| Message ID | 20240726023241.17295-1-zhujun2@cmss.chinamobile.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [v5] tools/testing:Fix the wrong format specifier | expand |
> The format specifier in fprintf is "%u", that "%u" should use > unsigned int type instead. * Please improve the change description with imperative wordings. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.10#n94 * Would you like to add any tags (like “Fixes” and “Cc”) accordingly? https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.10#n145 * How good does the summary phrase fit to your proposal? > the problem is discovered by reading code. How do you think about to detect such change possibilities by more systematic source code analyses? Regards, Markus
On 7/25/24 20:32, Zhu Jun wrote: > The format specifier in fprintf is "%u", that "%u" should use > unsigned int type instead.the problem is discovered by reading code. > > Signed-off-by: Zhu Jun <zhujun2@cmss.chinamobile.com> > --- > v1->v2: > modify commit info add how to find the problem in the log > v2->v3: > Seems this can use macro WTERMSIG like those above usage, rather than > changing the print format. > v3->v4: > Now the commit summary doesn't match the change you are making. > Also WTERMSIG() is incorrect for this conditional code path. > See comments below in the code path. > > I would leave the v2 code intact. How are you testing this change? > Please include the details in the change log. I am asking you to send the v2 change with updated change log to show how this change is tested. > v4->v5: > Compile the kernel for testing using make Running make isn't sufficient - please verify running the tests making sure there are no regressions. Since you are changing the harness, you want to make sure this change You are changing the harness. > > > tools/testing/selftests/kselftest_harness.h | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/tools/testing/selftests/kselftest_harness.h b/tools/testing/selftests/kselftest_harness.h > index e05ac8261046..675b8f43e148 100644 > --- a/tools/testing/selftests/kselftest_harness.h > +++ b/tools/testing/selftests/kselftest_harness.h > @@ -910,7 +910,7 @@ void __wait_for_test(struct __test_metadata *t) > .sa_flags = SA_SIGINFO, > }; > struct sigaction saved_action; > - int status; > + unsigned int status; I am asking you to send the v2 change with updated change log to show how this change is tested. fprintf(TH_LOG_STREAM, - "# %s: Test ended in some other way [%u]\n", + "# %s: Test ended in some other way [%d]\n", t->name, %d makes sense here in this message. thanks, -- Shuah
diff --git a/tools/testing/selftests/kselftest_harness.h b/tools/testing/selftests/kselftest_harness.h index e05ac8261046..675b8f43e148 100644 --- a/tools/testing/selftests/kselftest_harness.h +++ b/tools/testing/selftests/kselftest_harness.h @@ -910,7 +910,7 @@ void __wait_for_test(struct __test_metadata *t) .sa_flags = SA_SIGINFO, }; struct sigaction saved_action; - int status; + unsigned int status; if (sigaction(SIGALRM, &action, &saved_action)) { t->passed = 0;
The format specifier in fprintf is "%u", that "%u" should use unsigned int type instead.the problem is discovered by reading code. Signed-off-by: Zhu Jun <zhujun2@cmss.chinamobile.com> --- v1->v2: modify commit info add how to find the problem in the log v2->v3: Seems this can use macro WTERMSIG like those above usage, rather than changing the print format. v3->v4: Now the commit summary doesn't match the change you are making. Also WTERMSIG() is incorrect for this conditional code path. See comments below in the code path. I would leave the v2 code intact. How are you testing this change? Please include the details in the change log. v4->v5: Compile the kernel for testing using make tools/testing/selftests/kselftest_harness.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)