Commit | Line | Data |
---|---|---|
e4c4625f JM |
1 | /* |
2 | * librpma_gpspm: IO engine that uses PMDK librpma to write data, | |
3 | * based on General Purpose Server Persistency Method | |
4 | * | |
5 | * Copyright 2020-2021, Intel Corporation | |
6 | * | |
7 | * This program is free software; you can redistribute it and/or | |
8 | * modify it under the terms of the GNU General Public License, | |
9 | * version 2 as published by the Free Software Foundation.. | |
10 | * | |
11 | * This program is distributed in the hope that it will be useful, | |
12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
14 | * GNU General Public License for more details. | |
15 | */ | |
16 | ||
17 | #include "librpma_fio.h" | |
18 | ||
19 | #include <libpmem.h> | |
20 | ||
21 | /* Generated by the protocol buffer compiler from: librpma_gpspm_flush.proto */ | |
22 | #include "librpma_gpspm_flush.pb-c.h" | |
23 | ||
24 | #define MAX_MSG_SIZE (512) | |
25 | #define IO_U_BUF_LEN (2 * MAX_MSG_SIZE) | |
26 | #define SEND_OFFSET (0) | |
27 | #define RECV_OFFSET (SEND_OFFSET + MAX_MSG_SIZE) | |
28 | ||
29 | #define GPSPM_FLUSH_REQUEST__LAST \ | |
30 | { PROTOBUF_C_MESSAGE_INIT(&gpspm_flush_request__descriptor), 0, 0, 0 } | |
31 | ||
32 | /* | |
33 | * 'Flush_req_last' is the last flush request | |
34 | * the client has to send to server to indicate | |
35 | * that the client is done. | |
36 | */ | |
37 | static const GPSPMFlushRequest Flush_req_last = GPSPM_FLUSH_REQUEST__LAST; | |
38 | ||
39 | #define IS_NOT_THE_LAST_MESSAGE(flush_req) \ | |
40 | (flush_req->length != Flush_req_last.length || \ | |
41 | flush_req->offset != Flush_req_last.offset) | |
42 | ||
43 | /* client side implementation */ | |
44 | ||
45 | /* get next io_u message buffer in the round-robin fashion */ | |
46 | #define IO_U_NEXT_BUF_OFF_CLIENT(cd) \ | |
47 | (IO_U_BUF_LEN * ((cd->msg_curr++) % cd->msg_num)) | |
48 | ||
49 | struct client_data { | |
50 | /* memory for sending and receiving buffered */ | |
51 | char *io_us_msgs; | |
52 | ||
53 | /* resources for messaging buffer */ | |
54 | uint32_t msg_num; | |
55 | uint32_t msg_curr; | |
56 | struct rpma_mr_local *msg_mr; | |
57 | }; | |
58 | ||
59 | static inline int client_io_flush(struct thread_data *td, | |
60 | struct io_u *first_io_u, struct io_u *last_io_u, | |
61 | unsigned long long int len); | |
62 | ||
63 | static int client_get_io_u_index(struct rpma_completion *cmpl, | |
64 | unsigned int *io_u_index); | |
65 | ||
66 | static int client_init(struct thread_data *td) | |
67 | { | |
68 | struct librpma_fio_client_data *ccd; | |
69 | struct client_data *cd; | |
70 | uint32_t write_num; | |
71 | struct rpma_conn_cfg *cfg = NULL; | |
72 | int ret; | |
73 | ||
74 | /* | |
75 | * not supported: | |
76 | * - readwrite = read / trim / randread / randtrim / | |
77 | * / rw / randrw / trimwrite | |
78 | */ | |
79 | if (td_read(td) || td_trim(td)) { | |
80 | td_verror(td, EINVAL, "Not supported mode."); | |
81 | return -1; | |
82 | } | |
83 | ||
84 | /* allocate client's data */ | |
85 | cd = calloc(1, sizeof(*cd)); | |
86 | if (cd == NULL) { | |
87 | td_verror(td, errno, "calloc"); | |
88 | return -1; | |
89 | } | |
90 | ||
91 | /* | |
92 | * Calculate the required number of WRITEs and FLUSHes. | |
93 | * | |
94 | * Note: Each flush is a request (SEND) and response (RECV) pair. | |
95 | */ | |
96 | if (td_random(td)) { | |
97 | write_num = td->o.iodepth; /* WRITE * N */ | |
98 | cd->msg_num = td->o.iodepth; /* FLUSH * N */ | |
99 | } else { | |
100 | if (td->o.sync_io) { | |
101 | write_num = 1; /* WRITE */ | |
102 | cd->msg_num = 1; /* FLUSH */ | |
103 | } else { | |
104 | write_num = td->o.iodepth; /* WRITE * N */ | |
105 | /* | |
106 | * FLUSH * B where: | |
107 | * - B == ceil(iodepth / iodepth_batch) | |
108 | * which is the number of batches for N writes | |
109 | */ | |
110 | cd->msg_num = LIBRPMA_FIO_CEIL(td->o.iodepth, | |
111 | td->o.iodepth_batch); | |
112 | } | |
113 | } | |
114 | ||
115 | /* create a connection configuration object */ | |
116 | if ((ret = rpma_conn_cfg_new(&cfg))) { | |
117 | librpma_td_verror(td, ret, "rpma_conn_cfg_new"); | |
118 | goto err_free_cd; | |
119 | } | |
120 | ||
121 | /* | |
122 | * Calculate the required queue sizes where: | |
123 | * - the send queue (SQ) has to be big enough to accommodate | |
124 | * all io_us (WRITEs) and all flush requests (SENDs) | |
125 | * - the receive queue (RQ) has to be big enough to accommodate | |
126 | * all flush responses (RECVs) | |
127 | * - the completion queue (CQ) has to be big enough to accommodate all | |
128 | * success and error completions (sq_size + rq_size) | |
129 | */ | |
130 | if ((ret = rpma_conn_cfg_set_sq_size(cfg, write_num + cd->msg_num))) { | |
131 | librpma_td_verror(td, ret, "rpma_conn_cfg_set_sq_size"); | |
132 | goto err_cfg_delete; | |
133 | } | |
134 | if ((ret = rpma_conn_cfg_set_rq_size(cfg, cd->msg_num))) { | |
135 | librpma_td_verror(td, ret, "rpma_conn_cfg_set_rq_size"); | |
136 | goto err_cfg_delete; | |
137 | } | |
138 | if ((ret = rpma_conn_cfg_set_cq_size(cfg, write_num + cd->msg_num * 2))) { | |
139 | librpma_td_verror(td, ret, "rpma_conn_cfg_set_cq_size"); | |
140 | goto err_cfg_delete; | |
141 | } | |
142 | ||
143 | if (librpma_fio_client_init(td, cfg)) | |
144 | goto err_cfg_delete; | |
145 | ||
146 | ccd = td->io_ops_data; | |
147 | ||
148 | if (ccd->ws->direct_write_to_pmem && | |
149 | ccd->server_mr_flush_type == RPMA_FLUSH_TYPE_PERSISTENT && | |
150 | td->thread_number == 1) { | |
151 | /* XXX log_info mixes with the JSON output */ | |
152 | log_err( | |
153 | "Note: The server side supports Direct Write to PMem and it is equipped with PMem (direct_write_to_pmem).\n" | |
154 | "You can use librpma_client and librpma_server engines for better performance instead of GPSPM.\n"); | |
155 | } | |
156 | ||
157 | /* validate the server's RQ capacity */ | |
158 | if (cd->msg_num > ccd->ws->max_msg_num) { | |
159 | log_err( | |
160 | "server's RQ size (iodepth) too small to handle the client's workspace requirements (%u < %u)\n", | |
161 | ccd->ws->max_msg_num, cd->msg_num); | |
162 | goto err_cleanup_common; | |
163 | } | |
164 | ||
165 | if ((ret = rpma_conn_cfg_delete(&cfg))) { | |
166 | librpma_td_verror(td, ret, "rpma_conn_cfg_delete"); | |
167 | /* non fatal error - continue */ | |
168 | } | |
169 | ||
170 | ccd->flush = client_io_flush; | |
171 | ccd->get_io_u_index = client_get_io_u_index; | |
172 | ccd->client_data = cd; | |
173 | ||
174 | return 0; | |
175 | ||
176 | err_cleanup_common: | |
177 | librpma_fio_client_cleanup(td); | |
178 | ||
179 | err_cfg_delete: | |
180 | (void) rpma_conn_cfg_delete(&cfg); | |
181 | ||
182 | err_free_cd: | |
183 | free(cd); | |
184 | ||
185 | return -1; | |
186 | } | |
187 | ||
188 | static int client_post_init(struct thread_data *td) | |
189 | { | |
190 | struct librpma_fio_client_data *ccd = td->io_ops_data; | |
191 | struct client_data *cd = ccd->client_data; | |
192 | unsigned int io_us_msgs_size; | |
193 | int ret; | |
194 | ||
195 | /* message buffers initialization and registration */ | |
196 | io_us_msgs_size = cd->msg_num * IO_U_BUF_LEN; | |
197 | if ((ret = posix_memalign((void **)&cd->io_us_msgs, page_size, | |
198 | io_us_msgs_size))) { | |
199 | td_verror(td, ret, "posix_memalign"); | |
200 | return ret; | |
201 | } | |
202 | if ((ret = rpma_mr_reg(ccd->peer, cd->io_us_msgs, io_us_msgs_size, | |
203 | RPMA_MR_USAGE_SEND | RPMA_MR_USAGE_RECV, | |
204 | &cd->msg_mr))) { | |
205 | librpma_td_verror(td, ret, "rpma_mr_reg"); | |
206 | return ret; | |
207 | } | |
208 | ||
209 | return librpma_fio_client_post_init(td); | |
210 | } | |
211 | ||
212 | static void client_cleanup(struct thread_data *td) | |
213 | { | |
214 | struct librpma_fio_client_data *ccd = td->io_ops_data; | |
215 | struct client_data *cd; | |
216 | size_t flush_req_size; | |
217 | size_t io_u_buf_off; | |
218 | size_t send_offset; | |
219 | void *send_ptr; | |
220 | int ret; | |
221 | ||
222 | if (ccd == NULL) | |
223 | return; | |
224 | ||
225 | cd = ccd->client_data; | |
226 | if (cd == NULL) { | |
227 | librpma_fio_client_cleanup(td); | |
228 | return; | |
229 | } | |
230 | ||
231 | /* | |
232 | * Make sure all SEND completions are collected ergo there are free | |
233 | * slots in the SQ for the last SEND message. | |
234 | * | |
235 | * Note: If any operation will fail we still can send the termination | |
236 | * notice. | |
237 | */ | |
238 | (void) librpma_fio_client_io_complete_all_sends(td); | |
239 | ||
240 | /* prepare the last flush message and pack it to the send buffer */ | |
241 | flush_req_size = gpspm_flush_request__get_packed_size(&Flush_req_last); | |
242 | if (flush_req_size > MAX_MSG_SIZE) { | |
243 | log_err( | |
244 | "Packed flush request size is bigger than available send buffer space (%zu > %d\n", | |
245 | flush_req_size, MAX_MSG_SIZE); | |
246 | } else { | |
247 | io_u_buf_off = IO_U_NEXT_BUF_OFF_CLIENT(cd); | |
248 | send_offset = io_u_buf_off + SEND_OFFSET; | |
249 | send_ptr = cd->io_us_msgs + send_offset; | |
250 | (void) gpspm_flush_request__pack(&Flush_req_last, send_ptr); | |
251 | ||
252 | /* send the flush message */ | |
253 | if ((ret = rpma_send(ccd->conn, cd->msg_mr, send_offset, | |
254 | flush_req_size, RPMA_F_COMPLETION_ALWAYS, | |
255 | NULL))) | |
256 | librpma_td_verror(td, ret, "rpma_send"); | |
257 | ||
258 | ++ccd->op_send_posted; | |
259 | ||
260 | /* Wait for the SEND to complete */ | |
261 | (void) librpma_fio_client_io_complete_all_sends(td); | |
262 | } | |
263 | ||
264 | /* deregister the messaging buffer memory */ | |
265 | if ((ret = rpma_mr_dereg(&cd->msg_mr))) | |
266 | librpma_td_verror(td, ret, "rpma_mr_dereg"); | |
267 | ||
268 | free(ccd->client_data); | |
269 | ||
270 | librpma_fio_client_cleanup(td); | |
271 | } | |
272 | ||
273 | static inline int client_io_flush(struct thread_data *td, | |
274 | struct io_u *first_io_u, struct io_u *last_io_u, | |
275 | unsigned long long int len) | |
276 | { | |
277 | struct librpma_fio_client_data *ccd = td->io_ops_data; | |
278 | struct client_data *cd = ccd->client_data; | |
279 | size_t io_u_buf_off = IO_U_NEXT_BUF_OFF_CLIENT(cd); | |
280 | size_t send_offset = io_u_buf_off + SEND_OFFSET; | |
281 | size_t recv_offset = io_u_buf_off + RECV_OFFSET; | |
282 | void *send_ptr = cd->io_us_msgs + send_offset; | |
283 | void *recv_ptr = cd->io_us_msgs + recv_offset; | |
284 | GPSPMFlushRequest flush_req = GPSPM_FLUSH_REQUEST__INIT; | |
285 | size_t flush_req_size = 0; | |
286 | int ret; | |
287 | ||
288 | /* prepare a response buffer */ | |
289 | if ((ret = rpma_recv(ccd->conn, cd->msg_mr, recv_offset, MAX_MSG_SIZE, | |
290 | recv_ptr))) { | |
291 | librpma_td_verror(td, ret, "rpma_recv"); | |
292 | return -1; | |
293 | } | |
294 | ||
295 | /* prepare a flush message and pack it to a send buffer */ | |
296 | flush_req.offset = first_io_u->offset; | |
297 | flush_req.length = len; | |
298 | flush_req.op_context = last_io_u->index; | |
299 | flush_req_size = gpspm_flush_request__get_packed_size(&flush_req); | |
300 | if (flush_req_size > MAX_MSG_SIZE) { | |
301 | log_err( | |
302 | "Packed flush request size is bigger than available send buffer space (%" | |
303 | PRIu64 " > %d\n", flush_req_size, MAX_MSG_SIZE); | |
304 | return -1; | |
305 | } | |
306 | (void) gpspm_flush_request__pack(&flush_req, send_ptr); | |
307 | ||
308 | /* send the flush message */ | |
309 | if ((ret = rpma_send(ccd->conn, cd->msg_mr, send_offset, flush_req_size, | |
310 | RPMA_F_COMPLETION_ALWAYS, NULL))) { | |
311 | librpma_td_verror(td, ret, "rpma_send"); | |
312 | return -1; | |
313 | } | |
314 | ||
315 | ++ccd->op_send_posted; | |
316 | ||
317 | return 0; | |
318 | } | |
319 | ||
320 | static int client_get_io_u_index(struct rpma_completion *cmpl, | |
321 | unsigned int *io_u_index) | |
322 | { | |
323 | GPSPMFlushResponse *flush_resp; | |
324 | ||
325 | if (cmpl->op != RPMA_OP_RECV) | |
326 | return 0; | |
327 | ||
328 | /* unpack a response from the received buffer */ | |
329 | flush_resp = gpspm_flush_response__unpack(NULL, | |
330 | cmpl->byte_len, cmpl->op_context); | |
331 | if (flush_resp == NULL) { | |
332 | log_err("Cannot unpack the flush response buffer\n"); | |
333 | return -1; | |
334 | } | |
335 | ||
336 | memcpy(io_u_index, &flush_resp->op_context, sizeof(*io_u_index)); | |
337 | ||
338 | gpspm_flush_response__free_unpacked(flush_resp, NULL); | |
339 | ||
340 | return 1; | |
341 | } | |
342 | ||
343 | FIO_STATIC struct ioengine_ops ioengine_client = { | |
344 | .name = "librpma_gpspm_client", | |
345 | .version = FIO_IOOPS_VERSION, | |
346 | .init = client_init, | |
347 | .post_init = client_post_init, | |
348 | .get_file_size = librpma_fio_client_get_file_size, | |
349 | .open_file = librpma_fio_file_nop, | |
350 | .queue = librpma_fio_client_queue, | |
351 | .commit = librpma_fio_client_commit, | |
352 | .getevents = librpma_fio_client_getevents, | |
353 | .event = librpma_fio_client_event, | |
354 | .errdetails = librpma_fio_client_errdetails, | |
355 | .close_file = librpma_fio_file_nop, | |
356 | .cleanup = client_cleanup, | |
357 | .flags = FIO_DISKLESSIO, | |
358 | .options = librpma_fio_options, | |
359 | .option_struct_size = sizeof(struct librpma_fio_options_values), | |
360 | }; | |
361 | ||
362 | /* server side implementation */ | |
363 | ||
364 | #define IO_U_BUFF_OFF_SERVER(i) (i * IO_U_BUF_LEN) | |
365 | ||
366 | struct server_data { | |
367 | /* aligned td->orig_buffer */ | |
368 | char *orig_buffer_aligned; | |
369 | ||
370 | /* resources for messaging buffer from DRAM allocated by fio */ | |
371 | struct rpma_mr_local *msg_mr; | |
372 | ||
373 | uint32_t msg_sqe_available; /* # of free SQ slots */ | |
374 | ||
375 | /* in-memory queues */ | |
376 | struct rpma_completion *msgs_queued; | |
377 | uint32_t msg_queued_nr; | |
378 | }; | |
379 | ||
380 | static int server_init(struct thread_data *td) | |
381 | { | |
382 | struct librpma_fio_server_data *csd; | |
383 | struct server_data *sd; | |
384 | int ret = -1; | |
385 | ||
386 | if ((ret = librpma_fio_server_init(td))) | |
387 | return ret; | |
388 | ||
389 | csd = td->io_ops_data; | |
390 | ||
391 | /* allocate server's data */ | |
392 | sd = calloc(1, sizeof(*sd)); | |
393 | if (sd == NULL) { | |
394 | td_verror(td, errno, "calloc"); | |
395 | goto err_server_cleanup; | |
396 | } | |
397 | ||
398 | /* allocate in-memory queue */ | |
399 | sd->msgs_queued = calloc(td->o.iodepth, sizeof(*sd->msgs_queued)); | |
400 | if (sd->msgs_queued == NULL) { | |
401 | td_verror(td, errno, "calloc"); | |
402 | goto err_free_sd; | |
403 | } | |
404 | ||
405 | /* | |
406 | * Assure a single io_u buffer can store both SEND and RECV messages and | |
407 | * an io_us buffer allocation is page-size-aligned which is required | |
408 | * to register for RDMA. User-provided values are intentionally ignored. | |
409 | */ | |
410 | td->o.max_bs[DDIR_READ] = IO_U_BUF_LEN; | |
411 | td->o.mem_align = page_size; | |
412 | ||
413 | csd->server_data = sd; | |
414 | ||
415 | return 0; | |
416 | ||
417 | err_free_sd: | |
418 | free(sd); | |
419 | ||
420 | err_server_cleanup: | |
421 | librpma_fio_server_cleanup(td); | |
422 | ||
423 | return -1; | |
424 | } | |
425 | ||
426 | static int server_post_init(struct thread_data *td) | |
427 | { | |
428 | struct librpma_fio_server_data *csd = td->io_ops_data; | |
429 | struct server_data *sd = csd->server_data; | |
430 | size_t io_us_size; | |
431 | size_t io_u_buflen; | |
432 | int ret; | |
433 | ||
434 | /* | |
435 | * td->orig_buffer is not aligned. The engine requires aligned io_us | |
436 | * so FIO alignes up the address using the formula below. | |
437 | */ | |
438 | sd->orig_buffer_aligned = PTR_ALIGN(td->orig_buffer, page_mask) + | |
439 | td->o.mem_align; | |
440 | ||
441 | /* | |
442 | * XXX | |
443 | * Each io_u message buffer contains recv and send messages. | |
444 | * Aligning each of those buffers may potentially give | |
445 | * some performance benefits. | |
446 | */ | |
447 | io_u_buflen = td_max_bs(td); | |
448 | ||
449 | /* check whether io_u buffer is big enough */ | |
450 | if (io_u_buflen < IO_U_BUF_LEN) { | |
451 | log_err( | |
452 | "blocksize too small to accommodate assumed maximal request/response pair size (%" PRIu64 " < %d)\n", | |
453 | io_u_buflen, IO_U_BUF_LEN); | |
454 | return -1; | |
455 | } | |
456 | ||
457 | /* | |
458 | * td->orig_buffer_size beside the space really consumed by io_us | |
459 | * has paddings which can be omitted for the memory registration. | |
460 | */ | |
461 | io_us_size = (unsigned long long)io_u_buflen * | |
462 | (unsigned long long)td->o.iodepth; | |
463 | ||
464 | if ((ret = rpma_mr_reg(csd->peer, sd->orig_buffer_aligned, io_us_size, | |
465 | RPMA_MR_USAGE_SEND | RPMA_MR_USAGE_RECV, | |
466 | &sd->msg_mr))) { | |
467 | librpma_td_verror(td, ret, "rpma_mr_reg"); | |
468 | return -1; | |
469 | } | |
470 | ||
471 | return 0; | |
472 | } | |
473 | ||
474 | static void server_cleanup(struct thread_data *td) | |
475 | { | |
476 | struct librpma_fio_server_data *csd = td->io_ops_data; | |
477 | struct server_data *sd; | |
478 | int ret; | |
479 | ||
480 | if (csd == NULL) | |
481 | return; | |
482 | ||
483 | sd = csd->server_data; | |
484 | ||
485 | if (sd != NULL) { | |
486 | /* rpma_mr_dereg(messaging buffer from DRAM) */ | |
487 | if ((ret = rpma_mr_dereg(&sd->msg_mr))) | |
488 | librpma_td_verror(td, ret, "rpma_mr_dereg"); | |
489 | ||
490 | free(sd->msgs_queued); | |
491 | free(sd); | |
492 | } | |
493 | ||
494 | librpma_fio_server_cleanup(td); | |
495 | } | |
496 | ||
497 | static int prepare_connection(struct thread_data *td, | |
498 | struct rpma_conn_req *conn_req) | |
499 | { | |
500 | struct librpma_fio_server_data *csd = td->io_ops_data; | |
501 | struct server_data *sd = csd->server_data; | |
502 | int ret; | |
503 | int i; | |
504 | ||
505 | /* prepare buffers for a flush requests */ | |
506 | sd->msg_sqe_available = td->o.iodepth; | |
507 | for (i = 0; i < td->o.iodepth; i++) { | |
508 | size_t offset_recv_msg = IO_U_BUFF_OFF_SERVER(i) + RECV_OFFSET; | |
509 | if ((ret = rpma_conn_req_recv(conn_req, sd->msg_mr, | |
510 | offset_recv_msg, MAX_MSG_SIZE, | |
511 | (const void *)(uintptr_t)i))) { | |
512 | librpma_td_verror(td, ret, "rpma_conn_req_recv"); | |
513 | return ret; | |
514 | } | |
515 | } | |
516 | ||
517 | return 0; | |
518 | } | |
519 | ||
520 | static int server_open_file(struct thread_data *td, struct fio_file *f) | |
521 | { | |
522 | struct librpma_fio_server_data *csd = td->io_ops_data; | |
523 | struct rpma_conn_cfg *cfg = NULL; | |
524 | uint16_t max_msg_num = td->o.iodepth; | |
525 | int ret; | |
526 | ||
527 | csd->prepare_connection = prepare_connection; | |
528 | ||
529 | /* create a connection configuration object */ | |
530 | if ((ret = rpma_conn_cfg_new(&cfg))) { | |
531 | librpma_td_verror(td, ret, "rpma_conn_cfg_new"); | |
532 | return -1; | |
533 | } | |
534 | ||
535 | /* | |
536 | * Calculate the required queue sizes where: | |
537 | * - the send queue (SQ) has to be big enough to accommodate | |
538 | * all possible flush requests (SENDs) | |
539 | * - the receive queue (RQ) has to be big enough to accommodate | |
540 | * all flush responses (RECVs) | |
541 | * - the completion queue (CQ) has to be big enough to accommodate | |
542 | * all success and error completions (sq_size + rq_size) | |
543 | */ | |
544 | if ((ret = rpma_conn_cfg_set_sq_size(cfg, max_msg_num))) { | |
545 | librpma_td_verror(td, ret, "rpma_conn_cfg_set_sq_size"); | |
546 | goto err_cfg_delete; | |
547 | } | |
548 | if ((ret = rpma_conn_cfg_set_rq_size(cfg, max_msg_num))) { | |
549 | librpma_td_verror(td, ret, "rpma_conn_cfg_set_rq_size"); | |
550 | goto err_cfg_delete; | |
551 | } | |
552 | if ((ret = rpma_conn_cfg_set_cq_size(cfg, max_msg_num * 2))) { | |
553 | librpma_td_verror(td, ret, "rpma_conn_cfg_set_cq_size"); | |
554 | goto err_cfg_delete; | |
555 | } | |
556 | ||
557 | ret = librpma_fio_server_open_file(td, f, cfg); | |
558 | ||
559 | err_cfg_delete: | |
560 | (void) rpma_conn_cfg_delete(&cfg); | |
561 | ||
562 | return ret; | |
563 | } | |
564 | ||
565 | static int server_qe_process(struct thread_data *td, | |
566 | struct rpma_completion *cmpl) | |
567 | { | |
568 | struct librpma_fio_server_data *csd = td->io_ops_data; | |
569 | struct server_data *sd = csd->server_data; | |
570 | GPSPMFlushRequest *flush_req; | |
571 | GPSPMFlushResponse flush_resp = GPSPM_FLUSH_RESPONSE__INIT; | |
572 | size_t flush_resp_size = 0; | |
573 | size_t send_buff_offset; | |
574 | size_t recv_buff_offset; | |
575 | size_t io_u_buff_offset; | |
576 | void *send_buff_ptr; | |
577 | void *recv_buff_ptr; | |
578 | void *op_ptr; | |
579 | int msg_index; | |
580 | int ret; | |
581 | ||
582 | /* calculate SEND/RECV pair parameters */ | |
583 | msg_index = (int)(uintptr_t)cmpl->op_context; | |
584 | io_u_buff_offset = IO_U_BUFF_OFF_SERVER(msg_index); | |
585 | send_buff_offset = io_u_buff_offset + SEND_OFFSET; | |
586 | recv_buff_offset = io_u_buff_offset + RECV_OFFSET; | |
587 | send_buff_ptr = sd->orig_buffer_aligned + send_buff_offset; | |
588 | recv_buff_ptr = sd->orig_buffer_aligned + recv_buff_offset; | |
589 | ||
590 | /* unpack a flush request from the received buffer */ | |
591 | flush_req = gpspm_flush_request__unpack(NULL, cmpl->byte_len, | |
592 | recv_buff_ptr); | |
593 | if (flush_req == NULL) { | |
594 | log_err("cannot unpack the flush request buffer\n"); | |
595 | goto err_terminate; | |
596 | } | |
597 | ||
598 | if (IS_NOT_THE_LAST_MESSAGE(flush_req)) { | |
599 | op_ptr = csd->ws_ptr + flush_req->offset; | |
600 | pmem_persist(op_ptr, flush_req->length); | |
601 | } else { | |
602 | /* | |
603 | * This is the last message - the client is done. | |
604 | */ | |
605 | gpspm_flush_request__free_unpacked(flush_req, NULL); | |
606 | td->done = true; | |
607 | return 0; | |
608 | } | |
609 | ||
610 | /* initiate the next receive operation */ | |
611 | if ((ret = rpma_recv(csd->conn, sd->msg_mr, recv_buff_offset, | |
612 | MAX_MSG_SIZE, | |
613 | (const void *)(uintptr_t)msg_index))) { | |
614 | librpma_td_verror(td, ret, "rpma_recv"); | |
615 | goto err_free_unpacked; | |
616 | } | |
617 | ||
618 | /* prepare a flush response and pack it to a send buffer */ | |
619 | flush_resp.op_context = flush_req->op_context; | |
620 | flush_resp_size = gpspm_flush_response__get_packed_size(&flush_resp); | |
621 | if (flush_resp_size > MAX_MSG_SIZE) { | |
622 | log_err( | |
623 | "Size of the packed flush response is bigger than the available space of the send buffer (%" | |
624 | PRIu64 " > %i\n", flush_resp_size, MAX_MSG_SIZE); | |
625 | goto err_free_unpacked; | |
626 | } | |
627 | ||
628 | (void) gpspm_flush_response__pack(&flush_resp, send_buff_ptr); | |
629 | ||
630 | /* send the flush response */ | |
631 | if ((ret = rpma_send(csd->conn, sd->msg_mr, send_buff_offset, | |
632 | flush_resp_size, RPMA_F_COMPLETION_ALWAYS, NULL))) { | |
633 | librpma_td_verror(td, ret, "rpma_send"); | |
634 | goto err_free_unpacked; | |
635 | } | |
636 | --sd->msg_sqe_available; | |
637 | ||
638 | gpspm_flush_request__free_unpacked(flush_req, NULL); | |
639 | ||
640 | return 0; | |
641 | ||
642 | err_free_unpacked: | |
643 | gpspm_flush_request__free_unpacked(flush_req, NULL); | |
644 | ||
645 | err_terminate: | |
646 | td->terminate = true; | |
647 | ||
648 | return -1; | |
649 | } | |
650 | ||
651 | static inline int server_queue_process(struct thread_data *td) | |
652 | { | |
653 | struct librpma_fio_server_data *csd = td->io_ops_data; | |
654 | struct server_data *sd = csd->server_data; | |
655 | int ret; | |
656 | int i; | |
657 | ||
658 | /* min(# of queue entries, # of SQ entries available) */ | |
659 | uint32_t qes_to_process = min(sd->msg_queued_nr, sd->msg_sqe_available); | |
660 | if (qes_to_process == 0) | |
661 | return 0; | |
662 | ||
663 | /* process queued completions */ | |
664 | for (i = 0; i < qes_to_process; ++i) { | |
665 | if ((ret = server_qe_process(td, &sd->msgs_queued[i]))) | |
666 | return ret; | |
667 | } | |
668 | ||
669 | /* progress the queue */ | |
670 | for (i = 0; i < sd->msg_queued_nr - qes_to_process; ++i) { | |
671 | memcpy(&sd->msgs_queued[i], | |
672 | &sd->msgs_queued[qes_to_process + i], | |
673 | sizeof(sd->msgs_queued[i])); | |
674 | } | |
675 | ||
676 | sd->msg_queued_nr -= qes_to_process; | |
677 | ||
678 | return 0; | |
679 | } | |
680 | ||
681 | static int server_cmpl_process(struct thread_data *td) | |
682 | { | |
683 | struct librpma_fio_server_data *csd = td->io_ops_data; | |
684 | struct server_data *sd = csd->server_data; | |
685 | struct rpma_completion *cmpl = &sd->msgs_queued[sd->msg_queued_nr]; | |
686 | int ret; | |
687 | ||
688 | ret = rpma_conn_completion_get(csd->conn, cmpl); | |
689 | if (ret == RPMA_E_NO_COMPLETION) { | |
690 | /* lack of completion is not an error */ | |
691 | return 0; | |
692 | } else if (ret != 0) { | |
693 | librpma_td_verror(td, ret, "rpma_conn_completion_get"); | |
694 | goto err_terminate; | |
695 | } | |
696 | ||
697 | /* validate the completion */ | |
698 | if (cmpl->op_status != IBV_WC_SUCCESS) | |
699 | goto err_terminate; | |
700 | ||
701 | if (cmpl->op == RPMA_OP_RECV) | |
702 | ++sd->msg_queued_nr; | |
703 | else if (cmpl->op == RPMA_OP_SEND) | |
704 | ++sd->msg_sqe_available; | |
705 | ||
706 | return 0; | |
707 | ||
708 | err_terminate: | |
709 | td->terminate = true; | |
710 | ||
711 | return -1; | |
712 | } | |
713 | ||
714 | static enum fio_q_status server_queue(struct thread_data *td, struct io_u *io_u) | |
715 | { | |
716 | do { | |
717 | if (server_cmpl_process(td)) | |
718 | return FIO_Q_BUSY; | |
719 | ||
720 | if (server_queue_process(td)) | |
721 | return FIO_Q_BUSY; | |
722 | ||
723 | } while (!td->done); | |
724 | ||
725 | return FIO_Q_COMPLETED; | |
726 | } | |
727 | ||
728 | FIO_STATIC struct ioengine_ops ioengine_server = { | |
729 | .name = "librpma_gpspm_server", | |
730 | .version = FIO_IOOPS_VERSION, | |
731 | .init = server_init, | |
732 | .post_init = server_post_init, | |
733 | .open_file = server_open_file, | |
734 | .close_file = librpma_fio_server_close_file, | |
735 | .queue = server_queue, | |
736 | .invalidate = librpma_fio_file_nop, | |
737 | .cleanup = server_cleanup, | |
738 | .flags = FIO_SYNCIO, | |
739 | .options = librpma_fio_options, | |
740 | .option_struct_size = sizeof(struct librpma_fio_options_values), | |
741 | }; | |
742 | ||
743 | /* register both engines */ | |
744 | ||
745 | static void fio_init fio_librpma_gpspm_register(void) | |
746 | { | |
747 | register_ioengine(&ioengine_client); | |
748 | register_ioengine(&ioengine_server); | |
749 | } | |
750 | ||
751 | static void fio_exit fio_librpma_gpspm_unregister(void) | |
752 | { | |
753 | unregister_ioengine(&ioengine_client); | |
754 | unregister_ioengine(&ioengine_server); | |
755 | } |