On Wed, Jul 29, 2020 at 9:24 AM John Fastabend <john.fastab...@gmail.com> wrote:
>
[...]
>
> Signed-off-by: John Fastabend <john.fastab...@gmail.com>
Acked-by: Song Liu <songliubrav...@fb.com>
> ---
> .../testing/selftests/bpf/progs/test_tcpbpf_kern.c | 13 +++++++++++++
> 1 file changed, 13 insertions(+)
>
> diff --git a/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c
> b/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c
> index 1f1966e..f8b13682 100644
> --- a/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c
> +++ b/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c
> @@ -54,6 +54,7 @@ SEC("sockops")
> int bpf_testcb(struct bpf_sock_ops *skops)
> {
> char header[sizeof(struct ipv6hdr) + sizeof(struct tcphdr)];
> + struct bpf_sock_ops *reuse = skops;
> struct tcphdr *thdr;
> int good_call_rv = 0;
> int bad_call_rv = 0;
> @@ -62,6 +63,18 @@ int bpf_testcb(struct bpf_sock_ops *skops)
> int v = 0;
> int op;
>
> + /* Test reading fields in bpf_sock_ops using single register */
> + asm volatile (
> + "%[reuse] = *(u32 *)(%[reuse] +96)"
> + : [reuse] "+r"(reuse)
> + :);
> +
> + asm volatile (
> + "%[op] = *(u32 *)(%[skops] +96)"
> + : [op] "+r"(op)
> + : [skops] "r"(skops)
> + :);
> +
Shall we add a separate test for this? It does seem to fix in bpf_testcb().
> op = (int) skops->op;
>
> update_event_map(op);
>
