- 08 May, 2024 8 commits
-
-
Edward Liaw authored
Add KHDR_INCLUDES to the CFLAGS to pull in the kselftest harness dependencies (-D_GNU_SOURCE). Also, remove redefinitions of _GNU_SOURCE in the source code. Fixes: 80921623 ("selftests/harness: remove use of LINE_MAX") Reported-by:
kernel test robot <oliver.sang@intel.com> Closes: https://lore.kernel.org/oe-lkp/202404301040.3bea5782-oliver.sang@intel.comSigned-off-by:
Edward Liaw <edliaw@google.com> Reviewed-by:
Muhammad Usama Anjum <usama.anjum@collabora.com> Acked-by:
Dave Hansen <dave.hansen@linux.intel.com> Reviewed-by:
Jarkko Sakkinen <jarkko@kernel.org> Tested-by:
John Hubbard <jhubbard@nvidia.com> Signed-off-by:
Shuah Khan <skhan@linuxfoundation.org>
-
Edward Liaw authored
Add the -D_GNU_SOURCE flag to KHDR_INCLUDES so that it is defined in a central location. Commit 80921623 ("selftests/harness: remove use of LINE_MAX") introduced asprintf into kselftest_harness.h, which is a GNU extension and needs _GNU_SOURCE to either be defined prior to including headers or with the -D_GNU_SOURCE flag passed to the compiler. Fixed up commit log: Shuah Khan <skhan@linuxfoundation.org> Fixes: 80921623 ("selftests/harness: remove use of LINE_MAX") Reported-by:
Mark Brown <broonie@kernel.org> Reviewed-by:
-
John Hubbard authored
When building with clang, via: make LLVM=1 -C tools/testing/selftests ...two types of warnings occur: warning: absolute value function 'abs' given an argument of type 'long' but has parameter of type 'int' which may cause truncation of value warning: taking the absolute value of unsigned type 'unsigned long' has no effect Fix these by: a) using labs() in place of abs(), when long integers are involved, and b) Change to use signed integer data types, in places where subtraction is used (and could end up with negative values). c) Remove a duplicate abs() call in cmt_test.c. Cc: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Reviewed-by:Reinette Chatre <reinette.chatre@intel.com> Signed-off-by:
-
Masami Hiramatsu (Google) authored
Fix the below checkbashisms errors. Because of these errors, these tests will fail on dash shell. possible bashism in test.d/kprobe/kretprobe_entry_arg.tc line 14 ('function' is useless): function streq() { possible bashism in test.d/dynevent/fprobe_entry_arg.tc line 14 ('function' is useless): function streq() { Fixes: f6e2253a ("selftests/ftrace: Add test cases for entry args at function exit") Cc: stable@vger.kernel.org Signed-off-by:Masami Hiramatsu (Google) <mhiramat@kernel.org> Signed-off-by:
-
Masami Hiramatsu (Google) authored
Since the dynevent/add_remove_btfarg.tc test case forgets to ensure that fprobe is enabled for some structure field access tests which uses the fprobe, it fails if CONFIG_FPROBE=n or CONFIG_FPROBE_EVENTS=n. Fixes it to ensure the fprobe events are supported. Fixes: d892d3d3 ("selftests/ftrace: Add BTF fields access testcases") Cc: stable@vger.kernel.org Signed-off-by:
-
Amer Al Shanawany authored
Fix the following warnings by adding return check and error handling. test_execve.c: In function ‘do_tests’: test_execve.c:100:17: warning: ignoring return value of ‘capng_get_caps_process’ declared with attribute ‘warn_unused_result’ [-Wunused-result] 100 | capng_get_caps_process(); | ^~~~~~~~~~~~~~~~~~~~~~~~ validate_cap.c: In function ‘main’: validate_cap.c:47:9: warning: ignoring return value of ‘capng_get_caps_process’ declared with attribute ‘warn_unused_result’ [-Wunused-result] 47 | capng_get_caps_process(); | ^~~~~~~~~~~~~~~~~~~~~~~~ Signed-off-by:Amer Al Shanawany <amer.shanawany@gmail.com> Reviewed-by:
-
Amer Al Shanawany authored
fix compiler warning and errors when compiling statmount test. gcc 12.3 (Ubuntu 12.3.0-1ubuntu1~22.04) statmount_test.c:572:24: warning: implicit declaration of function ‘offsetof’ [-Wimplicit-function-declaration] 572 | #define str_off(memb) (offsetof(struct statmount, memb) / sizeof(uint32_t)) | ^~~~~~~~ statmount_test.c:598:51: note: in expansion of macro ‘str_off’ 598 | test_statmount_string(STATMOUNT_MNT_ROOT, str_off(mnt_root), "mount root"); | ^~~~~~~ statmount_test.c:18:1: note: ‘offsetof’ is defined in header ‘<stddef.h>’; did you forget to ‘#include <stddef.h>’? 17 | #include "../../kselftest.h" +++ |+#include <stddef.h> Signed-off-by: -
Lu Dai authored
In the function l5_test(), variable $tests is empty when there is no .mk file in the subsystem to be tested. It causes the following grep operation get stuck. This fix check the variable $tests, return when it is empty. Signed-off-by:
Lu Dai <dai.lu@exordes.com> Signed-off-by:
-
- 06 May, 2024 32 commits
-
-
Valentin Obst authored
Align the behavior for gcc and clang builds by interpreting unset `ARCH` and `CROSS_COMPILE` variables in `LLVM` builds as a sign that the user wants to build for the host architecture. This patch preserves the properties that setting the `ARCH` variable to an unknown value will trigger an error that complains about insufficient information, and that a set `CROSS_COMPILE` variable will override the target triple that is determined based on presence/absence of `ARCH`. When compiling with clang, i.e., `LLVM` is set, an unset `ARCH` variable in combination with an unset `CROSS_COMPILE` variable, i.e., compiling for the host architecture, leads to compilation failures since `lib.mk` can not determine the clang target triple. In this case, the following error message is displayed for each subsystem that does not set `ARCH` in its own Makefile before including `lib.mk` (lines wrapped at 75 chrs): make[1]: Entering directory '/mnt/build/linux/tools/testing/selftests/ sysctl' ../lib.mk:33: *** Specify CROSS_COMPILE or add '--target=' option to lib.mk. Stop. make[1]: Leaving directory '/mnt/build/linux/tools/testing/selftests/ sysctl' In the same scenario a gcc build would default to the host architecture, i.e., it would use plain `gcc`. Fixes: 795285ef ("selftests: Fix clang cross compilation") Reviewed-by:
Valentin Obst <kernel@valentinobst.de> Reviewed-by:
-
John Hubbard authored
First of all, in order to build with clang at all, one must first apply Valentin Obst's build fix for LLVM [1]. Once that is done, then when building with clang, via: make LLVM=1 -C tools/testing/selftests ...the following error occurs: clang: error: cannot specify -o when generating multiple output files This is because clang, unlike gcc, won't accept invocations of this form: clang file1.c header2.h Fix this by using selftests/lib.mk facilities for tracking local header file dependencies: add them to LOCAL_HDRS, leaving only the .c files to be passed to the compiler. [1] https://lore.kernel.org/all/20240329-selftests-libmk-llvm-rfc-v1-1-2f9ed7d1c49f@valentinobst.de/ Fixes: 8e289f45 ("selftests/resctrl: Add resctrl.h into build deps") Cc: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Signed-off-by: -
John Hubbard authored
First of all, in order to build with clang at all, one must first apply Valentin Obst's build fix for LLVM [1]. Once that is done, then when building with clang, via: make LLVM=1 -C tools/testing/selftests ...the following error occurs: clang: error: cannot specify -o when generating multiple output files This is because clang, unlike gcc, won't accept invocations of this form: clang file1.c header2.h While trying to fix this, I noticed that: a) selftests/lib.mk already avoids the problem, and b) The binderfs Makefile indavertently bypasses the selftests/lib.mk build system, and quitely uses Make's implicit build rules for .c files instead. The Makefile attempts to set up both a dependency and a source file, neither of which was needed, because lib.mk is able to automatically handle both. This line: binderfs_test: binderfs_test.c ...causes Make's implicit rules to run, which builds binderfs_test without ever looking at lib.mk. Fix this by simply deleting the "binderfs_test:" Makefile target and letting lib.mk handle it instead. [1] https://lore.kernel.org/all/20240329-selftests-libmk-llvm-rfc-v1-1-2f9ed7d1c49f@valentinobst.de/ Fixes: 6e29225a ("binderfs: port tests to test harness infrastructure") Cc: Christian Brauner <brauner@kernel.org> Signed-off-by:Christian Brauner <brauner@kernel.org> Signed-off-by:
-
Yo-Jung (Leo) Lin authored
Add extra colon to mark command in the next paragraph as codeblock Signed-off-by:
Yo-Jung (Leo) Lin <0xff07@gmail.com> Signed-off-by:
-
Nathan Chancellor authored
Commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn") marked functions that call exit() as __noreturn but it did not change the return type of these functions from 'void' to 'int' like it should have (since a noreturn function by definition cannot return an integer because it does not return...) because there were many tests that return the result of the ksft_exit functions, even though it has never been used due to calling exit(). Now that all uses of 'return ksft_exit...()' have been cleaned up properly, change the types of the ksft_exit...() functions to void to match their __noreturn nature. Reviewed-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
Nathan Chancellor <nathan@kernel.org> Signed-off-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the call to ksft_exit_pass(), as __noreturn prevents the compiler from warning that a caller of ksft_exit_pass() does not return a value because the program will terminate upon calling these functions. Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the calls to ksft_exit_...(), as __noreturn prevents the compiler from warning that a caller of the ksft_exit functions does not return a value because the program will terminate upon calling these functions. Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the call to ksft_exit_pass(), as __noreturn prevents the compiler from warning that a caller of ksft_exit_pass() does not return a value because the program will terminate upon calling these functions (which is what the comment alluded to as well). Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the calls to ksft_exit_skip(), as __noreturn prevents the compiler from warning that a caller of ksft_exit_skip() does not return a value because the program will terminate upon calling these functions. Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the calls to ksft_exit_{pass,fail}(), as __noreturn prevents the compiler from warning that a caller of the ksft_exit functions does not return a value because the program will terminate upon calling these functions. Just removing 'return' would have resulted in !ret ? ksft_exit_pass() : ksft_exit_fail(); so convert that into the more idiomatic if (ret) ksft_exit_fail(); ksft_exit_pass(); Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the calls to ksft_exit_...(), as __noreturn prevents the compiler from warning that a caller of the ksft_exit functions does not return a value because the program will terminate upon calling these functions. Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the calls to ksft_exit_pass(), as __noreturn prevents the compiler from warning that a caller of ksft_exit_pass() does not return a value because the program will terminate upon calling these functions. Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the calls to ksft_exit_...(), as __noreturn prevents the compiler from warning that a caller of the ksft_exit functions does not return a value because the program will terminate upon calling these functions. Reviewed-by:
-
Nathan Chancellor authored
After commit f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn"), ksft_exit_...() functions are marked as __noreturn, which means the return type should not be 'int' but 'void' because they are not returning anything (and never were since exit() has always been called). To facilitate updating the return type of these functions, remove 'return' before the calls to ksft_exit_{pass,fail}(), as __noreturn prevents the compiler from warning that a caller of the ksft_exit functions does not return a value because the program will terminate upon calling these functions. Just removing 'return' would have resulted in !ret ? ksft_exit_pass() : ksft_exit_fail(); so convert that into the more idiomatic if (ret) ksft_exit_fail(); ksft_exit_pass(); Reviewed-by:
-
Nícolas F. R. A. Prado authored
There is one use of bash specific syntax in the script. Change it to the equivalent POSIX syntax. This doesn't change functionality and allows the test to be run on shells other than bash. Reported-by:
Mike Looijmans <mike.looijmans@topic.nl> Closes: https://lore.kernel.org/all/efae4037-c22a-40be-8ba9-7c1c12ece042@topic.nl/ Fixes: 4a679c5a ("selftests: Add test to verify power supply properties") Signed-off-by:
Nícolas F. R. A. Prado <nfraprado@collabora.com> Reviewed-by:
-
Nícolas F. R. A. Prado authored
There are a couple uses of bash specific syntax in the script. Change them to the equivalent POSIX syntax. This doesn't change functionality and allows non-bash test scripts to make use of these helpers. Reported-by:
-
Muhammad Usama Anjum authored
This test outputs lots of information. Let's conform the core part of the test to TAP and leave the information printing messages for now. Include ktap_helpers.sh to print conformed logs. Use KSFT_* macros to return the correct exit code for the kselftest framework and CIs to understand the exit status. Signed-off-by:
Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by:
-
Muhammad Usama Anjum authored
Let the compilers (clang) know that this function would just call exit() and would never return. It is needed to avoid false positive static analysis errors. All similar functions calling exit() unconditionally have been marked as __noreturn. Signed-off-by:
-
Mark Brown authored
When logging an error from calling waitpid() on the child we print a misleading error message saying that the error we report was returned by the chilld. Fix this to say the error is from waitpid(). Applied after fixing merge conflict: Shuah Khan <skhan@linuxfoundation.org> Signed-off-by:
-
Mark Brown authored
When the child exits during the clone3() selftest we use WEXITSTATUS() to get the exit status from the process without first checking WIFEXITED() to see if the result will be valid. This can lead to incorrect results, for example if the child exits due to signal. Add a WIFEXTED() check and report any non-standard exit as a failure, using EXIT_FAILURE as the exit status for call_clone3() since we otherwise report 0 or negative errnos. Signed-off-by:
-
Mark Brown authored
Shuah reported a compiler warning with an Ubuntu GCC 13 build, I've been unable to reproduce it but hopefully this fixes the issue: clone3_set_tid.c:136:43: warning: format ‘%d’ expects argument of type ‘int’, but argument 3 has type ‘size_t’ {aka ‘long unsigned int’} [-Wformat=] Reported-by: -
Mark Brown authored
In order to facilitate debugging of issues from automated runs of the ftrace selftests turn on verbose logging by default when run from the kselftest runner. This is primarily used by automated systems where developers may not have direct access to the system so defaulting to providing diagnostic information which might help debug problems seems like a good idea. When tests pass no extra output is generated, when they fail a full log of the test run is provided. Since this really is rather verbose when there are a large number of test failures or output is slow (eg, with a serial console) this could substantially increase the run time for the tests which might present problems with timeout detection for affected systems, hopefully we keep the tests running well enough that this is not too much of an issue. Signed-off-by:
Steven Rostedt (Google) <rostedt@goodmis.org> Acked-by:
-
Mark Brown authored
When -v is specified ftracetest will dump logs of test execution to the console which if -K is also specified for KTAP output will result in output that is not properly KTAP formatted. All that's required for KTAP formatting is that anything we log have a '#' at the start of the line so we can improve things by washing the output through a simple read loop. This will help automated parsers when verbose mode is enabled. Signed-off-by:
-
Muhammad Usama Anjum authored
Use ksft_exit_fail_perror() to print the value of errno and its string form. This is the first user of the ksft_exit_fail_perror() and proves the usefulness of this API. Signed-off-by:
-
Muhammad Usama Anjum authored
Add a version of ksft_exit_fail_msg() which prints the errno and its string form with ease. There is no benefit of exit message without errno. Whenever some error occurs, instead of printing errno manually, this function would be very helpful. In the next TAP ports or new tests, this function will be used instead of ksft_exit_fail_msg() as it prints errno. Resolved merge conflict found in next between the following commits: f7d5bcd3 ("selftests: kselftest: Mark functions that unconditionally call exit() as __noreturn") f07041728422 ("selftests: add ksft_exit_fail_perror()") Reported-by:
Stephen Rothwell <sfr@canb.auug.org.au> Shuah Khan <skhan@linuxfoundation.org> Signed-off-by:
-
Muhammad Usama Anjum authored
The comment on top of the file is used by many developers to glance over all the available functions. Add the recently added ksft_perror() to it. Signed-off-by:
-
Mark Brown authored
The test results reported for the clone3_set_tid tests interact poorly with automation for running kselftest since the reported test names include TIDs dynamically allocated at runtime. A lot of automation for running kselftest will compare runs by looking at the test name to identify if the same test is being run so changing names make it look like the testsuite has been updated to include new tests. This makes the results display less clearly and breaks cases like bisection. Address this by providing a brief description of the tests and logging that along with the stable parameters for the test currently logged. The TIDs are already logged separately in existing logging except for the final test which has a new log message added. We also tweak the formatting of the logging of expected/actual values for clarity. There are still issues with the logging of skipped tests (many are simply not logged at all when skipped and all are logged with different names) but these are less disruptive since the skips are all based on not being run as root, a condition likely to be stable for a given test system. Acked-by:
-
Maciej Wieczor-Retman authored
Every test calls its cleanup function at the end of it's test function. After the cleanup function pointer is added to the test framework this can be simplified to executing the callback function at the end of the generic test running function. Make test cleanup functions static and call them from the end of run_single_test() from the resctrl_test's cleanup function pointer. Signed-off-by:
Maciej Wieczor-Retman <maciej.wieczor-retman@intel.com> Reviewed-by:
-
Maciej Wieczor-Retman authored
Ctrl-c handler isn't aware of what test is currently running. Because of that it executes all cleanups even if they aren't necessary. Since the ctrl-c handler uses the sa_sigaction system no parameters can be passed to it as function arguments. Add a global variable to make ctrl-c handler aware of the currently run test and only execute the correct cleanup callback. Signed-off-by:
-
Maciej Wieczor-Retman authored
Resctrl selftests use very similar functions to cleanup after themselves. This creates a lot of code duplication. Also not being hooked to the test framework means that ctrl-c handler isn't aware of what test is currently running and executes all cleanups even though only one is needed. Add a function pointer to the resctrl_test struct and attach to it cleanup functions from individual tests. Signed-off-by:
-
Muhammad Usama Anjum authored
Conform the layout, informational and status messages to TAP. No functional change is intended other than the layout of output messages. Improve the TAP messages as well. Reviewed-by:
T.J. Mercier <tjmercier@google.com> Signed-off-by:
-
Muhammad Usama Anjum authored
Conform the layout, informational and status messages to TAP. No functional change is intended other than the layout of output messages. Signed-off-by:
-
