On Wed, Dec 20, 2023 at 03:31:26PM -0800, Andrii Nakryiko wrote:
> This limitation was the reason to add btf_decl_tag("arg:ctx"), making
> the actual argument type not important, so that user can just define
> "generic" signature:
>
> __noinline int global_subprog(void *ctx __arg_ctx) { ... }
Just realized that we probably need to enforce in both libbpf doing
rewrite and in the kernel that __arg_ctx is either valid
'struct correct_type_for_bpf_prog *' or 'void *'.
Otherwise the user will get surprising behavior when
int foo(struct __sk_buff *ctx __arg_ctx)
{
ctx->len;
}
will get rewritten to 'struct pt_regs *ctx' based on prog type
while all asm instructions inside prog were compiled with 'struct __sk_buff'
and CO_RE performs relocations against that type.
> +static struct {
> + enum bpf_prog_type prog_type;
> + const char *ctx_name;
> +} global_ctx_map[] = {
> + { BPF_PROG_TYPE_CGROUP_DEVICE, "bpf_cgroup_dev_ctx" },
> + { BPF_PROG_TYPE_CGROUP_SKB, "__sk_buff" },
> + { BPF_PROG_TYPE_CGROUP_SOCK, "bpf_sock" },
> + { BPF_PROG_TYPE_CGROUP_SOCK_ADDR, "bpf_sock_addr" },
> + { BPF_PROG_TYPE_CGROUP_SOCKOPT, "bpf_sockopt" },
> + { BPF_PROG_TYPE_CGROUP_SYSCTL, "bpf_sysctl" },
> + { BPF_PROG_TYPE_FLOW_DISSECTOR, "__sk_buff" },
> + { BPF_PROG_TYPE_KPROBE, "bpf_user_pt_regs_t" },
> + { BPF_PROG_TYPE_LWT_IN, "__sk_buff" },
> + { BPF_PROG_TYPE_LWT_OUT, "__sk_buff" },
> + { BPF_PROG_TYPE_LWT_SEG6LOCAL, "__sk_buff" },
> + { BPF_PROG_TYPE_LWT_XMIT, "__sk_buff" },
> + { BPF_PROG_TYPE_NETFILTER, "bpf_nf_ctx" },
> + { BPF_PROG_TYPE_PERF_EVENT, "bpf_perf_event_data" },
> + { BPF_PROG_TYPE_RAW_TRACEPOINT, "bpf_raw_tracepoint_args" },
> + { BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, "bpf_raw_tracepoint_args" },
> + { BPF_PROG_TYPE_SCHED_ACT, "__sk_buff" },
> + { BPF_PROG_TYPE_SCHED_CLS, "__sk_buff" },
> + { BPF_PROG_TYPE_SK_LOOKUP, "bpf_sk_lookup" },
> + { BPF_PROG_TYPE_SK_MSG, "sk_msg_md" },
> + { BPF_PROG_TYPE_SK_REUSEPORT, "sk_reuseport_md" },
> + { BPF_PROG_TYPE_SK_SKB, "__sk_buff" },
> + { BPF_PROG_TYPE_SOCK_OPS, "bpf_sock_ops" },
> + { BPF_PROG_TYPE_SOCKET_FILTER, "__sk_buff" },
> + { BPF_PROG_TYPE_XDP, "xdp_md" },
We already share the .c files (like relo_core.c) between kernel and libbpf
let's share here as well to avoid copy paste.
All of the above is available in include/linux/bpf_types.h
> + /* clone fn/fn_proto, unless we already did it for another arg */
> + if (func_rec->type_id == orig_fn_id) {
It feels that body of this 'if' can be factored out as a separate helper function.
> -static int
> -bpf_object__load_progs(struct bpf_object *obj, int log_level)
> +static int bpf_object_load_progs(struct bpf_object *obj, int log_level)
pls keep __ convention.
