btrfs-progs/tests
David Sterba da64ae3a0f btrfs-progs: reorganize extent_buffer and fix alignment of data
Reported by UBSAN, the checksum code tries to access unaligned data that
come from the extent_buffer.

struct extent_buffer {
        struct cache_extent        cache_node;           /*     0    48 */
        u64                        start;                /*    48     8 */
        u64                        dev_bytenr;           /*    56     8 */
        /* --- cacheline 1 boundary (64 bytes) --- */
        u32                        len;                  /*    64     4 */

        /* XXX 4 bytes hole, try to pack */

        struct extent_io_tree *    tree;                 /*    72     8 */
        struct list_head           lru;                  /*    80    16 */
        struct list_head           recow;                /*    96    16 */
        int                        refs;                 /*   112     4 */
        u32                        flags;                /*   116     4 */
        int                        fd;                   /*   120     4 */
        char                       data[0];              /*   124     0 */

        /* size: 128, cachelines: 2, members: 11 */
        /* sum members: 120, holes: 1, sum holes: 4 */
        /* padding: 4 */
};

Add explicit alignment to data.

Reported-by: Lukas Lueg <lukas.lueg@gmail.com>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=156471
Signed-off-by: David Sterba <dsterba@suse.com>
2016-09-21 11:50:05 +02:00
..
cli-tests btrfs-progs: tests: add 003-fi-resize-args 2016-06-17 17:02:12 +02:00
convert-tests btrfs-progs: convert-test: Check if the ext2_save/image is read only 2016-08-19 13:53:28 +02:00
fsck-tests btrfs-progs: tests: add 021-partially-dropped-snapshot-case 2016-09-21 11:50:01 +02:00
fuzz-tests btrfs-progs: reorganize extent_buffer and fix alignment of data 2016-09-21 11:50:05 +02:00
misc-tests btrfs-progs: tests: add 015-dump-super-garbage 2016-09-05 12:20:23 +02:00
mkfs-tests btrfs-progs: typo review of strings and comments 2016-06-01 14:56:56 +02:00
clean-tests.sh btrfs-progs: tests: use /bin/bash for scripts 2016-07-04 13:45:15 +02:00
cli-tests.sh btrfs-progs: tests: unify test drivers 2016-06-03 16:53:27 +02:00
common btrfs-progs: convert-tests: Add support for custom test scripts 2016-06-03 15:22:18 +02:00
common.convert btrfs-progs: tests: post btrfs-convert verify permissions and ACLs 2016-09-21 11:47:24 +02:00
convert-tests.sh btrfs-progs: tests: print shorter test name in the output 2016-06-22 11:30:24 +02:00
fsck-tests.sh btrfs-progs: tests: unify test drivers 2016-06-03 16:53:27 +02:00
fuzz-tests.sh btrfs-progs: tests: unify test drivers 2016-06-03 16:53:27 +02:00
misc-tests.sh btrfs-progs: tests: unify test drivers 2016-06-03 16:53:27 +02:00
mkfs-tests.sh btrfs-progs: tests: unify test drivers 2016-06-03 16:53:27 +02:00
README.md btrfs-progs: tests README: fuzzed images 2016-09-05 10:05:08 +02:00
test-console.sh btrfs-progs: tests: use /bin/bash for scripts 2016-07-04 13:45:15 +02:00

Btrfs-progs tests

A testsuite covering functionality of btrfs-progs, ie. the checker, image, mkfs and similar tools. There are no special requirements on kernel features, the tests build on top of the core functionality like snapshots and device management. In some cases optional features are turned on by mkfs and the filesystem image could be mounted, such tests might fail if there's lack of support.

Quick start

Run the tests from the top directory:

$ make test
$ make test-fsck
$ make test-convert

or selectively from the tests/ directory:

$ ./fsck-tests.sh
$ ./misc-tests.sh

The verbose output of the tests is logged into a file named after the test category, eg. fsck-tests-results.txt.

Selective testing

The test are prefixed by a number for ordering and uniqueness. To run a particular test use:

$ make TEST=MASK test

where MASK is a glob expression that will execute only tests that match the MASK. Here the test number comes handy:

$ make TEST=001\* test-fsck
$ TEST=001\* ./fsck-tests.sh

will run the first test in fsck-tests subdirectory.

Test structure

tests/fsck-tests/:

  • tests targeted at bugs that are fixable by fsck

tests/convert-tests/:

  • coverage tests of ext2/3/4 and btrfs-convert options

tests/fuzz-tests/:

  • collection of fuzzed or crafted images
  • tests that are supposed to run various utilities on the images and not crash

tests/cli-tests/:

  • tests for command line interface, option coverage, weird option combinations that should not work
  • not necessary to do any functional testing, could be rather lightweight
  • functional tests should go to to other test dirs
  • the driver script will only execute ./test.sh in the test directory

tests/misc-tests/:

  • anything that does not fit to the above, the test driver script will only execute ./test.sh in the test directory

tests/common: tests/common.convert:

  • script with shell helpers, separated by functionality

tests/test.img:

  • default testing image, the file is never deleted by the scripts but truncated to 0 bytes, so it keeps it's permissions. It's eg. possible to host it on NFS, make it chmod a+w for root.

Other tuning, environment variables

Instrumentation

It's possible to wrap the tested commands to utilities that might do more checking or catch failures at runtime. This can be done by setting the INSTRUMENT environment variable:

INSTRUMENT=valgrind ./fuzz-tests.sh    # in tests/
make INSTRUMENT=valgrind test-fuzz     # in the top directory

The variable is prepended to the command unquoted, all sorts of shell tricks are possible.

Note: instrumentation is not applied to privileged commands (anything that uses the root helper).

Verbosity

Setting the variable TEST_LOG=tty will print all commands executed by some of the wrappers (run_check etc), other commands are not printed to the terminal (but the full output is in the log).

Permissions

Some commands require root privileges (to mount/umount, access loop devices). It is assumed that sudo will work in some way (no password, password asked and cached). Note that instrumentation is not applied in this case, for safety reasons. You need to modify the test script instead.

Cleanup

The tests are supposed to cleanup after themselves if they pass. In case of failure, the rest of the tests are skipped and intermediate files, mounts and loop devices are kept. This should help to investigate the test failure but at least the mounts and loop devices need to be cleaned before the next run.

This is partially done by the script clean-tests.sh, you may want to check the loop devices as they are managed on a per-test basis.

Prototyping tests, quick tests

There's a script test-console.sh that will run shell commands in a loop and logs the output with the testing environment set up.

New test

  1. Pick the category for the new test or fallback to misc-tests if not sure. For an easy start copy an existing test.sh script from some test that might be close to the purpose of your new test. The environment setup includes the common scripts and/or prepares the test devices. Other scripts contain examples how to do mkfs, mount, unmount, check, etc.

  2. Use the highest unused number in the sequence, write a short descriptive title and join by dashes -. This will become the directory name, eg. 012-subvolume-sync-must-wait.

  3. Write a short description of the bug and how it's tested to the comment at the begining of test.sh. You don't need to add the file to git yet.

  4. Write the test commands, comment anything that's not obvious.

  5. Test your test. Use the TEST variable to jump right to your test:

$ make TEST=012\* tests-misc           # from top directory
$ TEST=012\* ./misc-tests.sh           # from tests/
  1. The commit changelog should reference a commit that either introduced or fixed the bug (or both). Subject line of the shall mention the name of the new directory for ease of search, eg. btrfs-progs: tests: add 012-subvolume-sync-must-wait

Crafted/fuzzed images

Images that are create by fuzzing or specially crafted to trigger some error conditions should be added to the directory fuzz-tests/images, accompanied by a textual description of the source (bugzilla, mail), the reporter, brief description of the problem or the stack trace.

If you have a fix for the problem, please submit it prior to the test image, so the fuzz tests always succeed when run on random checked out. This helps bisectability.