Allow BPF TCP CCs to write app_limited
Subsystems (4)
v2 cover Allow BPF TCP CCs to write app_limited
This series allow BPF TCP CCs to write app_limited of struct tcp_sock. A built-in CC or one from a kernel module is already able to write to app_limited of struct tcp_sock. Until now, a BPF CC doesn't have write access to this member of struct tcp_sock.
v2:
- Merge the test of writing app_limited into the test of writing sk_pacing. (Martin KaFai Lau)
Yixin Shen (2):
bpf: allow a TCP CC to write app_limited selftests/bpf: test a BPF CC writing app_limited
net/ipv4/bpf_tcp_ca.c | 3 +++ .../selftests/bpf/progs/tcp_ca_write_sk_pacing.c | 13 ++++++++++++- 2 files changed, 15 insertions(+), 1 deletion(-)
Signature · 1 line
2.25.1
PATCH v2 1/2 bpf: allow a TCP CC to write app_limited
A CC that implements tcp_congestion_ops.cong_control() should be able to write app_limited. A built-in CC or one from a kernel module is already able to write to this member of struct tcp_sock.
For a BPF program, write access has not been allowed, yet.
--- net/ipv4/bpf_tcp_ca.c | 3 +++ 1 file changed, 3 insertions(+)
net/ipv4/bpf_tcp_ca.c +3 -0
@@ -113,6 +113,9 @@ static int bpf_tcp_ca_btf_struct_access(struct bpf_verifier_log *log,case offsetof(struct tcp_sock , ecn_flags): end = offsetofend(struct tcp_sock, ecn_flags);break ; +case offsetof(struct tcp_sock , app_limited): + end = offsetofend(struct tcp_sock, app_limited); +break ;default :bpf_log (log,"no write support to tcp_sock at off %d\n" , off);return -EACCES ;
Signature · 1 line
2.25.1
PATCH v2 2/2 selftests/bpf: test a BPF CC writing app_limited
In reply to Yixin ShenTest whether a TCP CC implemented in BPF is allowed to write app_limited in struct tcp_sock. This is already allowed for the built-in TCP CC.
--- .../selftests/bpf/progs/tcp_ca_write_sk_pacing.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-)
tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c +12 -1
@@ -16,6 +16,16 @@ static inline struct tcp_sock *tcp_sk(const struct sock *sk)return (struct tcp_sock *)sk; } +static inline unsigned int tcp_left_out (const struct tcp_sock *tp) +{ +return tp->sacked_out + tp->lost_out; +} + +static inline unsigned int tcp_packets_in_flight (const struct tcp_sock *tp) +{ +return tp->packets_out -tcp_left_out (tp) + tp->retrans_out; +} + SEC("struct_ops/write_sk_pacing_init")void BPF_PROG (write_sk_pacing_init ,struct sock *sk) {
@@ -31,11 +41,12 @@ SEC("struct_ops/write_sk_pacing_cong_control")void BPF_PROG (write_sk_pacing_cong_control ,struct sock *sk,const struct rate_sample *rs) { -const struct tcp_sock *tp =tcp_sk (sk); +struct tcp_sock *tp =tcp_sk (sk);unsigned long rate = ((tp->snd_cwnd * tp->mss_cache *USEC_PER_SEC ) <<3 ) / (tp->srtt_us ?:1U <<3 ); sk->sk_pacing_rate =min (rate, sk->sk_max_pacing_rate); + tp->app_limited = (tp->delivered +tcp_packets_in_flight (tp)) ?:1 ; }SEC ("struct_ops/write_sk_pacing_ssthresh" )
Signature · 1 line
2.25.1
This series was applied to bpf/bpf-next.git (master) by Martin KaFai Lau <[email protected]>:
Hello:
This series was applied to bpf/bpf-next.git (master) by Martin KaFai Lau <[email protected]>:
Quoted 11 lines · On Wed, 29 Mar 2023 07:35:56 +0000 you wrote:
This series allow BPF TCP CCs to write
app_limitedof structtcp_sock. A built-in CC or one from a kernel module is already able to write toapp_limitedof structtcp_sock. Until now, a BPF CC doesn't have write access to this member of structtcp_sock.v2:
- Merge the test of writingapp_limitedinto the test of writingsk_pacing. (Martin KaFai Lau)[...]
Here is the summary with links:
- [bpf-next,v2,1/2] bpf: allow a TCP CC to write app_limited
https://git.kernel.org/bpf/bpf-next/c/562dc56a8898 - [bpf-next,v2,2/2] selftests/bpf: test a BPF CC writing app_limited https://git.kernel.org/bpf/bpf-next/c/4239561b69fe
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html