast devices, and
discard progress is reported back to the user.
I gave up on doing fancy things like align discard according to
discard_alignment, checking for discard granularity and computing
estimate time. First of all, because it would require either new ioctl
to retrieve those information or use of libudev library, none of it
seems to be worth it. Regarding discard_granularity, I doubt there is
any sane device with discard granularity that big it would affect this.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
It is not true that 'nodiscard' is set as default, so remove this
sentence. The default is 'discard' and it is properly documented in man
page.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
User namespace xattrs are generally useful, and I think extN
is the only filesystem requiring a special mount option to
enable them, when xattrs are otherwise available. So this
change sets that mount option into the defaults, via a
mke2fs.conf option.
Note that if xattrs are config'd off, this will lead to a
mostly-harmless:
EXT4-fs (sdc1): (no)user_xattr options not supported
message at mount time...
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The forced fsck often comes at unexpected and inopportune moments,
and even enterprise customers are often caught by surprise when
this happens. Because a filesystem with an error condition will
be marked as requiring fsck anyway, I submit that the time-based
and mount-based checks are not particularly useful, and that
administrators can schedule fscks on their own time, or tune2fs
the enforced intervals if they so choose. This patch disables the
intervals by default, and I've added a new mkfs.conf option to
turn on the old behavior of random, unexpected, time-consuming
fscks at boot time. ;)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
When using the -v option, report a breakdown of the number of read,
write, and comparison errors that were found by badblocks.
Thanks to Ragnar Kjørstad for providing this patch.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If there was a bad block for block #0, badblocks would never switch
back testing blocks more efficiently. In addition, we were
double-incrementing the blocks to be tested in the read/write test due
to failure to remove code.
Thanks to Ragnar Kjørstad for pointing these problems out.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
With Direct I/O, the kernel can report 0 bytes read even though the
first block has no errors. So there are any errors, we need try to
read/write blocks one at a time and to get an accurate report.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Currently, e4defrag always does byte-swapping when it gets superblock
information, so the calculation of the best extents count is not
correct on little endian machine. This doesn't cause data corruption,
but it may confuse users by showing the wrong extent count. To solve
this problem, we use ext2fs_open() instead of get_superblock_info()
that is the original function.
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If a file gets deleted or truncated while e4defrag is trying to
operate on it, it's possible for it seg fault.
Addresses-Red-Hat-Bugzilla: #641926
Reported-by: Michal Piotrowski <mkkp4x4@gmail.com>
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If the file system size was not specified on the command line, we were
always using the usage type "floppy" since we didn't determine the
device size until after calling parse_fs_types(). Doh!
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Avoid memory leaks on error paths, and make sure we issue the correct
error messages in the case of (highly) unlikely errors.
Original patch submitted by Namhyung Kim <namhyung@gmail.com>, but
highly rewritten since then.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Check return value of some functions and exit if unhandled error occurred.
Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
In original code, 'huge' type could not be selected because it
always be caught for 'big' type. Change the ordering.
Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The commit 493024ea1d ("mke2fs: Fix up the
fs type and feature selection for 64-bit blocks") added 'big' and 'huge'
usage-type but was missing description in man page. Add it.
Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
... in the very unlikely case that e2p_os2string fails to allocate
memory.
Reported-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The fallocate() interface on 32-bit machines is defined to use off_t,
not loff_t (even though the system call interface is 64-bit clean).
This causes e4defrag to fail on files greater than 2GB. Fix this by
trying to use fallocate64(), and using the hard-coded syscall if it
does not exist.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Add the description of the size per one extent and the maximum extent size
in ext4 into e4defrag man page.
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If non-privileged user runs e4defrag, e4defrag returns an exit status
of 1 despite its success. This patch fixes this problem.
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
e4defrag uses st_blocks (struct stat) to calculate file blocks. However,
st_blocks also has meta data blocks in addition to file blocks. So, we
calculate file blocks by sum of the extent length.
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
e4defrag with -c option outputs "ratio" that means the levels of
fragmentation. However, it's difficult for users to understand, so we will
use size per extent instead of ratio.
Before:
# e4defrag -c /mnt/mp1/file
<File> now/best ratio
/mnt/mp1/file 6/1 0.00%
Total/best extents 6/1
Fragmentation ratio 0.00%
Fragmentation score 0.04
[0-30 no problem: 31-55 a little bit fragmented: 55- needs defrag]
This file(/mnt/mp1/file) does not need defragmentation.
Done.
After:
# e4defrag -c /mnt/mp1/file
<File> now/best size/ext
/mnt/mp1/file 6/1 16666 KB
Total/best extents 6/1
Average size per extent 16666 KB
Fragmentation score 0
[0-30 no problem: 31-55 a little bit fragmented: 56- needs defrag]
This file (/mnt/mp1/file) does not need defragmentation.
Done.
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Currently e4defrag relies on the EXT4_IOC_MOVE_EXT ioctl to perform online
defragmentation. However, this iotcl kernel patch is not available before
2.6.30-rc1. e4defrag shall fail without obvious reasons on systems running
older kernels. The patch adds more detailed error message addressing this
issue and prompts users with the minimal kernel version that is needed to
run e4defrag.
Signed-off-by: Peng Tao <bergwolf@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Akira Fujita merged a patch into 2.6.33 that adds a requirement that a
file being defragged must be opened with read and write access, so
e2fsprogs needs to satisfy that.
Signed-off-by: Darrick J. Wong <djwong@us.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When running dumpe2fs on a filesystem formatted with flex_bg, it
prints out the relative offsets for the bitmaps and inode table
badly on 64-bit systems, because the offset is computed as a
large positive number instead of being a negative numer (which
will not be printed at all):
Group 1: (Blocks 0x8000-0xffff) [INODE_UNINIT, ITABLE_ZEROED]
Block bitmap at 0x0102 (+4294934786), Inode bitmap at 0x0202 (+4294935042)
Inode table at 0x037e-0x03fa (+4294935422)
This commit prints out the relative offsets for flex_bg
groups as the offset within the reported group. This makes it
more clear where the metadata is located, rather than simply
printing some large negative number.
Group 1: (Blocks 0x8000-0xffff) [INODE_UNINIT, ITABLE_ZEROED]
Block bitmap at 0x0102 (bg #0 +258), Inode bitmap at 0x0202 (bg #0 +514)
Inode table at 0x037e-0x03fa (bg #0 +894)
Signed-off-by: Andreas Dilger <adilger@dilger.ca>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If the user passes a file system type which is not defined in
mke2fs.conf (i.e., mke2fs -t xfs ...) change mke2fs so that it prints
a warning and aborts the run. (There is an exception for ext2, since
that file system does not need a special definition in the fs_types
section of the /etc/mke2fs.conf file.)
In addition, print a warning if there are usage types (specified using
the -T option) which are not defined in /etc/mke2fs.conf.
Addresses-Debian-Bug: #594609
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
There is generic discard function in struct_io_manager, or in
unix_io_manager to be specific. So use this instead of
mke2fs_discard_blocks().
Since mke2fs_discard_blocks() is not used anymore (and should not be)
remove it.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
When the device have discard support and simultaneously discard zeroes
data (and it is properly advertised), then we can take advantage of such
behavior in several e2fsprogs tools.
Add new flag CHANNEL_FLAGS_DISCARD_ZEROES for struct_io_channel so
each io_manager can take advantage of this. The flag is properly set
according to BLKDISCARDZEROES ioctl in unix_open.
Also remove old mke2fs_discard_zeroes_data() function and substitute it
with helper which test this flag.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Allow to specify discard in mke2fs.conf. Also change the way how to
specify default value for lazy_itable_init. It is better to have all
this defaulting done in the same place so do it in definition (as we do
with discard).
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
It would be nice to have consistent "discard" options in every system
tool (mount, fsck, mkfs) taking advantage of discards. Also "discard"
and "nodiscard" is more descriptive instead of just "-K" and can be
easily defaulted and it is something we can not do with "-K".
With this commit you need to specify extended option like this:
./mke2fs -T <fstype> -E nodiscard <device>
in order make a filesystem without discarding the device first. And
./mke2fs -T <fstype> -E discard <device>
respectively.
-K option is with this commit deprecated and should not be used anymore.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If MKE2FS_DEVICE_SECTSIZE is set, then this will override the logical
sector size, which is the smallest sector size that can be written
atomically by the device. (Previously MKE2FS_DEVICE_SECTSIZE set the
physical sector size, which was incorrect given its historical usage.)
The environment variable MKE2FS_DEVICE_PHYS_SECTSIZE will set the
physical sector size, which is the actual sector size used by the
device in reality.
The logical sector size is always less than or equal to the physical
sector size; and writes smaller than the physical sector size but
greather than or equal to the logical sector size will cause a
read-modify-write cycle within the device firmware (or in some
abstract layer lower than the Linux block I/O subsystem, at any rate).
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If the device does not have an explicitly specified minimum io_size or
optimal io_size, and the physical sector size is greater than the
block size, then use the physical sector size as a better-than-nothing
hint.
This should help for SSD's that have a physical sector size of 8k or
16k (which are reportedly will be coming soon).
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
There will be SSD's out soon that have 8k or 16k phyiscal block sizes.
So don't enforce a requirement that the block size be less than the
physical block size unless the force option is given, and don't give a
warning if the user can't do anything about it (i.e., if the physical
block size is > than the page size).
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Add check for /sys/fs/ext4/features/lazy_itable_init. If this file
exists, it should be OK to skip initializing the inode table since the
kernel will do it at mount time.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Mistakes on the commandline can lead to odd error messages:
# mke2fs -t ext4 -E stride=128 stripe-width=512 /dev/sda1
mke2fs: invalid blocks count - /dev/sda1
Making it a bit more explicit is more obvious:
mke2fs: invalid blocks count '/dev/sda1' on device 'stripe-width=512'
(hint, the mistake was no comma separation for -E)
Reported-by: Adam Huffman <bloch@verdurin.com>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If a device supports discard -and- returns 0s for discarded blocks,
then we can skip the inode table initialization -and- the inode table
zeroing at mkfs time, and skip the lazy init as well since they are
already zeroed out.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
This bit of the mke2fs manpage is slightly confusing:
-b block-size
Specify the size of blocks in bytes. <snip>
If block-size is negative, then mke2fs will use heuristics
to determine the appropriate block size, with the constraint
that the block size will be at least block-size bytes.
because it sounds like the block size will be at least a negative
number. Clarify just what the negative sign means.
Reported-by: Chris Frost <chris@frostnet.net>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The getopt() function returns an int, not a char. On systems where the
default char is unsigned (like ppc), we get weird behavior where -1 is
truncated to 0xff but compared to (int)-1.
Also fix this same bug for two test programs, test_rel and iscan,
which aren't currently used at the moment.
Addresses-Gentoo-Bug: #299386
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
We need to defer setting the blocks count field in the fs_param
structure until it is known whether 64-bit feature will be enabled
(and whether the blocks count is valid).
We also add a new mke2fs.conf configuration parameter,
auto_64-bit_support which will automatically enable the 64-bit feature
if the number of blocks requires it.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Use 64-bit interfaces in mke2fs. This should be most most of whats
needed to support creating a 64-bit filesystem.
Signed-off-by: Jose R. Santos <jrs@us.ibm.com>
Signed-off-by: Valerie Aurora Henson <vaurora@redhat.com>
Signed-off-by: Nick Dokos <nicholas.dokos@hp.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Allow the uninit_bg feature to be set without requiring an fsck. The
first full fsck will require scanning all of the inode table blocks,
but subsequent fsck's will be fast. This allows flexibility over
requiring a full fsck after setting this feature, which is what
tune2fs previously mandated.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Some devices, notably 4k sector drives, may have a 512 logical
sector size, mapped onto a 4k physical sector size.
When mke2fs is ratcheting down the blocksize for small filesystems,
or when a blocksize is specified on the commandline, we should not
willingly go below the physical sector size of the device.
When a blocksize is specified, we -must- not go below
the logical sector size of the device.
Add a new library function, ext2fs_get_device_phys_sectsize()
to get the physical sector size if possible, and adjust the
logic in mke2fs to enforce the above rules.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
These options allow e2fsprogs to be built using symlinks instead of
hard links, and to be installed using symlinks instead of hard links,
respectively.
Addresses-Sourceforge-Bug: #1436294
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This for RH bug #572935 -
RFE: Misleading error message from mke2fs -J option
If the journal device UUID is typo'd or otherwise not found,
the error message looks like it's a usage() type of problem.
It'd be helpful to explicitly say that the device requested
could not be found.
Addresses-Red-Hat-Bug: #572935
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Just print the warning message in this case.
Addresses-Red-Hat-Bug: #569021
Addresses-Launchpad-Bug: #530071
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
mkfsing a plain file would lead to a warning about being unable
to determine geometry; we should just skip the topology-getting
if we see that we have a regular file.
This was breaking "make check" but I had missed it since I
inadvertently stopped running the checks during the Fedora
RPM build.
Also, add a newline to the warning.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
On 32-bit platforms where the file system block size is 8k or greater,
the calculation bpib*bpib*bpib* will overflow a 32-bit calculation,
leading to a divide by zero error. Fix this.
Thanks to Mikulas Patocka for pointing this out.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Sorry about that, the discard ioctl doesn't actually work
unless you open the file with write capabilities...
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Try calling the BLKDISCARD ioctl at mkfs time to pre-discard all blocks
on an ssd, or a thinly-provisioned storage device.
No real error checking; if it fails, it fails, and that's ok - it's
just an optimization. Also, it cannot work in conjunction with
the undo io manager, for obvious reasons.
Optionally disabled with a "-K" (mnemonic: Keep) option.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Try calling the BLKDISCARD ioctl at mkfs time to pre-discard all blocks
on an ssd, or a thinly-provisioned storage device.
No real error checking; if it fails, it fails, and that's ok - it's
just an optimization. Also, it cannot work in conjunction with
the undo io manager, for obvious reasons.
Optionally disabled with a "-K" (mnemonic: Keep) option.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The ext2fs_bg_flag* functions were confusing.
Currently we have this:
void ext2fs_bg_flags_set(ext2_filsys fs, dgrp_t group, __u16 bg_flags);
void ext2fs_bg_flags_clear(ext2_filsys fs, dgrp_t group,__u16 bg_flags);
(_set (unused) sets exactly bg_flags; _clear clears all and ignores bg_flags)
and these, which can twiddle individual bits in bg_flags:
void ext2fs_bg_flag_set(ext2_filsys fs, dgrp_t group, __u16 bg_flag);
void ext2fs_bg_flag_clear(ext2_filsys fs, dgrp_t group, __u16 bg_flag);
A better interface, after the patch below, is just:
ext2fs_bg_flags_zap(fs, group) /* zeros bg_flags */
ext2fs_bg_flags_set(fs, group, flags) /* adds flags to bg_flags */
ext2fs_bg_flags_clear(fs, group, flags) /* clears flags in bg_flags */
and remove the original ext2fs_bg_flags_set / ext2fs_bg_flags_clear.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Handle automatic selection of stride/stripe:
mke2fs 1.41.9 (22-Aug-2009)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=16 blocks, Stripe width=32 blocks
...
And warn on block device misalignment:
mke2fs 1.41.9 (22-Aug-2009)
/dev/sdc1 alignment is offset by 32256 bytes.
This may result in very poor performance, (re)-partitioning suggested.
Proceed anyway? (y,n)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
This commit forces the use of the system-provided blkid or uuid header
files if we are using the system-provided blkid or uuid libraries.
This avoids using the in-tree header files with the system libraries.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
fsck leaks fds when invoked with -R -A -M -a -t noopts=nofail
Signed-off-by: Cristian Rodríguez <crrodriguez@opensuse.org>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Add a more explicit description of how specifying the flex_bg file
system feature changes the layout of the per-block group metadata.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
With 64-bit file systems, mke2fs can take a long time to do things
other than write inode tables. I exported the mke2fs numeric progress
meter and used it for allocating group tables and the final file
system flush.
Signed-off-by: Valerie Aurora (Henson) <vaurora@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
ppc glibc seems to be missing sync_file_range, so we fell back
to the local define, and there ppc differs as well, so the
build was failing.
Thanks to Kyle for the patch w/ the tidy solution.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The feature name "extent" is documented in mke2fs.conf, although both
"extent" and "extents" are accepted by e2fsprogs.
Addreses-Debian-Bug: #540111
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When increasing inode size if we find that the new block
that we needed to increase the inode table size is a bad
block we fail. This make sure we don't end up with a corrupt
file system when doing inode resize on a file system having
bad blocks.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
With file system formated for RAID arrays we can have inode bitmap
and block bitmap after inode table. Make sure we move them around
properly when doing inode resize.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
This removes the metadata block bitmap and makes the error handling
simpler. It also check for the enospc with the correct number needed
blocks. Also added specific error messages. We need to run e2undo
only if we start modiyfing inode, group desc and inode table.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The extent list header gets printed before we fall back to bmap:
# filefrag -v /mnt/test/bar
Filesystem type is: 58465342
File size of /mnt/test/bar is 12288 (3 blocks, blocksize 4096)
ext logical physical expected length flags <---- HERE
Discontinuity: Block 2 is at 17 (was 16)
/mnt/test/bar: 2 extents found
so delay printing it until we know fiemap is working.
(though ideally it'd be nice to have the same verbose output
regardless of the interface we used, I think).
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The fragmentation count in the bmap case seems to be
off by one:
/mnt/test/bar: 0 extents found
Addresses-Debian-Bug: #540376
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Fix a bug in e2freefrag where if the last free extent is at the very
end of the filesystem, it would be disregarded.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If the file system has a non-zero s_first_data_block, as is the case
when the block size is 1kb, e2freefrag would incorrectly try to
reference invalid data blocks in the block allocation bitmap.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
"Free chunks" is confusing since it has nothing to do with the
chunksize; use "free extents" instead.
Also add a missing newline in an error message.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
e4defrag.c had a lot of stuff copied into it from other
places, redefinitions of existing interfaces, etc.
We should be able to remove most of this, as the tool only
works on recent kernels anyway, we should just pick up
definitions from recent kernel headers whenever possible.
I've left the local definitions of fallocate, fadvise
(changed to posix_fadvise) and sync_file_range, and
wrapped them in #ifdef configure-time tests - though
really it seems like only fallocate should be necessary
by now, and perhaps the others can be dropped.
We still need some Makefile work so that it won't try to
build e4defrag if the right pieces aren't there (and
if the local definitions won't work...)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The FIEMAP support added in e2fsprogs 1.41.6 broke the "perfection
would be XXX expects" calculation restore it.
Also fix some gcc -Wall warnings as well. (Cleaning up gcc -Wall is
what caused me to notice this regression).
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When used with -v and the targeted file has more than 144
extents(double of the length of fm_extents array provided by buf),
filefrag_fiemap loops and calls fiemap ioctl() multiple times to
calculate the actual number of extents in a file. Each call to fiemap
ioctl() uses fm_start as the starting logical offset. The patch fixes
fm_start in each loop( except for the first one) and makes the extent
calculation correct for files with more that 144 extents.
To produce the problem, first run filefrag -v on a highly fragmented
file. Then change the buf size in filefrag_fiemap to make it large
enough to have all the extent mapped in a single loop and run filefrag
-v after recompiling. The former will produce a much smaller extent
count because of the false fm_start used in the loop. And the two will
produce different extent output since the 145th extent.
Signed-off-by: Peng Tao <bergwolf@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Some people don't want to see the concise "kernel-style" make output.
This configure option allows build engines that want to see the full
set of commands executed by the makefile to get what they want. Most
people will find this more distracting than useful, unless they need
to debug the Makefiles.
(It is not necessary to rerun configure to enable this verbose make
output temprarily; if a developer wants to do a quick debug of a
directory's makefile, he or she can simply edit the definition of the
$(E) and $(Q) variables in the Makefile; instructions can be found in
the MCONFIG file which is included in at the beginning of every
Makefile.)
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The e2fsprogs makefiles were using the same Makefile variable
LIBCOM_ERR for the link-line arguments as well as the dependencies.
Since LIBCOM_ERR can now include non-file arguments such as
"-lpthread", we need to use a separate DEPLIBCOM_ERR variable that
only has build file dependencies.
Do the same thing for STATIC_LIBCOM_ERR and PROFILED_LIBCOM_ERR.
Addresses-Sourceforge-Patches: #2813809
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If for some reason the uuidd daemon or the process calling uuidd
exited unexpectely, the read_all() function would end up looping
forever, either in uuidd or in libuuid. Fix this terminating the loop
if no data can be read after five tries to read from the file
descriptor.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
In the event that file descriptors 0-2 are closed when uuidd is
started, the server socket could be created as a file descriptor that
will get closed when create_daemon() tries detaching the uuidd daemon
from its controlling tty. Avoid this case by using dup(2).
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Some terminal programs may print wierd characters when they see the
\001 or \002 characters. So filter them out if the -s option
(skip_mode) is enabled.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
lsattr doesn't return an error if you point it at a file that
doesn't exist.
This is slightly trickier because it can take more than one
file as an arg, but ls seems to report an error if any occurred,
so this does the same, it'll report the last error that was
encountered.
Addresses-RedHat-Bugzilla: #489841
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Previously e2fsprogs interpreted 0 for a rec_len of 65536 (which could
occur if the directory block is completely empty in 64k blocksize
filesystems), while the kernel interpreted 65535 to mean 65536. The
kernel will accept both to mean 65536, and encodes 65535 to be 65536.
This commit changes e2fsprogs to match.
We add the encoding agreed upon for 128k and 256k filesystems, but we
don't enable support for these larger block sizes, since they haven't
been fully tested.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When compile e2fsprogs git tree with gcc-wall option, we get some warnings about
e4defrag. This patch fixes them.
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
To make it easier to maintain changes and fixes to the e4defrag
program, check it into the e2fsprogs source tree.
Signed-off-by: Akira Fujita <a-fujita@rs.jp.nec.com>
Signed-off-by: Takashi Sato <t-sato@yk.jp.nec.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The e2fsprogs programs have historically just said that they operate
on ext2 and ext3 file system in their man pages. Update them to say
that they also operate on ext4 file systems.
Addresses-Launchpad-bug: #381854
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Tidy up the chattr(1) manpage to completely document all
available options, and differentiate those which are read-only
early in the manpage as well.
* Remove "I" from settable attribute list
* add "e" to 2nd list of settable attributes & descriptions
* Note that h/E/I/X/Z are readonly
* Correct "H" to "h" for huge file attribute description
* fix long_name for indexed directory in flags_array
Addresses-Red-Hat-Bugzilla: BZ#502971
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
This patch adds new option, +e to chattr. The +e option
is used to convert the ext3 format (non extent) file
to ext4 (extent) format. This can be used to migrate
the ext3 file system to ext4 file system.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The FIEMAP ioctl is more efficient and doesn't require root
privileges. So if it is available, use it in preference to repeated
FIBMAP calls.
Signed-off-by: Kalpak Shah <kalpak.shah@sun.com>
Signed-off-by: Andreas Dilger <adilger@sun.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Fixed a potential bug where by partial returns from the write(2)
system call could some bytes to be lost when writing to the log file.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Add an option to switch between the private (in-tree) libblkid and
public (in-system installed) library. The private version is still
enabled by default.
If --disable-libblkid is specified the findfs(8) program, which is a
variant of tune2fs, is also not built or installed.
Signed-off-by: Karel Zak <kzak@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Hurd doesn't define PATH_MAX, so calculate the exact size needed for
the tdb filename, and allocate it dynamically.
Addresses-Debian-Bug: #521602
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Fixed a potential bug caused by partial returns from the write system
call (especially possible for network connections).
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Nice for testing w/o needing to swizzle around system
libraries...
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Add a check to make sure the argument to the -m option (which
specifies the reserved ratio) is greater than zero.
Addresses-Debian-Bug: #517015
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This field tracks the lifetime amount of writes to the filesystem. It
will be updated by the kernel as well as by e2fsprogs programs which
write to the filesystem.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Mandriva apparently uses "mke3fs" as an alias for mkfs.ext3. I'm not
particularly fond of that practice, but we'll include it as legacy
support.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
If a filesystem is built with the stride extended-option (which is
often used in RAID filesystems to make sure the block and inode
allocation bitmaps don't end up hitting one disk platter harder than
the rest), this can cause tune2fs -I to corrupt the filesystem because
it fails to handle the case where the allocation bitmaps are located
after the inode table, where the inode table needs to grow. Handle
this case correctly.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
With flex_bg usually the inode table for most block groups are packed
right against each other, so expanding the inode table size needs
special handling that's not currently in tune2fs.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When running "tune2fs -I 256" on moderate to large filesystems, the
time required to run tune2fs can take many hours (20+ before some
users gave up in disgust). This was due to some O(n**2) and O(n*m)
algorithms in move_block() and inode_scan_and_fix(), respectively.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Since the block group checksums depend on the UUID, we need to update
the block group checksums when setting the UUID. We only do so if all
of the checksums are correct, however.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Always initialize the starting time so that badblocks -sw works.
Thanks Jelle de Jong (jelledejong at powercraft.nl) for reporting this
bug.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Thanks to A. Costa for pointing these out.
Addresses-Debian-Bug: #498100
Addresses-Debian-Bug: #498101
Addresses-Debian-Bug: #498102
Addresses-Debian-Bug: #498103
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>