From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-16.8 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 6DEBCC433F5 for ; Wed, 22 Sep 2021 02:20:28 +0000 (UTC) Received: from pdx1-mailman02.dreamhost.com (pdx1-mailman02.dreamhost.com [64.90.62.194]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 2BC2060F4C for ; Wed, 22 Sep 2021 02:20:28 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.1 mail.kernel.org 2BC2060F4C Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=infradead.org Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=lists.lustre.org Received: from pdx1-mailman02.dreamhost.com (localhost [IPv6:::1]) by pdx1-mailman02.dreamhost.com (Postfix) with ESMTP id 1F38821F246; Tue, 21 Sep 2021 19:20:24 -0700 (PDT) Received: from smtp3.ccs.ornl.gov (smtp3.ccs.ornl.gov [160.91.203.39]) by pdx1-mailman02.dreamhost.com (Postfix) with ESMTP id 57CF521CA53 for ; Tue, 21 Sep 2021 19:20:06 -0700 (PDT) Received: from star.ccs.ornl.gov (star.ccs.ornl.gov [160.91.202.134]) by smtp3.ccs.ornl.gov (Postfix) with ESMTP id 09080350; Tue, 21 Sep 2021 22:20:04 -0400 (EDT) Received: by star.ccs.ornl.gov (Postfix, from userid 2004) id F0BA2FF4BE; Tue, 21 Sep 2021 22:20:03 -0400 (EDT) From: James Simmons To: Andreas Dilger , Oleg Drokin , NeilBrown Date: Tue, 21 Sep 2021 22:19:40 -0400 Message-Id: <1632277201-6920-4-git-send-email-jsimmons@infradead.org> X-Mailer: git-send-email 1.8.3.1 In-Reply-To: <1632277201-6920-1-git-send-email-jsimmons@infradead.org> References: <1632277201-6920-1-git-send-email-jsimmons@infradead.org> Subject: [lustre-devel] [PATCH 03/24] lnet: introduce struct lnet_nid X-BeenThere: lustre-devel@lists.lustre.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: "For discussing Lustre software development." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Lustre Development List MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: lustre-devel-bounces@lists.lustre.org Sender: "lustre-devel" From: Mr NeilBrown LNet nids are currently limited to 4-bytes for addresses. This excludes the use of IPv6. In order to support IPv6, introduce 'struct lnet_nid' which can hold up to 128bit address and is extensible, and deprecate 'lnet_nid_t'. lnet_nid_it will eventually be removed. Where lnet_nid_t is often passed around by value, 'struct lnet_nid' will normally be passed around by reference as it is over twice as large. The net_type field, which currently has value up to 16, is now limited to 0-254 with 255 being used as a wildcard. The most significant byte is now a size field which gives the size of the whole nid minus 8. So zero is correct for current nids with 4-byte addresses. Where we still need to use 4-byte-address nids, we will use names containing "nid4". So "nid4" is a lnet_nid_t when "nid" is a struct lnet_nid. lnet_nid_to_nid4 converts a 'struct lnet_nid' to an lnet_nid_t. While lnet_nid_t is stored and often transmitted in host-endian format (and possibly byte-swapped on receipt), 'struct lnet_nid' is always stored in network-byte-order (i.e. big-endian). This is more common approach for network addresses. In this first instance, 'struct lnet_nid' is used for ni_nid in 'struct lnet_ni', and related support functions. In particular libcfs_nidstr() is introduced which parallels libcfs_nid2str(), but takes 'struct lnet_nid'. In cases were we need to have similar functions for old and new style nid, the new function is introduced with a slightly different name, such as libcfs_nid2str above, or LNET_NID_NET (like LNET_NIDNET). It will be confusing having both, but the plan is to remove the old names as soon as practical. WC-bug-id: https://jira.whamcloud.com/browse/LU-10391 Lustre-commit: 82a17076f880770a ("LU-10391 lnet: introduce struct lnet_nid") Signed-off-by: Mr NeilBrown Reviewed-on: https://review.whamcloud.com/42100 Reviewed-by: James Simmons Reviewed-by: Serguei Smirnov Reviewed-by: Chris Horn Reviewed-by: Oleg Drokin Signed-off-by: James Simmons --- include/linux/lnet/lib-lnet.h | 4 +- include/linux/lnet/lib-types.h | 2 +- include/uapi/linux/lnet/lnet-idl.h | 28 ++++++++++- include/uapi/linux/lnet/lnet-types.h | 66 ++++++++++++++++++++++++- include/uapi/linux/lnet/nidstr.h | 10 +++- net/lnet/klnds/o2iblnd/o2iblnd.c | 13 +++-- net/lnet/klnds/o2iblnd/o2iblnd_cb.c | 17 ++++--- net/lnet/klnds/socklnd/socklnd.c | 10 ++-- net/lnet/klnds/socklnd/socklnd_cb.c | 10 ++-- net/lnet/lnet/acceptor.c | 3 +- net/lnet/lnet/api-ni.c | 60 +++++++++++++--------- net/lnet/lnet/config.c | 5 +- net/lnet/lnet/lib-move.c | 96 +++++++++++++++++++++--------------- net/lnet/lnet/lib-msg.c | 11 +++-- net/lnet/lnet/lib-socket.c | 32 +++++++++--- net/lnet/lnet/lo.c | 3 +- net/lnet/lnet/net_fault.c | 2 +- net/lnet/lnet/nidstrings.c | 40 +++++++++++++++ net/lnet/lnet/router.c | 8 +-- net/lnet/lnet/router_proc.c | 2 +- net/lnet/lnet/udsp.c | 13 ++--- 21 files changed, 312 insertions(+), 123 deletions(-) diff --git a/include/linux/lnet/lib-lnet.h b/include/linux/lnet/lib-lnet.h index 37489ae..acc069d 100644 --- a/include/linux/lnet/lib-lnet.h +++ b/include/linux/lnet/lib-lnet.h @@ -110,7 +110,7 @@ if (ni->ni_status && ni->ni_status->ns_status != status) { CDEBUG(D_NET, "ni %s status changed from %#x to %#x\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), ni->ni_status->ns_status, status); ni->ni_status->ns_status = status; update = true; @@ -123,7 +123,7 @@ lnet_ni_get_status_locked(struct lnet_ni *ni) __must_hold(&ni->ni_lock) { - if (ni->ni_nid == LNET_NID_LO_0) + if (nid_is_lo0(&ni->ni_nid)) return LNET_NI_STATUS_UP; else if (atomic_read(&ni->ni_fatal_error_on)) return LNET_NI_STATUS_DOWN; diff --git a/include/linux/lnet/lib-types.h b/include/linux/lnet/lib-types.h index 85b0d54..80cf4f3 100644 --- a/include/linux/lnet/lib-types.h +++ b/include/linux/lnet/lib-types.h @@ -463,7 +463,7 @@ struct lnet_ni { u32 *ni_cpts; /* interface's NID */ - lnet_nid_t ni_nid; + struct lnet_nid ni_nid; /* instance-specific data */ void *ni_data; diff --git a/include/uapi/linux/lnet/lnet-idl.h b/include/uapi/linux/lnet/lnet-idl.h index 1e27f3b..3fc0df1 100644 --- a/include/uapi/linux/lnet/lnet-idl.h +++ b/include/uapi/linux/lnet/lnet-idl.h @@ -40,18 +40,42 @@ * These are sent in sender's byte order (i.e. receiver flips). */ -/** - * Address of an end-point in an LNet network. +/** Address of an end-point in an LNet network. * * A node can have multiple end-points and hence multiple addresses. * An LNet network can be a simple network (e.g. tcp0) or a network of * LNet networks connected by LNet routers. Therefore an end-point address * has two parts: network ID, and address within a network. + * The most-significant-byte in this format is always 0. A larger value + * would imply a larger nid with a larger address. * * \see LNET_NIDNET, LNET_NIDADDR, and LNET_MKNID. */ typedef __u64 lnet_nid_t; +/* + * Address of LNet end-point in extended form + * + * To support addresses larger than 32bits we have + * an extended nid which supports up to 128 bits + * of address and is extensible. + * If nid_size is 0, then the nid can be stored in an lnet_nid_t, + * and the first 8 bytes of the 'struct lnet_nid' are identical to + * the lnet_nid_t in big-endian format. + * If nid_type == 0xff, then all other fields should be ignored + * and this is an ANY wildcard address. In particular, the nid_size + * can be 0xff without making the address too big to fit. + */ +struct lnet_nid { + __u8 nid_size; /* total bytes - 8 */ + __u8 nid_type; + __be16 nid_num; + __be32 nid_addr[4]; +} __attribute__((packed)); + +#define NID_BYTES(nid) ((nid)->nid_size + 8) +#define NID_ADDR_BYTES(nid) ((nid)->nid_size + 4) + /** * ID of a process in a node. Shortened as PID to distinguish from * lnet_process_id, the global process ID. diff --git a/include/uapi/linux/lnet/lnet-types.h b/include/uapi/linux/lnet/lnet-types.h index 0c426ac..ba8a079 100644 --- a/include/uapi/linux/lnet/lnet-types.h +++ b/include/uapi/linux/lnet/lnet-types.h @@ -37,6 +37,9 @@ #include #include +#include +#include + /** \addtogroup lnet * @{ */ @@ -57,6 +60,15 @@ /** wildcard PID that matches any lnet_pid_t */ #define LNET_PID_ANY ((lnet_pid_t)(-1)) +static inline int LNET_NID_IS_ANY(const struct lnet_nid *nid) +{ + /* A NULL pointer can be used to mean "ANY" */ + return !nid || nid->nid_type == 0xFF; +} + +#define LNET_ANY_NID ((struct lnet_nid) \ + {0xFF, 0xFF, ~0, {~0, ~0, ~0, ~0} }) + #define LNET_PID_RESERVED 0xf0000000 /* reserved bits in PID */ #define LNET_PID_USERFLAG 0x80000000 /* set in userspace peers */ #define LNET_PID_LUSTRE 12345 @@ -86,7 +98,7 @@ static inline __u32 LNET_NETNUM(__u32 net) static inline __u32 LNET_NETTYP(__u32 net) { - return (net >> 16) & 0xffff; + return (net >> 16) & 0xff; } static inline __u32 LNET_MKNET(__u32 type, __u32 num) @@ -99,6 +111,58 @@ static inline __u32 LNET_MKNET(__u32 type, __u32 num) #define LNET_NET_ANY LNET_NIDNET(LNET_NID_ANY) +static inline int nid_is_nid4(const struct lnet_nid *nid) +{ + return NID_ADDR_BYTES(nid) == 4; +} + +/* LOLND may not be defined yet, so we cannot use an inline */ +#define nid_is_lo0(__nid) \ + ((__nid)->nid_type == LOLND && \ + nid_is_nid4(__nid) && \ + (__nid)->nid_num == 0 && \ + (__nid)->nid_addr[0] == 0) + +static inline __u32 LNET_NID_NET(const struct lnet_nid *nid) +{ + return LNET_MKNET(nid->nid_type, __be16_to_cpu(nid->nid_num)); +} + +static inline void lnet_nid4_to_nid(lnet_nid_t nid4, struct lnet_nid *nid) +{ + if (nid4 == LNET_NID_ANY) { + /* equal to setting to LNET_ANY_NID */ + memset(nid, 0xff, sizeof(*nid)); + return; + } + + nid->nid_size = 0; + nid->nid_type = LNET_NETTYP(LNET_NIDNET(nid4)); + nid->nid_num = __cpu_to_be16(LNET_NETNUM(LNET_NIDNET(nid4))); + nid->nid_addr[0] = __cpu_to_be32(LNET_NIDADDR(nid4)); + nid->nid_addr[1] = nid->nid_addr[2] = nid->nid_addr[3] = 0; +} + +static inline lnet_nid_t lnet_nid_to_nid4(const struct lnet_nid *nid) +{ + if (LNET_NID_IS_ANY(nid)) + return LNET_NID_ANY; + + return LNET_MKNID(LNET_NID_NET(nid), __be32_to_cpu(nid->nid_addr[0])); +} + +static inline int nid_same(const struct lnet_nid *n1, + const struct lnet_nid *n2) +{ + return n1->nid_size == n2->nid_size && + n1->nid_type == n2->nid_type && + n1->nid_num == n2->nid_num && + n1->nid_addr[0] == n2->nid_addr[0] && + n1->nid_addr[1] == n2->nid_addr[1] && + n1->nid_addr[2] == n2->nid_addr[2] && + n1->nid_addr[3] == n2->nid_addr[3]; +} + struct lnet_counters_health { __u32 lch_rst_alloc; __u32 lch_resend_count; diff --git a/include/uapi/linux/lnet/nidstr.h b/include/uapi/linux/lnet/nidstr.h index caf28e2..d5b9d69 100644 --- a/include/uapi/linux/lnet/nidstr.h +++ b/include/uapi/linux/lnet/nidstr.h @@ -62,7 +62,7 @@ enum { struct list_head; #define LNET_NIDSTR_COUNT 1024 /* # of nidstrings */ -#define LNET_NIDSTR_SIZE 32 /* size of each one (see below for usage) */ +#define LNET_NIDSTR_SIZE 64 /* size of each one (see below for usage) */ /* support decl needed by both kernel and user space */ char *libcfs_next_nidstring(void); @@ -90,6 +90,14 @@ static inline char *libcfs_nid2str(lnet_nid_t nid) LNET_NIDSTR_SIZE); } +char *libcfs_nidstr_r(const struct lnet_nid *nid, + char *buf, __kernel_size_t buf_size); +static inline char *libcfs_nidstr(const struct lnet_nid *nid) +{ + return libcfs_nidstr_r(nid, libcfs_next_nidstring(), + LNET_NIDSTR_SIZE); +} + __u32 libcfs_str2net(const char *str); lnet_nid_t libcfs_str2nid(const char *str); int libcfs_str2anynid(lnet_nid_t *nid, const char *str); diff --git a/net/lnet/klnds/o2iblnd/o2iblnd.c b/net/lnet/klnds/o2iblnd/o2iblnd.c index a4949d8..fd807c2 100644 --- a/net/lnet/klnds/o2iblnd/o2iblnd.c +++ b/net/lnet/klnds/o2iblnd/o2iblnd.c @@ -178,8 +178,7 @@ void kiblnd_pack_msg(struct lnet_ni *ni, struct kib_msg *msg, int version, { struct kib_net *net = ni->ni_data; - /* - * CAVEAT EMPTOR! all message fields not set here should have been + /* CAVEAT EMPTOR! all message fields not set here should have been * initialised previously. */ msg->ibm_magic = IBLND_MSG_MAGIC; @@ -188,7 +187,7 @@ void kiblnd_pack_msg(struct lnet_ni *ni, struct kib_msg *msg, int version, msg->ibm_credits = credits; /* ibm_nob */ msg->ibm_cksum = 0; - msg->ibm_srcnid = ni->ni_nid; + msg->ibm_srcnid = lnet_nid_to_nid4(&ni->ni_nid); msg->ibm_srcstamp = net->ibn_incarnation; msg->ibm_dstnid = dstnid; msg->ibm_dststamp = dststamp; @@ -397,7 +396,7 @@ struct kib_peer_ni *kiblnd_find_peer_locked(struct lnet_ni *ni, lnet_nid_t nid) * created. */ if (peer_ni->ibp_nid != nid || - peer_ni->ibp_ni->ni_nid != ni->ni_nid) + !nid_same(&peer_ni->ibp_ni->ni_nid, &ni->ni_nid)) continue; CDEBUG(D_NET, "got peer_ni [%p] -> %s (%d) version: %x\n", @@ -2201,7 +2200,7 @@ static int kiblnd_port_get_attr(struct kib_hca_dev *hdev) list_for_each_entry(net, &hdev->ibh_dev->ibd_nets, ibn_list) { if (val) CDEBUG(D_NETERROR, "Fatal device error for NI %s\n", - libcfs_nid2str(net->ibn_ni->ni_nid)); + libcfs_nidstr(&net->ibn_ni->ni_nid)); atomic_set(&net->ibn_ni->ni_fatal_error_on, val); } } @@ -2591,7 +2590,7 @@ static void kiblnd_shutdown(struct lnet_ni *ni) wait_var_event_warning(&net->ibn_npeers, atomic_read(&net->ibn_npeers) == 0, "%s: waiting for %d peers to disconnect\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), atomic_read(&net->ibn_npeers)); kiblnd_net_fini_pools(net); @@ -2906,7 +2905,7 @@ static int kiblnd_startup(struct lnet_ni *ni) } net->ibn_dev = ibdev; - ni->ni_nid = LNET_MKNID(LNET_NIDNET(ni->ni_nid), ibdev->ibd_ifip); + ni->ni_nid.nid_addr[0] = cpu_to_be32(ibdev->ibd_ifip); ni->ni_dev_cpt = ifaces[i].li_cpt; diff --git a/net/lnet/klnds/o2iblnd/o2iblnd_cb.c b/net/lnet/klnds/o2iblnd/o2iblnd_cb.c index 8ccd2ab..380374e 100644 --- a/net/lnet/klnds/o2iblnd/o2iblnd_cb.c +++ b/net/lnet/klnds/o2iblnd/o2iblnd_cb.c @@ -507,7 +507,7 @@ static int kiblnd_init_rdma(struct kib_conn *conn, struct kib_tx *tx, int type, } if (msg->ibm_srcnid != conn->ibc_peer->ibp_nid || - msg->ibm_dstnid != ni->ni_nid || + msg->ibm_dstnid != lnet_nid_to_nid4(&ni->ni_nid) || msg->ibm_srcstamp != conn->ibc_incarnation || msg->ibm_dststamp != net->ibn_incarnation) { CERROR("Stale rx from %s\n", @@ -2369,11 +2369,12 @@ static int kiblnd_map_tx(struct lnet_ni *ni, struct kib_tx *tx, } if (!ni || /* no matching net */ - ni->ni_nid != reqmsg->ibm_dstnid || /* right NET, wrong NID! */ + lnet_nid_to_nid4(&ni->ni_nid) != + reqmsg->ibm_dstnid || /* right NET, wrong NID! */ net->ibn_dev != ibdev) { /* wrong device */ CERROR("Can't accept conn from %s on %s (%s:%d:%pI4h): bad dst nid %s\n", libcfs_nid2str(nid), - !ni ? "NA" : libcfs_nid2str(ni->ni_nid), + ni ? libcfs_nidstr(&ni->ni_nid) : "NA", ibdev->ibd_ifname, ibdev->ibd_nnets, &ibdev->ibd_ifip, libcfs_nid2str(reqmsg->ibm_dstnid)); @@ -2490,8 +2491,8 @@ static int kiblnd_map_tx(struct lnet_ni *ni, struct kib_tx *tx, * the lower NID connection win so we can move forward. */ if (peer2->ibp_connecting && - nid < ni->ni_nid && peer2->ibp_races < - MAX_CONN_RACES_BEFORE_ABORT) { + nid < lnet_nid_to_nid4(&ni->ni_nid) && + peer2->ibp_races < MAX_CONN_RACES_BEFORE_ABORT) { peer2->ibp_races++; write_unlock_irqrestore(g_lock, flags); @@ -2924,7 +2925,7 @@ static int kiblnd_map_tx(struct lnet_ni *ni, struct kib_tx *tx, } read_lock_irqsave(&kiblnd_data.kib_global_lock, flags); - if (msg->ibm_dstnid == ni->ni_nid && + if (msg->ibm_dstnid == lnet_nid_to_nid4(&ni->ni_nid) && msg->ibm_dststamp == net->ibn_incarnation) rc = 0; else @@ -3568,13 +3569,13 @@ static int kiblnd_map_tx(struct lnet_ni *ni, struct kib_tx *tx, case IB_EVENT_PORT_ERR: case IB_EVENT_DEVICE_FATAL: CERROR("Fatal device error for NI %s\n", - libcfs_nid2str(conn->ibc_peer->ibp_ni->ni_nid)); + libcfs_nidstr(&conn->ibc_peer->ibp_ni->ni_nid)); atomic_set(&conn->ibc_peer->ibp_ni->ni_fatal_error_on, 1); return; case IB_EVENT_PORT_ACTIVE: CERROR("Port reactivated for NI %s\n", - libcfs_nid2str(conn->ibc_peer->ibp_ni->ni_nid)); + libcfs_nidstr(&conn->ibc_peer->ibp_ni->ni_nid)); atomic_set(&conn->ibc_peer->ibp_ni->ni_fatal_error_on, 0); return; diff --git a/net/lnet/klnds/socklnd/socklnd.c b/net/lnet/klnds/socklnd/socklnd.c index 96cb0e0..21569fb 100644 --- a/net/lnet/klnds/socklnd/socklnd.c +++ b/net/lnet/klnds/socklnd/socklnd.c @@ -949,7 +949,7 @@ struct ksock_peer_ni * * Am I already connecting to this guy? Resolve in * favour of higher NID... */ - if (peerid.nid < ni->ni_nid && + if (peerid.nid < lnet_nid_to_nid4(&ni->ni_nid) && ksocknal_connecting(peer_ni->ksnp_conn_cb, ((struct sockaddr *)&conn->ksnc_peeraddr))) { rc = EALREADY; @@ -1820,12 +1820,13 @@ static int ksocknal_push(struct lnet_ni *ni, struct lnet_process_id id) case IOC_LIBCFS_REGISTER_MYNID: /* Ignore if this is a noop */ - if (data->ioc_nid == ni->ni_nid) + if (nid_is_nid4(&ni->ni_nid) && + data->ioc_nid == lnet_nid_to_nid4(&ni->ni_nid)) return 0; CERROR("obsolete IOC_LIBCFS_REGISTER_MYNID: %s(%s)\n", libcfs_nid2str(data->ioc_nid), - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); return -EINVAL; case IOC_LIBCFS_PUSH_CONNECTION: @@ -2369,8 +2370,7 @@ static int ksocknal_device_event(struct notifier_block *unused, LASSERT(ksi); LASSERT(ksi->ksni_addr.ss_family == AF_INET); - ni->ni_nid = LNET_MKNID(LNET_NIDNET(ni->ni_nid), - ntohl(((struct sockaddr_in *)&ksi->ksni_addr)->sin_addr.s_addr)); + ni->ni_nid.nid_addr[0] = ((struct sockaddr_in *)&ksi->ksni_addr)->sin_addr.s_addr; list_add(&net->ksnn_list, &ksocknal_data.ksnd_nets); net->ksnn_ni = ni; ksocknal_data.ksnd_nnets++; diff --git a/net/lnet/klnds/socklnd/socklnd_cb.c b/net/lnet/klnds/socklnd/socklnd_cb.c index efec479..e6cd976 100644 --- a/net/lnet/klnds/socklnd/socklnd_cb.c +++ b/net/lnet/klnds/socklnd/socklnd_cb.c @@ -1579,7 +1579,7 @@ void ksocknal_write_callback(struct ksock_conn *conn) /* rely on caller to hold a ref on socket so it wouldn't disappear */ LASSERT(conn->ksnc_proto); - hello->kshm_src_nid = ni->ni_nid; + hello->kshm_src_nid = lnet_nid_to_nid4(&ni->ni_nid); hello->kshm_dst_nid = peer_nid; hello->kshm_src_pid = the_lnet.ln_pid; @@ -1628,7 +1628,7 @@ void ksocknal_write_callback(struct ksock_conn *conn) LASSERT(!active == !(conn->ksnc_type != SOCKLND_CONN_NONE)); timeout = active ? ksocknal_timeout() : - lnet_acceptor_timeout(); + lnet_acceptor_timeout(); rc = lnet_sock_read(sock, &hello->kshm_magic, sizeof(hello->kshm_magic), timeout); @@ -1672,7 +1672,9 @@ void ksocknal_write_callback(struct ksock_conn *conn) conn->ksnc_proto = &ksocknal_protocol_v1x; #endif hello->kshm_nips = 0; - ksocknal_send_hello(ni, conn, ni->ni_nid, hello); + ksocknal_send_hello(ni, conn, + lnet_nid_to_nid4(&ni->ni_nid), + hello); } CERROR("Unknown protocol version (%d.x expected) from %pIS\n", @@ -1709,7 +1711,7 @@ void ksocknal_write_callback(struct ksock_conn *conn) recv_id.pid = rpc_get_port((struct sockaddr *) &conn->ksnc_peeraddr) | LNET_PID_USERFLAG; - recv_id.nid = LNET_MKNID(LNET_NIDNET(ni->ni_nid), + recv_id.nid = LNET_MKNID(LNET_NID_NET(&ni->ni_nid), ntohl(((struct sockaddr_in *) &conn->ksnc_peeraddr)->sin_addr.s_addr)); } else { diff --git a/net/lnet/lnet/acceptor.c b/net/lnet/lnet/acceptor.c index 3708b89..243c34f 100644 --- a/net/lnet/lnet/acceptor.c +++ b/net/lnet/lnet/acceptor.c @@ -284,7 +284,8 @@ struct socket * ni = lnet_nid2ni_addref(cr.acr_nid); if (!ni || /* no matching net */ - ni->ni_nid != cr.acr_nid) { /* right NET, wrong NID! */ + lnet_nid_to_nid4(&ni->ni_nid) != cr.acr_nid) { + /* right NET, wrong NID! */ if (ni) lnet_ni_decref(ni); LCONSOLE_ERROR_MSG(0x120, diff --git a/net/lnet/lnet/api-ni.c b/net/lnet/lnet/api-ni.c index 41d2d26..9471edb 100644 --- a/net/lnet/lnet/api-ni.c +++ b/net/lnet/lnet/api-ni.c @@ -667,6 +667,17 @@ static void lnet_assert_wire_constants(void) BUILD_BUG_ON((int)sizeof(lnet_nid_t) != 8); BUILD_BUG_ON((int)sizeof(lnet_pid_t) != 4); + /* Checks for struct lnet_nid */ + BUILD_BUG_ON((int)sizeof(struct lnet_nid) != 20); + BUILD_BUG_ON((int)offsetof(struct lnet_nid, nid_size) != 0); + BUILD_BUG_ON((int)sizeof(((struct lnet_nid *)0)->nid_size) != 1); + BUILD_BUG_ON((int)offsetof(struct lnet_nid, nid_type) != 1); + BUILD_BUG_ON((int)sizeof(((struct lnet_nid *)0)->nid_type) != 1); + BUILD_BUG_ON((int)offsetof(struct lnet_nid, nid_num) != 2); + BUILD_BUG_ON((int)sizeof(((struct lnet_nid *)0)->nid_num) != 2); + BUILD_BUG_ON((int)offsetof(struct lnet_nid, nid_addr) != 4); + BUILD_BUG_ON((int)sizeof(((struct lnet_nid *)0)->nid_addr) != 16); + /* Checks for struct lnet_process_id_packed */ BUILD_BUG_ON((int)sizeof(struct lnet_process_id_packed) != 12); BUILD_BUG_ON((int)offsetof(struct lnet_process_id_packed, nid) != 0); @@ -1518,16 +1529,18 @@ struct lnet_net * } struct lnet_ni * -lnet_nid2ni_locked(lnet_nid_t nid, int cpt) +lnet_nid2ni_locked(lnet_nid_t nid4, int cpt) { struct lnet_net *net; struct lnet_ni *ni; + struct lnet_nid nid; LASSERT(cpt != LNET_LOCK_EX); + lnet_nid4_to_nid(nid4, &nid); list_for_each_entry(net, &the_lnet.ln_nets, net_list) { list_for_each_entry(ni, &net->net_ni_list, ni_netlist) { - if (ni->ni_nid == nid) + if (nid_same(&ni->ni_nid, &nid)) return ni; } } @@ -1826,7 +1839,9 @@ struct lnet_ping_buffer * ns = &pbuf->pb_info.pi_ni[i]; - ns->ns_nid = ni->ni_nid; + if (!nid_is_nid4(&ni->ni_nid)) + continue; + ns->ns_nid = lnet_nid_to_nid4(&ni->ni_nid); lnet_ni_lock(ni); ns->ns_status = lnet_ni_get_status_locked(ni); @@ -2142,7 +2157,7 @@ static void lnet_push_target_fini(void) ++i; if ((i & (-i)) == i) { CDEBUG(D_WARNING, "Waiting for zombie LNI %s\n", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); } schedule_timeout_uninterruptible(HZ); @@ -2167,7 +2182,7 @@ static void lnet_push_target_fini(void) if (!islo) CDEBUG(D_LNI, "Removed LNI %s\n", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); lnet_ni_free(ni); i = 2; @@ -2283,7 +2298,6 @@ static void lnet_push_target_fini(void) struct lnet_tx_queue *tq; int i; struct lnet_net *net = ni->ni_net; - u32 seed; mutex_lock(&the_lnet.ln_lnd_mutex); @@ -2339,18 +2353,12 @@ static void lnet_push_target_fini(void) tq->tq_credits = lnet_ni_tq_credits(ni); } - /* Nodes with small feet have little entropy. The NID for this - * node gives the most entropy in the low bits. - */ - seed = LNET_NIDADDR(ni->ni_nid); - add_device_randomness(&seed, sizeof(seed)); - atomic_set(&ni->ni_tx_credits, lnet_ni_tq_credits(ni) * ni->ni_ncpts); atomic_set(&ni->ni_healthv, LNET_MAX_HEALTH_VALUE); CDEBUG(D_LNI, "Added LNI %s [%d/%d/%d/%d]\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), ni->ni_net->net_tunables.lct_peer_tx_credits, lnet_ni_tq_credits(ni) * LNET_CPT_NUMBER, ni->ni_net->net_tunables.lct_peer_rtr_credits, @@ -2924,7 +2932,7 @@ void lnet_lib_exit(void) size_t min_size = 0; int i; - if (!ni || !cfg_ni || !tun) + if (!ni || !cfg_ni || !tun || !nid_is_nid4(&ni->ni_nid)) return; if (ni->ni_interface) { @@ -2933,7 +2941,7 @@ void lnet_lib_exit(void) sizeof(cfg_ni->lic_ni_intf)); } - cfg_ni->lic_nid = ni->ni_nid; + cfg_ni->lic_nid = lnet_nid_to_nid4(&ni->ni_nid); cfg_ni->lic_status = lnet_ni_get_status_locked(ni); cfg_ni->lic_dev_cpt = ni->ni_dev_cpt; @@ -2993,7 +3001,7 @@ void lnet_lib_exit(void) size_t min_size, tunable_size = 0; int i; - if (!ni || !config) + if (!ni || !config || !nid_is_nid4(&ni->ni_nid)) return; net_config = (struct lnet_ioctl_net_config *)config->cfg_bulk; @@ -3007,7 +3015,7 @@ void lnet_lib_exit(void) ni->ni_interface, sizeof(net_config->ni_interface)); - config->cfg_nid = ni->ni_nid; + config->cfg_nid = lnet_nid_to_nid4(&ni->ni_nid); config->cfg_config_u.cfg_net.net_peer_timeout = ni->ni_net->net_tunables.lct_peer_timeout; config->cfg_config_u.cfg_net.net_max_tx_credits = @@ -3287,7 +3295,7 @@ static int lnet_add_net_common(struct lnet_net *net, rc = lnet_udsp_apply_policies_on_ni(ni); if (rc) CERROR("Failed to apply UDSPs on ni %s\n", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); } lnet_net_unlock(LNET_LOCK_EX); @@ -3637,12 +3645,13 @@ u32 lnet_get_dlc_seq_locked(void) lnet_net_lock(LNET_LOCK_EX); list_for_each_entry(net, &the_lnet.ln_nets, net_list) { list_for_each_entry(ni, &net->net_ni_list, ni_netlist) { - if (ni->ni_nid == nid || all) { + if (all || (nid_is_nid4(&ni->ni_nid) && + lnet_nid_to_nid4(&ni->ni_nid) == nid)) { atomic_set(&ni->ni_healthv, value); if (list_empty(&ni->ni_recovery) && value < LNET_MAX_HEALTH_VALUE) { CERROR("manually adding local NI %s to recovery\n", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); list_add_tail(&ni->ni_recovery, &the_lnet.ln_mt_localNIRecovq); lnet_ni_addref_locked(ni, 0); @@ -3666,7 +3675,7 @@ u32 lnet_get_dlc_seq_locked(void) lnet_net_lock(LNET_LOCK_EX); list_for_each_entry(net, &the_lnet.ln_nets, net_list) { list_for_each_entry(ni, &net->net_ni_list, ni_netlist) { - if (ni->ni_nid != nid && !all) + if (lnet_nid_to_nid4(&ni->ni_nid) != nid && !all) continue; if (LNET_NETTYP(net->net_id) == SOCKLND) ni->ni_lnd_tunables.lnd_tun_u.lnd_sock.lnd_conns_per_peer = value; @@ -3729,7 +3738,9 @@ u32 lnet_get_dlc_seq_locked(void) lnet_net_lock(LNET_LOCK_EX); list_for_each_entry(ni, &the_lnet.ln_mt_localNIRecovq, ni_recovery) { - list->rlst_nid_array[i] = ni->ni_nid; + if (!nid_is_nid4(&ni->ni_nid)) + continue; + list->rlst_nid_array[i] = lnet_nid_to_nid4(&ni->ni_nid); i++; if (i >= LNET_MAX_SHOW_NUM_NID) break; @@ -4381,10 +4392,13 @@ void LNetDebugPeer(struct lnet_process_id id) list_for_each_entry(net, &the_lnet.ln_nets, net_list) { list_for_each_entry(ni, &net->net_ni_list, ni_netlist) { + if (!nid_is_nid4(&ni->ni_nid)) + /* FIXME this needs to be handled */ + continue; if (index-- != 0) continue; - id->nid = ni->ni_nid; + id->nid = lnet_nid_to_nid4(&ni->ni_nid); id->pid = the_lnet.ln_pid; rc = 0; break; diff --git a/net/lnet/lnet/config.c b/net/lnet/lnet/config.c index 0117611..0c833fe 100644 --- a/net/lnet/lnet/config.c +++ b/net/lnet/lnet/config.c @@ -375,7 +375,7 @@ struct lnet_net * if (ni->ni_interface) { LCONSOLE_ERROR_MSG(0x115, "%s: interface %s already set for net %s: rc = %d\n", iface, ni->ni_interface, - libcfs_net2str(LNET_NIDNET(ni->ni_nid)), + libcfs_net2str(LNET_NID_NET(&ni->ni_nid)), -EINVAL); return -EINVAL; } @@ -435,7 +435,8 @@ struct lnet_net * ni->ni_net = net; /* LND will fill in the address part of the NID */ - ni->ni_nid = LNET_MKNID(net->net_id, 0); + ni->ni_nid.nid_type = LNET_NETTYP(net->net_id); + ni->ni_nid.nid_num = cpu_to_be16(LNET_NETNUM(net->net_id)); /* Store net namespace in which current ni is being created */ if (current->nsproxy && current->nsproxy->net_ns) diff --git a/net/lnet/lnet/lib-move.c b/net/lnet/lnet/lib-move.c index 035bda3..c70ec37 100644 --- a/net/lnet/lnet/lib-move.c +++ b/net/lnet/lnet/lib-move.c @@ -537,7 +537,7 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, int rc; LASSERT(!in_interrupt()); - LASSERT(ni->ni_nid == LNET_NID_LO_0 || + LASSERT(nid_is_lo0(&ni->ni_nid) || (msg->msg_txcredit && msg->msg_peertxcredit)); rc = ni->ni_net->net_lnd->lnd_send(ni, priv, msg); @@ -648,7 +648,8 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, /* can't get here if we're sending to the loopback interface */ if (the_lnet.ln_loni) - LASSERT(lp->lpni_nid != the_lnet.ln_loni->ni_nid); + LASSERT(lp->lpni_nid != + lnet_nid_to_nid4(&the_lnet.ln_loni->ni_nid)); /* NB 'lp' is always the next hop */ if (!(msg->msg_target.pid & LNET_PID_USERFLAG) && @@ -1133,10 +1134,11 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, * preferred, then let's use it */ if (best_ni) { + /* FIXME need to handle large-addr nid */ lpni_is_preferred = lnet_peer_is_pref_nid_locked(lpni, - best_ni->ni_nid); + lnet_nid_to_nid4(&best_ni->ni_nid)); CDEBUG(D_NET, "%s lpni_is_preferred = %d\n", - libcfs_nid2str(best_ni->ni_nid), + libcfs_nidstr(&best_ni->ni_nid), lpni_is_preferred); } else { lpni_is_preferred = false; @@ -1514,9 +1516,9 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, if (best_ni) CDEBUG(D_NET, "compare ni %s [c:%d, d:%d, s:%d, p:%u, g:%u] with best_ni %s [c:%d, d:%d, s:%d, p:%u, g:%u]\n", - libcfs_nid2str(ni->ni_nid), ni_credits, distance, + libcfs_nidstr(&ni->ni_nid), ni_credits, distance, ni->ni_seq, ni_sel_prio, ni_dev_prio, - (best_ni) ? libcfs_nid2str(best_ni->ni_nid) + (best_ni) ? libcfs_nidstr(&best_ni->ni_nid) : "not selected", best_credits, shortest_distance, (best_ni) ? best_ni->ni_seq : 0, best_sel_prio, best_dev_prio); @@ -1561,7 +1563,7 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, } CDEBUG(D_NET, "selected best_ni %s\n", - (best_ni) ? libcfs_nid2str(best_ni->ni_nid) : "no selection"); + (best_ni) ? libcfs_nidstr(&best_ni->ni_nid) : "no selection"); return best_ni; } @@ -1620,11 +1622,12 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, /* No send credit hassles with LOLND */ lnet_ni_addref_locked(the_lnet.ln_loni, cpt); - msg->msg_hdr.dest_nid = cpu_to_le64(the_lnet.ln_loni->ni_nid); + msg->msg_hdr.dest_nid = + cpu_to_le64(lnet_nid_to_nid4(&the_lnet.ln_loni->ni_nid)); if (!msg->msg_routing) msg->msg_hdr.src_nid = - cpu_to_le64(the_lnet.ln_loni->ni_nid); - msg->msg_target.nid = the_lnet.ln_loni->ni_nid; + cpu_to_le64(lnet_nid_to_nid4(&the_lnet.ln_loni->ni_nid)); + msg->msg_target.nid = lnet_nid_to_nid4(&the_lnet.ln_loni->ni_nid); lnet_msg_commit(msg, cpt); msg->msg_txni = the_lnet.ln_loni; @@ -1655,7 +1658,7 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, CDEBUG(D_NET, "%s NI seq info: [%d:%d:%d:%u] %s LPNI seq info [%d:%d:%d:%u]\n", - libcfs_nid2str(best_ni->ni_nid), + libcfs_nidstr(&best_ni->ni_nid), best_ni->ni_seq, best_ni->ni_net->net_seq, atomic_read(&best_ni->ni_tx_credits), best_ni->ni_sel_priority, @@ -1719,7 +1722,8 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, * originator and set it here. */ if (!msg->msg_routing) - msg->msg_hdr.src_nid = cpu_to_le64(msg->msg_txni->ni_nid); + msg->msg_hdr.src_nid = + cpu_to_le64(lnet_nid_to_nid4(&msg->msg_txni->ni_nid)); if (routing) { msg->msg_target_is_router = 1; @@ -1757,7 +1761,7 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, if (!rc) CDEBUG(D_NET, "TRACE: %s(%s:%s) -> %s(%s:%s) %s : %s try# %d\n", libcfs_nid2str(msg->msg_hdr.src_nid), - libcfs_nid2str(msg->msg_txni->ni_nid), + libcfs_nidstr(&msg->msg_txni->ni_nid), libcfs_nid2str(sd->sd_src_nid), libcfs_nid2str(msg->msg_hdr.dest_nid), libcfs_nid2str(sd->sd_dst_nid), @@ -1775,9 +1779,10 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, if (!lnet_peer_is_multi_rail(lpni->lpni_peer_net->lpn_peer) && !lnet_msg_is_response(msg) && lpni->lpni_pref_nnids == 0) { CDEBUG(D_NET, "Setting preferred local NID %s on NMR peer %s\n", - libcfs_nid2str(lni->ni_nid), + libcfs_nidstr(&lni->ni_nid), libcfs_nid2str(lpni->lpni_nid)); - lnet_peer_ni_set_non_mr_pref_nid(lpni, lni->ni_nid); + lnet_peer_ni_set_non_mr_pref_nid(lpni, + lnet_nid_to_nid4(&lni->ni_nid)); } } @@ -1828,7 +1833,8 @@ void lnet_usr_translate_stats(struct lnet_ioctl_element_msg_stats *msg_stats, } if (sd->sd_best_lpni && - sd->sd_best_lpni->lpni_nid == the_lnet.ln_loni->ni_nid) + sd->sd_best_lpni->lpni_nid == + lnet_nid_to_nid4(&the_lnet.ln_loni->ni_nid)) return lnet_handle_lo_send(sd); else if (sd->sd_best_lpni) return lnet_handle_send(sd); @@ -1951,7 +1957,7 @@ struct lnet_ni * struct lnet_peer_ni *gwni = NULL; bool route_found = false; lnet_nid_t src_nid = (sd->sd_src_nid != LNET_NID_ANY) ? sd->sd_src_nid : - sd->sd_best_ni ? sd->sd_best_ni->ni_nid : + sd->sd_best_ni ? lnet_nid_to_nid4(&sd->sd_best_ni->ni_nid) : LNET_NID_ANY; int best_lpn_healthv = 0; u32 best_lpn_sel_prio = LNET_MAX_SELECTION_PRIORITY; @@ -2454,7 +2460,8 @@ struct lnet_ni * * network */ if (sd->sd_best_lpni && - sd->sd_best_lpni->lpni_nid == the_lnet.ln_loni->ni_nid) { + sd->sd_best_lpni->lpni_nid == + lnet_nid_to_nid4(&the_lnet.ln_loni->ni_nid)) { /* in case we initially started with a routed * destination, let's reset to local */ @@ -3195,7 +3202,7 @@ struct lnet_mt_event_info { lnet_net_unlock(0); CDEBUG(D_NET, "attempting to recover local ni: %s\n", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); lnet_ni_lock(ni); if (!(ni->ni_recovery_state & LNET_NI_RECOVERY_PENDING)) { @@ -3205,7 +3212,7 @@ struct lnet_mt_event_info { ev_info = kzalloc(sizeof(*ev_info), GFP_NOFS); if (!ev_info) { CERROR("out of memory. Can't recover %s\n", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); lnet_ni_lock(ni); ni->ni_recovery_state &= ~LNET_NI_RECOVERY_PENDING; @@ -3218,7 +3225,8 @@ struct lnet_mt_event_info { * We'll unlink the mdh in this case below. */ LNetInvalidateMDHandle(&ni->ni_ping_mdh); - nid = ni->ni_nid; + /* FIXME need to handle large-addr nid */ + nid = lnet_nid_to_nid4(&ni->ni_nid); /* remove the NI from the local queue and drop the * reference count to it while we're recovering @@ -3986,11 +3994,12 @@ void lnet_monitor_thr_stop(void) lnet_ni_recv(ni, msg->msg_private, NULL, 0, 0, 0, 0); msg->msg_receiving = 0; - rc = lnet_send(ni->ni_nid, msg, msg->msg_from); + /* FIXME need to handle large-addr nid */ + rc = lnet_send(lnet_nid_to_nid4(&ni->ni_nid), msg, msg->msg_from); if (rc < 0) { /* didn't get as far as lnet_ni_send() */ CERROR("%s: Unable to send REPLY for GET from %s: %d\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(info.mi_id), rc); lnet_finalize(msg, rc); @@ -4020,7 +4029,7 @@ void lnet_monitor_thr_stop(void) md = lnet_wire_handle2md(&hdr->msg.reply.dst_wmd); if (!md || !md->md_threshold || md->md_me) { CNETERR("%s: Dropping REPLY from %s for %s MD %#llx.%#llx\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(src), + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(src), !md ? "invalid" : "inactive", hdr->msg.reply.dst_wmd.wh_interface_cookie, hdr->msg.reply.dst_wmd.wh_object_cookie); @@ -4040,7 +4049,7 @@ void lnet_monitor_thr_stop(void) if (mlength < rlength && !(md->md_options & LNET_MD_TRUNCATE)) { CNETERR("%s: Dropping REPLY from %s length %d for MD %#llx would overflow (%d)\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(src), + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(src), rlength, hdr->msg.reply.dst_wmd.wh_object_cookie, mlength); lnet_res_unlock(cpt); @@ -4048,7 +4057,7 @@ void lnet_monitor_thr_stop(void) } CDEBUG(D_NET, "%s: Reply from %s of length %d/%d into md %#llx\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(src), + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(src), mlength, rlength, hdr->msg.reply.dst_wmd.wh_object_cookie); lnet_msg_attach_md(msg, md, 0, mlength); @@ -4088,7 +4097,7 @@ void lnet_monitor_thr_stop(void) /* Don't moan; this is expected */ CDEBUG(D_NET, "%s: Dropping ACK from %s to %s MD %#llx.%#llx\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(src), + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(src), !md ? "invalid" : "inactive", hdr->msg.ack.dst_wmd.wh_interface_cookie, hdr->msg.ack.dst_wmd.wh_object_cookie); @@ -4101,7 +4110,7 @@ void lnet_monitor_thr_stop(void) } CDEBUG(D_NET, "%s: ACK from %s into md %#llx\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(src), + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(src), hdr->msg.ack.dst_wmd.wh_object_cookie); lnet_msg_attach_md(msg, md, 0, 0); @@ -4213,12 +4222,13 @@ void lnet_monitor_thr_stop(void) dest_pid = le32_to_cpu(hdr->dest_pid); payload_length = le32_to_cpu(hdr->payload_length); - for_me = (ni->ni_nid == dest_nid); + /* FIXME handle large-addr nids */ + for_me = (lnet_nid_to_nid4(&ni->ni_nid) == dest_nid); cpt = lnet_cpt_of_nid(from_nid, ni); CDEBUG(D_NET, "TRACE: %s(%s) <- %s : %s\n", libcfs_nid2str(dest_nid), - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), libcfs_nid2str(src_nid), lnet_msgtyp2str(type)); @@ -4274,7 +4284,7 @@ void lnet_monitor_thr_stop(void) * or malicious so we chop them off at the knees :) */ if (!for_me) { - if (LNET_NIDNET(dest_nid) == LNET_NIDNET(ni->ni_nid)) { + if (LNET_NIDNET(dest_nid) == LNET_NID_NET(&ni->ni_nid)) { /* should have gone direct */ CERROR("%s, src %s: Bad dest nid %s (should have been sent direct)\n", libcfs_nid2str(from_nid), @@ -4324,8 +4334,9 @@ void lnet_monitor_thr_stop(void) goto drop; } + /* FIXME need to support large-addr nid */ if (!list_empty(&the_lnet.ln_drop_rules) && - lnet_drop_rule_match(hdr, ni->ni_nid, NULL)) { + lnet_drop_rule_match(hdr, lnet_nid_to_nid4(&ni->ni_nid), NULL)) { CDEBUG(D_NET, "%s, src %s, dst %s: Dropping %s to simulate silent message loss\n", libcfs_nid2str(from_nid), libcfs_nid2str(src_nid), libcfs_nid2str(dest_nid), lnet_msgtyp2str(type)); @@ -4368,7 +4379,9 @@ void lnet_monitor_thr_stop(void) } lnet_net_lock(cpt); - lpni = lnet_nid2peerni_locked(from_nid, ni->ni_nid, cpt); + /* FIXME support large-addr nid */ + lpni = lnet_nid2peerni_locked(from_nid, lnet_nid_to_nid4(&ni->ni_nid), + cpt); if (IS_ERR(lpni)) { lnet_net_unlock(cpt); CERROR("%s, src %s: Dropping %s (error %ld looking up sender)\n", @@ -4790,7 +4803,7 @@ struct lnet_msg * msg = kmem_cache_zalloc(lnet_msg_cachep, GFP_NOFS); if (!msg) { CERROR("%s: Dropping REPLY from %s: can't allocate msg\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(peer_id)); + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(peer_id)); goto drop; } @@ -4801,7 +4814,7 @@ struct lnet_msg * if (!getmd->md_threshold) { CERROR("%s: Dropping REPLY from %s for inactive MD %p\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(peer_id), + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(peer_id), getmd); lnet_res_unlock(cpt); goto drop; @@ -4810,7 +4823,7 @@ struct lnet_msg * LASSERT(!getmd->md_offset); CDEBUG(D_NET, "%s: Reply from %s md %p\n", - libcfs_nid2str(ni->ni_nid), libcfs_id2str(peer_id), getmd); + libcfs_nidstr(&ni->ni_nid), libcfs_id2str(peer_id), getmd); /* setup information for lnet_build_msg_event */ msg->msg_initiator = getmsg->msg_txpeer->lpni_peer_net->lpn_peer->lp_primary_nid; @@ -5032,7 +5045,8 @@ struct lnet_msg * cpt = lnet_net_lock_current(); while ((ni = lnet_get_next_ni_locked(NULL, ni))) { - if (ni->ni_nid == dstnid) { + /* FIXME support large-addr nid */ + if (lnet_nid_to_nid4(&ni->ni_nid) == dstnid) { if (srcnidp) *srcnidp = dstnid; if (orderp) { @@ -5046,7 +5060,7 @@ struct lnet_msg * return local_nid_dist_zero ? 0 : 1; } - if (!matched_dstnet && LNET_NIDNET(ni->ni_nid) == dstnet) { + if (!matched_dstnet && LNET_NID_NET(&ni->ni_nid) == dstnet) { matched_dstnet = true; /* We matched the destination net, but we may have * additional local NIs to inspect. @@ -5055,7 +5069,8 @@ struct lnet_msg * * they may be overwritten if we match local NI above. */ if (srcnidp) - *srcnidp = ni->ni_nid; + /* FIXME support large-addr nids */ + *srcnidp = lnet_nid_to_nid4(&ni->ni_nid); if (orderp) { /* Check if ni was originally created in @@ -5110,7 +5125,8 @@ struct lnet_msg * net = lnet_get_net_locked(shortest->lr_lnet); LASSERT(net); ni = lnet_get_next_ni_locked(net, NULL); - *srcnidp = ni->ni_nid; + /* FIXME support large-addr nids */ + *srcnidp = lnet_nid_to_nid4(&ni->ni_nid); } if (orderp) *orderp = order; diff --git a/net/lnet/lnet/lib-msg.c b/net/lnet/lnet/lib-msg.c index e471848..b1f684f 100644 --- a/net/lnet/lnet/lib-msg.c +++ b/net/lnet/lnet/lib-msg.c @@ -473,7 +473,7 @@ CDEBUG(D_NET, "%s added to recovery queue. ping count: %u next ping: %lld health :%d\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), ni->ni_ping_count, ni->ni_next_ping, atomic_read(&ni->ni_healthv)); @@ -796,10 +796,11 @@ /* if we're sending to the LOLND then the msg_txpeer will not be * set. So no need to sanity check it. */ - if (msg->msg_tx_committed && msg->msg_txni->ni_nid != LNET_NID_LO_0) + if (msg->msg_tx_committed && + !nid_is_lo0(&msg->msg_txni->ni_nid)) LASSERT(msg->msg_txpeer); else if (msg->msg_tx_committed && - msg->msg_txni->ni_nid == LNET_NID_LO_0) + nid_is_lo0(&msg->msg_txni->ni_nid)) lo = true; if (hstatus != LNET_MSG_STATUS_OK && @@ -827,7 +828,7 @@ LASSERT(ni); CDEBUG(D_NET, "health check: %s->%s: %s: %s\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), (lo) ? "self" : libcfs_nid2str(lpni->lpni_nid), lnet_msgtyp2str(msg->msg_type), lnet_health_error2str(hstatus)); @@ -1114,7 +1115,7 @@ CDEBUG(D_NET, "src %s(%s)->dst %s: %s simulate health error: %s\n", libcfs_nid2str(msg->msg_hdr.src_nid), - libcfs_nid2str(msg->msg_txni->ni_nid), + libcfs_nidstr(&msg->msg_txni->ni_nid), libcfs_nid2str(msg->msg_hdr.dest_nid), lnet_msgtyp2str(msg->msg_type), lnet_health_error2str(*hstatus)); diff --git a/net/lnet/lnet/lib-socket.c b/net/lnet/lnet/lib-socket.c index 317d3cf..7deb48a 100644 --- a/net/lnet/lnet/lib-socket.c +++ b/net/lnet/lnet/lib-socket.c @@ -235,9 +235,33 @@ int choose_ipv4_src(u32 *ret, int interface, u32 dst_ipaddr, struct net *ns) #if IS_ENABLED(CONFIG_IPV6) case AF_INET6: { struct sockaddr_in6 *sin6 = (void *)&locaddr; + int val = 0; sin6->sin6_family = AF_INET6; sin6->sin6_addr = in6addr_any; + + /* Make sure we get both IPv4 and IPv6 connections. + * This is the default, but it can be overridden so we + * force it back. + */ + /* From v5.7-rc6-2614-g5a892ff2facb when + * kernel_setsockopt() was removed until + * sockptr_t (above) there is no clean way to + * pass kernel address to setsockopt. We could + * use get_fs()/set_fs(), but in this particular + * situation there is an easier way. It depends + * on the fact that at least for these few + * kernels a NULL address to ipv6_setsockopt() + * is treated like the address of a zero. + */ + if (ipv6_only_sock(sock->sk) && !val) { + void *optval = NULL; + + sock->ops->setsockopt(sock, + IPPROTO_IPV6, IPV6_V6ONLY, + optval, sizeof(val)); + } + if (interface >= 0 && remaddr) { struct sockaddr_in6 *rem = (void *)remaddr; @@ -352,7 +376,6 @@ struct socket * lnet_sock_listen(int local_port, int backlog, struct net *ns) { struct socket *sock; - int val = 0; int rc; sock = lnet_sock_create(-1, NULL, local_port, ns); @@ -364,13 +387,6 @@ struct socket * return ERR_PTR(rc); } - /* Make sure we get both IPv4 and IPv6 connections. - * This is the default, but it can be overridden so - * we force it back. - */ - kernel_setsockopt(sock, IPPROTO_IPV6, IPV6_V6ONLY, - (char *)&val, sizeof(val)); - rc = kernel_listen(sock, backlog); if (!rc) return sock; diff --git a/net/lnet/lnet/lo.c b/net/lnet/lnet/lo.c index 4ddf1cd..3d3dcf8 100644 --- a/net/lnet/lnet/lo.c +++ b/net/lnet/lnet/lo.c @@ -40,7 +40,8 @@ LASSERT(!lntmsg->msg_routing); LASSERT(!lntmsg->msg_target_is_router); - return lnet_parse(ni, &lntmsg->msg_hdr, ni->ni_nid, lntmsg, 0); + return lnet_parse(ni, &lntmsg->msg_hdr, + lnet_nid_to_nid4(&ni->ni_nid), lntmsg, 0); } static int diff --git a/net/lnet/lnet/net_fault.c b/net/lnet/lnet/net_fault.c index 0d19da4..4c50eec 100644 --- a/net/lnet/lnet/net_fault.c +++ b/net/lnet/lnet/net_fault.c @@ -684,7 +684,7 @@ struct delay_daemon_data { list_del_init(&msg->msg_list); ni = msg->msg_txni; CDEBUG(D_NET, "TRACE: msg %p %s -> %s : %s\n", msg, - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), libcfs_nid2str(msg->msg_txpeer->lpni_nid), lnet_msgtyp2str(msg->msg_type)); lnet_ni_send(ni, msg); diff --git a/net/lnet/lnet/nidstrings.c b/net/lnet/lnet/nidstrings.c index 209da0f..6da43d5 100644 --- a/net/lnet/lnet/nidstrings.c +++ b/net/lnet/lnet/nidstrings.c @@ -906,6 +906,46 @@ int cfs_print_nidlist(char *buffer, int count, struct list_head *nidlist) } EXPORT_SYMBOL(libcfs_nid2str_r); +char * +libcfs_nidstr_r(const struct lnet_nid *nid, char *buf, size_t buf_size) +{ + u32 nnum = be16_to_cpu(nid->nid_num); + u32 lnd = nid->nid_type; + struct netstrfns *nf; + + if (LNET_NID_IS_ANY(nid)) { + strncpy(buf, "", buf_size); + buf[buf_size - 1] = '\0'; + return buf; + } + + nf = libcfs_lnd2netstrfns(lnd); + if (nf && nid_is_nid4(nid)) { + size_t addr_len; + + nf->nf_addr2str(ntohl(nid->nid_addr[0]), buf, buf_size); + addr_len = strlen(buf); + if (nnum == 0) + snprintf(buf + addr_len, buf_size - addr_len, "@%s", + nf->nf_name); + else + snprintf(buf + addr_len, buf_size - addr_len, "@%s%u", + nf->nf_name, nnum); + } else { + int l = 0; + int words = DIV_ROUND_UP(NID_ADDR_BYTES(nid), 4); + int i; + + for (i = 0; i < words && i < 4; i++) + l = snprintf(buf + l, buf_size - l, "%s%x", + i ? ":" : "", ntohl(nid->nid_addr[i])); + snprintf(buf + l, buf_size - l, "@<%u:%u>", lnd, nnum); + } + + return buf; +} +EXPORT_SYMBOL(libcfs_nidstr_r); + static struct netstrfns * libcfs_str2net_internal(const char *str, u32 *net) { diff --git a/net/lnet/lnet/router.c b/net/lnet/lnet/router.c index 9003d47..6335425 100644 --- a/net/lnet/lnet/router.c +++ b/net/lnet/lnet/router.c @@ -1691,21 +1691,21 @@ bool lnet_router_checker_active(void) LASSERT(!in_interrupt()); CDEBUG(D_NET, "%s notifying %s: %s\n", - !ni ? "userspace" : libcfs_nid2str(ni->ni_nid), + !ni ? "userspace" : libcfs_nidstr(&ni->ni_nid), libcfs_nid2str(nid), alive ? "up" : "down"); if (ni && - LNET_NIDNET(ni->ni_nid) != LNET_NIDNET(nid)) { + LNET_NID_NET(&ni->ni_nid) != LNET_NIDNET(nid)) { CWARN("Ignoring notification of %s %s by %s (different net)\n", libcfs_nid2str(nid), alive ? "birth" : "death", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); return -EINVAL; } /* can't do predictions... */ if (when > now) { CWARN("Ignoring prediction from %s of %s %s %lld seconds in the future\n", - !ni ? "userspace" : libcfs_nid2str(ni->ni_nid), + ni ? libcfs_nidstr(&ni->ni_nid) : "userspace", libcfs_nid2str(nid), alive ? "up" : "down", when - now); return -EINVAL; } diff --git a/net/lnet/lnet/router_proc.c b/net/lnet/lnet/router_proc.c index 43f70b6..6649f06 100644 --- a/net/lnet/lnet/router_proc.c +++ b/net/lnet/lnet/router_proc.c @@ -702,7 +702,7 @@ static int proc_lnet_nis(struct ctl_table *table, int write, s += scnprintf(s, tmpstr + tmpsiz - s, "%-24s %6s %5lld %4d %4d %4d %5d %5d %5d\n", - libcfs_nid2str(ni->ni_nid), stat, + libcfs_nidstr(&ni->ni_nid), stat, last_alive, *ni->ni_refs[i], ni->ni_net->net_tunables.lct_peer_tx_credits, ni->ni_net->net_tunables.lct_peer_rtr_credits, diff --git a/net/lnet/lnet/udsp.c b/net/lnet/lnet/udsp.c index 516db98..4495062 100644 --- a/net/lnet/lnet/udsp.c +++ b/net/lnet/lnet/udsp.c @@ -213,7 +213,7 @@ enum udsp_apply { struct lnet_ud_nid_descr *ni_match = udi->udi_match; u32 priority = (udi->udi_revert) ? -1 : udi->udi_priority; - rc = cfs_match_nid_net(ni->ni_nid, + rc = cfs_match_nid_net(lnet_nid_to_nid4(&ni->ni_nid), ni_match->ud_net_id.udn_net_type, &ni_match->ud_net_id.udn_net_num_range, &ni_match->ud_addr_range); @@ -221,7 +221,7 @@ enum udsp_apply { return 0; CDEBUG(D_NET, "apply udsp on ni %s\n", - libcfs_nid2str(ni->ni_nid)); + libcfs_nidstr(&ni->ni_nid)); /* Detected match. Set NIDs priority */ lnet_ni_set_sel_priority_locked(ni, priority); @@ -481,7 +481,7 @@ enum udsp_apply { ni_action->ud_net_id.udn_net_type) continue; list_for_each_entry(ni, &net->net_ni_list, ni_netlist) { - rc = cfs_match_nid_net(ni->ni_nid, + rc = cfs_match_nid_net(lnet_nid_to_nid4(&ni->ni_nid), ni_action->ud_net_id.udn_net_type, &ni_action->ud_net_id.udn_net_num_range, &ni_action->ud_addr_range); @@ -500,15 +500,16 @@ enum udsp_apply { } } CDEBUG(D_NET, "add nid %s as preferred for peer %s\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), libcfs_nid2str(lpni->lpni_nid)); /* match. Add to pref NIDs */ - rc = lnet_peer_add_pref_nid(lpni, ni->ni_nid); + rc = lnet_peer_add_pref_nid(lpni, + lnet_nid_to_nid4(&ni->ni_nid)); lnet_net_lock(LNET_LOCK_EX); /* success if EEXIST return */ if (rc && rc != -EEXIST) { CERROR("Failed to add %s to %s pref nid list\n", - libcfs_nid2str(ni->ni_nid), + libcfs_nidstr(&ni->ni_nid), libcfs_nid2str(lpni->lpni_nid)); return rc; } -- 1.8.3.1 _______________________________________________ lustre-devel mailing list lustre-devel@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org