Re: [PATCH bpf-next 4/5] bpf: Add bpf_get_func_ip helper for kprobe programs

[Yonghong Song <yhs@fb.com>]

On 6/30/21 4:58 PM, Masami Hiramatsu wrote:
> On Wed, 30 Jun 2021 10:47:01 -0700
> Yonghong Song <yhs@fb.com> wrote:
> 
>>
>>
>> On 6/29/21 12:29 PM, Jiri Olsa wrote:
>>> Adding bpf_get_func_ip helper for BPF_PROG_TYPE_KPROBE programs,
>>> so it's now possible to call bpf_get_func_ip from both kprobe and
>>> kretprobe programs.
>>>
>>> Taking the caller's address from 'struct kprobe::addr', which is
>>> defined for both kprobe and kretprobe.
>>>
>>> Signed-off-by: Jiri Olsa <jolsa@kernel.org>
>>> ---
>>>    include/uapi/linux/bpf.h       |  2 +-
>>>    kernel/bpf/verifier.c          |  2 ++
>>>    kernel/trace/bpf_trace.c       | 14 ++++++++++++++
>>>    kernel/trace/trace_kprobe.c    | 20 ++++++++++++++++++--
>>>    kernel/trace/trace_probe.h     |  5 +++++
>>>    tools/include/uapi/linux/bpf.h |  2 +-
>>>    6 files changed, 41 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>>> index 83e87ffdbb6e..4894f99a1993 100644
>>> --- a/include/uapi/linux/bpf.h
>>> +++ b/include/uapi/linux/bpf.h
>>> @@ -4783,7 +4783,7 @@ union bpf_attr {
>>>     *
>>>     * u64 bpf_get_func_ip(void *ctx)
>>>     * 	Description
>>> - * 		Get address of the traced function (for tracing programs).
>>> + * 		Get address of the traced function (for tracing and kprobe programs).
>>>     * 	Return
>>>     * 		Address of the traced function.
>>>     */
>>> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
>>> index 701ff7384fa7..b66e0a7104f8 100644
>>> --- a/kernel/bpf/verifier.c
>>> +++ b/kernel/bpf/verifier.c
>>> @@ -5979,6 +5979,8 @@ static bool has_get_func_ip(struct bpf_verifier_env *env)
>>>    			return -ENOTSUPP;
>>>    		}
>>>    		return 0;
>>> +	} else if (type == BPF_PROG_TYPE_KPROBE) {
>>> +		return 0;
>>>    	}
>>>    
>>>    	verbose(env, "func %s#%d not supported for program type %d\n",
>>> diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
>>> index 9edd3b1a00ad..1a5bddce9abd 100644
>>> --- a/kernel/trace/bpf_trace.c
>>> +++ b/kernel/trace/bpf_trace.c
>>> @@ -961,6 +961,18 @@ static const struct bpf_func_proto bpf_get_func_ip_proto_tracing = {
>>>    	.arg1_type	= ARG_PTR_TO_CTX,
>>>    };
>>>    
>>> +BPF_CALL_1(bpf_get_func_ip_kprobe, struct pt_regs *, regs)
>>> +{
>>> +	return trace_current_kprobe_addr();
>>> +}
>>> +
>>> +static const struct bpf_func_proto bpf_get_func_ip_proto_kprobe = {
>>> +	.func		= bpf_get_func_ip_kprobe,
>>> +	.gpl_only	= true,
>>> +	.ret_type	= RET_INTEGER,
>>> +	.arg1_type	= ARG_PTR_TO_CTX,
>>> +};
>>> +
>>>    const struct bpf_func_proto *
>>>    bpf_tracing_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
>>>    {
>>> @@ -1092,6 +1104,8 @@ kprobe_prog_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
>>>    	case BPF_FUNC_override_return:
>>>    		return &bpf_override_return_proto;
>>>    #endif
>>> +	case BPF_FUNC_get_func_ip:
>>> +		return &bpf_get_func_ip_proto_kprobe;
>>>    	default:
>>>    		return bpf_tracing_func_proto(func_id, prog);
>>>    	}
>>> diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c
>>> index ea6178cb5e33..b07d5888db14 100644
>>> --- a/kernel/trace/trace_kprobe.c
>>> +++ b/kernel/trace/trace_kprobe.c
>>> @@ -1570,6 +1570,18 @@ static int kretprobe_event_define_fields(struct trace_event_call *event_call)
>>>    }
>>>    
>>>    #ifdef CONFIG_PERF_EVENTS
>>> +/* Used by bpf get_func_ip helper */
>>> +DEFINE_PER_CPU(u64, current_kprobe_addr) = 0;
>>
>> Didn't check other architectures. But this should work
>> for x86 where if nested kprobe happens, the second
>> kprobe will not call kprobe handlers.
> 
> No problem, other architecture also does not call nested kprobes handlers.
> However, you don't need this because you can use kprobe_running()
> in kprobe context.
> 
> kp = kprobe_running();
> if (kp)
> 	return kp->addr;
> 
> BTW, I'm not sure why don't you use instruction_pointer(regs)?

How about kretprobe? I guess kp->addr should still point to
function address but instruction_pointer(regs) does not?

> 
> Thank you,
> 
>>
>> This essentially is to provide an additional parameter to
>> bpf program. Andrii is developing a mechanism to
>> save arbitrary data in *current task_struct*, which
>> might be used here to save current_kprobe_addr, we can
>> save one per cpu variable.
>>
>>> +
>>> +u64 trace_current_kprobe_addr(void)
>>> +{
>>> +	return *this_cpu_ptr(&current_kprobe_addr);
>>> +}
>>> +
>>> +static void trace_current_kprobe_set(struct trace_kprobe *tk)
>>> +{
>>> +	__this_cpu_write(current_kprobe_addr, (u64) tk->rp.kp.addr);
>>> +}
>>>    
>>>    /* Kprobe profile handler */
>>>    static int
>>> @@ -1585,6 +1597,7 @@ kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs)
>>>    		unsigned long orig_ip = instruction_pointer(regs);
>>>    		int ret;
>>>    
>>> +		trace_current_kprobe_set(tk);
>>>    		ret = trace_call_bpf(call, regs);
>>>    
>>>    		/*
>>> @@ -1631,8 +1644,11 @@ kretprobe_perf_func(struct trace_kprobe *tk, struct kretprobe_instance *ri,
>>>    	int size, __size, dsize;
>>>    	int rctx;
>>>    
>>> -	if (bpf_prog_array_valid(call) && !trace_call_bpf(call, regs))
>>> -		return;
>>> +	if (bpf_prog_array_valid(call)) {
>>> +		trace_current_kprobe_set(tk);
>>> +		if (!trace_call_bpf(call, regs))
>>> +			return;
>>> +	}
>>>    
>>>    	head = this_cpu_ptr(call->perf_events);
>>>    	if (hlist_empty(head))
>> [...]
> 
>