| Message ID | 1436195511-32314-2-git-send-email-ubraun@linux.vnet.ibm.com |
|---|---|
| State | Changes Requested, archived |
| Delegated to: | David Miller |
| Headers | show |
On Mon, 2015-07-06 at 17:11 +0200, Ursula Braun wrote: > From: Ursula Braun <ursula.braun@de.ibm.com> > > The SMC-R protocol defines dynamic discovery of peers. This is done by > implementing experimental TCP options as defined in RFC6994. The TCP code > needs to be extended to support RFC6994. > > Setting the TCP experimental option for SMC-R [2] will be triggered from > kernel exploiters like the new SMC-R socket family by setting a new > flag "syn_smc" on struct tcp_sock of the connecting and the listening > socket. If the client peer is SMC-R capable, flag syn_smc is kept on the > connecting socket after the 3-way TCP handshake, otherwise it is reset. > If the server peer is SMC-R capable, the new connected TCP socket has > the new flag set, otherwise not. > > Code snippet client: > tcp_sk(sock->sk)->syn_smc = 1; > rc = kernel_connect(sock, addr, alen, flags); > if (tcp_sk(sock->sk)->syn_smc) { > /* switch to smc for this connection */ > > Code snippet server: > tcp_sk(sock->sk)->syn_smc = 1; > rc = kernel_listen(sock, backlog); > rc = kernel_accept(sock, &newsock, 0); > if (tcp_sk(newsock->sk)->syn_smc) { > /* switch to smc for this connection */ > > References: > [1] Shared Use of TCP Experimental Options RFC 6994: > https://tools.ietf.org/rfc/rfc6994.txt > [2] IANA ExID SMCR: > http://www.iana.org/assignments/tcp-parameters/tcp-parameters.xhtml#tcp-exids > > This patch has already been posted in June 2013, but Dave Miller has > postponed applying till the user of the new flags, ie. the entire SMC-R > protocol stack is implemented. > > Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com> > struct tcp_out_options { > u16 options; /* bit field of OPTION_* */ > @@ -544,6 +545,14 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, > } > ptr += (len + 3) >> 2; > } > + > + if (unlikely(OPTION_SMC & options)) { > + *ptr++ = htonl((TCPOPT_NOP << 24) | > + (TCPOPT_NOP << 16) | > + (TCPOPT_EXP << 8) | > + (TCPOLEN_EXP_SMC_BASE)); > + *ptr++ = htonl(TCPOPT_SMC_MAGIC); > + } > } I am concerned about adding an additional conditional branch in TCP write fast path, on all hosts, while SMC seems to be available only for some hardware class. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Eric, understood, would it be acceptable if the SMC-specific hooks in the TCP-code are enclosed with "#ifdef CONFIG_SMC ... #endif"? Regards, Ursula On Mon, 2015-07-06 at 18:08 +0200, Eric Dumazet wrote: > On Mon, 2015-07-06 at 17:11 +0200, Ursula Braun wrote: > > From: Ursula Braun <ursula.braun@de.ibm.com> > > > > The SMC-R protocol defines dynamic discovery of peers. This is done by > > implementing experimental TCP options as defined in RFC6994. The TCP code > > needs to be extended to support RFC6994. > > > > Setting the TCP experimental option for SMC-R [2] will be triggered from > > kernel exploiters like the new SMC-R socket family by setting a new > > flag "syn_smc" on struct tcp_sock of the connecting and the listening > > socket. If the client peer is SMC-R capable, flag syn_smc is kept on the > > connecting socket after the 3-way TCP handshake, otherwise it is reset. > > If the server peer is SMC-R capable, the new connected TCP socket has > > the new flag set, otherwise not. > > > > Code snippet client: > > tcp_sk(sock->sk)->syn_smc = 1; > > rc = kernel_connect(sock, addr, alen, flags); > > if (tcp_sk(sock->sk)->syn_smc) { > > /* switch to smc for this connection */ > > > > Code snippet server: > > tcp_sk(sock->sk)->syn_smc = 1; > > rc = kernel_listen(sock, backlog); > > rc = kernel_accept(sock, &newsock, 0); > > if (tcp_sk(newsock->sk)->syn_smc) { > > /* switch to smc for this connection */ > > > > References: > > [1] Shared Use of TCP Experimental Options RFC 6994: > > https://tools.ietf.org/rfc/rfc6994.txt > > [2] IANA ExID SMCR: > > http://www.iana.org/assignments/tcp-parameters/tcp-parameters.xhtml#tcp-exids > > > > This patch has already been posted in June 2013, but Dave Miller has > > postponed applying till the user of the new flags, ie. the entire SMC-R > > protocol stack is implemented. > > > > Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com> > > > > struct tcp_out_options { > > u16 options; /* bit field of OPTION_* */ > > @@ -544,6 +545,14 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, > > } > > ptr += (len + 3) >> 2; > > } > > + > > + if (unlikely(OPTION_SMC & options)) { > > + *ptr++ = htonl((TCPOPT_NOP << 24) | > > + (TCPOPT_NOP << 16) | > > + (TCPOPT_EXP << 8) | > > + (TCPOLEN_EXP_SMC_BASE)); > > + *ptr++ = htonl(TCPOPT_SMC_MAGIC); > > + } > > } > > > I am concerned about adding an additional conditional branch in TCP > write fast path, on all hosts, while SMC seems to be available only for > some hardware class. > > > > -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, 2015-07-07 at 15:57 +0200, Ursula Braun wrote: > Eric, > > understood, would it be acceptable if the SMC-specific hooks in the > TCP-code are enclosed with "#ifdef CONFIG_SMC ... #endif"? If this CONFIG_SMC is enabled only on relevant builds, I guess it would be ok. (Try to use helpers in include files to avoid spreading new #ifdef in C files) -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/include/linux/tcp.h b/include/linux/tcp.h index 48c3696..de0d67c 100644 --- a/include/linux/tcp.h +++ b/include/linux/tcp.h @@ -90,6 +90,7 @@ struct tcp_options_received { sack_ok : 4, /* SACK seen on SYN packet */ snd_wscale : 4, /* Window scaling received from sender */ rcv_wscale : 4; /* Window scaling to send to receiver */ + u8 smc_capability:1; /* SMC capability */ u8 num_sacks; /* Number of SACK blocks */ u16 user_mss; /* mss requested by user in ioctl */ u16 mss_clamp; /* Maximal mss, negotiated at connection setup */ @@ -99,6 +100,7 @@ static inline void tcp_clear_options(struct tcp_options_received *rx_opt) { rx_opt->tstamp_ok = rx_opt->sack_ok = 0; rx_opt->wscale_ok = rx_opt->snd_wscale = 0; + rx_opt->smc_capability = 0; } /* This is the max number of SACKS that we'll generate and process. It's safe @@ -207,7 +209,8 @@ struct tcp_sock { syn_fastopen_exp:1,/* SYN includes Fast Open exp. option */ 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? */ + is_cwnd_limited:1,/* forward progress limited by snd_cwnd? */ + syn_smc:1; /* SYN includes SMC */ u32 tlp_high_seq; /* snd_nxt at the time of TLP retransmit. */ /* RTT measurement */ diff --git a/include/net/request_sock.h b/include/net/request_sock.h index 87935ca..dee47d2 100644 --- a/include/net/request_sock.h +++ b/include/net/request_sock.h @@ -55,7 +55,8 @@ struct request_sock { struct sock *rsk_listener; u16 mss; u8 num_retrans; /* number of retransmits */ - u8 cookie_ts:1; /* syncookie: encode tcpopts in timestamp */ + u8 cookie_ts:1, /* syncookie: encode tcpopts in timestamp */ + smc_capability:1; u8 num_timeout:7; /* number of timeouts */ /* The following two fields can be easily recomputed I think -AK */ u32 window_clamp; /* window clamp at creation time */ diff --git a/include/net/tcp.h b/include/net/tcp.h index 950cfec..882e8d5 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -185,6 +185,7 @@ void tcp_time_wait(struct sock *sk, int state, int timeo); * experimental options. See draft-ietf-tcpm-experimental-options-00.txt */ #define TCPOPT_FASTOPEN_MAGIC 0xF989 +#define TCPOPT_SMC_MAGIC 0xE2D4C3D9 /* * TCP option lengths @@ -197,6 +198,7 @@ void tcp_time_wait(struct sock *sk, int state, int timeo); #define TCPOLEN_MD5SIG 18 #define TCPOLEN_FASTOPEN_BASE 2 #define TCPOLEN_EXP_FASTOPEN_BASE 4 +#define TCPOLEN_EXP_SMC_BASE 6 /* But this is what stacks really send out. */ #define TCPOLEN_TSTAMP_ALIGNED 12 @@ -207,6 +209,7 @@ void tcp_time_wait(struct sock *sk, int state, int timeo); #define TCPOLEN_SACK_PERBLOCK 8 #define TCPOLEN_MD5SIG_ALIGNED 20 #define TCPOLEN_MSS_ALIGNED 4 +#define TCPOLEN_EXP_SMC_BASE_ALIGNED 8 /* Flags in tp->nonagle */ #define TCP_NAGLE_OFF 1 /* Nagle's algo is disabled */ diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 684f095..0cde982 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -3743,6 +3743,11 @@ void tcp_parse_options(const struct sk_buff *skb, tcp_parse_fastopen_option(opsize - TCPOLEN_EXP_FASTOPEN_BASE, ptr + 2, th->syn, foc, true); + else if (th->syn && !(opsize & 1) && + opsize >= TCPOLEN_EXP_SMC_BASE && + get_unaligned_be32(ptr) == + TCPOPT_SMC_MAGIC) + opt_rx->smc_capability = 1; break; } @@ -5554,6 +5559,9 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, * is initialized. */ tp->copied_seq = tp->rcv_nxt; + if (tp->syn_smc && !tp->rx_opt.smc_capability) + tp->syn_smc = 0; + smp_mb(); tcp_finish_connect(sk, skb); @@ -6000,6 +6008,7 @@ static void tcp_openreq_init(struct request_sock *req, req->rcv_wnd = 0; /* So that tcp_send_synack() knows! */ req->cookie_ts = 0; + req->smc_capability = 0; tcp_rsk(req)->rcv_isn = TCP_SKB_CB(skb)->seq; tcp_rsk(req)->rcv_nxt = TCP_SKB_CB(skb)->seq + 1; tcp_rsk(req)->snt_synack = tcp_time_stamp; @@ -6140,6 +6149,9 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops, /* Note: tcp_v6_init_req() might override ir_iif for link locals */ inet_rsk(req)->ir_iif = sk->sk_bound_dev_if; + if (tmp_opt.smc_capability) + req->smc_capability = 1; + af_ops->init_req(req, sk, skb); if (security_inet_conn_request(sk, skb, req)) diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index 4bc00cb..3a19d97 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -445,6 +445,10 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, struct tcp_request_sock *treq = tcp_rsk(req); struct inet_connection_sock *newicsk = inet_csk(newsk); struct tcp_sock *newtp = tcp_sk(newsk); + struct tcp_sock *oldtp = tcp_sk(sk); + + if (oldtp->syn_smc && !req->smc_capability) + newtp->syn_smc = 0; /* Now setup tcp_sock */ newtp->pred_flags = 0; diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index b1c218d..acb6d8d 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -426,6 +426,7 @@ static inline bool tcp_urg_mode(const struct tcp_sock *tp) #define OPTION_MD5 (1 << 2) #define OPTION_WSCALE (1 << 3) #define OPTION_FAST_OPEN_COOKIE (1 << 8) +#define OPTION_SMC (1 << 9) struct tcp_out_options { u16 options; /* bit field of OPTION_* */ @@ -544,6 +545,14 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, } ptr += (len + 3) >> 2; } + + if (unlikely(OPTION_SMC & options)) { + *ptr++ = htonl((TCPOPT_NOP << 24) | + (TCPOPT_NOP << 16) | + (TCPOPT_EXP << 8) | + (TCPOLEN_EXP_SMC_BASE)); + *ptr++ = htonl(TCPOPT_SMC_MAGIC); + } } /* Compute TCP options for SYN packets. This is not the final @@ -611,6 +620,15 @@ static unsigned int tcp_syn_options(struct sock *sk, struct sk_buff *skb, } } + if (tp->syn_smc) { + u32 need = TCPOLEN_EXP_SMC_BASE_ALIGNED; + + if (remaining >= need) { + opts->options |= OPTION_SMC; + remaining -= need; + } + } + return MAX_TCP_OPTION_SPACE - remaining; } @@ -623,6 +641,7 @@ static unsigned int tcp_synack_options(struct sock *sk, struct tcp_fastopen_cookie *foc) { struct inet_request_sock *ireq = inet_rsk(req); + struct tcp_sock *tp = tcp_sk(sk); unsigned int remaining = MAX_TCP_OPTION_SPACE; #ifdef CONFIG_TCP_MD5SIG @@ -672,6 +691,15 @@ static unsigned int tcp_synack_options(struct sock *sk, } } + if (tp->syn_smc && req->smc_capability) { + u32 need = TCPOLEN_EXP_SMC_BASE_ALIGNED; + + if (remaining >= need) { + opts->options |= OPTION_SMC; + remaining -= need; + } + } + return MAX_TCP_OPTION_SPACE - remaining; }