In the future, btrfs_util_[gs]et_subvolume_flags() might be useful, but
since these are the only subvolume flags we've defined in all this time,
this will do for now.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This gets the the information in `btrfs subvolume show` from the root
item.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We can just walk up root backrefs with BTRFS_IOC_TREE_SEARCH and inode
paths with BTRFS_IOC_INO_LOOKUP.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Doing the ioctl() directly isn't too bad, but passing in a full path is
more convenient than opening the parent and passing the path component.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
These are the most trivial helpers in the library and will be used to
implement several of the more involved functions.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Systems with older kernels won't have these available, and the copies in
btrfs-progs aren't quite compatible, so for now, let's just copy these
in. We can potentially deduplicate some of this in the future.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
These were broken when the patch series got shuffled around.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Just as kernel find_free_dev_extent(), allow it to return maximum hole
size for us to build device list for later chunk allocator rework.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
As part of the effort to unify code and behavior between btrfs-progs and
kernel, copy the btrfs_raid_array from kernel to btrfs-progs.
So later we can use the btrfs_raid_array[] to get needed raid info other
than manually do if-else branches.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We want to hide struct btrfs_qgroup_inherit from the user because that
comes from the Btrfs UAPI headers. Instead, wrap it in a struct
btrfs_util_qgroup_inherit and provide helpers to manipulate it. This
will be used for subvolume and snapshot creation.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The C libbtrfsutil library isn't very useful for scripting, so we also
want bindings for Python. Writing unit tests in Python is also much
easier than doing so in C. Only Python 3 is supported; if someone really
wants Python 2 support, they can write their own bindings. This commit
is just the scaffolding.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently, users wishing to manage Btrfs filesystems programatically
have to shell out to btrfs-progs and parse the output. This isn't ideal.
The goal of libbtrfsutil is to provide a library version of as many of
the operations of btrfs-progs as possible and to migrate btrfs-progs to
use it.
Rather than simply refactoring the existing btrfs-progs code, the code
has to be written from scratch for a couple of reasons:
* A lot of the btrfs-progs code was not designed with a nice library API
in mind in terms of reusability, naming, and error reporting.
* libbtrfsutil is licensed under the LGPL, whereas btrfs-progs is under
the GPL, which makes it dubious to directly copy or move the code.
Eventually, most of the low-level btrfs-progs code should either live in
libbtrfsutil or the shared kernel/userspace filesystem code, and
btrfs-progs will just be the CLI wrapper.
This first commit just includes the build system changes, license,
README, and error reporting helper.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Add additional bound checks to prevent memory corruption on incorrect
usage of subvol_strip_mountpoint. Assert sane return value by properly
comparing the mount point to the full_path before stripping it off.
Mitigates issue: "btrfs send -p" fails if source and parent subvolumes
are on different mountpoints (memory corruption):
https://github.com/kdave/btrfs-progs/issues/96
Note that this does not properly fix this bug, but prevents a possible
security issue by unexpected usage of "btrfs send -p".
Issue: #96
Pull-request: #98
Signed-off-by: Axel Burri <axel@tty0.ch>
Signed-off-by: David Sterba <dsterba@suse.com>
Right now the pkg-config command is hard-coded in configure.ac, which may
result in build errors on system and cross environments that have prefixed
toolchains, e.g. /usr/x86_64-pc-linux-gnu/bin/x86_64-pc-linux-gnu-pkg-config.
Please see the attached patch, it has been written a while ago but it seems it
hasn't been submitted for upstream inclusion.
0001-configure.ac-Consistently-use-PKG_CONFIG.txt
Submitted by Timo Gurr.
Author: Wulf C. Krueger <philantrop@exherbo.org>
Issue: #101
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_show_qgroups frees the filter and comparer in case it succeeds.
This makes the caller slightly more complicated so move the freeing up
one level.
Issue: #20
Signed-off-by: David Sterba <dsterba@suse.com>
Exposed by convert-test with D=asan.
Unlike btrfs, ext2fs_close() still leaves its ext2_filsys parameter
filled with allocated pointers.
It needs ext2fs_free() to free those pointers.
Issue: #92
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The bug is exposed by mkfs test case 009, with D=asan.
We are leaking memory of parent_dir_entry->path() which ,except the
rootdir, is allocated by strdup().
Before fixing it, unifiy the allocation of parent_dir_entry() to heap
allocation.
Then fix it by adding "free(parent_dir_entry->path);" in
traverse_directory() and error handler.
Issue: #92
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This bug is exposed by fsck-test with D=asan, hit by test case 020, with
the following error report:
=================================================================
==10740==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x621000061580 at pc 0x56051f0db6cd bp 0x7ffe170f3e20 sp 0x7ffe170f3e10
READ of size 1 at 0x621000061580 thread T0
#0 0x56051f0db6cc in btrfs_extent_inline_ref_type /home/adam/btrfs/btrfs-progs/ctree.h:1727
#1 0x56051f13b669 in build_roots_info_cache /home/adam/btrfs/btrfs-progs/cmds-check.c:14306
#2 0x56051f13c86a in repair_root_items /home/adam/btrfs/btrfs-progs/cmds-check.c:14450
#3 0x56051f13ea89 in cmd_check /home/adam/btrfs/btrfs-progs/cmds-check.c:14965
#4 0x56051efe75bb in main /home/adam/btrfs/btrfs-progs/btrfs.c:302
#5 0x7f04ddbb0f49 in __libc_start_main (/usr/lib/libc.so.6+0x20f49)
#6 0x56051efe68c9 in _start (/home/adam/btrfs/btrfs-progs/btrfs+0x5b8c9)
0x621000061580 is located 0 bytes to the right of 4224-byte region [0x621000060500,0x621000061580)
allocated by thread T0 here:
#0 0x7f04ded50ce1 in __interceptor_calloc /build/gcc/src/gcc/libsanitizer/asan/asan_malloc_linux.cc:70
#1 0x56051f04685e in __alloc_extent_buffer /home/adam/btrfs/btrfs-progs/extent_io.c:553
#2 0x56051f047563 in alloc_extent_buffer /home/adam/btrfs/btrfs-progs/extent_io.c:687
#3 0x56051efff1d1 in btrfs_find_create_tree_block /home/adam/btrfs/btrfs-progs/disk-io.c:187
#4 0x56051f000133 in read_tree_block /home/adam/btrfs/btrfs-progs/disk-io.c:327
#5 0x56051efeddb8 in read_node_slot /home/adam/btrfs/btrfs-progs/ctree.c:652
#6 0x56051effb0d9 in btrfs_next_leaf /home/adam/btrfs/btrfs-progs/ctree.c:2853
#7 0x56051f13b343 in build_roots_info_cache /home/adam/btrfs/btrfs-progs/cmds-check.c:14267
#8 0x56051f13c86a in repair_root_items /home/adam/btrfs/btrfs-progs/cmds-check.c:14450
#9 0x56051f13ea89 in cmd_check /home/adam/btrfs/btrfs-progs/cmds-check.c:14965
#10 0x56051efe75bb in main /home/adam/btrfs/btrfs-progs/btrfs.c:302
#11 0x7f04ddbb0f49 in __libc_start_main (/usr/lib/libc.so.6+0x20f49)
It's completely possible that one extent/metadata item has no inline
reference, while build_roots_info_cache() doesn't have such check.
Fix it by checking @iref against item end to avoid such problem.
Issue: #92
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
dump-tree prints wrong sequence number and the flags numbers,
as we misplaced the printf args. This patch fixes it.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Make the TOP variable more configurable, allow to set it to any path
where to find binaries when the testsuite is exported, or fallback to
system binaries.
There's now more code duplication, the logic is now more complex so it's
left open coded for clarity. Further cleanups are possible.
Signed-off-by: David Sterba <dsterba@suse.com>
Use TEST_TOP as base for tests to reference any files, this will be used
for git and external testsuite.
INTERNAL_BIN is needed for referencing binaries that could reside in
different paths in git vs external testsuite.
Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com>
[ add quotes around sourced files, update changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
Move the testsuite to tests/ and make the tarball generation more
deterministic. As there could be many random temporary files left in the
test directories, we can't just copy them. Use 'git ls-tree' to
filter just what we want, this needs a slight extension of the file list
specification.
Signed-off-by: David Sterba <dsterba@suse.com>
Export the testsuite files to a separate tar. Since fsck tests depend
on btrfs-corrupt-block, and misc tests depends on both
btrfs-corrupt-block and fssum, so set it as prerequisites for package
commad.
Because, althougth fssum can be generated by source that are all in
tests directory, and has no rely on the btrfs's structure. But
btrfs-corrupt-block deeply relys on btrfs's structure. For consistency,
at the present stage, generete the two when create test tar.
Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com>
[ applied without changes, the generated tarball will be different from
the one after the follow up commits ]
Signed-off-by: David Sterba <dsterba@suse.com>
Since tests/misc-tests/005-convert-progress-thread-crash/test.sh need
the btrfs-convert for prerequisite. So add the dependency in Makefile.
Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since tests/fuzz-tests/002-simple-image/test.sh need the btrfs-image for
prerequisite. So add the dependency in Makefile.
Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since tests/cli-tests/002-balance-full-no-filters/test.sh need
the mkfs.btrfs for prerequisite.
So add the dependency in Makefile.
Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Enable docker in travis and start build test(s) before our testsuite.
Now for musl libc, so we don't accidentally break the build anymore.
Issue: #29
Signed-off-by: David Sterba <dsterba@suse.com>
Build environment with musl as the libc, based on Alpine Linux.
To build the docker image:
- cd travis/images
- cp test-build $dir
- cd $dir
- docker build -t $tag .
- docker push $tag
We're going to have more images that will share the prep script so it
has to be copied to the directory before making the image. No other way
around that.
Signed-off-by: David Sterba <dsterba@suse.com>
The test-build resides inside the docker image and is supposed to be
called from outside, downloads the branch and calls build-default.
Buid-default will run up to make.
Signed-off-by: David Sterba <dsterba@suse.com>
We've been using asciidoc that's written in python2, which is going to
be phased out and deprecated next year. There's a replacement,
asciidoctor. Add a configure-time detection which tool is available,
update Documentation/Makefile.in.
The original asciidoc tool is still preferred as it produces slightly
better output. The file asciidoc.conf does not have a direct equivalten
in asciidoct and would need to be replaced by extension written in ruby.
The differences:
- the <literal> are not automatically underlined and are less visible in
the generated manual page, but it's still acceptable
- the inline CSS for the html output looks subjectively worse, is less
compact and colourful
Issue: #89
Signed-off-by: David Sterba <dsterba@suse.com>
Omar Sandoval
Qu Wenruo
David Sterba
Axel Burri
Anand Jain
Gu Jinxiang
Su Yue