BPF packet scheduler
Subsystems (6)
v2 cover BPF packet scheduler
v2:
- split into more small patches.
- change all parameters of mptcp_sched_ops from sk to msk:
void (*init)(struct mptcp_sock *msk);
void (*release)(struct mptcp_sock *msk);
struct sock * (*get_subflow)(struct mptcp_sock *msk); - add tests in bpf_tcp_ca.c, instead of adding a new one.
Depends on the bpf patches queued in the patchwork:
c42b1c51caa4 selftests: bpf: add bpf_first test49747d1d690e selftests: bpf: add bpf_first scheduler857cc219d855 mptcp: add bpf_mptcp_sched_ops287b994479ea mptcp: add mptcp_get_subflow wrappera626e160359a mptcp: add sched for struct mptcp_sock7613acc6e70e mptcp: add a new sysctl scheduler70ff4322bee0 mptcp: register default scheduler50a302c839de mptcp: add struct mptcp_sched_ops2b41b6b9a948 selftests: bpf: verify first subflow of mptcp_sock 91e52c32a832 selftests: bpf: verify ca_name of struct mptcp_sock 07a66e354126 Squash to "selftests: bpf: test bpf_skc_to_mptcp_sock" 168ee1d1a118 selftests: bpf: test bpf_skc_to_mptcp_sock7a773e6af914 Squash to "selftests: bpf: add MPTCP test base" a1933407b912 bpf: add bpf_skc_to_mptcp_sock_proto0095d8ca9a67 Revert "bpf: add 'bpf_mptcp_sock' structure and helper" 2856d8c8020a Revert "selftests: bpf: add bpf_mptcp_sock() verifier tests" df0e2a4c394f (origin/export, origin/HEAD) DO-NOT-MERGE: mptcp: enabled by default
v1:
Addressed to the commends in the RFC version:
https://patchwork.kernel.org/project/mptcp/cover/[email protected]/
Geliang Tang (8):
mptcp: add struct mptcp_sched_ops
mptcp: register default scheduler
mptcp: add a new sysctl scheduler
mptcp: add sched for struct mptcp_sock
mptcp: add mptcp_get_subflow wrapper
mptcp: add bpf_mptcp_sched_ops
selftests: bpf: add bpf_first scheduler
selftests: bpf: add bpf_first test
Documentation/networking/mptcp-sysctl.rst | 8 ++ include/net/mptcp.h | 13 +++ kernel/bpf/bpf_struct_ops_types.h | 4 + net/mptcp/Makefile | 2 +- net/mptcp/bpf.c | 102 ++++++++++++++++++ net/mptcp/ctrl.c | 14 +++ net/mptcp/protocol.c | 12 ++- net/mptcp/protocol.h | 9 ++ net/mptcp/sched.c | 89 +++++++++++++++ tools/testing/selftests/bpf/bpf_tcp_helpers.h | 12 +++ .../selftests/bpf/prog_tests/bpf_tcp_ca.c | 42 +++++++- tools/testing/selftests/bpf/progs/bpf_first.c | 24 +++++ 12 files changed, 322 insertions(+), 9 deletions(-) create mode 100644 net/mptcp/sched.c create mode 100644 tools/testing/selftests/bpf/progs/bpf_first.c
Signature · 1 line
2.34.1
PATCH v2 1/8 mptcp: add struct mptcp_sched_ops
This patch added struct mptcp_sched_ops. And define the scheduler register, unregister and find functions.
--- include/net/mptcp.h | 13 ++++++++++ net/mptcp/Makefile | 2 +- net/mptcp/protocol.h | 3 +++ net/mptcp/sched.c | 56 ++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 73 insertions(+), 1 deletion(-) create mode 100644 net/mptcp/sched.c
include/net/mptcp.h +13 -0
@@ -95,6 +95,19 @@ struct mptcp_out_options {#endif }; +#define MPTCP_SCHED_NAME_MAX 16 + +struct mptcp_sched_ops { +struct sock * (*get_subflow)(struct mptcp_sock *msk); + +char name[MPTCP_SCHED_NAME_MAX ]; +struct module *owner; +struct list_head list; + +void (*init)(struct mptcp_sock *msk); +void (*release)(struct mptcp_sock *msk); +}____cacheline_aligned_in_smp ; +#ifdef CONFIG_MPTCP extern struct request_sock_ops mptcp_subflow_request_sock_ops;
net/mptcp/Makefile +1 -1
@@ -3,7 +3,7 @@ obj-$(CONFIG_MPTCP) += mptcp.o ccflags-y += -DDEBUG mptcp-y := protocol.o subflow.o options.o token.o crypto.o ctrl.o pm.o diag.o \ - mib.o pm_netlink.o sockopt.o + mib.o pm_netlink.o sockopt.o sched.o obj-$(CONFIG_SYN_COOKIES) += syncookies.o obj-$(CONFIG_INET_MPTCP_DIAG) += mptcp_diag.o
net/mptcp/protocol.h +3 -0
@@ -605,6 +605,9 @@ int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock);void mptcp_info2sockaddr (const struct mptcp_addr_info *info,struct sockaddr_storage *addr,unsigned short family); +struct mptcp_sched_ops *mptcp_sched_find (const char *name); +int mptcp_register_scheduler (struct mptcp_sched_ops *sched); +void mptcp_unregister_scheduler (struct mptcp_sched_ops *sched);static inline bool __mptcp_subflow_active (struct mptcp_subflow_context *subflow) {
net/mptcp/sched.c new file +56 -0
56 lines not on this page · raw message
Signature · 1 line
2.34.1
PATCH v2 2/8 mptcp: register default scheduler
In reply to Geliang TangThis patch defined the default packet scheduler mptcp_sched_default, registered it in mptcp_sched_init().
--- net/mptcp/protocol.c | 3 ++- net/mptcp/protocol.h | 2 ++ net/mptcp/sched.c | 11 +++++++++++ 3 files changed, 15 insertions(+), 1 deletion(-)
net/mptcp/protocol.c +2 -1
@@ -1424,7 +1424,7 @@ bool mptcp_subflow_active(struct mptcp_subflow_context *subflow)* returns the subflow that will transmit the next DSS * additionally updates the rtx timeout */ -static struct sock *mptcp_subflow_get_send (struct mptcp_sock *msk) +struct sock *mptcp_subflow_get_send (struct mptcp_sock *msk) {struct subflow_send_info send_info[SSK_MODE_MAX ];struct mptcp_subflow_context *subflow;
@@ -3743,6 +3743,7 @@ void __init mptcp_proto_init(void)mptcp_subflow_init ();mptcp_pm_init (); +mptcp_sched_init ();mptcp_token_init ();if (proto_register (&mptcp_prot,MPTCP_USE_SLAB ) !=0 )
net/mptcp/protocol.h +2 -0
@@ -608,6 +608,8 @@ void mptcp_info2sockaddr(const struct mptcp_addr_info *info,struct mptcp_sched_ops *mptcp_sched_find (const char *name);int mptcp_register_scheduler (struct mptcp_sched_ops *sched);void mptcp_unregister_scheduler (struct mptcp_sched_ops *sched); +struct sock *mptcp_subflow_get_send (struct mptcp_sock *msk); +void mptcp_sched_init (void );static inline bool __mptcp_subflow_active (struct mptcp_subflow_context *subflow) {
net/mptcp/sched.c +11 -0
@@ -54,3 +54,14 @@ void mptcp_unregister_scheduler(struct mptcp_sched_ops *sched)synchronize_rcu (); } + +static struct mptcp_sched_ops mptcp_sched_default = { + .get_subflow = mptcp_subflow_get_send, + .name ="default" , + .owner =THIS_MODULE , +}; + +void mptcp_sched_init (void ) +{ +mptcp_register_scheduler (&mptcp_sched_default); +}
Signature · 1 line
2.34.1
PATCH v2 3/8 mptcp: add a new sysctl scheduler
In reply to Geliang TangThis patch added a new sysctl, named scheduler, to support for selection of different schedulers.
--- Documentation/networking/mptcp-sysctl.rst | 8 ++++++++ net/mptcp/ctrl.c | 14 ++++++++++++++ net/mptcp/protocol.h | 1 + 3 files changed, 23 insertions(+)
Documentation/networking/mptcp-sysctl.rst +8 -0
@@ -75,3 +75,11 @@ stale_loss_cnt - INTEGER This is a per-namespace sysctl. Default: 4 + +scheduler - STRING + Select the scheduler of your choice. + + Support for selection of different schedulers. This is a per-namespace + sysctl. + + Default: "default"
net/mptcp/ctrl.c +14 -0
@@ -32,6 +32,7 @@ struct mptcp_pernet {u8 checksum_enabled;u8 allow_join_initial_addr_port;u8 pm_type; +char scheduler[MPTCP_SCHED_NAME_MAX ]; };static struct mptcp_pernet *mptcp_get_pernet (const struct net *net)
@@ -69,6 +70,11 @@ int mptcp_get_pm_type(const struct net *net)return mptcp_get_pernet (net)->pm_type; } +const char *mptcp_get_scheduler (const struct net *net) +{ +return mptcp_get_pernet (net)->scheduler; +} +static void mptcp_pernet_set_defaults (struct mptcp_pernet *pernet) { pernet->mptcp_enabled =1 ;
@@ -77,6 +83,7 @@ static void mptcp_pernet_set_defaults(struct mptcp_pernet *pernet) pernet->allow_join_initial_addr_port =1 ; pernet->stale_loss_cnt =4 ; pernet->pm_type =MPTCP_PM_TYPE_KERNEL ; +strcpy (pernet->scheduler,"default" ); }#ifdef CONFIG_SYSCTL
@@ -128,6 +135,12 @@ static struct ctl_table mptcp_sysctl_table[] = { .extra1 = SYSCTL_ZERO, .extra2 = &mptcp_pm_type_max }, + { + .procname = "scheduler", + .maxlen = MPTCP_SCHED_NAME_MAX, + .mode = 0644, + .proc_handler = proc_dostring, + }, {} };
@@ -149,6 +162,7 @@ static int mptcp_pernet_new_table(struct net *net, struct mptcp_pernet *pernet) table[3 ].data = &pernet->allow_join_initial_addr_port; table[4 ].data = &pernet->stale_loss_cnt; table[5 ].data = &pernet->pm_type; + table[6 ].data = &pernet->scheduler; hdr =register_net_sysctl (net,MPTCP_SYSCTL_PATH , table);if (!hdr)
net/mptcp/protocol.h +1 -0
@@ -583,6 +583,7 @@ int mptcp_is_checksum_enabled(const struct net *net);int mptcp_allow_join_id0 (const struct net *net);unsigned int mptcp_stale_loss_cnt (const struct net *net);int mptcp_get_pm_type (const struct net *net); +const char *mptcp_get_scheduler (const struct net *net);void mptcp_subflow_fully_established (struct mptcp_subflow_context *subflow,struct mptcp_options_received *mp_opt);bool __mptcp_retransmit_pending_data (struct sock *sk);
Signature · 1 line
2.34.1
PATCH v2 4/8 mptcp: add sched for struct mptcp_sock
In reply to Geliang TangThis patch added a new struct member sched in struct mptcp_sock. Set it in mptcp_set_sched() with the scheduler sysctl.
--- net/mptcp/protocol.c | 3 +++ net/mptcp/protocol.h | 2 ++ net/mptcp/sched.c | 16 ++++++++++++++++ 3 files changed, 21 insertions(+)
net/mptcp/protocol.c +3 -0
@@ -2609,6 +2609,7 @@ static int mptcp_init_sock(struct sock *sk)* propagate the correct value */ mptcp_ca_reset (sk); +mptcp_set_sched (sk);sk_sockets_allocated_inc (sk); sk->sk_rcvbuf =sock_net (sk)->ipv4.sysctl_tcp_rmem[1 ];
@@ -2760,6 +2761,7 @@ static void __mptcp_destroy_sock(struct sock *sk)sk_stop_timer (sk, &msk->sk.icsk_retransmit_timer);sk_stop_timer (sk, &sk->sk_timer); msk->pm.status =0 ; + msk->sched =NULL ;/* clears msk->subflow, allowing the following loop to close * even the initial subflow
@@ -2933,6 +2935,7 @@ struct sock *mptcp_sk_clone(const struct sock *sk, msk->snd_una = msk->write_seq; msk->wnd_end = msk->snd_nxt + req->rsk_rcv_wnd; msk->setsockopt_seq =mptcp_sk (sk)->setsockopt_seq; + msk->sched =mptcp_sk (sk)->sched;if (mp_opt->suboptions &OPTIONS_MPTCP_MPC ) { msk->can_ack = true;
net/mptcp/protocol.h +2 -0
@@ -286,6 +286,7 @@ struct mptcp_sock {struct socket *subflow;/* outgoing connect/listener/!mp_capable */ struct sock *first;struct mptcp_pm_data pm; +struct mptcp_sched_ops *sched;struct {u32 space;/* bytes copied in last measurement window */ u32 copied;/* bytes copied in this measurement window */
@@ -609,6 +610,7 @@ void mptcp_info2sockaddr(const struct mptcp_addr_info *info,struct mptcp_sched_ops *mptcp_sched_find (const char *name);int mptcp_register_scheduler (struct mptcp_sched_ops *sched);void mptcp_unregister_scheduler (struct mptcp_sched_ops *sched); +void mptcp_set_sched (struct sock *sk);struct sock *mptcp_subflow_get_send (struct mptcp_sock *msk);void mptcp_sched_init (void );
net/mptcp/sched.c +16 -0
@@ -55,6 +55,22 @@ void mptcp_unregister_scheduler(struct mptcp_sched_ops *sched)synchronize_rcu (); } +void mptcp_set_sched (struct sock *sk) +{ +struct mptcp_sock *msk =mptcp_sk (sk); +struct net *net =sock_net (sk); + + msk->sched =mptcp_sched_find (mptcp_get_scheduler (net)); +if (!msk->sched) { +pr_debug ("sched %s not found" ,mptcp_get_scheduler (net)); +return ; + } + +pr_debug ("sched=%s" , msk->sched->name); +if (msk->sched->init) + msk->sched->init (msk); +} +static struct mptcp_sched_ops mptcp_sched_default = { .get_subflow = mptcp_subflow_get_send, .name ="default" ,
Signature · 1 line
2.34.1
PATCH v2 5/8 mptcp: add mptcp_get_subflow wrapper
In reply to Geliang TangThis patch defined a new wrapper mptcp_get_subflow(), used it instead of mptcp_subflow_get_send().
--- net/mptcp/protocol.c | 6 +++--- net/mptcp/protocol.h | 1 + net/mptcp/sched.c | 6 ++++++ 3 files changed, 10 insertions(+), 3 deletions(-)
net/mptcp/protocol.c +3 -3
@@ -1567,7 +1567,7 @@ void __mptcp_push_pending(struct sock *sk, unsigned int flags)int ret =0 ; prev_ssk = ssk; - ssk =mptcp_subflow_get_send (msk); + ssk =mptcp_get_subflow (msk);/* First check. If the ssk has changed since * the last round, release prev_ssk
@@ -1634,7 +1634,7 @@ static void __mptcp_subflow_push_pending(struct sock *sk, struct sock *ssk)* check for a different subflow usage only after * spooling the first chunk of data */ - xmit_ssk = first ? ssk :mptcp_subflow_get_send (mptcp_sk (sk)); + xmit_ssk = first ? ssk :mptcp_get_subflow (mptcp_sk (sk));if (!xmit_ssk) goto out;if (xmit_ssk != ssk) {
@@ -3073,7 +3073,7 @@ void __mptcp_check_push(struct sock *sk, struct sock *ssk)return ;if (!sock_owned_by_user (sk)) { -struct sock *xmit_ssk =mptcp_subflow_get_send (mptcp_sk (sk)); +struct sock *xmit_ssk =mptcp_get_subflow (mptcp_sk (sk));if (xmit_ssk == ssk)__mptcp_subflow_push_pending (sk, ssk);
net/mptcp/protocol.h +1 -0
@@ -613,6 +613,7 @@ void mptcp_unregister_scheduler(struct mptcp_sched_ops *sched);void mptcp_set_sched (struct sock *sk);struct sock *mptcp_subflow_get_send (struct mptcp_sock *msk);void mptcp_sched_init (void ); +struct sock *mptcp_get_subflow (struct mptcp_sock *msk);static inline bool __mptcp_subflow_active (struct mptcp_subflow_context *subflow) {
net/mptcp/sched.c +6 -0
@@ -81,3 +81,9 @@ void mptcp_sched_init(void) {mptcp_register_scheduler (&mptcp_sched_default); } + +struct sock *mptcp_get_subflow (struct mptcp_sock *msk) +{ +return INDIRECT_CALL_INET_1 (msk->sched->get_subflow, + mptcp_subflow_get_send, msk); +}
Signature · 1 line
2.34.1
PATCH v2 6/8 mptcp: add bpf_mptcp_sched_ops
In reply to Geliang TangThis patch implemented a new struct bpf_struct_ops, bpf_mptcp_sched_ops. Register and unregister the mptcp scheduler in .reg and .unreg.
--- kernel/bpf/bpf_struct_ops_types.h | 4 ++ net/mptcp/bpf.c | 102 ++++++++++++++++++++++++++++++ 2 files changed, 106 insertions(+)
kernel/bpf/bpf_struct_ops_types.h +4 -0
@@ -8,5 +8,9 @@ BPF_STRUCT_OPS_TYPE(bpf_dummy_ops)#ifdef CONFIG_INET #include <net/tcp.h> BPF_STRUCT_OPS_TYPE(tcp_congestion_ops) +#ifdef CONFIG_MPTCP +#include <net/mptcp.h> +BPF_STRUCT_OPS_TYPE(mptcp_sched_ops) +#endif #endif #endif
net/mptcp/bpf.c +102 -0
110 lines not on this page · raw message
Signature · 1 line
2.34.1
PATCH v2 7/8 selftests: bpf: add bpf_first scheduler
In reply to Geliang TangThis patch implemented the simplest MPTCP scheduler, named bpf_first, which always picks the first subflow to send data.
--- tools/testing/selftests/bpf/bpf_tcp_helpers.h | 12 ++++++++++ tools/testing/selftests/bpf/progs/bpf_first.c | 24 +++++++++++++++++++ 2 files changed, 36 insertions(+) create mode 100644 tools/testing/selftests/bpf/progs/bpf_first.c
tools/testing/selftests/bpf/bpf_tcp_helpers.h +12 -0
@@ -197,6 +197,18 @@ struct tcp_congestion_ops {void *owner; }; +#define MPTCP_SCHED_NAME_MAX 16 + +struct mptcp_sched_ops { +char name[MPTCP_SCHED_NAME_MAX ]; + +void (*init)(struct mptcp_sock *msk); +void (*release)(struct mptcp_sock *msk); + +struct sock * (*get_subflow)(struct mptcp_sock *msk); +void *owner; +}; +#define min (a, b) ((a) < (b) ? (a) : (b))#define max (a, b) ((a) > (b) ? (a) : (b))#define min_not_zero (x, y) ({ \
tools/testing/selftests/bpf/progs/bpf_first.c new file +24 -0
@@ -0,0 +1,24 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Copyright (c) 2022, SUSE. */ + +#include <linux/bpf.h> +#include "bpf_tcp_helpers.h" + +char _license[] SEC("license") = "GPL"; + +SEC("struct_ops/mptcp_sched_first_init") +void BPF_PROG (mptcp_sched_first_init ,struct mptcp_sock *msk) +{ +} + +struct sock *BPF_STRUCT_OPS (bpf_first_get_subflow ,struct mptcp_sock *msk) +{ +return msk->first; +} + +SEC(".struct_ops") +struct mptcp_sched_ops first = { + .init = (void *)mptcp_sched_first_init, + .get_subflow = (void *)bpf_first_get_subflow, + .name ="bpf_first" , +};
Signature · 1 line
2.34.1
PATCH v2 8/8 selftests: bpf: add bpf_first test
In reply to Geliang TangThis patch added the MPTCP sched testcases. Use sysctl to set net.mptcp.scheduler in the tests.
--- .../selftests/bpf/prog_tests/bpf_tcp_ca.c | 42 +++++++++++++++++-- 1 file changed, 38 insertions(+), 4 deletions(-)
tools/testing/selftests/bpf/prog_tests/bpf_tcp_ca.c +38 -4
@@ -9,6 +9,7 @@#include "bpf_cubic.skel.h" #include "bpf_tcp_nogpl.skel.h" #include "bpf_dctcp_release.skel.h" +#include "bpf_first.skel.h" #define min (a, b) ((a) < (b) ? (a) : (b))
@@ -16,6 +17,10 @@#define ENOTSUPP 524#endif +#ifndef IPPROTO_MPTCP +#define IPPROTO_MPTCP 262 +#endif +static const unsigned int total_bytes =10 *1024 *1024 ;static int expected_stg =0xeB9F ;static int stop, duration;
@@ -85,21 +90,26 @@ static void do_test(const char *tcp_ca, const struct bpf_map *sk_stg_map)socklen_t addrlen =sizeof (sa6);void *thread_ret;char batch[1500 ]; +int proto =0 ;int err;WRITE_ONCE (stop,0 ); - lfd =socket (AF_INET6 ,SOCK_STREAM ,0 ); +if (!strcmp (tcp_ca,"bpf_first" )) + proto =IPPROTO_MPTCP ; + + lfd =socket (AF_INET6 ,SOCK_STREAM , proto);if (CHECK (lfd ==-1 ,"socket" ,"errno:%d\n" , errno))return ; - fd =socket (AF_INET6 ,SOCK_STREAM ,0 ); + fd =socket (AF_INET6 ,SOCK_STREAM , proto);if (CHECK (fd ==-1 ,"socket" ,"errno:%d\n" , errno)) {close (lfd);return ; } -if (settcpca (lfd, tcp_ca) ||settcpca (fd, tcp_ca) || -settimeo (lfd,0 ) ||settimeo (fd,0 )) +if (!proto && + (settcpca (lfd, tcp_ca) ||settcpca (fd, tcp_ca) || +settimeo (lfd,0 ) ||settimeo (fd,0 ))) goto done;/* bind, listen and start server thread to accept */
@@ -324,6 +334,28 @@ static void test_rel_setsockopt(void)bpf_dctcp_release__destroy (rel_skel); } +static void test_first (void ) +{ +struct bpf_first *first_skel; +struct bpf_link *link; + + first_skel =bpf_first__open_and_load (); +if (CHECK (!first_skel,"bpf_first__open_and_load" ,"failed\n" )) +return ; + + link =bpf_map__attach_struct_ops (first_skel->maps.first); +if (!ASSERT_OK_PTR (link,"bpf_map__attach_struct_ops" )) { +bpf_first__destroy (first_skel); +return ; + } + +system ("sysctl -q net.mptcp.scheduler=bpf_first" ); +do_test ("bpf_first" ,NULL ); + +bpf_link__destroy (link); +bpf_first__destroy (first_skel); +} +void test_bpf_tcp_ca (void ) {if (test__start_subtest ("dctcp" ))
@@ -336,4 +368,6 @@ void test_bpf_tcp_ca(void)test_dctcp_fallback ();if (test__start_subtest ("rel_setsockopt" ))test_rel_setsockopt (); +if (test__start_subtest ("first" )) +test_first (); }
Signature · 1 line
2.34.1