[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <b699df99-9375-9d7c-6e84-5e2131a93ddd@linux.ibm.com>
Date: Fri, 18 Aug 2023 21:41:20 +0200
From: Jan Karcher <jaka@...ux.ibm.com>
To: Guangguan Wang <guangguan.wang@...ux.alibaba.com>,
wenjia@...ux.ibm.com, kgraul@...ux.ibm.com,
tonylu@...ux.alibaba.com, davem@...emloft.net, edumazet@...gle.com,
kuba@...nel.org, pabeni@...hat.com
Cc: horms@...nel.org, alibuda@...ux.alibaba.com,
guwen@...ux.alibaba.com, linux-s390@...r.kernel.org,
netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH net-next v2 5/6] net/smc: support max links per lgr
negotiation in clc handshake
On 17/08/2023 15:20, Guangguan Wang wrote:
> Support max links per lgr negotiation in clc handshake for SMCR v2.1,
> which is one of smc v2.1 features. Server makes decision for the final
> value of max links based on the client preferred max links and
> self-preferred max links. Here use the minimum value of the client
> preferred max links and server preferred max links.
>
> Client Server
> Proposal(max links(client preferred))
> -------------------------------------->
>
> Accept(max links(accepted value))
> accepted value=min(client preferred, server preferred)
> <-------------------------------------
>
> Confirm(max links(accepted value))
> ------------------------------------->
>
> Signed-off-by: Guangguan Wang <guangguan.wang@...ux.alibaba.com>
> Reviewed-by: Tony Lu <tonylu@...ux.alibaba.com>
Thank you for your contribution, Guangguan.
Reviewed-by: Jan Karcher <jaka@...ux.ibm.com>
> ---
> net/smc/af_smc.c | 44 +++++++++++++++++++++++++++-----------------
> net/smc/smc_clc.c | 17 ++++++++++++++++-
> net/smc/smc_clc.h | 7 +++++--
> net/smc/smc_core.c | 5 +++++
> net/smc/smc_core.h | 12 ++++++++++++
> net/smc/smc_llc.c | 21 +++++++++++++++++----
> 6 files changed, 82 insertions(+), 24 deletions(-)
>
> diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
> index 8f041e871ddd..4d7152c8a80d 100644
> --- a/net/smc/af_smc.c
> +++ b/net/smc/af_smc.c
> @@ -610,20 +610,22 @@ static int smcr_clnt_conf_first_link(struct smc_sock *smc)
> smc_llc_link_active(link);
> smcr_lgr_set_type(link->lgr, SMC_LGR_SINGLE);
>
> - /* optional 2nd link, receive ADD LINK request from server */
> - qentry = smc_llc_wait(link->lgr, NULL, SMC_LLC_WAIT_TIME,
> - SMC_LLC_ADD_LINK);
> - if (!qentry) {
> - struct smc_clc_msg_decline dclc;
> -
> - rc = smc_clc_wait_msg(smc, &dclc, sizeof(dclc),
> - SMC_CLC_DECLINE, CLC_WAIT_TIME_SHORT);
> - if (rc == -EAGAIN)
> - rc = 0; /* no DECLINE received, go with one link */
> - return rc;
> + if (link->lgr->max_links > 1) {
> + /* optional 2nd link, receive ADD LINK request from server */
> + qentry = smc_llc_wait(link->lgr, NULL, SMC_LLC_WAIT_TIME,
> + SMC_LLC_ADD_LINK);
> + if (!qentry) {
> + struct smc_clc_msg_decline dclc;
> +
> + rc = smc_clc_wait_msg(smc, &dclc, sizeof(dclc),
> + SMC_CLC_DECLINE, CLC_WAIT_TIME_SHORT);
> + if (rc == -EAGAIN)
> + rc = 0; /* no DECLINE received, go with one link */
> + return rc;
> + }
> + smc_llc_flow_qentry_clr(&link->lgr->llc_flow_lcl);
> + smc_llc_cli_add_link(link, qentry);
> }
> - smc_llc_flow_qentry_clr(&link->lgr->llc_flow_lcl);
> - smc_llc_cli_add_link(link, qentry);
> return 0;
> }
>
> @@ -1212,6 +1214,7 @@ static int smc_connect_rdma(struct smc_sock *smc,
> memcpy(ini->peer_gid, aclc->r0.lcl.gid, SMC_GID_SIZE);
> memcpy(ini->peer_mac, aclc->r0.lcl.mac, ETH_ALEN);
> ini->max_conns = SMC_CONN_PER_LGR_MAX;
> + ini->max_links = SMC_LINKS_ADD_LNK_MAX;
>
> reason_code = smc_connect_rdma_v2_prepare(smc, aclc, ini);
> if (reason_code)
> @@ -1856,10 +1859,12 @@ static int smcr_serv_conf_first_link(struct smc_sock *smc)
> smc_llc_link_active(link);
> smcr_lgr_set_type(link->lgr, SMC_LGR_SINGLE);
>
> - down_write(&link->lgr->llc_conf_mutex);
> - /* initial contact - try to establish second link */
> - smc_llc_srv_add_link(link, NULL);
> - up_write(&link->lgr->llc_conf_mutex);
> + if (link->lgr->max_links > 1) {
> + down_write(&link->lgr->llc_conf_mutex);
> + /* initial contact - try to establish second link */
> + smc_llc_srv_add_link(link, NULL);
> + up_write(&link->lgr->llc_conf_mutex);
> + }
> return 0;
> }
>
> @@ -2463,6 +2468,11 @@ static void smc_listen_work(struct work_struct *work)
> goto out_decl;
> }
>
> + /* fce smc release version is needed in smc_listen_rdma_finish,
> + * so save fce info here.
> + */
> + smc_conn_save_peer_info_fce(new_smc, cclc);
> +
> /* finish worker */
> if (!ini->is_smcd) {
> rc = smc_listen_rdma_finish(new_smc, cclc,
> diff --git a/net/smc/smc_clc.c b/net/smc/smc_clc.c
> index 60cfe37625b8..9573920c1773 100644
> --- a/net/smc/smc_clc.c
> +++ b/net/smc/smc_clc.c
> @@ -433,8 +433,10 @@ static int smc_clc_fill_fce(struct smc_clc_first_contact_ext_v2x *fce,
> }
>
> if (ini->release_nr >= SMC_RELEASE_1) {
> - if (!ini->is_smcd)
> + if (!ini->is_smcd) {
> fce->max_conns = ini->max_conns;
> + fce->max_links = ini->max_links;
> + }
> }
>
> out:
> @@ -942,6 +944,7 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini)
> if (smcr_indicated(ini->smc_type_v2)) {
> memcpy(v2_ext->roce, ini->smcrv2.ib_gid_v2, SMC_GID_SIZE);
> v2_ext->max_conns = SMC_CONN_PER_LGR_PREFER;
> + v2_ext->max_links = SMC_LINKS_PER_LGR_MAX_PREFER;
> }
>
> pclc_base->hdr.length = htons(plen);
> @@ -1174,6 +1177,7 @@ int smc_clc_srv_v2x_features_validate(struct smc_clc_msg_proposal *pclc,
> struct smc_clc_v2_extension *pclc_v2_ext;
>
> ini->max_conns = SMC_CONN_PER_LGR_MAX;
> + ini->max_links = SMC_LINKS_ADD_LNK_MAX;
>
> if ((!(ini->smcd_version & SMC_V2) && !(ini->smcr_version & SMC_V2)) ||
> ini->release_nr < SMC_RELEASE_1)
> @@ -1187,6 +1191,10 @@ int smc_clc_srv_v2x_features_validate(struct smc_clc_msg_proposal *pclc,
> ini->max_conns = min_t(u8, pclc_v2_ext->max_conns, SMC_CONN_PER_LGR_PREFER);
> if (ini->max_conns < SMC_CONN_PER_LGR_MIN)
> return SMC_CLC_DECL_MAXCONNERR;
> +
> + ini->max_links = min_t(u8, pclc_v2_ext->max_links, SMC_LINKS_PER_LGR_MAX_PREFER);
> + if (ini->max_links < SMC_LINKS_ADD_LNK_MIN)
> + return SMC_CLC_DECL_MAXLINKERR;
> }
>
> return 0;
> @@ -1205,6 +1213,11 @@ int smc_clc_clnt_v2x_features_validate(struct smc_clc_first_contact_ext *fce,
> if (fce_v2x->max_conns < SMC_CONN_PER_LGR_MIN)
> return SMC_CLC_DECL_MAXCONNERR;
> ini->max_conns = fce_v2x->max_conns;
> +
> + if (fce_v2x->max_links > SMC_LINKS_ADD_LNK_MAX ||
> + fce_v2x->max_links < SMC_LINKS_ADD_LNK_MIN)
> + return SMC_CLC_DECL_MAXLINKERR;
> + ini->max_links = fce_v2x->max_links;
> }
>
> return 0;
> @@ -1233,6 +1246,8 @@ int smc_clc_v2x_features_confirm_check(struct smc_clc_msg_accept_confirm *cclc,
> if (!ini->is_smcd) {
> if (fce_v2x->max_conns != ini->max_conns)
> return SMC_CLC_DECL_MAXCONNERR;
> + if (fce_v2x->max_links != ini->max_links)
> + return SMC_CLC_DECL_MAXLINKERR;
> }
>
> return 0;
> diff --git a/net/smc/smc_clc.h b/net/smc/smc_clc.h
> index 464b93b46047..c5c8e7db775a 100644
> --- a/net/smc/smc_clc.h
> +++ b/net/smc/smc_clc.h
> @@ -47,6 +47,7 @@
> #define SMC_CLC_DECL_NOUEID 0x03030008 /* peer sent no UEID */
> #define SMC_CLC_DECL_RELEASEERR 0x03030009 /* release version negotiate failed */
> #define SMC_CLC_DECL_MAXCONNERR 0x0303000a /* max connections negotiate failed */
> +#define SMC_CLC_DECL_MAXLINKERR 0x0303000b /* max links negotiate failed */
> #define SMC_CLC_DECL_MODEUNSUPP 0x03040000 /* smc modes do not match (R or D)*/
> #define SMC_CLC_DECL_RMBE_EC 0x03050000 /* peer has eyecatcher in RMBE */
> #define SMC_CLC_DECL_OPTUNSUPP 0x03060000 /* fastopen sockopt not supported */
> @@ -136,7 +137,8 @@ struct smc_clc_v2_extension {
> struct smc_clnt_opts_area_hdr hdr;
> u8 roce[16]; /* RoCEv2 GID */
> u8 max_conns;
> - u8 reserved[15];
> + u8 max_links;
> + u8 reserved[14];
> u8 user_eids[][SMC_MAX_EID_LEN];
> };
>
> @@ -239,7 +241,8 @@ struct smc_clc_first_contact_ext {
> struct smc_clc_first_contact_ext_v2x {
> struct smc_clc_first_contact_ext fce_v2_base;
> u8 max_conns; /* for SMC-R only */
> - u8 reserved3[3];
> + u8 max_links; /* for SMC-R only */
> + u8 reserved3[2];
> __be32 vendor_exp_options;
> u8 reserved4[8];
> } __packed; /* format defined in
> diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c
> index a8091a3e6cdd..1e1475084bb4 100644
> --- a/net/smc/smc_core.c
> +++ b/net/smc/smc_core.c
> @@ -896,10 +896,12 @@ static int smc_lgr_create(struct smc_sock *smc, struct smc_init_info *ini)
> memcpy(lgr->nexthop_mac, ini->smcrv2.nexthop_mac,
> ETH_ALEN);
> lgr->max_conns = ini->max_conns;
> + lgr->max_links = ini->max_links;
> } else {
> ibdev = ini->ib_dev;
> ibport = ini->ib_port;
> lgr->max_conns = SMC_CONN_PER_LGR_MAX;
> + lgr->max_links = SMC_LINKS_ADD_LNK_MAX;
> }
> memcpy(lgr->pnet_id, ibdev->pnetid[ibport - 1],
> SMC_MAX_PNETID_LEN);
> @@ -1667,6 +1669,9 @@ void smcr_port_add(struct smc_ib_device *smcibdev, u8 ibport)
> !rdma_dev_access_netns(smcibdev->ibdev, lgr->net))
> continue;
>
> + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1)
> + continue;
> +
> /* trigger local add link processing */
> link = smc_llc_usable_link(lgr);
> if (link)
> diff --git a/net/smc/smc_core.h b/net/smc/smc_core.h
> index 32b199477ef3..120027d40469 100644
> --- a/net/smc/smc_core.h
> +++ b/net/smc/smc_core.h
> @@ -173,6 +173,15 @@ struct smc_link {
> */
> #define SMC_LINKS_PER_LGR_MAX 3
> #define SMC_SINGLE_LINK 0
> +#define SMC_LINKS_ADD_LNK_MIN 1 /* min. # of links per link group */
> +#define SMC_LINKS_ADD_LNK_MAX 2 /* max. # of links per link group, also is the
> + * default value for smc-r v1.0 and v2.0
> + */
> +#define SMC_LINKS_PER_LGR_MAX_PREFER 2 /* Preferred max links per link group used for
> + * SMC-R v2.1 and later negotiation, vendors or
> + * distrubutions may modify it to a value between
> + * 1-2 as needed.
> + */
>
> /* tx/rx buffer list element for sndbufs list and rmbs list of a lgr */
> struct smc_buf_desc {
> @@ -342,6 +351,8 @@ struct smc_link_group {
> struct net *net;
> u8 max_conns;
> /* max conn can be assigned to lgr */
> + u8 max_links;
> + /* max links can be added in lgr */
> };
> struct { /* SMC-D */
> u64 peer_gid;
> @@ -387,6 +398,7 @@ struct smc_init_info {
> u8 smc_type_v2;
> u8 release_nr;
> u8 max_conns;
> + u8 max_links;
> u8 first_contact_peer;
> u8 first_contact_local;
> unsigned short vlan_id;
> diff --git a/net/smc/smc_llc.c b/net/smc/smc_llc.c
> index 5347b62f1518..018ce8133b02 100644
> --- a/net/smc/smc_llc.c
> +++ b/net/smc/smc_llc.c
> @@ -59,8 +59,6 @@ struct smc_llc_msg_confirm_link { /* type 0x01 */
> #define SMC_LLC_FLAG_ADD_LNK_REJ 0x40
> #define SMC_LLC_REJ_RSN_NO_ALT_PATH 1
>
> -#define SMC_LLC_ADD_LNK_MAX_LINKS 2
> -
> struct smc_llc_msg_add_link { /* type 0x02 */
> struct smc_llc_hdr hd;
> u8 sender_mac[ETH_ALEN];
> @@ -472,10 +470,12 @@ int smc_llc_send_confirm_link(struct smc_link *link,
> hton24(confllc->sender_qp_num, link->roce_qp->qp_num);
> confllc->link_num = link->link_id;
> memcpy(confllc->link_uid, link->link_uid, SMC_LGR_ID_SIZE);
> - confllc->max_links = SMC_LLC_ADD_LNK_MAX_LINKS;
> + confllc->max_links = SMC_LINKS_ADD_LNK_MAX;
> if (link->lgr->smc_version == SMC_V2 &&
> - link->lgr->peer_smc_release >= SMC_RELEASE_1)
> + link->lgr->peer_smc_release >= SMC_RELEASE_1) {
> confllc->max_conns = link->lgr->max_conns;
> + confllc->max_links = link->lgr->max_links;
> + }
> /* send llc message */
> rc = smc_wr_tx_send(link, pend);
> put_out:
> @@ -1045,6 +1045,11 @@ int smc_llc_cli_add_link(struct smc_link *link, struct smc_llc_qentry *qentry)
> goto out_reject;
> }
>
> + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1) {
> + rc = 0;
> + goto out_reject;
> + }
> +
> ini->vlan_id = lgr->vlan_id;
> if (lgr->smc_version == SMC_V2) {
> ini->check_smcrv2 = true;
> @@ -1169,6 +1174,9 @@ static void smc_llc_cli_add_link_invite(struct smc_link *link,
> lgr->type == SMC_LGR_ASYMMETRIC_PEER)
> goto out;
>
> + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1)
> + goto out;
> +
> ini = kzalloc(sizeof(*ini), GFP_KERNEL);
> if (!ini)
> goto out;
> @@ -1414,6 +1422,11 @@ int smc_llc_srv_add_link(struct smc_link *link,
> goto out;
> }
>
> + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1) {
> + rc = 0;
> + goto out;
> + }
> +
> /* ignore client add link recommendation, start new flow */
> ini->vlan_id = lgr->vlan_id;
> if (lgr->smc_version == SMC_V2) {
Powered by blists - more mailing lists