Hi Mark,
On Mon, May 05, 2025 at 07:50:31PM +0900, Mark Brown wrote:
On Fri, Apr 11, 2025 at 11:00:36AM +0200, Thomas Weißschuh wrote:
Usage of longjmp() was added to ensure that teardown is always run in commit 63e6b2a42342 ("selftests/harness: Run TEARDOWN for ASSERT failures") However instead of calling longjmp() to the teardown handler it is easier to just call the teardown handler directly from __bail(). Any potential duplicate teardown invocations are harmless as the actual handler will only ever be executed once since commit fff37bd32c76 ("selftests/harness: Fix fixture teardown").
I didn't fully verify (I'm on holiday right now) but it looks like this, which is in -next as 8a37733a874f:
@@ -757,7 +749,7 @@ */ #define OPTIONAL_HANDLER(_assert) \ for (; _metadata->trigger; _metadata->trigger = \
__bail(_assert, _metadata))
__bail(_assert, _metadata, self, variant))#define is_signed_type(var) (!!(((__typeof__(var))(-1)) < (__typeof__(var))1))
breaks the build of several selftests when building for arm64:
Thanks for the report. These tests are doing funny things with the harness internals. I'll drop these patches from -next.
In file included from seccomp_bpf.c:56: seccomp_bpf.c: In function ‘kill_thread_or_group’: ../kselftest_harness.h:755:52: error: ‘self’ undeclared (first use in this function) 755 | __bail(_assert, _metadata, self, variant)) | ^~~~ ../kselftest_harness.h:803:14: note: in expansion of macro ‘OPTIONAL_HANDLER’ 803 | } while (0); OPTIONAL_HANDLER(_assert) | ^~~~~~~~~~~~~~~~ ../kselftest_harness.h:516:9: note: in expansion of macro ‘__EXPECT’ 516 | __EXPECT(expected, #expected, seen, #seen, ==, 1) | ^~~~~~~~ seccomp_bpf.c:844:9: note: in expansion of macro ‘ASSERT_EQ’ 844 | ASSERT_EQ(0, prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) { | ^~~~~~~~~
Full build log at:
https://builds.sirena.org.uk/407f60a151df3c44397e5afc0111eb9b026c38d3/arm64/defconfig/build.log