Re: [PATCH] Remove kretprobe_trampoline_holder.
From: Michael Ellerman <mpe@ellerman.id.au>
Date: 2016-03-30 00:09:04
On Tue, 2016-03-29 at 15:34 -0300, Thiago Jung Bauermann wrote:
Am Dienstag, 29 März 2016, 10:45:57 schrieb Michael Ellerman:quoted
On Mon, 2016-03-28 at 17:29 -0300, Thiago Jung Bauermann wrote:quoted
If I do s/_do_fork/._do_fork/ in kprobe_ftrace.tc then all ftrace kprobe tests pass:OK. We fixed that in 'perf probe', but not if you're using the sysfs file directly. Do you want to write a patch for ftracetest to try and handle it? I guess you'd try "_do_fork" and if that fails then try "._do_fork", and maybe only if uname -m says you're running on ppc64?I did write a patch yesterday (included below for reference), but then I noticed that the other ftrace tests use _do_fork and they work fine (I guess because of the fix you mentioned). I think that ideally the ftrace filter mechanism should work with dot symbols as well as regular symbols. I think this could work by creating a mechanism analogous to the ARCH_HAS_SYSCALL_MATCH_SYM_NAME one in trace_syscalls.c. Ftrace_match_record could call it to adjust the symbol name (like kprobe_lookup_name) before calling ftrace_match. But I’m wondering if it’s really worth the effort and maybe patching the testcase is enough? Also, I don’t know whether my idea would have any side effects.
It'd be nice if it worked properly. Reusing kprobe_lookup_name() looks like it would be the right fix, given this is "kprobe events". cheers