When it fails to allocate fragment, it does not free and return error.
And check the pointer inappropriately.
Fixed merge conflicts with
commit 618887768b ("kunit: update NULL vs IS_ERR() tests")
Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: YoungJun.park <her0gyugyu@gmail.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
This KUnit next update for Linux 6.2-rc1 consists of several enhancements,
fixes, clean-ups, documentation updates, improvements to logging and KTAP
compliance of KUnit test output:
- log numbers in decimal and hex
- parse KTAP compliant test output
- allow conditionally exposing static symbols to tests
when KUNIT is enabled
- make static symbols visible during kunit testing
- clean-ups to remove unused structure definition
-----BEGIN PGP SIGNATURE-----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=UQNn
-----END PGP SIGNATURE-----
Merge tag 'linux-kselftest-kunit-next-6.2-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest
Pull KUnit updates from Shuah Khan:
"Several enhancements, fixes, clean-ups, documentation updates,
improvements to logging and KTAP compliance of KUnit test output:
- log numbers in decimal and hex
- parse KTAP compliant test output
- allow conditionally exposing static symbols to tests when KUNIT is
enabled
- make static symbols visible during kunit testing
- clean-ups to remove unused structure definition"
* tag 'linux-kselftest-kunit-next-6.2-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest: (29 commits)
Documentation: dev-tools: Clarify requirements for result description
apparmor: test: make static symbols visible during kunit testing
kunit: add macro to allow conditionally exposing static symbols to tests
kunit: tool: make parser preserve whitespace when printing test log
Documentation: kunit: Fix "How Do I Use This" / "Next Steps" sections
kunit: tool: don't include KTAP headers and the like in the test log
kunit: improve KTAP compliance of KUnit test output
kunit: tool: parse KTAP compliant test output
mm: slub: test: Use the kunit_get_current_test() function
kunit: Use the static key when retrieving the current test
kunit: Provide a static key to check if KUnit is actively running tests
kunit: tool: make --json do nothing if --raw_ouput is set
kunit: tool: tweak error message when no KTAP found
kunit: remove KUNIT_INIT_MEM_ASSERTION macro
Documentation: kunit: Remove redundant 'tips.rst' page
Documentation: KUnit: reword description of assertions
Documentation: KUnit: make usage.rst a superset of tips.rst, remove duplication
kunit: eliminate KUNIT_INIT_*_ASSERT_STRUCT macros
kunit: tool: remove redundant file.close() call in unit test
kunit: tool: unit tests all check parser errors, standardize formatting a bit
...
Change KUnit test output to better comply with KTAP v1 specifications
found here: https://kernel.org/doc/html/latest/dev-tools/ktap.html.
1) Use "KTAP version 1" instead of "TAP version 14" as test output header
2) Remove '-' between test number and test name on test result lines
2) Add KTAP version lines to each subtest header as well
Note that the new KUnit output still includes the “# Subtest” line now
located after the KTAP version line. This does not completely match the
KTAP v1 spec but since it is classified as a diagnostic line, it is not
expected to be disruptive or break any existing parsers. This
“# Subtest” line comes from the TAP 14 spec
(https://testanything.org/tap-version-14-specification.html) and it is
used to define the test name before the results.
Original output:
TAP version 14
1..1
# Subtest: kunit-test-suite
1..3
ok 1 - kunit_test_1
ok 2 - kunit_test_2
ok 3 - kunit_test_3
# kunit-test-suite: pass:3 fail:0 skip:0 total:3
# Totals: pass:3 fail:0 skip:0 total:3
ok 1 - kunit-test-suite
New output:
KTAP version 1
1..1
KTAP version 1
# Subtest: kunit-test-suite
1..3
ok 1 kunit_test_1
ok 2 kunit_test_2
ok 3 kunit_test_3
# kunit-test-suite: pass:3 fail:0 skip:0 total:3
# Totals: pass:3 fail:0 skip:0 total:3
ok 1 kunit-test-suite
Signed-off-by: Rae Moar <rmoar@google.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Tested-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
KUnit does a few expensive things when enabled. This hasn't been a
problem because KUnit was only enabled on test kernels, but with a few
people enabling (but not _using_) KUnit on production systems, we need a
runtime way of handling this.
Provide a 'kunit_running' static key (defaulting to false), which allows
us to hide any KUnit code behind a static branch. This should reduce the
performance impact (on other code) of having KUnit enabled to a single
NOP when no tests are running.
Note that, while it looks unintuitive, tests always run entirely within
__kunit_test_suites_init(), so it's safe to decrement the static key at
the end of this function, rather than in __kunit_test_suites_exit(),
which is only there to clean up results in debugfs.
Signed-off-by: David Gow <davidgow@google.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Augment the example_all_expect_macros_test with the KUNIT_EXPECT_MEMEQ
and KUNIT_EXPECT_MEMNEQ macros by creating a test with memory block
assertions.
Signed-off-by: Maíra Canal <mairacanal@riseup.net>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Currently, in order to compare memory blocks in KUnit, the KUNIT_EXPECT_EQ
or KUNIT_EXPECT_FALSE macros are used in conjunction with the memcmp
function, such as:
KUNIT_EXPECT_EQ(test, memcmp(foo, bar, size), 0);
Although this usage produces correct results for the test cases, when
the expectation fails, the error message is not very helpful,
indicating only the return of the memcmp function.
Therefore, create a new set of macros KUNIT_EXPECT_MEMEQ and
KUNIT_EXPECT_MEMNEQ that compare memory blocks until a specified size.
In case of expectation failure, those macros print the hex dump of the
memory blocks, making it easier to debug test failures for memory blocks.
That said, the expectation
KUNIT_EXPECT_EQ(test, memcmp(foo, bar, size), 0);
would translate to the expectation
KUNIT_EXPECT_MEMEQ(test, foo, bar, size);
Signed-off-by: Maíra Canal <mairacanal@riseup.net>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
When KUNIT_EXPECT_EQ() or KUNIT_ASSERT_EQ() log a failure, they log the
two values being compared, with numerical values logged in decimal.
In some cases, decimal output is painful to consume, and hexadecimal
output would be more helpful. For example, this is the case for tests
I'm currently developing for the arm64 insn encoding/decoding code,
where comparing two 32-bit instruction opcodes results in output such
as:
| # test_insn_add_shifted_reg: EXPECTATION FAILED at arch/arm64/lib/test_insn.c:2791
| Expected obj_insn == gen_insn, but
| obj_insn == 2332164128
| gen_insn == 1258422304
To make this easier to consume, this patch logs the values in both
decimal and hexadecimal:
| # test_insn_add_shifted_reg: EXPECTATION FAILED at arch/arm64/lib/test_insn.c:2791
| Expected obj_insn == gen_insn, but
| obj_insn == 2332164128 (0x8b020020)
| gen_insn == 1258422304 (0x4b020020)
As can be seen from the example, having hexadecimal makes it
significantly easier for a human to spot which specific bits are
incorrect.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Brendan Higgins <brendan.higgins@linux.dev>
Cc: David Gow <davidgow@google.com>
Cc: linux-kselftest@vger.kernel.org
Cc: kunit-dev@googlegroups.com
Acked-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The alloc_string_stream() functions were changed from returning NULL on
error to returning error pointers so these caller needs to be updated
as well.
Fixes: 78b1c6584f ("kunit: string-stream: Simplify resource use")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Each calll to a KUNIT_EXPECT_*() macro creates a local variable which
contains a struct kunit_assert.
Normally, we'd hope the compiler would be able to optimize this away,
but we've seen cases where it hasn't, see
https://groups.google.com/g/kunit-dev/c/i3fZXgvBrfA/m/GbrMNej2BAAJ.
In changes like commit 21957f90b2 ("kunit: split out part of
kunit_assert into a static const"), we've moved more and more parts out
of struct kunit_assert and its children types (kunit_binary_assert).
This patch removes the final field and gets us to:
sizeof(struct kunit_assert) == 0
sizeof(struct kunit_binary_assert) == 24 (on UML x86_64).
This also reduces the amount of macro plumbing going on at the cost of
passing in one more arg to the base KUNIT_ASSERTION macro and
kunit_do_failed_assertion().
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The real kfree() function will silently return when given a NULL.
So a user might reasonably think they can write the following code:
char *buffer = NULL;
if (param->use_buffer) buffer = kunit_kzalloc(test, 10, GFP_KERNEL);
...
kunit_kfree(test, buffer);
As-is, kunit_kfree() will mark the test as FAILED when buffer is NULL.
(And in earlier times, it would segfault).
Let's match the semantics of kfree().
Suggested-by: David Gow <davidgow@google.com>
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
kunit_kfree() can only work on data ("resources") allocated by KUnit.
Currently for code like this,
> void *ptr = kmalloc(4, GFP_KERNEL);
> kunit_kfree(test, ptr);
kunit_kfree() will segfault.
It'll try and look up the kunit_resource associated with `ptr` and get a
NULL back, but it won't check for this. This means we also segfault if
you double-free.
Change kunit_kfree() so it'll notice these invalid pointers and respond
by failing the test.
Implementation: kunit_destroy_resource() does what kunit_kfree() does,
but is more generic and returns -ENOENT when it can't find the resource.
Sadly, unlike just letting it crash, this means we don't get a stack
trace. But kunit_kfree() is so infrequently used it shouldn't be hard to
track down the bad callsite anyways.
After this change, the above code gives:
> # example_simple_test: EXPECTATION FAILED at lib/kunit/test.c:702
> kunit_kfree: 00000000626ec200 already freed or not allocated by kunit
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
kunit_kfree() exists to clean up allocations from kunit_kmalloc() and
friends early instead of waiting for this to happen automatically at the
end of the test.
But it can be used on *anything* registered with the kunit resource API.
E.g. the last 2 statements are equivalent:
struct kunit_resource *res = something();
kfree(res->data);
kunit_put_resource(res);
The problem is that there could be multiple resources that point to the
same `data`.
E.g. you can have a named resource acting as a pseudo-global variable in
a test. If you point it to data allocated with kunit_kmalloc(), then
calling `kunit_kfree(ptr)` has the chance to delete either the named
resource or to kfree `ptr`.
Which one it does depends on the order the resources are registered as
kunit_kfree() will delete resources in LIFO order.
So this patch restricts kunit_kfree() to only working on resources
created by kunit_kmalloc(). Calling it is therefore guaranteed to free
the memory, not do anything else.
Note: kunit_resource_instance_match() wasn't used outside of KUnit, so
it should be safe to remove from the public interface. It's also
generally dangerous, as shown above, and shouldn't be used.
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
We already store the `struct kunit *test` in the string_stream object
itself, so we need don't need to store a copy of this pointer in every
fragment in the stream.
Drop it, getting string_stream_fragment down the bare minimum: a
list_head and the `char *` with the actual fragment.
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Currently, KUnit's string streams are themselves "KUnit resources".
This is redundant since the stream itself is already allocated with
kunit_kzalloc() and will thus be freed automatically at the end of the
test.
string-stream is only used internally within KUnit, and isn't using the
extra features that resources provide like reference counting, being
able to locate them dynamically as "test-local variables", etc.
Indeed, the resource's refcount is never incremented when the
pointer is returned. The fact that it's always manually destroyed is
more evidence that the reference counting is unused.
Signed-off-by: David Gow <davidgow@google.com>
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
This patch adds the kunit.enable module parameter that will need to be
set to true in addition to KUNIT being enabled for KUnit tests to run.
The default value is true giving backwards compatibility. However, for
the production+testing use case the new config option
KUNIT_DEFAULT_ENABLED can be set to N requiring the tester to opt-in
by passing kunit.enable=1 to the kernel.
Signed-off-by: Joe Fradley <joefradley@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
It's possible that memory allocation for 'filtered' will fail, but for the
copy of the suite to succeed. In this case, the copy could be leaked.
Properly free 'copy' in the error case for the allocation of 'filtered'
failing.
Note that there may also have been a similar issue in
kunit_filter_subsuites, before it was removed in "kunit: flatten
kunit_suite*** to kunit_suite** in .kunit_test_suites".
This was reported by clang-analyzer via the kernel test robot, here:
https://lore.kernel.org/all/c8073b8e-7b9e-0830-4177-87c12f16349c@intel.com/
And by smatch via Dan Carpenter and the kernel test robot:
https://lore.kernel.org/all/202207101328.ASjx88yj-lkp@intel.com/
Fixes: a02353f491 ("kunit: bail out of test filtering logic quicker if OOM")
Reported-by: kernel test robot <yujie.liu@intel.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
We currently store kunit suites in the .kunit_test_suites ELF section as
a `struct kunit_suite***` (modulo some `const`s).
For every test file, we store a struct kunit_suite** NULL-terminated array.
This adds quite a bit of complexity to the test filtering code in the
executor.
Instead, let's just make the .kunit_test_suites section contain a single
giant array of struct kunit_suite pointers, which can then be directly
manipulated. This array is not NULL-terminated, and so none of the test
filtering code needs to NULL-terminate anything.
Tested-by: Maíra Canal <maira.canal@usp.br>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Co-developed-by: David Gow <davidgow@google.com>
Signed-off-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Currently, KUnit runs built-in tests and tests loaded from modules
differently. For built-in tests, the kunit_test_suite{,s}() macro adds a
list of suites in the .kunit_test_suites linker section. However, for
kernel modules, a module_init() function is used to run the test suites.
This causes problems if tests are included in a module which already
defines module_init/exit_module functions, as they'll conflict with the
kunit-provided ones.
This change removes the kunit-defined module inits, and instead parses
the kunit tests from their own section in the module. After module init,
we call __kunit_test_suites_init() on the contents of that section,
which prepares and runs the suite.
This essentially unifies the module- and non-module kunit init formats.
Tested-by: Maíra Canal <maira.canal@usp.br>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Signed-off-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
kmemdup() is easier than kmalloc() + memcpy(), per lkp bot.
Also make the input `suite` as const since we're now always making
copies after commit a127b154a8 ("kunit: tool: allow filtering test
cases via glob").
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Make KUnit trigger the new TAINT_TEST taint when any KUnit test is run.
Due to KUnit tests not being intended to run on production systems, and
potentially causing problems (or security issues like leaking kernel
addresses), the kernel's state should not be considered safe for
production use after KUnit tests are run.
This both marks KUnit modules as test modules using MODULE_INFO() and
manually taints the kernel when tests are run (which catches builtin
tests).
Acked-by: Luis Chamberlain <mcgrof@kernel.org>
Tested-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: David Gow <davidgow@google.com>
Tested-by: Maíra Canal <mairacanal@riseup.net>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The existing logic happens to work fine on UML, but is not correct when
running on other arches.
1. We didn't initialize `int err`, and kunit_filter_suites() doesn't
explicitly set it to 0 on success. So we had false "failures".
Note: it doesn't happen on UML, causing this to get overlooked.
2. If we error out, we do not call kunit_handle_shutdown().
This makes kunit.py timeout when using a non-UML arch, since the QEMU
process doesn't ever exit.
Fixes: a02353f491 ("kunit: bail out of test filtering logic quicker if OOM")
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The `kunit_do_failed_assertion` function passes its
`struct kunit_assert` argument to `kunit_fail`. This one,
in turn, calls its `format` field passing the assert again
as a `const` pointer.
Therefore, the whole chain may be made `const`.
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
When filtering what tests to run (suites and/or cases) via
kunit.filter_glob (e.g. kunit.py run <glob>), we allocate copies of
suites.
These allocations can fail, and we largely don't handle that.
Note: realistically, this probably doesn't matter much.
We're not allocating much memory and this happens early in boot, so if
we can't do that, then there's likely far bigger problems.
This patch makes us immediately bail out from the top-level function
(kunit_filter_suites) with -ENOMEM if any of the underlying kmalloc()
calls return NULL.
Implementation note: we used to return NULL pointers from some functions
to indicate either that all suites/tests were filtered out or there was
an error allocating the new array.
We'll log a short error in this case and not run any tests or print a
TAP header. From a kunit.py user's perspective, they'll get a message
about missing/invalid TAP output and have to dig into the test.log to
see it. Since hitting this error seems so unlikely, it's probably fine
to not invent a way to plumb this error message more visibly.
See also: https://lore.kernel.org/linux-kselftest/20220329103919.2376818-1-lv.ruyi@zte.com.cn/
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reported-by: Zeal Robot <zealci@zte.com.cn>
Reported-by: Lv Ruyi <lv.ruyi@zte.com.cn>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
KUnit's test-managed resources can be created in two ways:
- Using the kunit_add_resource() family of functions, which accept a
struct kunit_resource pointer, typically allocated statically or on
the stack during the test.
- Using the kunit_alloc_resource() family of functions, which allocate a
struct kunit_resource using kzalloc() behind the scenes.
Both of these families of functions accept a 'free' function to be
called when the resource is finally disposed of.
At present, KUnit will kfree() the resource if this 'free' function is
specified, and will not if it is NULL. However, this can lead
kunit_alloc_resource() to leak memory (if no 'free' function is passed
in), or kunit_add_resource() to incorrectly kfree() memory which was
allocated by some other means (on the stack, as part of a larger
allocation, etc), if a 'free' function is provided.
Instead, always kfree() if the resource was allocated with
kunit_alloc_resource(), and never kfree() if it was passed into
kunit_add_resource() by the user. (If the user of kunit_add_resource()
wishes the resource be kfree()ed, they can call kfree() on the resource
from within the 'free' function.
This is implemented by adding a 'should_free' member to
struct kunit_resource and setting it appropriately. To facilitate this,
the various resource add/alloc functions have been refactored somewhat,
making them all call a __kunit_add_resource() helper after setting the
'should_free' member appropriately. In the process, all other functions
have been made static inline functions.
Signed-off-by: David Gow <davidgow@google.com>
Tested-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Commit 6d2426b2f2 ("kunit: Support skipped tests") switched to using
`enum kunit_status` to track the result of running a test/suite since we
now have more than just pass/fail.
This callsite wasn't updated, silently converting to enum to a bool and
then back.
Fixes: 6d2426b2f2 ("kunit: Support skipped tests")
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
KUnit has support for setup/cleanup logic for each test case in a suite.
But it lacks the ability to specify setup/cleanup for the entire suite
itself.
This can be used to do setup that is too expensive or cumbersome to do
for each test.
Or it can be used to do simpler things like log debug information after
the suite completes.
It's a fairly common feature, so the lack of it is noticeable.
Some examples in other frameworks and languages:
* https://docs.python.org/3/library/unittest.html#setupclass-and-teardownclass
* https://google.github.io/googletest/reference/testing.html#Test::SetUpTestSuite
Meta:
This is very similar to this patch here: https://lore.kernel.org/linux-kselftest/20210805043503.20252-3-bvanassche@acm.org/
The changes from that patch:
* pass in `struct kunit *` so users can do stuff like
`kunit_info(suite, "debug message")`
* makes sure the init failure is bubbled up as a failure
* updates kunit-example-test.c to use a suite init
* Updates kunit/usage.rst to mention the new support
* some minor cosmetic things
* use `suite_{init,exit}` instead of `{init/exit}_suite`
* make suite init error message more consistent w/ test init
* etc.
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
These names sound more general than they are.
The _end() function increments a `static int kunit_suite_counter`, so it
can only safely be called on suites, aka top-level subtests.
It would need to have a separate counter for each level of subtest to be
generic enough.
So rename it to make it clear it's only appropriate for suites.
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The kunit_remove_resource() function is used to unlink a resource from
the list of resources in the test, making it no longer show up in
kunit_find_resource().
However, this could lead to a race condition if two threads called
kunit_remove_resource() on the same resource at the same time: the
resource would be removed from the list twice (causing a crash at the
second list_del()), and the refcount for the resource would be
decremented twice (instead of once, for the reference held by the
resource list).
Fix both problems, the first by using list_del_init(), and the second by
checking if the resource has already been removed using list_empty(),
and only decrementing its refcount if it has not.
Also add a KUnit test for the kunit_remove_resource() function which
tests this behaviour.
Reported-by: Daniel Latypov <dlatypov@google.com>
Signed-off-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
We've split out the declarations from include/kunit/test.h into
resource.h.
This patch splits out the definitions as well for consistency.
A side effect of this is git blame won't properly track history by
default, users need to run
$ git blame -L ,1 -C13 lib/kunit/resource.c
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Replace the NULL checks with the more specific and idiomatic NULL macros.
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
This KUnit update for Linux 5.18-rc1 consists of:
- changes to decrease macro layering string, integer, EQ/NE asserts
- remove unused macros
- several cleanups and fixes
- new list tests for list_del_init_careful(), list_is_head() and
list_entry_is_head()
-----BEGIN PGP SIGNATURE-----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=e1/u
-----END PGP SIGNATURE-----
Merge tag 'linux-kselftest-kunit-5.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest
Pull KUnit updates from Shuah Khan:
- changes to decrease macro layering string, integer, EQ/NE asserts
- remove unused macros
- several cleanups and fixes
- new list tests for list_del_init_careful(), list_is_head() and
list_entry_is_head()
* tag 'linux-kselftest-kunit-5.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:
list: test: Add a test for list_entry_is_head()
list: test: Add a test for list_is_head()
list: test: Add test for list_del_init_careful()
kunit: cleanup assertion macro internal variables
kunit: factor out str constants from binary assertion structs
kunit: consolidate KUNIT_INIT_BINARY_ASSERT_STRUCT macros
kunit: remove va_format from kunit_assert
kunit: tool: drop mostly unused KunitResult.result field
kunit: decrease macro layering for EQ/NE asserts
kunit: decrease macro layering for integer asserts
kunit: reduce layering in string assertion macros
kunit: drop unused intermediate macros for ptr inequality checks
kunit: make KUNIT_EXPECT_EQ() use KUNIT_EXPECT_EQ_MSG(), etc.
kunit: drop unused assert_type from kunit_assert and clean up macros
kunit: split out part of kunit_assert into a static const
kunit: factor out kunit_base_assert_format() call into kunit_fail()
kunit: drop unused kunit* field in kunit_assert
kunit: move check if assertion passed into the macros
kunit: add example test case showing off all the expect macros
In function kunit_test_timeout, it is declared "300 * MSEC_PER_SEC"
represent 5min. However, it is wrong when dealing with arm64 whose
default HZ = 250, or some other situations. Use msecs_to_jiffies to fix
this, and kunit_test_timeout will work as desired.
Link: https://lkml.kernel.org/r/20220309083753.1561921-3-liupeng256@huawei.com
Fixes: 5f3e062089 ("kunit: test: add support for test abort")
Signed-off-by: Peng Liu <liupeng256@huawei.com>
Reviewed-by: Marco Elver <elver@google.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Tested-by: Brendan Higgins <brendanhiggins@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Wang Kefeng <wangkefeng.wang@huawei.com>
Cc: David Gow <davidgow@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Patch series "kunit: fix a UAF bug and do some optimization", v2.
This series is to fix UAF (use after free) when running kfence test case
test_gfpzero, which is time costly. This UAF bug can be easily triggered
by setting CONFIG_KFENCE_NUM_OBJECTS = 65535. Furthermore, some
optimization for kunit tests has been done.
This patch (of 3):
Kunit will create a new thread to run an actual test case, and the main
process will wait for the completion of the actual test thread until
overtime. The variable "struct kunit test" has local property in function
kunit_try_catch_run, and will be used in the test case thread. Task
kunit_try_catch_run will free "struct kunit test" when kunit runs
overtime, but the actual test case is still run and an UAF bug will be
triggered.
The above problem has been both observed in a physical machine and qemu
platform when running kfence kunit tests. The problem can be triggered
when setting CONFIG_KFENCE_NUM_OBJECTS = 65535. Under this setting, the
test case test_gfpzero will cost hours and kunit will run to overtime.
The follows show the panic log.
BUG: unable to handle page fault for address: ffffffff82d882e9
Call Trace:
kunit_log_append+0x58/0xd0
...
test_alloc.constprop.0.cold+0x6b/0x8a [kfence_test]
test_gfpzero.cold+0x61/0x8ab [kfence_test]
kunit_try_run_case+0x4c/0x70
kunit_generic_run_threadfn_adapter+0x11/0x20
kthread+0x166/0x190
ret_from_fork+0x22/0x30
Kernel panic - not syncing: Fatal exception
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
Ubuntu-1.8.2-1ubuntu1 04/01/2014
To solve this problem, the test case thread should be stopped when the
kunit frame runs overtime. The stop signal will send in function
kunit_try_catch_run, and test_gfpzero will handle it.
Link: https://lkml.kernel.org/r/20220309083753.1561921-1-liupeng256@huawei.com
Link: https://lkml.kernel.org/r/20220309083753.1561921-2-liupeng256@huawei.com
Signed-off-by: Peng Liu <liupeng256@huawei.com>
Reviewed-by: Marco Elver <elver@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Tested-by: Brendan Higgins <brendanhiggins@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Wang Kefeng <wangkefeng.wang@huawei.com>
Cc: Daniel Latypov <dlatypov@google.com>
Cc: David Gow <davidgow@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If the compiler doesn't optimize them away, each kunit assertion (use of
KUNIT_EXPECT_EQ, etc.) can use 88 bytes of stack space in the worst and
most common case. This has led to compiler warnings and a suggestion
from Linus to move data from the structs into static const's where
possible [1].
This builds upon [2] which did so for the base struct kunit_assert type.
That only reduced sizeof(struct kunit_binary_assert) from 88 to 64.
Given these are by far the most commonly used asserts, this patch
factors out the textual representations of the operands and comparator
into another static const, saving 16 more bytes.
In detail, KUNIT_EXPECT_EQ(test, 2 + 2, 5) yields the following struct
(struct kunit_binary_assert) {
.assert = <struct kunit_assert>,
.operation = "==",
.left_text = "2 + 2",
.left_value = 4,
.right_text = "5",
.right_value = 5,
}
After this change
static const struct kunit_binary_assert_text __text = {
.operation = "==",
.left_text = "2 + 2",
.right_text = "5",
};
(struct kunit_binary_assert) {
.assert = <struct kunit_assert>,
.text = &__text,
.left_value = 4,
.right_value = 5,
}
This also DRYs the code a bit more since these str fields were repeated
for the string and pointer versions of kunit_binary_assert.
Note: we could name the kunit_binary_assert_text fields left/right
instead of left_text/right_text. But that would require changing the
macros a bit since they have args called "left" and "right" which would
be substituted in `.left = #left` as `.2 + 2 = \"2 + 2\"`.
[1] https://groups.google.com/g/kunit-dev/c/i3fZXgvBrfA/m/VULQg1z6BAAJ
[2] https://lore.kernel.org/linux-kselftest/20220113165931.451305-6-dlatypov@google.com/
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The concern is that having a lot of redundant fields in kunit_assert can
blow up stack usage if the compiler doesn't optimize them away [1].
The comment on this field implies that it was meant to be initialized
when the expect/assert was declared, but this only happens when we run
kunit_do_failed_assertion().
We don't need to access it outside of that function, so move it out of
the struct and make it a local variable there.
This change also takes the chance to reduce the number of macros by
inlining the now simplified KUNIT_INIT_ASSERT_STRUCT() macro.
[1] https://groups.google.com/g/kunit-dev/c/i3fZXgvBrfA/m/VULQg1z6BAAJ
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
This is per Linus's suggestion in [1].
The issue there is that every KUNIT_EXPECT/KUNIT_ASSERT puts a
kunit_assert object onto the stack. Normally we rely on compilers to
elide this, but when that doesn't work out, this blows up the stack
usage of kunit test functions.
We can move some data off the stack by making it static.
This change introduces a new `struct kunit_loc` to hold the file and
line number and then just passing assert_type (EXPECT or ASSERT) as an
argument.
In [1], it was suggested to also move out the format string as well, but
users could theoretically craft a format string at runtime, so we can't.
This change leaves a copy of `assert_type` in kunit_assert for now
because cleaning up all the macros to not pass it around is a bit more
involved.
Here's an example of the expanded code for KUNIT_FAIL():
if (__builtin_expect(!!(!(false)), 0)) {
static const struct kunit_loc loc = { .file = ... };
struct kunit_fail_assert __assertion = { .assert = { .type ... };
kunit_do_failed_assertion(test, &loc, KUNIT_EXPECTATION, &__assertion.assert, ...);
};
[1] https://groups.google.com/g/kunit-dev/c/i3fZXgvBrfA/m/VULQg1z6BAAJ
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
We call this function first thing for all the assertion `format()`
functions.
This is the part that prints the file and line number and assertion type
(EXPECTATION, ASSERTION).
Having it as part of the format functions lets us have the flexibility
to not print that information (or print it differently) for new
assertion types, but I think this we don't need that.
And in the future, we'd like to consider factoring that data (file,
line#, type) out of the kunit_assert struct and into a `static`
variable, as Linus suggested [1], so we'd need to extract it anyways.
[1] https://groups.google.com/g/kunit-dev/c/i3fZXgvBrfA/m/VULQg1z6BAAJ
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Currently the code always calls kunit_do_assertion() even though it does
nothing when `pass` is true.
This change moves the `if(!(pass))` check into the macro instead
and renames the function to kunit_do_failed_assertion().
I feel this a bit easier to read and understand.
This has the potential upside of avoiding a function call that does
nothing most of the time (assuming your tests are passing) but comes
with the downside of generating a bit more code and branches. We try to
mitigate the branches by tagging them with `unlikely()`.
This also means we don't have to initialize structs that we don't need,
which will become a tiny bit more expensive if we switch over to using
static variables to try and reduce stack usage. (There's runtime code
to check if the variable has been initialized yet or not).
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Currently, these macros are only really documented near the bottom of
https://www.kernel.org/doc/html/latest/dev-tools/kunit/api/test.html#c.KUNIT_FAIL.
E.g. it's likely someone might just not realize that
KUNIT_EXPECT_STREQ() exists and instead use KUNIT_EXPECT_FALSE(strcmp())
or similar.
This can also serve as a basic smoketest that the KUnit assert machinery
still works for all the macros.
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Pull signal/exit/ptrace updates from Eric Biederman:
"This set of changes deletes some dead code, makes a lot of cleanups
which hopefully make the code easier to follow, and fixes bugs found
along the way.
The end-game which I have not yet reached yet is for fatal signals
that generate coredumps to be short-circuit deliverable from
complete_signal, for force_siginfo_to_task not to require changing
userspace configured signal delivery state, and for the ptrace stops
to always happen in locations where we can guarantee on all
architectures that the all of the registers are saved and available on
the stack.
Removal of profile_task_ext, profile_munmap, and profile_handoff_task
are the big successes for dead code removal this round.
A bunch of small bug fixes are included, as most of the issues
reported were small enough that they would not affect bisection so I
simply added the fixes and did not fold the fixes into the changes
they were fixing.
There was a bug that broke coredumps piped to systemd-coredump. I
dropped the change that caused that bug and replaced it entirely with
something much more restrained. Unfortunately that required some
rebasing.
Some successes after this set of changes: There are few enough calls
to do_exit to audit in a reasonable amount of time. The lifetime of
struct kthread now matches the lifetime of struct task, and the
pointer to struct kthread is no longer stored in set_child_tid. The
flag SIGNAL_GROUP_COREDUMP is removed. The field group_exit_task is
removed. Issues where task->exit_code was examined with
signal->group_exit_code should been examined were fixed.
There are several loosely related changes included because I am
cleaning up and if I don't include them they will probably get lost.
The original postings of these changes can be found at:
https://lkml.kernel.org/r/87a6ha4zsd.fsf@email.froward.int.ebiederm.orghttps://lkml.kernel.org/r/87bl1kunjj.fsf@email.froward.int.ebiederm.orghttps://lkml.kernel.org/r/87r19opkx1.fsf_-_@email.froward.int.ebiederm.org
I trimmed back the last set of changes to only the obviously correct
once. Simply because there was less time for review than I had hoped"
* 'signal-for-v5.17' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace: (44 commits)
ptrace/m68k: Stop open coding ptrace_report_syscall
ptrace: Remove unused regs argument from ptrace_report_syscall
ptrace: Remove second setting of PT_SEIZED in ptrace_attach
taskstats: Cleanup the use of task->exit_code
exit: Use the correct exit_code in /proc/<pid>/stat
exit: Fix the exit_code for wait_task_zombie
exit: Coredumps reach do_group_exit
exit: Remove profile_handoff_task
exit: Remove profile_task_exit & profile_munmap
signal: clean up kernel-doc comments
signal: Remove the helper signal_group_exit
signal: Rename group_exit_task group_exec_task
coredump: Stop setting signal->group_exit_task
signal: Remove SIGNAL_GROUP_COREDUMP
signal: During coredumps set SIGNAL_GROUP_EXIT in zap_process
signal: Make coredump handling explicit in complete_signal
signal: Have prepare_signal detect coredumps using signal->core_state
signal: Have the oom killer detect coredumps using signal->core_state
exit: Move force_uaccess back into do_exit
exit: Guarantee make_task_dead leaks the tsk when calling do_task_exit
...
Currently, the results for individial parameters in a parameterised test
are simply output as (K)TAP diagnostic lines.
As kunit_tool now supports nested subtests, report each parameter as its
own subtest.
For example, here's what the output now looks like:
# Subtest: inode_test_xtimestamp_decoding
ok 1 - 1901-12-13 Lower bound of 32bit < 0 timestamp, no extra bits
ok 2 - 1969-12-31 Upper bound of 32bit < 0 timestamp, no extra bits
ok 3 - 1970-01-01 Lower bound of 32bit >=0 timestamp, no extra bits
ok 4 - 2038-01-19 Upper bound of 32bit >=0 timestamp, no extra bits
ok 5 - 2038-01-19 Lower bound of 32bit <0 timestamp, lo extra sec bit on
ok 6 - 2106-02-07 Upper bound of 32bit <0 timestamp, lo extra sec bit on
ok 7 - 2106-02-07 Lower bound of 32bit >=0 timestamp, lo extra sec bit on
ok 8 - 2174-02-25 Upper bound of 32bit >=0 timestamp, lo extra sec bit on
ok 9 - 2174-02-25 Lower bound of 32bit <0 timestamp, hi extra sec bit on
ok 10 - 2242-03-16 Upper bound of 32bit <0 timestamp, hi extra sec bit on
ok 11 - 2242-03-16 Lower bound of 32bit >=0 timestamp, hi extra sec bit on
ok 12 - 2310-04-04 Upper bound of 32bit >=0 timestamp, hi extra sec bit on
ok 13 - 2310-04-04 Upper bound of 32bit>=0 timestamp, hi extra sec bit 1. 1 ns
ok 14 - 2378-04-22 Lower bound of 32bit>= timestamp. Extra sec bits 1. Max ns
ok 15 - 2378-04-22 Lower bound of 32bit >=0 timestamp. All extra sec bits on
ok 16 - 2446-05-10 Upper bound of 32bit >=0 timestamp. All extra sec bits on
# inode_test_xtimestamp_decoding: pass:16 fail:0 skip:0 total:16
ok 1 - inode_test_xtimestamp_decoding
Signed-off-by: David Gow <davidgow@google.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
It's possible that a parameterised test could end up with zero
parameters. At the moment, the test function will nevertheless be called
with NULL as the parameter. Instead, don't try to run the test code, and
just mark the test as SKIPped.
Reported-by: Daniel Latypov <dlatypov@google.com>
Signed-off-by: David Gow <davidgow@google.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Update complete_and_exit to call kthread_exit instead of do_exit.
Change the name to reflect this change in functionality. All of the
users of complete_and_exit are causing the current kthread to exit so
this change makes it clear what is happening.
Move the implementation of kthread_complete_and_exit from
kernel/exit.c to to kernel/kthread.c. As this function is kthread
specific it makes most sense to live with the kthread functions.
There are no functional change.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
There are some KUnit tests (KFENCE, Thunderbolt) which, for various
reasons, do not use the kunit_test_suite() macro and end up running
before the KUnit executor runs its tests. This means that their results
are printed separately, and they aren't included in the suite count used
by the executor.
This causes the executor output to be invalid TAP, however, as the suite
numbers used are no-longer 1-based, and don't match the test plan.
kunit_tool, therefore, prints a large number of warnings.
While it'd be nice to fix the tests to run in the executor, in the
meantime, reset the suite counter to 1 in __kunit_test_suites_exit.
Not only does this fix the executor, it means that if there are multiple
calls to __kunit_test_suites_init() across different tests, they'll each
get their own numbering.
kunit_tool likes this better: even if it's lacking the results for those
tests which don't use the executor (due to the lack of TAP header), the
output for the other tests is valid.
Signed-off-by: David Gow <davidgow@google.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Context:
It's difficult to map a given .kunitconfig => set of enabled tests.
Letting kunit.py figure that out would be useful.
This patch:
* is intended to be an implementation detail used only by kunit.py
* adds a kunit.action module param with one valid non-null value, "list"
* for the "list" action, it simply prints out "<suite>.<test>"
* leaves the kunit.py changes to make use of this for another patch.
Note: kunit.filter_glob is respected for this and all future actions.
Hack: we print a TAP header (but no test plan) to allow kunit.py to
use the same code to pick up KUnit output that it does for normal tests.
Since this is intended to be an implementation detail, it seems fine for
now. Maybe in the future we output each test as SKIPPED or the like.
Go with a more generic "action" param, since it seems like we might
eventually have more modes besides just running or listing tests, e.g.
* perhaps a benchmark mode that reruns test cases and reports timing
* perhaps a deflake mode that reruns test cases that failed
* perhaps a mode where we randomize test order to try and catch
hermeticity bugs like "test a only passes if run after test b"
Tested:
$ ./tools/testing/kunit/kunit.py run --kernel_arg=kunit.action=list --raw_output=kunit
...
TAP version 14
1..1
example.example_simple_test
example.example_skip_test
example.example_mark_skipped_test
reboot: System halted
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
When a user filters by a suite and not a test, e.g.
$ ./tools/testing/kunit/kunit.py run 'suite_name'
it hits this code
const int len = strlen(filter_glob);
...
parsed->suite_glob = kmalloc(len, GFP_KERNEL);
which fails to allocate space for the terminating NULL.
Somehow, it seems like we can't easily reproduce this under UML, so the
existing `parse_filter_test()` didn't catch this.
Fix this by allocating `len + 1` and switch to kzalloc() just to be a
bit more defensive. We're only going to run this code once per kernel
boot, and it should never be very long.
Also update the unit tests to be a bit more cautious.
This bug showed up as a NULL pointer dereference here:
> KUNIT_EXPECT_STREQ(test, (const char *)filtered.start[0][0]->name, "suite0");
`filtered.start[0][0]` was NULL, and `name` is at offset 0 in the struct,
so `...->name` was also NULL.
Fixes: 3b29021ddd10 ("kunit: tool: allow filtering test cases via glob")
Reported-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Acked-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Commit 1d71307a6f ("kunit: add unit test for filtering suites by
names") introduced the ability to filter which suites we run via glob.
This change extends it so we can also filter individual test cases
inside of suites as well.
This is quite useful when, e.g.
* trying to run just the tests cases you've just added or are working on
* trying to debug issues with test hermeticity
Examples:
$ ./tools/testing/kunit/kunit.py run --kunitconfig=lib/kunit '*exec*.parse*'
...
============================================================
======== [PASSED] kunit_executor_test ========
[PASSED] parse_filter_test
============================================================
Testing complete. 1 tests run. 0 failed. 0 crashed.
$ ./tools/testing/kunit/kunit.py run --kunitconfig=lib/kunit '*.no_matching_tests'
...
[ERROR] no tests run!
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
This test assumes that the declared kunit_suite object is the exact one
which is being executed, which KUnit will not guarantee [1].
Specifically, `suite->log` is not initialized until a suite object is
executed. So if KUnit makes a copy of the suite and runs that instead,
this test dereferences an invalid pointer and (hopefully) segfaults.
N.B. since we no longer assume this, we can no longer verify that
`suite->log` is *not* allocated during normal execution.
An alternative to this patch that would allow us to test that would
require exposing an API for the current test to get its current suite.
Exposing that for one internal kunit test seems like overkill, and
grants users more footguns (e.g. reusing a test case in multiple suites
and changing behavior based on the suite name, dynamically modifying the
setup/cleanup funcs, storing/reading stuff out of the suite->log, etc.).
[1] In a subsequent patch, KUnit will allow running subsets of test
cases within a suite by making a copy of the suite w/ the filtered test
list. But there are other reasons KUnit might execute a copy, e.g. if it
ever wants to support parallel execution of different suites, recovering
from errors and restarting suites
Signed-off-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The reference counting issue happens in the normal path of
kfree_at_end(). When kunit_alloc_and_get_resource() is invoked, the
function forgets to handle the returned resource object, whose refcount
increased inside, causing a refcount leak.
Fix this issue by calling kunit_alloc_resource() instead of
kunit_alloc_and_get_resource().
Fixed the following when applying:
Shuah Khan <skhan@linuxfoundation.org>
CHECK: Alignment should match open parenthesis
+ kunit_alloc_resource(test, NULL, kfree_res_free, GFP_KERNEL,
(void *)to_free);
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
YoungJun.park
Linus Torvalds
Rae Moar
David Gow
Maíra Canal
Mark Rutland
Dan Carpenter
Daniel Latypov
Joe Fradley
Jeremy Kerr
Miguel Ojeda
Ricardo Ribalda
Peng Liu
Eric W. Biederman
Xiyu Yang