From c0338e609e6e8aff8a7052c90cff83a6bc792ebc Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:17 -0700 Subject: selftests/bpf: Remove bpf_tracing_net.h usages from two networking tests This patch removes the bpf_tracing_net.h usage from the networking tests, fib_lookup and test_lwt_redirect. Instead of using the (copied) macro TC_ACT_SHOT and ETH_HLEN from bpf_tracing_net.h, they can directly use the ones defined in the network header files under linux/. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-2-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/fib_lookup.c | 2 +- tools/testing/selftests/bpf/progs/test_lwt_redirect.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/fib_lookup.c b/tools/testing/selftests/bpf/progs/fib_lookup.c index c4514dd58c62..7b5dd2214ff4 100644 --- a/tools/testing/selftests/bpf/progs/fib_lookup.c +++ b/tools/testing/selftests/bpf/progs/fib_lookup.c @@ -3,8 +3,8 @@ #include #include +#include #include -#include "bpf_tracing_net.h" struct bpf_fib_lookup fib_params = {}; int fib_lookup_ret = 0; diff --git a/tools/testing/selftests/bpf/progs/test_lwt_redirect.c b/tools/testing/selftests/bpf/progs/test_lwt_redirect.c index 8c895122f293..83439b87b766 100644 --- a/tools/testing/selftests/bpf/progs/test_lwt_redirect.c +++ b/tools/testing/selftests/bpf/progs/test_lwt_redirect.c @@ -3,7 +3,7 @@ #include #include #include -#include "bpf_tracing_net.h" +#include /* We don't care about whether the packet can be received by network stack. * Just care if the packet is sent to the correct device at correct direction -- cgit From cbaec46df6c08a2fab6be03d093d3d6ce74adc9a Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:18 -0700 Subject: selftests/bpf: Add a few tcp helper functions and macros to bpf_tracing_net.h This patch adds a few tcp related helper functions to bpf_tracing_net.h. They will be useful for both tcp-cc and network tracing related bpf progs. They have already been in the bpf_tcp_helpers.h. This change is needed to retire the bpf_tcp_helpers.h and consolidate all tests to vmlinux.h (i.e. bpf_tracing_net.h). Some of the helpers (tcp_sk and inet_csk) are also defined in bpf_cc_cubic.c and they are removed. While at it, remove the vmlinux.h from bpf_cc_cubic.c. bpf_tracing_net.h (which has vmlinux.h after this patch) is enough and will be consistent with the other tcp-cc tests in the later patches. The other TCP_* macro additions will be needed for the bpf_dctcp changes in the later patch. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-3-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/bpf_cc_cubic.c | 14 +------- .../testing/selftests/bpf/progs/bpf_tracing_net.h | 41 ++++++++++++++++++++++ 2 files changed, 42 insertions(+), 13 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c b/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c index 5c7697c70e2a..2004be380683 100644 --- a/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c +++ b/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c @@ -13,11 +13,9 @@ * kernel functions. */ -#include "vmlinux.h" - +#include "bpf_tracing_net.h" #include #include -#include "bpf_tracing_net.h" #define BPF_STRUCT_OPS(name, args...) \ SEC("struct_ops/"#name) \ @@ -40,16 +38,6 @@ extern __u32 tcp_reno_undo_cwnd(struct sock *sk) __ksym; extern void cubictcp_acked(struct sock *sk, const struct ack_sample *sample) __ksym; extern void cubictcp_cong_avoid(struct sock *sk, __u32 ack, __u32 acked) __ksym; -static struct inet_connection_sock *inet_csk(const struct sock *sk) -{ - return (struct inet_connection_sock *)sk; -} - -static struct tcp_sock *tcp_sk(const struct sock *sk) -{ - return (struct tcp_sock *)sk; -} - static bool before(__u32 seq1, __u32 seq2) { return (__s32)(seq1-seq2) < 0; diff --git a/tools/testing/selftests/bpf/progs/bpf_tracing_net.h b/tools/testing/selftests/bpf/progs/bpf_tracing_net.h index f9ec630dfcd5..ba4ca0334586 100644 --- a/tools/testing/selftests/bpf/progs/bpf_tracing_net.h +++ b/tools/testing/selftests/bpf/progs/bpf_tracing_net.h @@ -2,6 +2,9 @@ #ifndef __BPF_TRACING_NET_H__ #define __BPF_TRACING_NET_H__ +#include +#include + #define AF_INET 2 #define AF_INET6 10 @@ -46,6 +49,13 @@ #define TCP_CA_NAME_MAX 16 #define TCP_NAGLE_OFF 1 +#define TCP_ECN_OK 1 +#define TCP_ECN_QUEUE_CWR 2 +#define TCP_ECN_DEMAND_CWR 4 +#define TCP_ECN_SEEN 8 + +#define TCP_CONG_NEEDS_ECN 0x2 + #define ICSK_TIME_RETRANS 1 #define ICSK_TIME_PROBE0 3 #define ICSK_TIME_LOSS_PROBE 5 @@ -129,4 +139,35 @@ #define tcp_jiffies32 ((__u32)bpf_jiffies64()) +static inline struct inet_connection_sock *inet_csk(const struct sock *sk) +{ + return (struct inet_connection_sock *)sk; +} + +static inline void *inet_csk_ca(const struct sock *sk) +{ + return (void *)inet_csk(sk)->icsk_ca_priv; +} + +static inline struct tcp_sock *tcp_sk(const struct sock *sk) +{ + return (struct tcp_sock *)sk; +} + +static inline bool tcp_in_slow_start(const struct tcp_sock *tp) +{ + return tp->snd_cwnd < tp->snd_ssthresh; +} + +static inline bool tcp_is_cwnd_limited(const struct sock *sk) +{ + const struct tcp_sock *tp = tcp_sk(sk); + + /* If in slow start, ensure cwnd grows to twice what was ACKed. */ + if (tcp_in_slow_start(tp)) + return tp->snd_cwnd < 2 * tp->max_packets_out; + + return !!BPF_CORE_READ_BITFIELD(tp, is_cwnd_limited); +} + #endif -- cgit From cc5b18ce1714160be3e0e3b9440a6306dc87e5c4 Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:19 -0700 Subject: selftests/bpf: Reuse the tcp_sk() from the bpf_tracing_net.h This patch removes the individual tcp_sk implementations from the tcp-cc tests. The tcp_sk() implementation from the bpf_tracing_net.h is reused instead. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-4-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c | 8 +------- tools/testing/selftests/bpf/progs/tcp_ca_update.c | 8 +------- tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c | 8 +------- 3 files changed, 3 insertions(+), 21 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c b/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c index 7bb872fb22dd..d6467fcb1deb 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c @@ -1,17 +1,11 @@ // SPDX-License-Identifier: GPL-2.0 -#include "vmlinux.h" - +#include "bpf_tracing_net.h" #include #include char _license[] SEC("license") = "GPL"; -static inline struct tcp_sock *tcp_sk(const struct sock *sk) -{ - return (struct tcp_sock *)sk; -} - SEC("struct_ops/incompl_cong_ops_ssthresh") __u32 BPF_PROG(incompl_cong_ops_ssthresh, struct sock *sk) { diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_update.c b/tools/testing/selftests/bpf/progs/tcp_ca_update.c index b93a0ed33057..8581cad321b6 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_update.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_update.c @@ -1,7 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 -#include "vmlinux.h" - +#include "bpf_tracing_net.h" #include #include @@ -10,11 +9,6 @@ char _license[] SEC("license") = "GPL"; int ca1_cnt = 0; int ca2_cnt = 0; -static inline struct tcp_sock *tcp_sk(const struct sock *sk) -{ - return (struct tcp_sock *)sk; -} - SEC("struct_ops/ca_update_1_init") void BPF_PROG(ca_update_1_init, struct sock *sk) { diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c b/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c index 0724a79cec78..4a369439335e 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c @@ -1,7 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 -#include "vmlinux.h" - +#include "bpf_tracing_net.h" #include #include @@ -11,11 +10,6 @@ char _license[] SEC("license") = "GPL"; #define min(a, b) ((a) < (b) ? (a) : (b)) -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; -- cgit From 7d3851a31832bf8dc776a78494b788518734ad0f Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:20 -0700 Subject: selftests/bpf: Sanitize the SEC and inline usages in the bpf-tcp-cc tests It is needed to remove the BPF_STRUCT_OPS usages from the tcp-cc tests because it is defined in bpf_tcp_helpers.h which is going to be retired. While at it, this patch consolidates all tcp-cc struct_ops programs to use the SEC("struct_ops") + BPF_PROG(). It also removes the unnecessary __always_inline usages from the tcp-cc tests. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-5-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/bpf_cc_cubic.c | 28 ++++++++------- tools/testing/selftests/bpf/progs/bpf_cubic.c | 42 +++++++++++----------- tools/testing/selftests/bpf/progs/bpf_dctcp.c | 26 +++++++------- .../selftests/bpf/progs/bpf_dctcp_release.c | 3 +- tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c | 3 +- .../selftests/bpf/progs/tcp_ca_incompl_cong_ops.c | 4 +-- tools/testing/selftests/bpf/progs/tcp_ca_kfunc.c | 22 ++++++------ .../selftests/bpf/progs/tcp_ca_unsupp_cong_op.c | 2 +- tools/testing/selftests/bpf/progs/tcp_ca_update.c | 10 +++--- .../selftests/bpf/progs/tcp_ca_write_sk_pacing.c | 12 +++---- 10 files changed, 77 insertions(+), 75 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c b/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c index 2004be380683..1654a530aa3d 100644 --- a/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c +++ b/tools/testing/selftests/bpf/progs/bpf_cc_cubic.c @@ -17,10 +17,6 @@ #include #include -#define BPF_STRUCT_OPS(name, args...) \ -SEC("struct_ops/"#name) \ -BPF_PROG(name, args) - #define USEC_PER_SEC 1000000UL #define TCP_PACING_SS_RATIO (200) #define TCP_PACING_CA_RATIO (120) @@ -114,18 +110,21 @@ static bool tcp_may_raise_cwnd(const struct sock *sk, const int flag) return flag & FLAG_DATA_ACKED; } -void BPF_STRUCT_OPS(bpf_cubic_init, struct sock *sk) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_init, struct sock *sk) { cubictcp_init(sk); } -void BPF_STRUCT_OPS(bpf_cubic_cwnd_event, struct sock *sk, enum tcp_ca_event event) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_cwnd_event, struct sock *sk, enum tcp_ca_event event) { cubictcp_cwnd_event(sk, event); } -void BPF_STRUCT_OPS(bpf_cubic_cong_control, struct sock *sk, __u32 ack, int flag, - const struct rate_sample *rs) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_cong_control, struct sock *sk, __u32 ack, int flag, + const struct rate_sample *rs) { struct tcp_sock *tp = tcp_sk(sk); @@ -151,23 +150,26 @@ void BPF_STRUCT_OPS(bpf_cubic_cong_control, struct sock *sk, __u32 ack, int flag tcp_update_pacing_rate(sk); } -__u32 BPF_STRUCT_OPS(bpf_cubic_recalc_ssthresh, struct sock *sk) +SEC("struct_ops") +__u32 BPF_PROG(bpf_cubic_recalc_ssthresh, struct sock *sk) { return cubictcp_recalc_ssthresh(sk); } -void BPF_STRUCT_OPS(bpf_cubic_state, struct sock *sk, __u8 new_state) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_state, struct sock *sk, __u8 new_state) { cubictcp_state(sk, new_state); } -void BPF_STRUCT_OPS(bpf_cubic_acked, struct sock *sk, - const struct ack_sample *sample) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_acked, struct sock *sk, const struct ack_sample *sample) { cubictcp_acked(sk, sample); } -__u32 BPF_STRUCT_OPS(bpf_cubic_undo_cwnd, struct sock *sk) +SEC("struct_ops") +__u32 BPF_PROG(bpf_cubic_undo_cwnd, struct sock *sk) { return tcp_reno_undo_cwnd(sk); } diff --git a/tools/testing/selftests/bpf/progs/bpf_cubic.c b/tools/testing/selftests/bpf/progs/bpf_cubic.c index c997e3e3d3fb..53a98b609e5f 100644 --- a/tools/testing/selftests/bpf/progs/bpf_cubic.c +++ b/tools/testing/selftests/bpf/progs/bpf_cubic.c @@ -91,7 +91,7 @@ struct bictcp { __u32 curr_rtt; /* the minimum rtt of current round */ }; -static inline void bictcp_reset(struct bictcp *ca) +static void bictcp_reset(struct bictcp *ca) { ca->cnt = 0; ca->last_max_cwnd = 0; @@ -112,7 +112,7 @@ extern unsigned long CONFIG_HZ __kconfig; #define USEC_PER_SEC 1000000UL #define USEC_PER_JIFFY (USEC_PER_SEC / HZ) -static __always_inline __u64 div64_u64(__u64 dividend, __u64 divisor) +static __u64 div64_u64(__u64 dividend, __u64 divisor) { return dividend / divisor; } @@ -120,7 +120,7 @@ static __always_inline __u64 div64_u64(__u64 dividend, __u64 divisor) #define div64_ul div64_u64 #define BITS_PER_U64 (sizeof(__u64) * 8) -static __always_inline int fls64(__u64 x) +static int fls64(__u64 x) { int num = BITS_PER_U64 - 1; @@ -153,12 +153,12 @@ static __always_inline int fls64(__u64 x) return num + 1; } -static __always_inline __u32 bictcp_clock_us(const struct sock *sk) +static __u32 bictcp_clock_us(const struct sock *sk) { return tcp_sk(sk)->tcp_mstamp; } -static __always_inline void bictcp_hystart_reset(struct sock *sk) +static void bictcp_hystart_reset(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); struct bictcp *ca = inet_csk_ca(sk); @@ -169,8 +169,7 @@ static __always_inline void bictcp_hystart_reset(struct sock *sk) ca->sample_cnt = 0; } -/* "struct_ops/" prefix is a requirement */ -SEC("struct_ops/bpf_cubic_init") +SEC("struct_ops") void BPF_PROG(bpf_cubic_init, struct sock *sk) { struct bictcp *ca = inet_csk_ca(sk); @@ -184,8 +183,7 @@ void BPF_PROG(bpf_cubic_init, struct sock *sk) tcp_sk(sk)->snd_ssthresh = initial_ssthresh; } -/* "struct_ops" prefix is a requirement */ -SEC("struct_ops/bpf_cubic_cwnd_event") +SEC("struct_ops") void BPF_PROG(bpf_cubic_cwnd_event, struct sock *sk, enum tcp_ca_event event) { if (event == CA_EVENT_TX_START) { @@ -230,7 +228,7 @@ static const __u8 v[] = { * Newton-Raphson iteration. * Avg err ~= 0.195% */ -static __always_inline __u32 cubic_root(__u64 a) +static __u32 cubic_root(__u64 a) { __u32 x, b, shift; @@ -263,8 +261,7 @@ static __always_inline __u32 cubic_root(__u64 a) /* * Compute congestion window to use. */ -static __always_inline void bictcp_update(struct bictcp *ca, __u32 cwnd, - __u32 acked) +static void bictcp_update(struct bictcp *ca, __u32 cwnd, __u32 acked) { __u32 delta, bic_target, max_cnt; __u64 offs, t; @@ -377,8 +374,8 @@ tcp_friendliness: ca->cnt = max(ca->cnt, 2U); } -/* Or simply use the BPF_STRUCT_OPS to avoid the SEC boiler plate. */ -void BPF_STRUCT_OPS(bpf_cubic_cong_avoid, struct sock *sk, __u32 ack, __u32 acked) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_cong_avoid, struct sock *sk, __u32 ack, __u32 acked) { struct tcp_sock *tp = tcp_sk(sk); struct bictcp *ca = inet_csk_ca(sk); @@ -397,7 +394,8 @@ void BPF_STRUCT_OPS(bpf_cubic_cong_avoid, struct sock *sk, __u32 ack, __u32 acke tcp_cong_avoid_ai(tp, ca->cnt, acked); } -__u32 BPF_STRUCT_OPS(bpf_cubic_recalc_ssthresh, struct sock *sk) +SEC("struct_ops") +__u32 BPF_PROG(bpf_cubic_recalc_ssthresh, struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); struct bictcp *ca = inet_csk_ca(sk); @@ -414,7 +412,8 @@ __u32 BPF_STRUCT_OPS(bpf_cubic_recalc_ssthresh, struct sock *sk) return max((tp->snd_cwnd * beta) / BICTCP_BETA_SCALE, 2U); } -void BPF_STRUCT_OPS(bpf_cubic_state, struct sock *sk, __u8 new_state) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_state, struct sock *sk, __u8 new_state) { if (new_state == TCP_CA_Loss) { bictcp_reset(inet_csk_ca(sk)); @@ -433,7 +432,7 @@ void BPF_STRUCT_OPS(bpf_cubic_state, struct sock *sk, __u8 new_state) * We apply another 100% factor because @rate is doubled at this point. * We cap the cushion to 1ms. */ -static __always_inline __u32 hystart_ack_delay(struct sock *sk) +static __u32 hystart_ack_delay(struct sock *sk) { unsigned long rate; @@ -444,7 +443,7 @@ static __always_inline __u32 hystart_ack_delay(struct sock *sk) div64_ul((__u64)GSO_MAX_SIZE * 4 * USEC_PER_SEC, rate)); } -static __always_inline void hystart_update(struct sock *sk, __u32 delay) +static void hystart_update(struct sock *sk, __u32 delay) { struct tcp_sock *tp = tcp_sk(sk); struct bictcp *ca = inet_csk_ca(sk); @@ -492,8 +491,8 @@ static __always_inline void hystart_update(struct sock *sk, __u32 delay) int bpf_cubic_acked_called = 0; -void BPF_STRUCT_OPS(bpf_cubic_acked, struct sock *sk, - const struct ack_sample *sample) +SEC("struct_ops") +void BPF_PROG(bpf_cubic_acked, struct sock *sk, const struct ack_sample *sample) { const struct tcp_sock *tp = tcp_sk(sk); struct bictcp *ca = inet_csk_ca(sk); @@ -524,7 +523,8 @@ void BPF_STRUCT_OPS(bpf_cubic_acked, struct sock *sk, extern __u32 tcp_reno_undo_cwnd(struct sock *sk) __ksym; -__u32 BPF_STRUCT_OPS(bpf_cubic_undo_cwnd, struct sock *sk) +SEC("struct_ops") +__u32 BPF_PROG(bpf_cubic_undo_cwnd, struct sock *sk) { return tcp_reno_undo_cwnd(sk); } diff --git a/tools/testing/selftests/bpf/progs/bpf_dctcp.c b/tools/testing/selftests/bpf/progs/bpf_dctcp.c index 460682759aed..b74dbb121384 100644 --- a/tools/testing/selftests/bpf/progs/bpf_dctcp.c +++ b/tools/testing/selftests/bpf/progs/bpf_dctcp.c @@ -48,8 +48,7 @@ struct dctcp { static unsigned int dctcp_shift_g = 4; /* g = 1/2^4 */ static unsigned int dctcp_alpha_on_init = DCTCP_MAX_ALPHA; -static __always_inline void dctcp_reset(const struct tcp_sock *tp, - struct dctcp *ca) +static void dctcp_reset(const struct tcp_sock *tp, struct dctcp *ca) { ca->next_seq = tp->snd_nxt; @@ -57,7 +56,7 @@ static __always_inline void dctcp_reset(const struct tcp_sock *tp, ca->old_delivered_ce = tp->delivered_ce; } -SEC("struct_ops/dctcp_init") +SEC("struct_ops") void BPF_PROG(dctcp_init, struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); @@ -104,7 +103,7 @@ void BPF_PROG(dctcp_init, struct sock *sk) dctcp_reset(tp, ca); } -SEC("struct_ops/dctcp_ssthresh") +SEC("struct_ops") __u32 BPF_PROG(dctcp_ssthresh, struct sock *sk) { struct dctcp *ca = inet_csk_ca(sk); @@ -114,7 +113,7 @@ __u32 BPF_PROG(dctcp_ssthresh, struct sock *sk) return max(tp->snd_cwnd - ((tp->snd_cwnd * ca->dctcp_alpha) >> 11U), 2U); } -SEC("struct_ops/dctcp_update_alpha") +SEC("struct_ops") void BPF_PROG(dctcp_update_alpha, struct sock *sk, __u32 flags) { const struct tcp_sock *tp = tcp_sk(sk); @@ -144,7 +143,7 @@ void BPF_PROG(dctcp_update_alpha, struct sock *sk, __u32 flags) } } -static __always_inline void dctcp_react_to_loss(struct sock *sk) +static void dctcp_react_to_loss(struct sock *sk) { struct dctcp *ca = inet_csk_ca(sk); struct tcp_sock *tp = tcp_sk(sk); @@ -153,7 +152,7 @@ static __always_inline void dctcp_react_to_loss(struct sock *sk) tp->snd_ssthresh = max(tp->snd_cwnd >> 1U, 2U); } -SEC("struct_ops/dctcp_state") +SEC("struct_ops") void BPF_PROG(dctcp_state, struct sock *sk, __u8 new_state) { if (new_state == TCP_CA_Recovery && @@ -164,7 +163,7 @@ void BPF_PROG(dctcp_state, struct sock *sk, __u8 new_state) */ } -static __always_inline void dctcp_ece_ack_cwr(struct sock *sk, __u32 ce_state) +static void dctcp_ece_ack_cwr(struct sock *sk, __u32 ce_state) { struct tcp_sock *tp = tcp_sk(sk); @@ -179,9 +178,8 @@ static __always_inline void dctcp_ece_ack_cwr(struct sock *sk, __u32 ce_state) * S: 0 <- last pkt was non-CE * 1 <- last pkt was CE */ -static __always_inline -void dctcp_ece_ack_update(struct sock *sk, enum tcp_ca_event evt, - __u32 *prior_rcv_nxt, __u32 *ce_state) +static void dctcp_ece_ack_update(struct sock *sk, enum tcp_ca_event evt, + __u32 *prior_rcv_nxt, __u32 *ce_state) { __u32 new_ce_state = (evt == CA_EVENT_ECN_IS_CE) ? 1 : 0; @@ -201,7 +199,7 @@ void dctcp_ece_ack_update(struct sock *sk, enum tcp_ca_event evt, dctcp_ece_ack_cwr(sk, new_ce_state); } -SEC("struct_ops/dctcp_cwnd_event") +SEC("struct_ops") void BPF_PROG(dctcp_cwnd_event, struct sock *sk, enum tcp_ca_event ev) { struct dctcp *ca = inet_csk_ca(sk); @@ -220,7 +218,7 @@ void BPF_PROG(dctcp_cwnd_event, struct sock *sk, enum tcp_ca_event ev) } } -SEC("struct_ops/dctcp_cwnd_undo") +SEC("struct_ops") __u32 BPF_PROG(dctcp_cwnd_undo, struct sock *sk) { const struct dctcp *ca = inet_csk_ca(sk); @@ -230,7 +228,7 @@ __u32 BPF_PROG(dctcp_cwnd_undo, struct sock *sk) extern void tcp_reno_cong_avoid(struct sock *sk, __u32 ack, __u32 acked) __ksym; -SEC("struct_ops/dctcp_reno_cong_avoid") +SEC("struct_ops") void BPF_PROG(dctcp_cong_avoid, struct sock *sk, __u32 ack, __u32 acked) { tcp_reno_cong_avoid(sk, ack, acked); diff --git a/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c b/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c index d836f7c372f0..a946b070bb06 100644 --- a/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c +++ b/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c @@ -13,7 +13,8 @@ char _license[] SEC("license") = "GPL"; const char cubic[] = "cubic"; -void BPF_STRUCT_OPS(dctcp_nouse_release, struct sock *sk) +SEC("struct_ops") +void BPF_PROG(dctcp_nouse_release, struct sock *sk) { bpf_setsockopt(sk, SOL_TCP, TCP_CONGESTION, (void *)cubic, sizeof(cubic)); diff --git a/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c b/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c index 2ecd833dcd41..633164e704dd 100644 --- a/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c +++ b/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c @@ -8,7 +8,8 @@ char _license[] SEC("license") = "X"; -void BPF_STRUCT_OPS(nogpltcp_init, struct sock *sk) +SEC("struct_ops") +void BPF_PROG(nogpltcp_init, struct sock *sk) { } diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c b/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c index d6467fcb1deb..0016c90e9c13 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_incompl_cong_ops.c @@ -6,13 +6,13 @@ char _license[] SEC("license") = "GPL"; -SEC("struct_ops/incompl_cong_ops_ssthresh") +SEC("struct_ops") __u32 BPF_PROG(incompl_cong_ops_ssthresh, struct sock *sk) { return tcp_sk(sk)->snd_ssthresh; } -SEC("struct_ops/incompl_cong_ops_undo_cwnd") +SEC("struct_ops") __u32 BPF_PROG(incompl_cong_ops_undo_cwnd, struct sock *sk) { return tcp_sk(sk)->snd_cwnd; diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_kfunc.c b/tools/testing/selftests/bpf/progs/tcp_ca_kfunc.c index 52b610357309..f95862f570b7 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_kfunc.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_kfunc.c @@ -27,7 +27,7 @@ extern void cubictcp_state(struct sock *sk, u8 new_state) __ksym; extern void cubictcp_cwnd_event(struct sock *sk, enum tcp_ca_event event) __ksym; extern void cubictcp_acked(struct sock *sk, const struct ack_sample *sample) __ksym; -SEC("struct_ops/init") +SEC("struct_ops") void BPF_PROG(init, struct sock *sk) { bbr_init(sk); @@ -35,38 +35,38 @@ void BPF_PROG(init, struct sock *sk) cubictcp_init(sk); } -SEC("struct_ops/in_ack_event") +SEC("struct_ops") void BPF_PROG(in_ack_event, struct sock *sk, u32 flags) { dctcp_update_alpha(sk, flags); } -SEC("struct_ops/cong_control") +SEC("struct_ops") void BPF_PROG(cong_control, struct sock *sk, u32 ack, int flag, const struct rate_sample *rs) { bbr_main(sk, ack, flag, rs); } -SEC("struct_ops/cong_avoid") +SEC("struct_ops") void BPF_PROG(cong_avoid, struct sock *sk, u32 ack, u32 acked) { cubictcp_cong_avoid(sk, ack, acked); } -SEC("struct_ops/sndbuf_expand") +SEC("struct_ops") u32 BPF_PROG(sndbuf_expand, struct sock *sk) { return bbr_sndbuf_expand(sk); } -SEC("struct_ops/undo_cwnd") +SEC("struct_ops") u32 BPF_PROG(undo_cwnd, struct sock *sk) { bbr_undo_cwnd(sk); return dctcp_cwnd_undo(sk); } -SEC("struct_ops/cwnd_event") +SEC("struct_ops") void BPF_PROG(cwnd_event, struct sock *sk, enum tcp_ca_event event) { bbr_cwnd_event(sk, event); @@ -74,7 +74,7 @@ void BPF_PROG(cwnd_event, struct sock *sk, enum tcp_ca_event event) cubictcp_cwnd_event(sk, event); } -SEC("struct_ops/ssthresh") +SEC("struct_ops") u32 BPF_PROG(ssthresh, struct sock *sk) { bbr_ssthresh(sk); @@ -82,13 +82,13 @@ u32 BPF_PROG(ssthresh, struct sock *sk) return cubictcp_recalc_ssthresh(sk); } -SEC("struct_ops/min_tso_segs") +SEC("struct_ops") u32 BPF_PROG(min_tso_segs, struct sock *sk) { return bbr_min_tso_segs(sk); } -SEC("struct_ops/set_state") +SEC("struct_ops") void BPF_PROG(set_state, struct sock *sk, u8 new_state) { bbr_set_state(sk, new_state); @@ -96,7 +96,7 @@ void BPF_PROG(set_state, struct sock *sk, u8 new_state) cubictcp_state(sk, new_state); } -SEC("struct_ops/pkts_acked") +SEC("struct_ops") void BPF_PROG(pkts_acked, struct sock *sk, const struct ack_sample *sample) { cubictcp_acked(sk, sample); diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_unsupp_cong_op.c b/tools/testing/selftests/bpf/progs/tcp_ca_unsupp_cong_op.c index c06f4a41c21a..54f916a931c6 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_unsupp_cong_op.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_unsupp_cong_op.c @@ -7,7 +7,7 @@ char _license[] SEC("license") = "GPL"; -SEC("struct_ops/unsupp_cong_op_get_info") +SEC("struct_ops") size_t BPF_PROG(unsupp_cong_op_get_info, struct sock *sk, u32 ext, int *attr, union tcp_cc_info *info) { diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_update.c b/tools/testing/selftests/bpf/progs/tcp_ca_update.c index 8581cad321b6..e4bd82bc0d01 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_update.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_update.c @@ -9,31 +9,31 @@ char _license[] SEC("license") = "GPL"; int ca1_cnt = 0; int ca2_cnt = 0; -SEC("struct_ops/ca_update_1_init") +SEC("struct_ops") void BPF_PROG(ca_update_1_init, struct sock *sk) { ca1_cnt++; } -SEC("struct_ops/ca_update_2_init") +SEC("struct_ops") void BPF_PROG(ca_update_2_init, struct sock *sk) { ca2_cnt++; } -SEC("struct_ops/ca_update_cong_control") +SEC("struct_ops") void BPF_PROG(ca_update_cong_control, struct sock *sk, const struct rate_sample *rs) { } -SEC("struct_ops/ca_update_ssthresh") +SEC("struct_ops") __u32 BPF_PROG(ca_update_ssthresh, struct sock *sk) { return tcp_sk(sk)->snd_ssthresh; } -SEC("struct_ops/ca_update_undo_cwnd") +SEC("struct_ops") __u32 BPF_PROG(ca_update_undo_cwnd, struct sock *sk) { return tcp_sk(sk)->snd_cwnd; diff --git a/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c b/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c index 4a369439335e..a58b5194fc89 100644 --- a/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c +++ b/tools/testing/selftests/bpf/progs/tcp_ca_write_sk_pacing.c @@ -10,17 +10,17 @@ char _license[] SEC("license") = "GPL"; #define min(a, b) ((a) < (b) ? (a) : (b)) -static inline unsigned int tcp_left_out(const struct tcp_sock *tp) +static 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) +static 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") +SEC("struct_ops") void BPF_PROG(write_sk_pacing_init, struct sock *sk) { #ifdef ENABLE_ATOMICS_TESTS @@ -31,7 +31,7 @@ void BPF_PROG(write_sk_pacing_init, struct sock *sk) #endif } -SEC("struct_ops/write_sk_pacing_cong_control") +SEC("struct_ops") void BPF_PROG(write_sk_pacing_cong_control, struct sock *sk, const struct rate_sample *rs) { @@ -43,13 +43,13 @@ void BPF_PROG(write_sk_pacing_cong_control, struct sock *sk, tp->app_limited = (tp->delivered + tcp_packets_in_flight(tp)) ?: 1; } -SEC("struct_ops/write_sk_pacing_ssthresh") +SEC("struct_ops") __u32 BPF_PROG(write_sk_pacing_ssthresh, struct sock *sk) { return tcp_sk(sk)->snd_ssthresh; } -SEC("struct_ops/write_sk_pacing_undo_cwnd") +SEC("struct_ops") __u32 BPF_PROG(write_sk_pacing_undo_cwnd, struct sock *sk) { return tcp_sk(sk)->snd_cwnd; -- cgit From b1d87ae9b0d3d91767d85183e40c96f4229a6c21 Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:21 -0700 Subject: selftests/bpf: Rename tcp-cc private struct in bpf_cubic and bpf_dctcp The "struct bictcp" and "struct dctcp" are private to the bpf prog and they are stored in the private buffer in inet_csk(sk)->icsk_ca_priv. Hence, there is no bpf CO-RE required. The same struct name exists in the vmlinux.h. To reuse vmlinux.h, they need to be renamed such that the bpf prog logic will be immuned from the kernel tcp-cc changes. This patch adds a "bpf_" prefix to them. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-6-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/bpf_cubic.c | 20 ++++++++++---------- tools/testing/selftests/bpf/progs/bpf_dctcp.c | 16 ++++++++-------- 2 files changed, 18 insertions(+), 18 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/bpf_cubic.c b/tools/testing/selftests/bpf/progs/bpf_cubic.c index 53a98b609e5f..53872e2d2c52 100644 --- a/tools/testing/selftests/bpf/progs/bpf_cubic.c +++ b/tools/testing/selftests/bpf/progs/bpf_cubic.c @@ -70,7 +70,7 @@ static const __u64 cube_factor = (__u64)(1ull << (10+3*BICTCP_HZ)) / (bic_scale * 10); /* BIC TCP Parameters */ -struct bictcp { +struct bpf_bictcp { __u32 cnt; /* increase cwnd by 1 after ACKs */ __u32 last_max_cwnd; /* last maximum snd_cwnd */ __u32 last_cwnd; /* the last snd_cwnd */ @@ -91,7 +91,7 @@ struct bictcp { __u32 curr_rtt; /* the minimum rtt of current round */ }; -static void bictcp_reset(struct bictcp *ca) +static void bictcp_reset(struct bpf_bictcp *ca) { ca->cnt = 0; ca->last_max_cwnd = 0; @@ -161,7 +161,7 @@ static __u32 bictcp_clock_us(const struct sock *sk) static void bictcp_hystart_reset(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); - struct bictcp *ca = inet_csk_ca(sk); + struct bpf_bictcp *ca = inet_csk_ca(sk); ca->round_start = ca->last_ack = bictcp_clock_us(sk); ca->end_seq = tp->snd_nxt; @@ -172,7 +172,7 @@ static void bictcp_hystart_reset(struct sock *sk) SEC("struct_ops") void BPF_PROG(bpf_cubic_init, struct sock *sk) { - struct bictcp *ca = inet_csk_ca(sk); + struct bpf_bictcp *ca = inet_csk_ca(sk); bictcp_reset(ca); @@ -187,7 +187,7 @@ SEC("struct_ops") void BPF_PROG(bpf_cubic_cwnd_event, struct sock *sk, enum tcp_ca_event event) { if (event == CA_EVENT_TX_START) { - struct bictcp *ca = inet_csk_ca(sk); + struct bpf_bictcp *ca = inet_csk_ca(sk); __u32 now = tcp_jiffies32; __s32 delta; @@ -261,7 +261,7 @@ static __u32 cubic_root(__u64 a) /* * Compute congestion window to use. */ -static void bictcp_update(struct bictcp *ca, __u32 cwnd, __u32 acked) +static void bictcp_update(struct bpf_bictcp *ca, __u32 cwnd, __u32 acked) { __u32 delta, bic_target, max_cnt; __u64 offs, t; @@ -378,7 +378,7 @@ SEC("struct_ops") void BPF_PROG(bpf_cubic_cong_avoid, struct sock *sk, __u32 ack, __u32 acked) { struct tcp_sock *tp = tcp_sk(sk); - struct bictcp *ca = inet_csk_ca(sk); + struct bpf_bictcp *ca = inet_csk_ca(sk); if (!tcp_is_cwnd_limited(sk)) return; @@ -398,7 +398,7 @@ SEC("struct_ops") __u32 BPF_PROG(bpf_cubic_recalc_ssthresh, struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); - struct bictcp *ca = inet_csk_ca(sk); + struct bpf_bictcp *ca = inet_csk_ca(sk); ca->epoch_start = 0; /* end of epoch */ @@ -446,7 +446,7 @@ static __u32 hystart_ack_delay(struct sock *sk) static void hystart_update(struct sock *sk, __u32 delay) { struct tcp_sock *tp = tcp_sk(sk); - struct bictcp *ca = inet_csk_ca(sk); + struct bpf_bictcp *ca = inet_csk_ca(sk); __u32 threshold; if (hystart_detect & HYSTART_ACK_TRAIN) { @@ -495,7 +495,7 @@ SEC("struct_ops") void BPF_PROG(bpf_cubic_acked, struct sock *sk, const struct ack_sample *sample) { const struct tcp_sock *tp = tcp_sk(sk); - struct bictcp *ca = inet_csk_ca(sk); + struct bpf_bictcp *ca = inet_csk_ca(sk); __u32 delay; bpf_cubic_acked_called = 1; diff --git a/tools/testing/selftests/bpf/progs/bpf_dctcp.c b/tools/testing/selftests/bpf/progs/bpf_dctcp.c index b74dbb121384..a8673b7ff35d 100644 --- a/tools/testing/selftests/bpf/progs/bpf_dctcp.c +++ b/tools/testing/selftests/bpf/progs/bpf_dctcp.c @@ -35,7 +35,7 @@ struct { #define DCTCP_MAX_ALPHA 1024U -struct dctcp { +struct bpf_dctcp { __u32 old_delivered; __u32 old_delivered_ce; __u32 prior_rcv_nxt; @@ -48,7 +48,7 @@ struct dctcp { static unsigned int dctcp_shift_g = 4; /* g = 1/2^4 */ static unsigned int dctcp_alpha_on_init = DCTCP_MAX_ALPHA; -static void dctcp_reset(const struct tcp_sock *tp, struct dctcp *ca) +static void dctcp_reset(const struct tcp_sock *tp, struct bpf_dctcp *ca) { ca->next_seq = tp->snd_nxt; @@ -60,7 +60,7 @@ SEC("struct_ops") void BPF_PROG(dctcp_init, struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); - struct dctcp *ca = inet_csk_ca(sk); + struct bpf_dctcp *ca = inet_csk_ca(sk); int *stg; if (!(tp->ecn_flags & TCP_ECN_OK) && fallback[0]) { @@ -106,7 +106,7 @@ void BPF_PROG(dctcp_init, struct sock *sk) SEC("struct_ops") __u32 BPF_PROG(dctcp_ssthresh, struct sock *sk) { - struct dctcp *ca = inet_csk_ca(sk); + struct bpf_dctcp *ca = inet_csk_ca(sk); struct tcp_sock *tp = tcp_sk(sk); ca->loss_cwnd = tp->snd_cwnd; @@ -117,7 +117,7 @@ SEC("struct_ops") void BPF_PROG(dctcp_update_alpha, struct sock *sk, __u32 flags) { const struct tcp_sock *tp = tcp_sk(sk); - struct dctcp *ca = inet_csk_ca(sk); + struct bpf_dctcp *ca = inet_csk_ca(sk); /* Expired RTT */ if (!before(tp->snd_una, ca->next_seq)) { @@ -145,7 +145,7 @@ void BPF_PROG(dctcp_update_alpha, struct sock *sk, __u32 flags) static void dctcp_react_to_loss(struct sock *sk) { - struct dctcp *ca = inet_csk_ca(sk); + struct bpf_dctcp *ca = inet_csk_ca(sk); struct tcp_sock *tp = tcp_sk(sk); ca->loss_cwnd = tp->snd_cwnd; @@ -202,7 +202,7 @@ static void dctcp_ece_ack_update(struct sock *sk, enum tcp_ca_event evt, SEC("struct_ops") void BPF_PROG(dctcp_cwnd_event, struct sock *sk, enum tcp_ca_event ev) { - struct dctcp *ca = inet_csk_ca(sk); + struct bpf_dctcp *ca = inet_csk_ca(sk); switch (ev) { case CA_EVENT_ECN_IS_CE: @@ -221,7 +221,7 @@ void BPF_PROG(dctcp_cwnd_event, struct sock *sk, enum tcp_ca_event ev) SEC("struct_ops") __u32 BPF_PROG(dctcp_cwnd_undo, struct sock *sk) { - const struct dctcp *ca = inet_csk_ca(sk); + const struct bpf_dctcp *ca = inet_csk_ca(sk); return max(tcp_sk(sk)->snd_cwnd, ca->loss_cwnd); } -- cgit From a824c9a8a4d9a654d62674a8425c0f1abc9c3d33 Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:22 -0700 Subject: selftests/bpf: Use bpf_tracing_net.h in bpf_cubic This patch uses bpf_tracing_net.h (i.e. vmlinux.h) in bpf_cubic. This will allow to retire the bpf_tcp_helpers.h and consolidate tcp-cc tests to vmlinux.h. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-7-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/bpf_cubic.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/bpf_cubic.c b/tools/testing/selftests/bpf/progs/bpf_cubic.c index 53872e2d2c52..d665b8a15cc4 100644 --- a/tools/testing/selftests/bpf/progs/bpf_cubic.c +++ b/tools/testing/selftests/bpf/progs/bpf_cubic.c @@ -14,14 +14,22 @@ * "ca->ack_cnt / delta" operation. */ -#include -#include -#include -#include "bpf_tcp_helpers.h" +#include "bpf_tracing_net.h" +#include char _license[] SEC("license") = "GPL"; #define clamp(val, lo, hi) min((typeof(val))max(val, lo), hi) +#define min(a, b) ((a) < (b) ? (a) : (b)) +#define max(a, b) ((a) > (b) ? (a) : (b)) +static bool before(__u32 seq1, __u32 seq2) +{ + return (__s32)(seq1-seq2) < 0; +} +#define after(seq2, seq1) before(seq1, seq2) + +extern __u32 tcp_slow_start(struct tcp_sock *tp, __u32 acked) __ksym; +extern void tcp_cong_avoid_ai(struct tcp_sock *tp, __u32 w, __u32 acked) __ksym; #define BICTCP_BETA_SCALE 1024 /* Scale factor beta calculation * max_cwnd = snd_cwnd * beta -- cgit From 6ad4e6e94697e960630594907666bc09e78a3b8a Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:23 -0700 Subject: selftests/bpf: Use bpf_tracing_net.h in bpf_dctcp This patch uses bpf_tracing_net.h (i.e. vmlinux.h) in bpf_dctcp. This will allow to retire the bpf_tcp_helpers.h and consolidate tcp-cc tests to vmlinux.h. It will have a dup on min/max macros with the bpf_cubic. It could be further refactored in the future. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-8-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/bpf_dctcp.c | 22 +++++++++++++++------- 1 file changed, 15 insertions(+), 7 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/bpf_dctcp.c b/tools/testing/selftests/bpf/progs/bpf_dctcp.c index a8673b7ff35d..3c9ffe340312 100644 --- a/tools/testing/selftests/bpf/progs/bpf_dctcp.c +++ b/tools/testing/selftests/bpf/progs/bpf_dctcp.c @@ -6,15 +6,23 @@ * the kernel BPF logic. */ -#include -#include -#include -#include -#include -#include +#include "bpf_tracing_net.h" #include #include -#include "bpf_tcp_helpers.h" + +#ifndef EBUSY +#define EBUSY 16 +#endif +#define min(a, b) ((a) < (b) ? (a) : (b)) +#define max(a, b) ((a) > (b) ? (a) : (b)) +#define min_not_zero(x, y) ({ \ + typeof(x) __x = (x); \ + typeof(y) __y = (y); \ + __x == 0 ? __y : ((__y == 0) ? __x : min(__x, __y)); }) +static bool before(__u32 seq1, __u32 seq2) +{ + return (__s32)(seq1-seq2) < 0; +} char _license[] SEC("license") = "GPL"; -- cgit From 6eee55aa769c241182da73a391980f51edba27dc Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:24 -0700 Subject: selftests/bpf: Remove bpf_tcp_helpers.h usages from other misc bpf tcp-cc tests This patch removed the final few bpf_tcp_helpers.h usages in some misc bpf tcp-cc tests and replace it with bpf_tracing_net.h (i.e. vmlinux.h) Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-9-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/bpf_dctcp_release.c | 7 +------ tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c | 5 +---- 2 files changed, 2 insertions(+), 10 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c b/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c index a946b070bb06..c91763f248b2 100644 --- a/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c +++ b/tools/testing/selftests/bpf/progs/bpf_dctcp_release.c @@ -1,14 +1,9 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright (c) 2021 Facebook */ -#include -#include -#include -#include -#include +#include "bpf_tracing_net.h" #include #include -#include "bpf_tcp_helpers.h" char _license[] SEC("license") = "GPL"; const char cubic[] = "cubic"; diff --git a/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c b/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c index 633164e704dd..8a7a4c1b54e8 100644 --- a/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c +++ b/tools/testing/selftests/bpf/progs/bpf_tcp_nogpl.c @@ -1,10 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 -#include -#include -#include +#include "bpf_tracing_net.h" #include -#include "bpf_tcp_helpers.h" char _license[] SEC("license") = "X"; -- cgit From c075c9c4af289bb5956b0164283a85cf9c293c8e Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:25 -0700 Subject: selftests/bpf: Remove the bpf_tcp_helpers.h usages from other non tcp-cc tests The patch removes the remaining bpf_tcp_helpers.h usages in the non tcp-cc networking tests. It either replaces it with bpf_tracing_net.h or just removed it because the test is not actually using any kernel sockets. For the later, the missing macro (mainly SOL_TCP) is defined locally. An exception is the test_sock_fields which is testing the "struct bpf_sock" type instead of the kernel sock type. Whenever "vmlinux.h" is used instead, it hits a verifier error on doing arithmetic on the sock_common pointer: ; return !a6[0] && !a6[1] && !a6[2] && a6[3] == bpf_htonl(1); @ test_sock_fields.c:54 21: (61) r2 = *(u32 *)(r1 +28) ; R1_w=sock_common() R2_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff)) 22: (56) if w2 != 0x0 goto pc-6 ; R2_w=0 23: (b7) r3 = 28 ; R3_w=28 24: (bf) r2 = r1 ; R1_w=sock_common() R2_w=sock_common() 25: (0f) r2 += r3 R2 pointer arithmetic on sock_common prohibited Hence, instead of including bpf_tracing_net.h, the test_sock_fields test defines a tcp_sock with one lsndtime field in it. Another highlight is, in sockopt_qos_to_cc.c, the tcp_cc_eq() is replaced by bpf_strncmp(). tcp_cc_eq() was a workaround in bpf_tcp_helpers.h before bpf_strncmp had been added. The SOL_IPV6 addition to bpf_tracing_net.h is needed by the test_tcpbpf_kern test. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-10-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/bpf_tracing_net.h | 1 + tools/testing/selftests/bpf/progs/connect4_prog.c | 6 ++++-- tools/testing/selftests/bpf/progs/mptcp_sock.c | 4 ++-- tools/testing/selftests/bpf/progs/sockopt_qos_to_cc.c | 16 ++++++---------- .../selftests/bpf/progs/test_btf_skc_cls_ingress.c | 16 +++++----------- tools/testing/selftests/bpf/progs/test_sock_fields.c | 5 ++++- tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c | 13 +------------ 7 files changed, 23 insertions(+), 38 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/bpf_tracing_net.h b/tools/testing/selftests/bpf/progs/bpf_tracing_net.h index ba4ca0334586..59843b430f76 100644 --- a/tools/testing/selftests/bpf/progs/bpf_tracing_net.h +++ b/tools/testing/selftests/bpf/progs/bpf_tracing_net.h @@ -25,6 +25,7 @@ #define IP_TOS 1 +#define SOL_IPV6 41 #define IPV6_TCLASS 67 #define IPV6_AUTOFLOWLABEL 70 diff --git a/tools/testing/selftests/bpf/progs/connect4_prog.c b/tools/testing/selftests/bpf/progs/connect4_prog.c index 7ef49ec04838..bec529da7c9d 100644 --- a/tools/testing/selftests/bpf/progs/connect4_prog.c +++ b/tools/testing/selftests/bpf/progs/connect4_prog.c @@ -14,8 +14,6 @@ #include #include -#include "bpf_tcp_helpers.h" - #define SRC_REWRITE_IP4 0x7f000004U #define DST_REWRITE_IP4 0x7f000001U #define DST_REWRITE_PORT4 4444 @@ -32,6 +30,10 @@ #define IFNAMSIZ 16 #endif +#ifndef SOL_TCP +#define SOL_TCP 6 +#endif + __attribute__ ((noinline)) __weak int do_bind(struct bpf_sock_addr *ctx) { diff --git a/tools/testing/selftests/bpf/progs/mptcp_sock.c b/tools/testing/selftests/bpf/progs/mptcp_sock.c index 91a0d7eff2ac..f3acb90588c7 100644 --- a/tools/testing/selftests/bpf/progs/mptcp_sock.c +++ b/tools/testing/selftests/bpf/progs/mptcp_sock.c @@ -2,9 +2,9 @@ /* Copyright (c) 2020, Tessares SA. */ /* Copyright (c) 2022, SUSE. */ -#include +#include "bpf_tracing_net.h" #include -#include "bpf_tcp_helpers.h" +#include char _license[] SEC("license") = "GPL"; __u32 token = 0; diff --git a/tools/testing/selftests/bpf/progs/sockopt_qos_to_cc.c b/tools/testing/selftests/bpf/progs/sockopt_qos_to_cc.c index 83753b00a556..5c3614333b01 100644 --- a/tools/testing/selftests/bpf/progs/sockopt_qos_to_cc.c +++ b/tools/testing/selftests/bpf/progs/sockopt_qos_to_cc.c @@ -1,24 +1,20 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright (c) 2021 Facebook */ -#include -#include -#include -#include -#include -#include "bpf_tcp_helpers.h" +#include "bpf_tracing_net.h" char _license[] SEC("license") = "GPL"; __s32 page_size = 0; +const char cc_reno[TCP_CA_NAME_MAX] = "reno"; +const char cc_cubic[TCP_CA_NAME_MAX] = "cubic"; + SEC("cgroup/setsockopt") int sockopt_qos_to_cc(struct bpf_sockopt *ctx) { void *optval_end = ctx->optval_end; int *optval = ctx->optval; char buf[TCP_CA_NAME_MAX]; - char cc_reno[TCP_CA_NAME_MAX] = "reno"; - char cc_cubic[TCP_CA_NAME_MAX] = "cubic"; if (ctx->level != SOL_IPV6 || ctx->optname != IPV6_TCLASS) goto out; @@ -29,11 +25,11 @@ int sockopt_qos_to_cc(struct bpf_sockopt *ctx) if (bpf_getsockopt(ctx->sk, SOL_TCP, TCP_CONGESTION, &buf, sizeof(buf))) return 0; - if (!tcp_cc_eq(buf, cc_cubic)) + if (bpf_strncmp(buf, sizeof(buf), cc_cubic)) return 0; if (*optval == 0x2d) { - if (bpf_setsockopt(ctx->sk, SOL_TCP, TCP_CONGESTION, &cc_reno, + if (bpf_setsockopt(ctx->sk, SOL_TCP, TCP_CONGESTION, (void *)&cc_reno, sizeof(cc_reno))) return 0; } diff --git a/tools/testing/selftests/bpf/progs/test_btf_skc_cls_ingress.c b/tools/testing/selftests/bpf/progs/test_btf_skc_cls_ingress.c index e2bea4da194b..f0759efff6ef 100644 --- a/tools/testing/selftests/bpf/progs/test_btf_skc_cls_ingress.c +++ b/tools/testing/selftests/bpf/progs/test_btf_skc_cls_ingress.c @@ -1,19 +1,13 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright (c) 2020 Facebook */ -#include -#include -#include -#include -#include -#include -#include -#include -#include - +#include "bpf_tracing_net.h" #include #include -#include "bpf_tcp_helpers.h" + +#ifndef ENOENT +#define ENOENT 2 +#endif struct sockaddr_in6 srv_sa6 = {}; __u16 listen_tp_sport = 0; diff --git a/tools/testing/selftests/bpf/progs/test_sock_fields.c b/tools/testing/selftests/bpf/progs/test_sock_fields.c index f75e531bf36f..196844be349c 100644 --- a/tools/testing/selftests/bpf/progs/test_sock_fields.c +++ b/tools/testing/selftests/bpf/progs/test_sock_fields.c @@ -7,7 +7,6 @@ #include #include -#include "bpf_tcp_helpers.h" enum bpf_linum_array_idx { EGRESS_LINUM_IDX, @@ -42,6 +41,10 @@ struct { __type(value, struct bpf_spinlock_cnt); } sk_pkt_out_cnt10 SEC(".maps"); +struct tcp_sock { + __u32 lsndtime; +} __attribute__((preserve_access_index)); + struct bpf_tcp_sock listen_tp = {}; struct sockaddr_in6 srv_sa6 = {}; struct bpf_tcp_sock cli_tp = {}; diff --git a/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c b/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c index a3f3f43fc195..6935f32eeb8f 100644 --- a/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c +++ b/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c @@ -1,18 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include +#include "bpf_tracing_net.h" #include #include -#include "bpf_tcp_helpers.h" #include "test_tcpbpf.h" struct tcpbpf_globals global = {}; -- cgit From 6a650816b098a15c4690a22e3889858264d01aa8 Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 9 May 2024 10:50:26 -0700 Subject: selftests/bpf: Retire bpf_tcp_helpers.h The previous patches have consolidated the tests to use bpf_tracing_net.h (i.e. vmlinux.h) instead of bpf_tcp_helpers.h. This patch can finally retire the bpf_tcp_helpers.h from the repository. Signed-off-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20240509175026.3423614-11-martin.lau@linux.dev Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/bpf_tcp_helpers.h | 241 -------------------------- 1 file changed, 241 deletions(-) delete mode 100644 tools/testing/selftests/bpf/bpf_tcp_helpers.h diff --git a/tools/testing/selftests/bpf/bpf_tcp_helpers.h b/tools/testing/selftests/bpf/bpf_tcp_helpers.h deleted file mode 100644 index 82a7c9de95f9..000000000000 --- a/tools/testing/selftests/bpf/bpf_tcp_helpers.h +++ /dev/null @@ -1,241 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __BPF_TCP_HELPERS_H -#define __BPF_TCP_HELPERS_H - -#include -#include -#include -#include -#include - -#define BPF_STRUCT_OPS(name, args...) \ -SEC("struct_ops/"#name) \ -BPF_PROG(name, args) - -#ifndef SOL_TCP -#define SOL_TCP 6 -#endif - -#ifndef TCP_CA_NAME_MAX -#define TCP_CA_NAME_MAX 16 -#endif - -#define tcp_jiffies32 ((__u32)bpf_jiffies64()) - -struct sock_common { - unsigned char skc_state; - __u16 skc_num; -} __attribute__((preserve_access_index)); - -enum sk_pacing { - SK_PACING_NONE = 0, - SK_PACING_NEEDED = 1, - SK_PACING_FQ = 2, -}; - -struct sock { - struct sock_common __sk_common; -#define sk_state __sk_common.skc_state - unsigned long sk_pacing_rate; - __u32 sk_pacing_status; /* see enum sk_pacing */ -} __attribute__((preserve_access_index)); - -struct inet_sock { - struct sock sk; -} __attribute__((preserve_access_index)); - -struct inet_connection_sock { - struct inet_sock icsk_inet; - __u8 icsk_ca_state:6, - icsk_ca_setsockopt:1, - icsk_ca_dst_locked:1; - struct { - __u8 pending; - } icsk_ack; - __u64 icsk_ca_priv[104 / sizeof(__u64)]; -} __attribute__((preserve_access_index)); - -struct request_sock { - struct sock_common __req_common; -} __attribute__((preserve_access_index)); - -struct tcp_sock { - struct inet_connection_sock inet_conn; - - __u32 rcv_nxt; - __u32 snd_nxt; - __u32 snd_una; - __u32 window_clamp; - __u8 ecn_flags; - __u32 delivered; - __u32 delivered_ce; - __u32 snd_cwnd; - __u32 snd_cwnd_cnt; - __u32 snd_cwnd_clamp; - __u32 snd_ssthresh; - __u8 syn_data:1, /* SYN includes data */ - syn_fastopen:1, /* SYN includes Fast Open option */ - syn_fastopen_exp:1,/* SYN includes Fast Open exp. option */ - syn_fastopen_ch:1, /* Active TFO re-enabling probe */ - syn_data_acked:1,/* data in SYN is acked by SYN-ACK */ - save_syn:1, /* Save headers of SYN packet */ - is_cwnd_limited:1,/* forward progress limited by snd_cwnd? */ - syn_smc:1; /* SYN includes SMC */ - __u32 max_packets_out; - __u32 lsndtime; - __u32 prior_cwnd; - __u64 tcp_mstamp; /* most recent packet received/sent */ - bool is_mptcp; -} __attribute__((preserve_access_index)); - -static __always_inline struct inet_connection_sock *inet_csk(const struct sock *sk) -{ - return (struct inet_connection_sock *)sk; -} - -static __always_inline void *inet_csk_ca(const struct sock *sk) -{ - return (void *)inet_csk(sk)->icsk_ca_priv; -} - -static __always_inline struct tcp_sock *tcp_sk(const struct sock *sk) -{ - return (struct tcp_sock *)sk; -} - -static __always_inline bool before(__u32 seq1, __u32 seq2) -{ - return (__s32)(seq1-seq2) < 0; -} -#define after(seq2, seq1) before(seq1, seq2) - -#define TCP_ECN_OK 1 -#define TCP_ECN_QUEUE_CWR 2 -#define TCP_ECN_DEMAND_CWR 4 -#define TCP_ECN_SEEN 8 - -enum inet_csk_ack_state_t { - ICSK_ACK_SCHED = 1, - ICSK_ACK_TIMER = 2, - ICSK_ACK_PUSHED = 4, - ICSK_ACK_PUSHED2 = 8, - ICSK_ACK_NOW = 16 /* Send the next ACK immediately (once) */ -}; - -enum tcp_ca_event { - CA_EVENT_TX_START = 0, - CA_EVENT_CWND_RESTART = 1, - CA_EVENT_COMPLETE_CWR = 2, - CA_EVENT_LOSS = 3, - CA_EVENT_ECN_NO_CE = 4, - CA_EVENT_ECN_IS_CE = 5, -}; - -struct ack_sample { - __u32 pkts_acked; - __s32 rtt_us; - __u32 in_flight; -} __attribute__((preserve_access_index)); - -struct rate_sample { - __u64 prior_mstamp; /* starting timestamp for interval */ - __u32 prior_delivered; /* tp->delivered at "prior_mstamp" */ - __s32 delivered; /* number of packets delivered over interval */ - long interval_us; /* time for tp->delivered to incr "delivered" */ - __u32 snd_interval_us; /* snd interval for delivered packets */ - __u32 rcv_interval_us; /* rcv interval for delivered packets */ - long rtt_us; /* RTT of last (S)ACKed packet (or -1) */ - int losses; /* number of packets marked lost upon ACK */ - __u32 acked_sacked; /* number of packets newly (S)ACKed upon ACK */ - __u32 prior_in_flight; /* in flight before this ACK */ - bool is_app_limited; /* is sample from packet with bubble in pipe? */ - bool is_retrans; /* is sample from retransmission? */ - bool is_ack_delayed; /* is this (likely) a delayed ACK? */ -} __attribute__((preserve_access_index)); - -#define TCP_CA_NAME_MAX 16 -#define TCP_CONG_NEEDS_ECN 0x2 - -struct tcp_congestion_ops { - char name[TCP_CA_NAME_MAX]; - __u32 flags; - - /* initialize private data (optional) */ - void (*init)(struct sock *sk); - /* cleanup private data (optional) */ - void (*release)(struct sock *sk); - - /* return slow start threshold (required) */ - __u32 (*ssthresh)(struct sock *sk); - /* do new cwnd calculation (required) */ - void (*cong_avoid)(struct sock *sk, __u32 ack, __u32 acked); - /* call before changing ca_state (optional) */ - void (*set_state)(struct sock *sk, __u8 new_state); - /* call when cwnd event occurs (optional) */ - void (*cwnd_event)(struct sock *sk, enum tcp_ca_event ev); - /* call when ack arrives (optional) */ - void (*in_ack_event)(struct sock *sk, __u32 flags); - /* new value of cwnd after loss (required) */ - __u32 (*undo_cwnd)(struct sock *sk); - /* hook for packet ack accounting (optional) */ - void (*pkts_acked)(struct sock *sk, const struct ack_sample *sample); - /* override sysctl_tcp_min_tso_segs */ - __u32 (*min_tso_segs)(struct sock *sk); - /* returns the multiplier used in tcp_sndbuf_expand (optional) */ - __u32 (*sndbuf_expand)(struct sock *sk); - /* call when packets are delivered to update cwnd and pacing rate, - * after all the ca_state processing. (optional) - */ - void (*cong_control)(struct sock *sk, const struct rate_sample *rs); - void *owner; -}; - -#define min(a, b) ((a) < (b) ? (a) : (b)) -#define max(a, b) ((a) > (b) ? (a) : (b)) -#define min_not_zero(x, y) ({ \ - typeof(x) __x = (x); \ - typeof(y) __y = (y); \ - __x == 0 ? __y : ((__y == 0) ? __x : min(__x, __y)); }) - -static __always_inline bool tcp_in_slow_start(const struct tcp_sock *tp) -{ - return tp->snd_cwnd < tp->snd_ssthresh; -} - -static __always_inline bool tcp_is_cwnd_limited(const struct sock *sk) -{ - const struct tcp_sock *tp = tcp_sk(sk); - - /* If in slow start, ensure cwnd grows to twice what was ACKed. */ - if (tcp_in_slow_start(tp)) - return tp->snd_cwnd < 2 * tp->max_packets_out; - - return !!BPF_CORE_READ_BITFIELD(tp, is_cwnd_limited); -} - -static __always_inline bool tcp_cc_eq(const char *a, const char *b) -{ - int i; - - for (i = 0; i < TCP_CA_NAME_MAX; i++) { - if (a[i] != b[i]) - return false; - if (!a[i]) - break; - } - - return true; -} - -extern __u32 tcp_slow_start(struct tcp_sock *tp, __u32 acked) __ksym; -extern void tcp_cong_avoid_ai(struct tcp_sock *tp, __u32 w, __u32 acked) __ksym; - -struct mptcp_sock { - struct inet_connection_sock sk; - - __u32 token; - struct sock *first; - char ca_name[TCP_CA_NAME_MAX]; -} __attribute__((preserve_access_index)); - -#endif -- cgit