libbpf: Fix unneeded extra initialization in bpf_map_batch_common
authorBrian Vazquez <brianvv@google.com>
Thu, 16 Jan 2020 04:59:18 +0000 (20:59 -0800)
committerDaniel Borkmann <daniel@iogearbox.net>
Thu, 16 Jan 2020 14:31:52 +0000 (15:31 +0100)
bpf_attr doesn't required to be declared with '= {}' as memset is used
in the code.

Fixes: 2ab3d86ea1859 ("libbpf: Add libbpf support to batch ops")
Reported-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Brian Vazquez <brianvv@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20200116045918.75597-1-brianvv@google.com
tools/lib/bpf/bpf.c

index 317727d6121494ecd4512254c3dde8b2ee96a3ba..ed42b006533c455463467414265d8ec19065c54d 100644 (file)
@@ -457,7 +457,7 @@ static int bpf_map_batch_common(int cmd, int fd, void  *in_batch,
                                __u32 *count,
                                const struct bpf_map_batch_opts *opts)
 {
-       union bpf_attr attr = {};
+       union bpf_attr attr;
        int ret;
 
        if (!OPTS_VALID(opts, bpf_map_batch_opts))