On 11/18/19 7:17 PM, Andrii Nakryiko wrote:
On Thu, Nov 14, 2019 at 5:05 PM Daniel Borkmann <daniel@xxxxxxxxxxxxx> wrote:
Add initial poke table data structures and management to the BPF
prog that can later be used by JITs. Also add an instance of poke
specific data for tail call maps. Plan for later work is to extend
this also for BPF static keys.
Signed-off-by: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
---
looks good, just one more minor naming nit
Acked-by: Andrii Nakryiko <andriin@xxxxxx>
include/linux/bpf.h | 20 ++++++++++++++++++++
include/linux/filter.h | 10 ++++++++++
kernel/bpf/core.c | 34 ++++++++++++++++++++++++++++++++++
3 files changed, 64 insertions(+)
diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index 40337fa0e463..0ff06a0d0058 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -484,6 +484,24 @@ struct bpf_func_info_aux {
bool unreliable;
};
+enum bpf_jit_poke_reason {
+ BPF_POKE_REASON_TAIL_CALL,
+};
+
+/* Descriptor of pokes pointing /into/ the JITed image. */
+struct bpf_jit_poke_descriptor {
+ void *ip;
+ union {
+ struct {
+ struct bpf_map *map;
+ u32 key;
+ } tc;
tc is a bit overloaded abbreviation, tail_call would be super-clear, though ;)
Ok, sure, will include it into the non-rfc version.
+ };
+ u8 ip_stable;
+ u8 adj_off;
+ u16 reason;
+};
+
[...]
