Re: [PATCH net-next v4 3/4] bpf: add helper bpf_perf_prog_read_value

[Daniel Borkmann]

On 09/19/2017 09:04 AM, Yonghong Song wrote:
[...]

  #ifdef CONFIG_CGROUP_PERF
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 2c68b9e..ba77022 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -590,6 +590,13 @@ union bpf_attr {
   *     @buf: buf to fill
   *     @buf_size: size of the buf
   *     Return: 0 on success or negative error code
+ *
+ * int bpf_perf_prog_read_value(ctx, buf, buf_size)
+ *     read perf prog attached perf event counter and enabled/running time
+ *     @ctx: pointer to ctx
+ *     @buf: buf to fill
+ *     @buf_size: size of the buf
+ *     Return : 0 on success or negative error code
   */
  #define __BPF_FUNC_MAPPER(FN)         \
        FN(unspec),                     \
@@ -647,6 +654,7 @@ union bpf_attr {
        FN(sk_redirect_map),            \
        FN(sock_map_update),            \
        FN(perf_event_read_value),              \
+       FN(perf_prog_read_value),               \

(Same here.)

  /* integer value in 'imm' field of BPF_CALL instruction selects which helper
   * function eBPF program intends to call
diff --git a/kernel/events/core.c b/kernel/events/core.c
index 2d5bbe5..d039086 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -8081,6 +8081,7 @@ static void bpf_overflow_handler(struct perf_event *event,
        struct bpf_perf_event_data_kern ctx = {
                .data = data,
                .regs = regs,
+               .event = event,
        };
        int ret = 0;

diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
index 39ce5d9..596b5c9 100644
--- a/kernel/trace/bpf_trace.c
+++ b/kernel/trace/bpf_trace.c
@@ -603,6 +603,18 @@ BPF_CALL_3(bpf_get_stackid_tp, void *, tp_buff, struct 
bpf_map *, map,
                               flags, 0, 0);
  }

+BPF_CALL_3(bpf_perf_prog_read_value_tp, void *, ctx, struct 
bpf_perf_event_value *,
+       buf, u32, size)

Nit: indent

+{
+       struct bpf_perf_event_data_kern *kctx = (struct 
bpf_perf_event_data_kern *)ctx;

Why having the arg as void * and have this detour instead of having
struct bpf_perf_event_data_kern * right in the helper signature as
argument?

+       if (size != sizeof(struct bpf_perf_event_value))

unlikely()

+               return -EINVAL;
+
+       return perf_event_read_local(kctx->event, &buf->counter, &buf->enabled,
+                                    &buf->running);
+}

bpf_perf_prog_read_value_proto_tp would go right underneath here,
and bpf_get_stackid_proto_tp below the previous helper above.

  static const struct bpf_func_proto bpf_get_stackid_proto_tp = {
        .func           = bpf_get_stackid_tp,
        .gpl_only       = true,
@@ -612,6 +624,15 @@ static const struct bpf_func_proto 
bpf_get_stackid_proto_tp = {
        .arg3_type      = ARG_ANYTHING,
  };

+static const struct bpf_func_proto bpf_perf_prog_read_value_proto_tp = {
+         .func           = bpf_perf_prog_read_value_tp,
+         .gpl_only       = true,
+         .ret_type       = RET_INTEGER,
+         .arg1_type      = ARG_PTR_TO_CTX,
+         .arg2_type      = ARG_PTR_TO_UNINIT_MEM,

Same on error path.

+         .arg3_type      = ARG_CONST_SIZE,
+};
+
  static const struct bpf_func_proto *tp_prog_func_proto(enum bpf_func_id 
func_id)
  {
        switch (func_id) {
@@ -619,6 +640,8 @@ static const struct bpf_func_proto *tp_prog_func_proto(enum 
bpf_func_id func_id)
                return &bpf_perf_event_output_proto_tp;
        case BPF_FUNC_get_stackid:
                return &bpf_get_stackid_proto_tp;
+       case BPF_FUNC_perf_prog_read_value:
+               return &bpf_perf_prog_read_value_proto_tp;
        default:
                return tracing_func_proto(func_id);
        }