> diff --git a/tools/testing/selftests/bpf/prog_tests/exceptions.c
> b/tools/testing/selftests/bpf/prog_tests/exceptions.c
> index 84ab73e08b0e6..94a62fcd4543d 100644
> --- a/tools/testing/selftests/bpf/prog_tests/exceptions.c
> +++ b/tools/testing/selftests/bpf/prog_tests/exceptions.c
> @@ -128,7 +128,7 @@ static void test_exceptions_success(void)
>
> bpf_program__fd(skel->progs.exception_ext_mod_cb_runtime),
> "exception_cb_mod"), "set_attach_target"))
> goto done;
> - }), "FENTRY/FEXIT programs cannot attach to exception callback", 0);
> + }), "FENTRY/FEXIT/FSESSION programs cannot attach to exception
> callback", 0);
These four selftest locations now expect the string
"FENTRY/FEXIT/FSESSION programs cannot attach to exception callback",
but the verifier message they match against was not updated. In
bpf_check_attach_target() the format string still reads:
prog_extension ? "Extension" : "FENTRY/FEXIT");
so the kernel produces "FENTRY/FEXIT programs cannot attach to
exception callback". The RUN_EXT macro uses strstr() to match the
expected string against the kernel log buffer, and since
"FENTRY/FEXIT/FSESSION" is not a substring of "FENTRY/FEXIT programs",
strstr() returns NULL and the ASSERT_OK_PTR fails.
Should the verifier message also be updated to include FSESSION, to
match the selftest expectation? As it stands these four subtests will
fail.
> [ ... ]
---
AI reviewed your patch. Please fix the bug or email reply why it's not a bug.
See: https://github.com/kernel-patches/vmtest/blob/master/ci/claude/README.md
CI run summary: https://github.com/kernel-patches/bpf/actions/runs/24117471086
