dantezy

@jordalgo Hi, I have added semantic_analyser tests and fixed the ci.

Could I ask what platform you used to discover this problem? I reproduced this on X86 vm by `kill -19 pid`, but saw nothing when `kill -18 pid`. BTW I...

I'm working on x86 platform. ```c // include/uapi/asm/signal.h #define SIGCHLD 17 #define SIGCONT 18 #define SIGSTOP 19 ``` And ```c // include/linux/signal.h * | SIGCHLD | ignore | * |...

@danobi @jordalgo Are there any unit tests for the `path` helper?

@jordalgo thanks! I will do it after work.

```shell ==================== bpftrace --info: ==================== System OS: Linux 6.10.5-arch1-1 #1 SMP PREEMPT_DYNAMIC Thu, 15 Aug 2024 00:25:30 +0000 Arch: x86_64 Build version: v0.21.0-111-g9996 LLVM: 18.1.8 unsafe probe: no bfd: yes...

I think I found the reason why the test failed on my machine: Since kernel 021a160abf62 ("fs: use __fput_sync in close(2)") , the `close` syscall wouldn't invoke `flip_close` any more....

What version of kernel do you use? `bpf_setsockopt` was introduced since v4.13. See https://ebpf-docs.dylanreimerink.nl/linux/helper-function/bpf_setsockopt/

That's because when we `attach_kprobe`: ``` attach_kprobe(bcc/__init__.py)->bpf_attach_kprobe(src/cc/libbpf.c)->bpf_attach_probe(src/cc/libbpf.c)->bpf_try_perf_event_open_with_probe ``` And `bpf_try_perf_event_open_with_probe` won't write anything into `/sys/kernel/debug/tracing/%s_events`. In fact, `/sys/kernel/debug/tracing/%s_events` will be written by `create_probe_event` if `bpf_try_perf_event_open_with_probe` failed. This is an old...

Could you describe the env with more detail? I cannot reproduce this problem with: CPU: Intel i7-8700K 3.7GHz Mem: 32G Kernel version: 6.9.10-arch1-1 I ran `go test -exec sudo -run...