Message ID | 01772bc1455ae16600796ac78c6cc9fff34f95ff.1662568410.git.dxu@dxuuu.xyz (mailing list archive) |
---|---|
State | Accepted |
Commit | 896f07c07da01aa7cee820a23c2bce1d8e9fe1e6 |
Delegated to: | BPF |
Headers | show |
Series | Support direct writes to nf_conn:mark | expand |
diff --git a/net/ipv4/bpf_tcp_ca.c b/net/ipv4/bpf_tcp_ca.c index 85a9e500c42d..6da16ae6a962 100644 --- a/net/ipv4/bpf_tcp_ca.c +++ b/net/ipv4/bpf_tcp_ca.c @@ -124,7 +124,7 @@ static int bpf_tcp_ca_btf_struct_access(struct bpf_verifier_log *log, return -EACCES; } - return NOT_INIT; + return 0; } BPF_CALL_2(bpf_tcp_send_ack, struct tcp_sock *, tp, u32, rcv_nxt)
Returning a bpf_reg_type only makes sense in the context of a BPF_READ. For writes, prefer to explicitly return 0 for clarity. Note that is non-functional change as it just so happened that NOT_INIT == 0. Signed-off-by: Daniel Xu <dxu@dxuuu.xyz> --- net/ipv4/bpf_tcp_ca.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)