X-Git-Url: https://git.kernel.dk/?a=blobdiff_plain;f=engines%2Frdma.c;h=f192f432738da6e8fa448766348ae6e4578fee5d;hb=95625c6d5ad7a4428ebbf55230016136054c7e44;hp=5f0adee5377c61a84080169811eebc16ad0c8b61;hpb=c44212635b91a2c983806761911c8d4ca956050a;p=fio.git diff --git a/engines/rdma.c b/engines/rdma.c index 5f0adee5..f192f432 100644 --- a/engines/rdma.c +++ b/engines/rdma.c @@ -30,7 +30,7 @@ #include #include #include -#include +#include #include #include #include @@ -41,9 +41,9 @@ #include "../fio.h" #include "../hash.h" +#include "../optgroup.h" #include -#include #define FIO_RDMA_MAX_IO_DEPTH 512 @@ -59,6 +59,7 @@ struct rdmaio_options { struct thread_data *td; unsigned int port; enum rdma_io_mode verb; + char *bindname; }; static int str_hostname_cb(void *data, const char *input) @@ -81,6 +82,16 @@ static struct fio_option options[] = { .category = FIO_OPT_C_ENGINE, .group = FIO_OPT_G_RDMA, }, + { + .name = "bindname", + .lname = "rdma engine bindname", + .type = FIO_OPT_STR_STORE, + .off1 = offsetof(struct rdmaio_options, bindname), + .help = "Bind for RDMA IO engine", + .def = "", + .category = FIO_OPT_C_ENGINE, + .group = FIO_OPT_G_RDMA, + }, { .name = "port", .lname = "rdma engine port", @@ -115,7 +126,7 @@ static struct fio_option options[] = { }, { .ival = "recv", .oval = FIO_RDMA_CHA_RECV, - .help = "Posted Recieve", + .help = "Posted Receive", }, }, .category = FIO_OPT_C_ENGINE, @@ -190,7 +201,7 @@ struct rdmaio_data { static int client_recv(struct thread_data *td, struct ibv_wc *wc) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; unsigned int max_bs; if (wc->byte_len != sizeof(rd->recv_buf)) { @@ -215,7 +226,7 @@ static int client_recv(struct thread_data *td, struct ibv_wc *wc) rd->rmt_nr = ntohl(rd->recv_buf.nr); for (i = 0; i < rd->rmt_nr; i++) { - rd->rmt_us[i].buf = ntohll(rd->recv_buf.rmt_us[i].buf); + rd->rmt_us[i].buf = be64_to_cpu(rd->recv_buf.rmt_us[i].buf); rd->rmt_us[i].rkey = ntohl(rd->recv_buf.rmt_us[i].rkey); rd->rmt_us[i].size = ntohl(rd->recv_buf.rmt_us[i].size); @@ -231,7 +242,7 @@ static int client_recv(struct thread_data *td, struct ibv_wc *wc) static int server_recv(struct thread_data *td, struct ibv_wc *wc) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; unsigned int max_bs; if (wc->wr_id == FIO_RDMA_MAX_IO_DEPTH) { @@ -256,7 +267,7 @@ static int server_recv(struct thread_data *td, struct ibv_wc *wc) static int cq_event_handler(struct thread_data *td, enum ibv_wc_opcode opcode) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct ibv_wc wc; struct rdma_io_u_data *r_io_u_d; int ret; @@ -367,7 +378,7 @@ static int cq_event_handler(struct thread_data *td, enum ibv_wc_opcode opcode) */ static int rdma_poll_wait(struct thread_data *td, enum ibv_wc_opcode opcode) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct ibv_cq *ev_cq; void *ev_ctx; int ret; @@ -404,7 +415,7 @@ again: static int fio_rdmaio_setup_qp(struct thread_data *td) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct ibv_qp_init_attr init_attr; int qp_depth = td->o.iodepth * 2; /* 2 times of io depth */ @@ -414,7 +425,7 @@ static int fio_rdmaio_setup_qp(struct thread_data *td) rd->pd = ibv_alloc_pd(rd->cm_id->verbs); if (rd->pd == NULL) { - log_err("fio: ibv_alloc_pd fail\n"); + log_err("fio: ibv_alloc_pd fail: %m\n"); return 1; } @@ -423,7 +434,7 @@ static int fio_rdmaio_setup_qp(struct thread_data *td) else rd->channel = ibv_create_comp_channel(rd->cm_id->verbs); if (rd->channel == NULL) { - log_err("fio: ibv_create_comp_channel fail\n"); + log_err("fio: ibv_create_comp_channel fail: %m\n"); goto err1; } @@ -437,12 +448,12 @@ static int fio_rdmaio_setup_qp(struct thread_data *td) rd->cq = ibv_create_cq(rd->cm_id->verbs, qp_depth, rd, rd->channel, 0); if (rd->cq == NULL) { - log_err("fio: ibv_create_cq failed\n"); + log_err("fio: ibv_create_cq failed: %m\n"); goto err2; } if (ibv_req_notify_cq(rd->cq, 0) != 0) { - log_err("fio: ibv_create_cq failed\n"); + log_err("fio: ibv_req_notify_cq failed: %m\n"); goto err3; } @@ -458,13 +469,13 @@ static int fio_rdmaio_setup_qp(struct thread_data *td) if (rd->is_client == 0) { if (rdma_create_qp(rd->child_cm_id, rd->pd, &init_attr) != 0) { - log_err("fio: rdma_create_qp failed\n"); + log_err("fio: rdma_create_qp failed: %m\n"); goto err3; } rd->qp = rd->child_cm_id->qp; } else { if (rdma_create_qp(rd->cm_id, rd->pd, &init_attr) != 0) { - log_err("fio: rdma_create_qp failed\n"); + log_err("fio: rdma_create_qp failed: %m\n"); goto err3; } rd->qp = rd->cm_id->qp; @@ -484,19 +495,19 @@ err1: static int fio_rdmaio_setup_control_msg_buffers(struct thread_data *td) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; rd->recv_mr = ibv_reg_mr(rd->pd, &rd->recv_buf, sizeof(rd->recv_buf), IBV_ACCESS_LOCAL_WRITE); if (rd->recv_mr == NULL) { - log_err("fio: recv_buf reg_mr failed\n"); + log_err("fio: recv_buf reg_mr failed: %m\n"); return 1; } rd->send_mr = ibv_reg_mr(rd->pd, &rd->send_buf, sizeof(rd->send_buf), 0); if (rd->send_mr == NULL) { - log_err("fio: send_buf reg_mr failed\n"); + log_err("fio: send_buf reg_mr failed: %m\n"); ibv_dereg_mr(rd->recv_mr); return 1; } @@ -528,7 +539,7 @@ static int get_next_channel_event(struct thread_data *td, struct rdma_event_channel *channel, enum rdma_cm_event_type wait_event) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct rdma_cm_event *event; int ret; @@ -560,7 +571,7 @@ static int get_next_channel_event(struct thread_data *td, static int fio_rdmaio_prep(struct thread_data *td, struct io_u *io_u) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct rdma_io_u_data *r_io_u_d; r_io_u_d = io_u->engine_data; @@ -603,7 +614,7 @@ static int fio_rdmaio_prep(struct thread_data *td, struct io_u *io_u) static struct io_u *fio_rdmaio_event(struct thread_data *td, int event) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct io_u *io_u; int i; @@ -621,7 +632,7 @@ static struct io_u *fio_rdmaio_event(struct thread_data *td, int event) static int fio_rdmaio_getevents(struct thread_data *td, unsigned int min, unsigned int max, const struct timespec *t) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; enum ibv_wc_opcode comp_opcode; struct ibv_cq *ev_cq; void *ev_ctx; @@ -683,7 +694,7 @@ again: static int fio_rdmaio_send(struct thread_data *td, struct io_u **io_us, unsigned int nr) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct ibv_send_wr *bad_wr; #if 0 enum ibv_wc_opcode comp_opcode; @@ -730,7 +741,7 @@ static int fio_rdmaio_send(struct thread_data *td, struct io_u **io_us, } if (ibv_post_send(rd->qp, &r_io_u_d->sq_wr, &bad_wr) != 0) { - log_err("fio: ibv_post_send fail\n"); + log_err("fio: ibv_post_send fail: %m\n"); return -1; } @@ -746,7 +757,7 @@ static int fio_rdmaio_send(struct thread_data *td, struct io_u **io_us, static int fio_rdmaio_recv(struct thread_data *td, struct io_u **io_us, unsigned int nr) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct ibv_recv_wr *bad_wr; struct rdma_io_u_data *r_io_u_d; int i; @@ -758,7 +769,7 @@ static int fio_rdmaio_recv(struct thread_data *td, struct io_u **io_us, r_io_u_d = io_us[i]->engine_data; if (ibv_post_recv(rd->qp, &r_io_u_d->rq_wr, &bad_wr) != 0) { - log_err("fio: ibv_post_recv fail\n"); + log_err("fio: ibv_post_recv fail: %m\n"); return 1; } } @@ -766,7 +777,7 @@ static int fio_rdmaio_recv(struct thread_data *td, struct io_u **io_us, || (rd->rdma_protocol == FIO_RDMA_MEM_WRITE)) { /* re-post the rq_wr */ if (ibv_post_recv(rd->qp, &rd->rq_wr, &bad_wr) != 0) { - log_err("fio: ibv_post_recv fail\n"); + log_err("fio: ibv_post_recv fail: %m\n"); return 1; } @@ -780,9 +791,10 @@ static int fio_rdmaio_recv(struct thread_data *td, struct io_u **io_us, return i; } -static int fio_rdmaio_queue(struct thread_data *td, struct io_u *io_u) +static enum fio_q_status fio_rdmaio_queue(struct thread_data *td, + struct io_u *io_u) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; fio_ro_check(td, io_u); @@ -800,8 +812,8 @@ static int fio_rdmaio_queue(struct thread_data *td, struct io_u *io_u) static void fio_rdmaio_queued(struct thread_data *td, struct io_u **io_us, unsigned int nr) { - struct rdmaio_data *rd = td->io_ops->data; - struct timeval now; + struct rdmaio_data *rd = td->io_ops_data; + struct timespec now; unsigned int i; if (!fio_fill_issue_time(td)) @@ -823,7 +835,7 @@ static void fio_rdmaio_queued(struct thread_data *td, struct io_u **io_us, static int fio_rdmaio_commit(struct thread_data *td) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct io_u **io_us; int ret; @@ -855,7 +867,7 @@ static int fio_rdmaio_commit(struct thread_data *td) static int fio_rdmaio_connect(struct thread_data *td, struct fio_file *f) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct rdma_conn_param conn_param; struct ibv_send_wr *bad_wr; @@ -865,7 +877,7 @@ static int fio_rdmaio_connect(struct thread_data *td, struct fio_file *f) conn_param.retry_count = 10; if (rdma_connect(rd->cm_id, &conn_param) != 0) { - log_err("fio: rdma_connect fail\n"); + log_err("fio: rdma_connect fail: %m\n"); return 1; } @@ -880,7 +892,7 @@ static int fio_rdmaio_connect(struct thread_data *td, struct fio_file *f) rd->send_buf.nr = htonl(td->o.iodepth); if (ibv_post_send(rd->qp, &rd->sq_wr, &bad_wr) != 0) { - log_err("fio: ibv_post_send fail"); + log_err("fio: ibv_post_send fail: %m\n"); return 1; } @@ -906,7 +918,7 @@ static int fio_rdmaio_connect(struct thread_data *td, struct fio_file *f) static int fio_rdmaio_accept(struct thread_data *td, struct fio_file *f) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct rdma_conn_param conn_param; struct ibv_send_wr *bad_wr; int ret = 0; @@ -917,7 +929,7 @@ static int fio_rdmaio_accept(struct thread_data *td, struct fio_file *f) conn_param.initiator_depth = 1; if (rdma_accept(rd->child_cm_id, &conn_param) != 0) { - log_err("fio: rdma_accept\n"); + log_err("fio: rdma_accept: %m\n"); return 1; } @@ -931,7 +943,7 @@ static int fio_rdmaio_accept(struct thread_data *td, struct fio_file *f) ret = rdma_poll_wait(td, IBV_WC_RECV) < 0; if (ibv_post_send(rd->qp, &rd->sq_wr, &bad_wr) != 0) { - log_err("fio: ibv_post_send fail"); + log_err("fio: ibv_post_send fail: %m\n"); return 1; } @@ -951,7 +963,7 @@ static int fio_rdmaio_open_file(struct thread_data *td, struct fio_file *f) static int fio_rdmaio_close_file(struct thread_data *td, struct fio_file *f) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct ibv_send_wr *bad_wr; /* unregister rdma buffer */ @@ -964,7 +976,7 @@ static int fio_rdmaio_close_file(struct thread_data *td, struct fio_file *f) || (rd->rdma_protocol == FIO_RDMA_MEM_READ))) { if (ibv_post_send(rd->qp, &rd->sq_wr, &bad_wr) != 0) { - log_err("fio: ibv_post_send fail"); + log_err("fio: ibv_post_send fail: %m\n"); return 1; } @@ -1004,30 +1016,53 @@ static int fio_rdmaio_close_file(struct thread_data *td, struct fio_file *f) return 0; } +static int aton(struct thread_data *td, const char *host, + struct sockaddr_in *addr) +{ + if (inet_aton(host, &addr->sin_addr) != 1) { + struct hostent *hent; + + hent = gethostbyname(host); + if (!hent) { + td_verror(td, errno, "gethostbyname"); + return 1; + } + + memcpy(&addr->sin_addr, hent->h_addr, 4); + } + return 0; +} + static int fio_rdmaio_setup_connect(struct thread_data *td, const char *host, unsigned short port) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; + struct rdmaio_options *o = td->eo; + struct sockaddr_storage addrb; struct ibv_recv_wr *bad_wr; int err; rd->addr.sin_family = AF_INET; rd->addr.sin_port = htons(port); - if (inet_aton(host, &rd->addr.sin_addr) != 1) { - struct hostent *hent; + err = aton(td, host, &rd->addr); + if (err) + return err; - hent = gethostbyname(host); - if (!hent) { - td_verror(td, errno, "gethostbyname"); - return 1; - } + /* resolve route */ + if (o->bindname && strlen(o->bindname)) { + addrb.ss_family = AF_INET; + err = aton(td, o->bindname, (struct sockaddr_in *)&addrb); + if (err) + return err; + err = rdma_resolve_addr(rd->cm_id, (struct sockaddr *)&addrb, + (struct sockaddr *)&rd->addr, 2000); - memcpy(&rd->addr.sin_addr, hent->h_addr, 4); + } else { + err = rdma_resolve_addr(rd->cm_id, NULL, + (struct sockaddr *)&rd->addr, 2000); } - /* resolve route */ - err = rdma_resolve_addr(rd->cm_id, NULL, (struct sockaddr *)&rd->addr, 2000); if (err != 0) { log_err("fio: rdma_resolve_addr: %d\n", err); return 1; @@ -1071,24 +1106,34 @@ static int fio_rdmaio_setup_connect(struct thread_data *td, const char *host, static int fio_rdmaio_setup_listen(struct thread_data *td, short port) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; + struct rdmaio_options *o = td->eo; struct ibv_recv_wr *bad_wr; + int state = td->runstate; + + td_set_runstate(td, TD_SETTING_UP); rd->addr.sin_family = AF_INET; - rd->addr.sin_addr.s_addr = htonl(INADDR_ANY); rd->addr.sin_port = htons(port); + if (!o->bindname || !strlen(o->bindname)) + rd->addr.sin_addr.s_addr = htonl(INADDR_ANY); + else + rd->addr.sin_addr.s_addr = htonl(*o->bindname); + /* rdma_listen */ if (rdma_bind_addr(rd->cm_id, (struct sockaddr *)&rd->addr) != 0) { - log_err("fio: rdma_bind_addr fail\n"); + log_err("fio: rdma_bind_addr fail: %m\n"); return 1; } if (rdma_listen(rd->cm_id, 3) != 0) { - log_err("fio: rdma_listen fail\n"); + log_err("fio: rdma_listen fail: %m\n"); return 1; } + log_info("fio: waiting for connection\n"); + /* wait for CONNECT_REQUEST */ if (get_next_channel_event (td, rd->cm_channel, RDMA_CM_EVENT_CONNECT_REQUEST) != 0) { @@ -1104,10 +1149,11 @@ static int fio_rdmaio_setup_listen(struct thread_data *td, short port) /* post recv buf */ if (ibv_post_recv(rd->qp, &rd->rq_wr, &bad_wr) != 0) { - log_err("fio: ibv_post_recv fail\n"); + log_err("fio: ibv_post_recv fail: %m\n"); return 1; } + td_set_runstate(td, state); return 0; } @@ -1149,7 +1195,8 @@ static int compat_options(struct thread_data *td) { // The original RDMA engine had an ugly / seperator // on the filename for it's options. This function - // retains backwards compatibility with it.100 + // retains backwards compatibility with it. Note we do not + // support setting the bindname option is this legacy mode. struct rdmaio_options *o = td->eo; char *modep, *portp; @@ -1200,10 +1247,9 @@ bad_host: static int fio_rdmaio_init(struct thread_data *td) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; struct rdmaio_options *o = td->eo; - unsigned int max_bs; - int ret, i; + int ret; if (td_rw(td)) { log_err("fio: rdma connections must be read OR write\n"); @@ -1231,13 +1277,13 @@ static int fio_rdmaio_init(struct thread_data *td) rd->cm_channel = rdma_create_event_channel(); if (!rd->cm_channel) { - log_err("fio: rdma_create_event_channel fail\n"); + log_err("fio: rdma_create_event_channel fail: %m\n"); return 1; } ret = rdma_create_id(rd->cm_channel, &rd->cm_id, rd, RDMA_PS_TCP); if (ret) { - log_err("fio: rdma_create_id fail\n"); + log_err("fio: rdma_create_id fail: %m\n"); return 1; } @@ -1264,12 +1310,20 @@ static int fio_rdmaio_init(struct thread_data *td) if (td_read(td)) { /* READ as the server */ rd->is_client = 0; + td->flags |= TD_F_NO_PROGRESS; /* server rd->rdma_buf_len will be setup after got request */ ret = fio_rdmaio_setup_listen(td, o->port); } else { /* WRITE as the client */ rd->is_client = 1; ret = fio_rdmaio_setup_connect(td, td->o.filename, o->port); } + return ret; +} +static int fio_rdmaio_post_init(struct thread_data *td) +{ + unsigned int max_bs; + int i; + struct rdmaio_data *rd = td->io_ops_data; max_bs = max(td->o.max_bs[DDIR_READ], td->o.max_bs[DDIR_WRITE]); rd->send_buf.max_bs = htonl(max_bs); @@ -1287,12 +1341,12 @@ static int fio_rdmaio_init(struct thread_data *td) IBV_ACCESS_REMOTE_READ | IBV_ACCESS_REMOTE_WRITE); if (io_u->mr == NULL) { - log_err("fio: ibv_reg_mr io_u failed\n"); + log_err("fio: ibv_reg_mr io_u failed: %m\n"); return 1; } rd->send_buf.rmt_us[i].buf = - htonll((uint64_t) (unsigned long)io_u->buf); + cpu_to_be64((uint64_t) (unsigned long)io_u->buf); rd->send_buf.rmt_us[i].rkey = htonl(io_u->mr->rkey); rd->send_buf.rmt_us[i].size = htonl(max_bs); @@ -1303,12 +1357,12 @@ static int fio_rdmaio_init(struct thread_data *td) rd->send_buf.nr = htonl(i); - return ret; + return 0; } static void fio_rdmaio_cleanup(struct thread_data *td) { - struct rdmaio_data *rd = td->io_ops->data; + struct rdmaio_data *rd = td->io_ops_data; if (rd) free(rd); @@ -1324,12 +1378,12 @@ static int fio_rdmaio_setup(struct thread_data *td) td->o.open_files++; } - if (!td->io_ops->data) { + if (!td->io_ops_data) { rd = malloc(sizeof(*rd)); memset(rd, 0, sizeof(*rd)); init_rand_seed(&rd->rand_state, (unsigned int) GOLDEN_RATIO_PRIME, 0); - td->io_ops->data = rd; + td->io_ops_data = rd; } return 0; @@ -1340,6 +1394,7 @@ static struct ioengine_ops ioengine_rw = { .version = FIO_IOOPS_VERSION, .setup = fio_rdmaio_setup, .init = fio_rdmaio_init, + .post_init = fio_rdmaio_post_init, .prep = fio_rdmaio_prep, .queue = fio_rdmaio_queue, .commit = fio_rdmaio_commit,