Pu Lehui <[email protected]> writes: > On 2026/7/23 13:41, Feng Jiang wrote: >> Enable verifier_may_goto_1 (raw instruction tests), stream_cond_break >> (250ms timeout path), and the may_goto_interaction fastcall test on >> riscv64 now that the JIT supports timed may_goto. >> >> Reviewed-by: Pu Lehui <[email protected]> >> Signed-off-by: Feng Jiang <[email protected]> >> --- >> tools/testing/selftests/bpf/progs/stream.c | 1 + >> tools/testing/selftests/bpf/progs/verifier_bpf_fastcall.c | 3 ++- >> tools/testing/selftests/bpf/progs/verifier_may_goto_1.c | 4 ++++ >> 3 files changed, 7 insertions(+), 1 deletion(-) >> >> diff --git a/tools/testing/selftests/bpf/progs/stream.c >> b/tools/testing/selftests/bpf/progs/stream.c >> index 92ba1d72e0ec..8d8e53d37266 100644 >> --- a/tools/testing/selftests/bpf/progs/stream.c >> +++ b/tools/testing/selftests/bpf/progs/stream.c >> @@ -64,6 +64,7 @@ SEC("syscall") >> __arch_x86_64 >> __arch_arm64 >> __arch_s390x >> +__arch_riscv64 >> __success __retval(0) >> __stderr("ERROR: Timeout detected for may_goto instruction") > > > This test fails because it relies on arch_bpf_stack_walk to fetch the > prog. It's fine to leave it on though, since support is on the way. > > https://lore.kernel.org/bpf/[email protected]/
I'm inclined to do the same here -- but only since it's RISC-V, where the tests are not run by any automation yet. In general I *don't* like that a series introduces a failing selftest merely because another unmerged series is expected to fix it later. Björn
