From patchwork Thu Jun 18 23:25:32 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Hal Rosenstock X-Patchwork-Id: 6641821 Return-Path: X-Original-To: patchwork-linux-rdma@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork1.web.kernel.org (Postfix) with ESMTP id 22C0B9F399 for ; Thu, 18 Jun 2015 23:25:33 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 567CE208D5 for ; Thu, 18 Jun 2015 23:25:31 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 1C886208CC for ; Thu, 18 Jun 2015 23:25:29 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751110AbbFRXZZ (ORCPT ); Thu, 18 Jun 2015 19:25:25 -0400 Received: from mail-wi0-f179.google.com ([209.85.212.179]:34574 "EHLO mail-wi0-f179.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751001AbbFRXZY (ORCPT ); Thu, 18 Jun 2015 19:25:24 -0400 Received: by wicnd19 with SMTP id nd19so3616824wic.1 for ; Thu, 18 Jun 2015 16:25:22 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:message-id:date:from:user-agent:mime-version:to :cc:subject:content-type:content-transfer-encoding; bh=CCACEKhyD5VBZf9GCilVi4EqFy52P1sqoGC1fIj5fTk=; b=X7bND2+31LLzsEXCb0XAa9PRMpcYRePHvUyIFL7CVGt1mB8fSiVGRM7zl9A2AYEc0G xUbuVR+wOVGE9RsPsBY5VAH0CRWYcPKBpVUKHM8QcjLPR9W+uX2Kw9kcQv8D2h24ZnCb p3agq8ibIOBZNYu+iw+HOvybAO5i5/0AV2PdU7h09sICGK6/zwxyUch+vfDqseo6kKxp nrw3sT+EVFpXRLcDT2PT6IUjqQpFo1HOxfvqB9fOC+QP2EVoBZqHxB4yU5+3h6toQTVL 2FwC/u1rIlJirNeXkPqw6e06PrFrxQTSm2pmxdTpeDwjHAn/+FdI+ri3PNQdGusBIrIT VxcQ== X-Gm-Message-State: ALoCoQkSa+b1ICcJRKljQuL/5iG/Br3ETlN0jZR2kIkKRO8TnxRIIUkqyPhCOKJVxG9873aC6xVu X-Received: by 10.194.5.103 with SMTP id r7mr19513354wjr.47.1434669922493; Thu, 18 Jun 2015 16:25:22 -0700 (PDT) Received: from [192.168.1.102] (c-71-234-225-223.hsd1.ct.comcast.net. [71.234.225.223]) by mx.google.com with ESMTPSA id ez19sm995673wid.19.2015.06.18.16.25.20 (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Thu, 18 Jun 2015 16:25:21 -0700 (PDT) Message-ID: <5583536C.3010504@dev.mellanox.co.il> Date: Thu, 18 Jun 2015 19:25:32 -0400 From: Hal Rosenstock User-Agent: Mozilla/5.0 (Windows NT 6.1; rv:9.0) Gecko/20111222 Thunderbird/9.0.1 MIME-Version: 1.0 To: Doug Ledford CC: "linux-rdma (linux-rdma@vger.kernel.org)" , Liran Liss , Ira Weiny Subject: [PATCHv2 RFC] IB: Add rdma_cap_ib_switch helper and use where appropriate Sender: linux-rdma-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org X-Spam-Status: No, score=-7.2 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=ham version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP Persuant to Liran's comments on node_type on linux-rdma mailing list: In an effort to reform the RDMA core and ULPs to minimize use of node_type in struct ib_device, an additional bit is added to struct ib_device for is_switch (IB switch). This is needed to be initialized by any IB switch device driver. This is a NEW requirement on such device drivers which are all "out of tree". In addition, an ib_switch helper was added to ib_verbs.h based on the is_switch device bit rather than node_type (although those should be consistent). The RDMA core (MAD, SMI, agent, sa_query, multicast, sysfs) as well as (IPoIB and SRP) ULPs are updated where appropriate to use this new helper. In some cases, the helper is now used under the covers of using rdma_[start end]_port rather than the open coding previously used. Signed-off-by: Hal Rosenstock Reviewed-by: Sean Hefty Reviewed-By: Jason Gunthorpe Reviewed-by: Ira Weiny Tested-by: Ira Weiny --- Changes since v1: Per Jason and Sean's comments, used Ira's rdma_[start end]_port routines where needed and made those routines in ib_verbs.h use the new is_switch helper. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" 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/drivers/infiniband/core/agent.c b/drivers/infiniband/core/agent.c index c7dcfe4..0429040 100644 --- a/drivers/infiniband/core/agent.c +++ b/drivers/infiniband/core/agent.c @@ -88,7 +88,7 @@ void agent_send_response(const struct ib_mad_hdr *mad_hdr, const struct ib_grh * struct ib_ah *ah; struct ib_mad_send_wr_private *mad_send_wr; - if (device->node_type == RDMA_NODE_IB_SWITCH) + if (rdma_cap_ib_switch(device)) port_priv = ib_get_agent_port(device, 0); else port_priv = ib_get_agent_port(device, port_num); @@ -122,7 +122,7 @@ void agent_send_response(const struct ib_mad_hdr *mad_hdr, const struct ib_grh * memcpy(send_buf->mad, mad_hdr, resp_mad_len); send_buf->ah = ah; - if (device->node_type == RDMA_NODE_IB_SWITCH) { + if (rdma_cap_ib_switch(device)) { mad_send_wr = container_of(send_buf, struct ib_mad_send_wr_private, send_buf); diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index a4b1466..c82d751 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -769,7 +769,7 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, bool opa = rdma_cap_opa_mad(mad_agent_priv->qp_info->port_priv->device, mad_agent_priv->qp_info->port_priv->port_num); - if (device->node_type == RDMA_NODE_IB_SWITCH && + if (rdma_cap_ib_switch(device) && smp->mgmt_class == IB_MGMT_CLASS_SUBN_DIRECTED_ROUTE) port_num = send_wr->wr.ud.port_num; else @@ -787,7 +787,8 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, if ((opa_get_smp_direction(opa_smp) ? opa_smp->route.dr.dr_dlid : opa_smp->route.dr.dr_slid) == OPA_LID_PERMISSIVE && - opa_smi_handle_dr_smp_send(opa_smp, device->node_type, + opa_smi_handle_dr_smp_send(opa_smp, + rdma_cap_ib_switch(device), port_num) == IB_SMI_DISCARD) { ret = -EINVAL; dev_err(&device->dev, "OPA Invalid directed route\n"); @@ -810,7 +811,7 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, } else { if ((ib_get_smp_direction(smp) ? smp->dr_dlid : smp->dr_slid) == IB_LID_PERMISSIVE && - smi_handle_dr_smp_send(smp, device->node_type, port_num) == + smi_handle_dr_smp_send(smp, rdma_cap_ib_switch(device), port_num) == IB_SMI_DISCARD) { ret = -EINVAL; dev_err(&device->dev, "Invalid directed route\n"); @@ -2030,7 +2031,7 @@ static enum smi_action handle_ib_smi(const struct ib_mad_port_private *port_priv struct ib_smp *smp = (struct ib_smp *)recv->mad; if (smi_handle_dr_smp_recv(smp, - port_priv->device->node_type, + rdma_cap_ib_switch(port_priv->device), port_num, port_priv->device->phys_port_cnt) == IB_SMI_DISCARD) @@ -2042,13 +2043,13 @@ static enum smi_action handle_ib_smi(const struct ib_mad_port_private *port_priv if (retsmi == IB_SMI_SEND) { /* don't forward */ if (smi_handle_dr_smp_send(smp, - port_priv->device->node_type, + rdma_cap_ib_switch(port_priv->device), port_num) == IB_SMI_DISCARD) return IB_SMI_DISCARD; if (smi_check_local_smp(smp, port_priv->device) == IB_SMI_DISCARD) return IB_SMI_DISCARD; - } else if (port_priv->device->node_type == RDMA_NODE_IB_SWITCH) { + } else if (rdma_cap_ib_switch(port_priv->device)) { /* forward case for switches */ memcpy(response, recv, mad_priv_size(response)); response->header.recv_wc.wc = &response->header.wc; @@ -2115,7 +2116,7 @@ handle_opa_smi(struct ib_mad_port_private *port_priv, struct opa_smp *smp = (struct opa_smp *)recv->mad; if (opa_smi_handle_dr_smp_recv(smp, - port_priv->device->node_type, + rdma_cap_ib_switch(port_priv->device), port_num, port_priv->device->phys_port_cnt) == IB_SMI_DISCARD) @@ -2127,7 +2128,7 @@ handle_opa_smi(struct ib_mad_port_private *port_priv, if (retsmi == IB_SMI_SEND) { /* don't forward */ if (opa_smi_handle_dr_smp_send(smp, - port_priv->device->node_type, + rdma_cap_ib_switch(port_priv->device), port_num) == IB_SMI_DISCARD) return IB_SMI_DISCARD; @@ -2135,7 +2136,7 @@ handle_opa_smi(struct ib_mad_port_private *port_priv, IB_SMI_DISCARD) return IB_SMI_DISCARD; - } else if (port_priv->device->node_type == RDMA_NODE_IB_SWITCH) { + } else if (rdma_cap_ib_switch(port_priv->device)) { /* forward case for switches */ memcpy(response, recv, mad_priv_size(response)); response->header.recv_wc.wc = &response->header.wc; @@ -2235,7 +2236,7 @@ static void ib_mad_recv_done_handler(struct ib_mad_port_private *port_priv, goto out; } - if (port_priv->device->node_type == RDMA_NODE_IB_SWITCH) + if (rdma_cap_ib_switch(port_priv->device)) port_num = wc->port_num; else port_num = port_priv->port_num; @@ -3297,17 +3298,11 @@ static int ib_mad_port_close(struct ib_device *device, int port_num) static void ib_mad_init_device(struct ib_device *device) { - int start, end, i; + int start, i; - if (device->node_type == RDMA_NODE_IB_SWITCH) { - start = 0; - end = 0; - } else { - start = 1; - end = device->phys_port_cnt; - } + start = rdma_start_port(device); - for (i = start; i <= end; i++) { + for (i = start; i <= rdma_end_port(device); i++) { if (!rdma_cap_ib_mad(device, i)) continue; @@ -3342,17 +3337,9 @@ error: static void ib_mad_remove_device(struct ib_device *device) { - int start, end, i; - - if (device->node_type == RDMA_NODE_IB_SWITCH) { - start = 0; - end = 0; - } else { - start = 1; - end = device->phys_port_cnt; - } + int i; - for (i = start; i <= end; i++) { + for (i = rdma_start_port(device); i <= rdma_end_port(device); i++) { if (!rdma_cap_ib_mad(device, i)) continue; diff --git a/drivers/infiniband/core/multicast.c b/drivers/infiniband/core/multicast.c index 1244f02..2cb865c 100644 --- a/drivers/infiniband/core/multicast.c +++ b/drivers/infiniband/core/multicast.c @@ -812,12 +812,8 @@ static void mcast_add_one(struct ib_device *device) if (!dev) return; - if (device->node_type == RDMA_NODE_IB_SWITCH) - dev->start_port = dev->end_port = 0; - else { - dev->start_port = 1; - dev->end_port = device->phys_port_cnt; - } + dev->start_port = rdma_start_port(device); + dev->end_port = rdma_end_port(device); for (i = 0; i <= dev->end_port - dev->start_port; i++) { if (!rdma_cap_ib_mcast(device, dev->start_port + i)) diff --git a/drivers/infiniband/core/opa_smi.h b/drivers/infiniband/core/opa_smi.h index 62d91bf..3bfab35 100644 --- a/drivers/infiniband/core/opa_smi.h +++ b/drivers/infiniband/core/opa_smi.h @@ -39,12 +39,12 @@ #include "smi.h" -enum smi_action opa_smi_handle_dr_smp_recv(struct opa_smp *smp, u8 node_type, +enum smi_action opa_smi_handle_dr_smp_recv(struct opa_smp *smp, bool is_switch, int port_num, int phys_port_cnt); int opa_smi_get_fwd_port(struct opa_smp *smp); extern enum smi_forward_action opa_smi_check_forward_dr_smp(struct opa_smp *smp); extern enum smi_action opa_smi_handle_dr_smp_send(struct opa_smp *smp, - u8 node_type, int port_num); + bool is_switch, int port_num); /* * Return IB_SMI_HANDLE if the SMP should be handled by the local SMA/SM diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index 0fae850..ca919f4 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -1156,12 +1156,8 @@ static void ib_sa_add_one(struct ib_device *device) int s, e, i; int count = 0; - if (device->node_type == RDMA_NODE_IB_SWITCH) - s = e = 0; - else { - s = 1; - e = device->phys_port_cnt; - } + s = rdma_start_port(device); + e = rdma_end_port(device); sa_dev = kzalloc(sizeof *sa_dev + (e - s + 1) * sizeof (struct ib_sa_port), diff --git a/drivers/infiniband/core/smi.c b/drivers/infiniband/core/smi.c index 368a561..f19b238 100644 --- a/drivers/infiniband/core/smi.c +++ b/drivers/infiniband/core/smi.c @@ -41,7 +41,7 @@ #include "smi.h" #include "opa_smi.h" -static enum smi_action __smi_handle_dr_smp_send(u8 node_type, int port_num, +static enum smi_action __smi_handle_dr_smp_send(bool is_switch, int port_num, u8 *hop_ptr, u8 hop_cnt, const u8 *initial_path, const u8 *return_path, @@ -64,7 +64,7 @@ static enum smi_action __smi_handle_dr_smp_send(u8 node_type, int port_num, /* C14-9:2 */ if (*hop_ptr && *hop_ptr < hop_cnt) { - if (node_type != RDMA_NODE_IB_SWITCH) + if (!is_switch) return IB_SMI_DISCARD; /* return_path set when received */ @@ -77,7 +77,7 @@ static enum smi_action __smi_handle_dr_smp_send(u8 node_type, int port_num, if (*hop_ptr == hop_cnt) { /* return_path set when received */ (*hop_ptr)++; - return (node_type == RDMA_NODE_IB_SWITCH || + return (is_switch || dr_dlid_is_permissive ? IB_SMI_HANDLE : IB_SMI_DISCARD); } @@ -96,7 +96,7 @@ static enum smi_action __smi_handle_dr_smp_send(u8 node_type, int port_num, /* C14-13:2 */ if (2 <= *hop_ptr && *hop_ptr <= hop_cnt) { - if (node_type != RDMA_NODE_IB_SWITCH) + if (!is_switch) return IB_SMI_DISCARD; (*hop_ptr)--; @@ -108,7 +108,7 @@ static enum smi_action __smi_handle_dr_smp_send(u8 node_type, int port_num, if (*hop_ptr == 1) { (*hop_ptr)--; /* C14-13:3 -- SMPs destined for SM shouldn't be here */ - return (node_type == RDMA_NODE_IB_SWITCH || + return (is_switch || dr_slid_is_permissive ? IB_SMI_HANDLE : IB_SMI_DISCARD); } @@ -127,9 +127,9 @@ static enum smi_action __smi_handle_dr_smp_send(u8 node_type, int port_num, * Return IB_SMI_DISCARD if the SMP should be discarded */ enum smi_action smi_handle_dr_smp_send(struct ib_smp *smp, - u8 node_type, int port_num) + bool is_switch, int port_num) { - return __smi_handle_dr_smp_send(node_type, port_num, + return __smi_handle_dr_smp_send(is_switch, port_num, &smp->hop_ptr, smp->hop_cnt, smp->initial_path, smp->return_path, @@ -139,9 +139,9 @@ enum smi_action smi_handle_dr_smp_send(struct ib_smp *smp, } enum smi_action opa_smi_handle_dr_smp_send(struct opa_smp *smp, - u8 node_type, int port_num) + bool is_switch, int port_num) { - return __smi_handle_dr_smp_send(node_type, port_num, + return __smi_handle_dr_smp_send(is_switch, port_num, &smp->hop_ptr, smp->hop_cnt, smp->route.dr.initial_path, smp->route.dr.return_path, @@ -152,7 +152,7 @@ enum smi_action opa_smi_handle_dr_smp_send(struct opa_smp *smp, OPA_LID_PERMISSIVE); } -static enum smi_action __smi_handle_dr_smp_recv(u8 node_type, int port_num, +static enum smi_action __smi_handle_dr_smp_recv(bool is_switch, int port_num, int phys_port_cnt, u8 *hop_ptr, u8 hop_cnt, const u8 *initial_path, @@ -173,7 +173,7 @@ static enum smi_action __smi_handle_dr_smp_recv(u8 node_type, int port_num, /* C14-9:2 -- intermediate hop */ if (*hop_ptr && *hop_ptr < hop_cnt) { - if (node_type != RDMA_NODE_IB_SWITCH) + if (!is_switch) return IB_SMI_DISCARD; return_path[*hop_ptr] = port_num; @@ -188,7 +188,7 @@ static enum smi_action __smi_handle_dr_smp_recv(u8 node_type, int port_num, return_path[*hop_ptr] = port_num; /* hop_ptr updated when sending */ - return (node_type == RDMA_NODE_IB_SWITCH || + return (is_switch || dr_dlid_is_permissive ? IB_SMI_HANDLE : IB_SMI_DISCARD); } @@ -208,7 +208,7 @@ static enum smi_action __smi_handle_dr_smp_recv(u8 node_type, int port_num, /* C14-13:2 */ if (2 <= *hop_ptr && *hop_ptr <= hop_cnt) { - if (node_type != RDMA_NODE_IB_SWITCH) + if (!is_switch) return IB_SMI_DISCARD; /* hop_ptr updated when sending */ @@ -224,8 +224,7 @@ static enum smi_action __smi_handle_dr_smp_recv(u8 node_type, int port_num, return IB_SMI_HANDLE; } /* hop_ptr updated when sending */ - return (node_type == RDMA_NODE_IB_SWITCH ? - IB_SMI_HANDLE : IB_SMI_DISCARD); + return (is_switch ? IB_SMI_HANDLE : IB_SMI_DISCARD); } /* C14-13:4 -- hop_ptr = 0 -> give to SM */ @@ -238,10 +237,10 @@ static enum smi_action __smi_handle_dr_smp_recv(u8 node_type, int port_num, * Adjust information for a received SMP * Return IB_SMI_DISCARD if the SMP should be dropped */ -enum smi_action smi_handle_dr_smp_recv(struct ib_smp *smp, u8 node_type, +enum smi_action smi_handle_dr_smp_recv(struct ib_smp *smp, bool is_switch, int port_num, int phys_port_cnt) { - return __smi_handle_dr_smp_recv(node_type, port_num, phys_port_cnt, + return __smi_handle_dr_smp_recv(is_switch, port_num, phys_port_cnt, &smp->hop_ptr, smp->hop_cnt, smp->initial_path, smp->return_path, @@ -254,10 +253,10 @@ enum smi_action smi_handle_dr_smp_recv(struct ib_smp *smp, u8 node_type, * Adjust information for a received SMP * Return IB_SMI_DISCARD if the SMP should be dropped */ -enum smi_action opa_smi_handle_dr_smp_recv(struct opa_smp *smp, u8 node_type, +enum smi_action opa_smi_handle_dr_smp_recv(struct opa_smp *smp, bool is_switch, int port_num, int phys_port_cnt) { - return __smi_handle_dr_smp_recv(node_type, port_num, phys_port_cnt, + return __smi_handle_dr_smp_recv(is_switch, port_num, phys_port_cnt, &smp->hop_ptr, smp->hop_cnt, smp->route.dr.initial_path, smp->route.dr.return_path, diff --git a/drivers/infiniband/core/smi.h b/drivers/infiniband/core/smi.h index aff96ba..33c91c8 100644 --- a/drivers/infiniband/core/smi.h +++ b/drivers/infiniband/core/smi.h @@ -51,12 +51,12 @@ enum smi_forward_action { IB_SMI_FORWARD /* SMP should be forwarded (for switches only) */ }; -enum smi_action smi_handle_dr_smp_recv(struct ib_smp *smp, u8 node_type, +enum smi_action smi_handle_dr_smp_recv(struct ib_smp *smp, bool is_switch, int port_num, int phys_port_cnt); int smi_get_fwd_port(struct ib_smp *smp); extern enum smi_forward_action smi_check_forward_dr_smp(struct ib_smp *smp); extern enum smi_action smi_handle_dr_smp_send(struct ib_smp *smp, - u8 node_type, int port_num); + bool is_switch, int port_num); /* * Return IB_SMI_HANDLE if the SMP should be handled by the local SMA/SM diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c index ed6b6c8..0b84a9c 100644 --- a/drivers/infiniband/core/sysfs.c +++ b/drivers/infiniband/core/sysfs.c @@ -870,7 +870,7 @@ int ib_device_register_sysfs(struct ib_device *device, goto err_put; } - if (device->node_type == RDMA_NODE_IB_SWITCH) { + if (rdma_cap_ib_switch(device)) { ret = add_port(device, 0, port_callback); if (ret) goto err_put; diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index da149c2..0d8336e 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -1684,7 +1684,7 @@ static void ipoib_add_one(struct ib_device *device) struct list_head *dev_list; struct net_device *dev; struct ipoib_dev_priv *priv; - int s, e, p; + int p; int count = 0; dev_list = kmalloc(sizeof *dev_list, GFP_KERNEL); @@ -1693,15 +1693,7 @@ static void ipoib_add_one(struct ib_device *device) INIT_LIST_HEAD(dev_list); - if (device->node_type == RDMA_NODE_IB_SWITCH) { - s = 0; - e = 0; - } else { - s = 1; - e = device->phys_port_cnt; - } - - for (p = s; p <= e; ++p) { + for (p = rdma_start_port(device); p <= rdma_end_port(device); ++p) { if (!rdma_protocol_ib(device, p)) continue; dev = ipoib_add_port("ib%d", device, p); diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index eada8f7..ef10f6c 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -3379,7 +3379,7 @@ static void srp_add_one(struct ib_device *device) struct srp_device *srp_dev; struct ib_device_attr *dev_attr; struct srp_host *host; - int mr_page_shift, s, e, p; + int mr_page_shift, p; u64 max_pages_per_mr; dev_attr = kmalloc(sizeof *dev_attr, GFP_KERNEL); @@ -3443,15 +3443,7 @@ static void srp_add_one(struct ib_device *device) if (IS_ERR(srp_dev->mr)) goto err_pd; - if (device->node_type == RDMA_NODE_IB_SWITCH) { - s = 0; - e = 0; - } else { - s = 1; - e = device->phys_port_cnt; - } - - for (p = s; p <= e; ++p) { + for (p = rdma_start_port(device); p <= rdma_end_port(device); ++p) { host = srp_add_port(srp_dev, p); if (host) list_add_tail(&host->list, &srp_dev->dev_list); diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 986fddb..b0f898e 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1745,6 +1745,7 @@ struct ib_device { char node_desc[64]; __be64 node_guid; u32 local_dma_lkey; + u16 is_switch:1; u8 node_type; u8 phys_port_cnt; @@ -1824,6 +1825,20 @@ enum rdma_link_layer rdma_port_get_link_layer(struct ib_device *device, u8 port_num); /** + * rdma_cap_ib_switch - Check if the device is IB switch + * @device: Device to check + * + * Device driver is responsible for setting is_switch bit on + * in ib_device structure at init time. + * + * Return: true if the device is IB switch. + */ +static inline bool rdma_cap_ib_switch(const struct ib_device *device) +{ + return device->is_switch; +} + +/** * rdma_start_port - Return the first valid port number for the device * specified * @@ -1833,7 +1848,7 @@ enum rdma_link_layer rdma_port_get_link_layer(struct ib_device *device, */ static inline u8 rdma_start_port(const struct ib_device *device) { - return (device->node_type == RDMA_NODE_IB_SWITCH) ? 0 : 1; + return rdma_cap_ib_switch(device) ? 0 : 1; } /** @@ -1846,8 +1861,7 @@ static inline u8 rdma_start_port(const struct ib_device *device) */ static inline u8 rdma_end_port(const struct ib_device *device) { - return (device->node_type == RDMA_NODE_IB_SWITCH) ? - 0 : device->phys_port_cnt; + return rdma_cap_ib_switch(device) ? 0 : device->phys_port_cnt; } static inline bool rdma_protocol_ib(const struct ib_device *device, u8 port_num)