On Wed, Feb 22, 2023 at 7:07 PM Alexei Starovoitov
<alexei.starovoitov@xxxxxxxxx> wrote:
>
> diff --git a/tools/lib/bpf/bpf_helpers.h b/tools/lib/bpf/bpf_helpers.h
> index 7d12d3e620cc..affc0997f937 100644
> --- a/tools/lib/bpf/bpf_helpers.h
> +++ b/tools/lib/bpf/bpf_helpers.h
> @@ -176,6 +176,7 @@ enum libbpf_tristate {
> #define __ksym __attribute__((section(".ksyms")))
> #define __kptr_untrusted __attribute__((btf_type_tag("kptr_untrusted")))
> #define __kptr __attribute__((btf_type_tag("kptr")))
> +#define __kptr_rcu __attribute__((btf_type_tag("kptr_rcu")))
Realized that the mechanism can work without requiring bpf prog
to use this new tag.
The kernel can determine whether __kptr is RCU or not
via rcu_protected_object().
So BPF_KPTR_RCU vs BPF_KPTR_REF will be kernel internal distinction.
Eventually all __kptr kernel objects will be RCU anyway.
I'll respin.
