1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
|
/*
// Copyright (c) 2010-2017 Intel Corporation
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
*/
#include <string.h>
#include <stdio.h>
#include <rte_version.h>
#include "prox_port_cfg.h"
#include "prox_malloc.h"
#include "task_init.h"
#include "rx_pkt.h"
#include "tx_pkt.h"
#include "log.h"
#include "quit.h"
#include "lconf.h"
#include "thread_generic.h"
#include "prox_assert.h"
#if RTE_VERSION < RTE_VERSION_NUM(1,8,0,0)
#define RTE_CACHE_LINE_SIZE CACHE_LINE_SIZE
#endif
static unsigned first_task = 1;
LIST_HEAD(,task_init) head;
void reg_task(struct task_init* t)
{
// PROX_PANIC(t->handle == NULL, "No handle function specified for task with name %d\n", t->mode);
if (t->thread_x == NULL)
t->thread_x = thread_generic;
if (first_task) {
first_task = 0;
LIST_INIT(&head);
}
LIST_INSERT_HEAD(&head, t, entries);
}
struct task_init *to_task_init(const char *mode_str, const char *sub_mode_str)
{
struct task_init *cur_t;
LIST_FOREACH(cur_t, &head, entries) {
if (!strcmp(mode_str, cur_t->mode_str) &&
!strcmp(sub_mode_str, cur_t->sub_mode_str)) {
return cur_t;
}
}
return NULL;
}
static int compare_strcmp(const void *a, const void *b)
{
return strcmp(*(const char * const *)a, *(const char * const *)b);
}
int task_is_master(struct task_args *targ)
{
return (targ->lconf->id == prox_cfg.master);
}
void tasks_list(void)
{
struct task_init *cur_t;
char buf[sizeof(cur_t->mode_str) + sizeof(cur_t->sub_mode_str) + 4];
int nb_modes = 1; /* master */
LIST_FOREACH(cur_t, &head, entries) {
++nb_modes;
}
char **modes = calloc(nb_modes, sizeof(*modes));
char **cur_m = modes;
*cur_m++ = strdup("master");
LIST_FOREACH(cur_t, &head, entries) {
snprintf(buf, sizeof(buf), "%s%s%s",
cur_t->mode_str,
(cur_t->sub_mode_str[0] == 0) ? "" : " / ",
cur_t->sub_mode_str);
*cur_m++ = strdup(buf);
}
qsort(modes, nb_modes, sizeof(*modes), compare_strcmp);
plog_info("=== List of supported task modes / sub modes ===\n");
for (cur_m = modes; nb_modes; ++cur_m, --nb_modes) {
plog_info("\t%s\n", *cur_m);
free(*cur_m);
}
free(modes);
}
static size_t calc_memsize(struct task_args *targ, size_t task_size)
{
size_t memsize = task_size;
memsize += sizeof(struct task_base_aux);
if (targ->nb_rxports != 0) {
memsize += 2 * sizeof(uint8_t)*targ->nb_rxports;
}
if (targ->nb_rxrings != 0 || targ->tx_opt_ring_task) {
memsize += sizeof(struct rte_ring *)*targ->nb_rxrings;
}
if (targ->nb_txrings != 0) {
memsize += sizeof(struct rte_ring *) * targ->nb_txrings;
memsize = RTE_ALIGN_CEIL(memsize, RTE_CACHE_LINE_SIZE);
memsize += sizeof(struct ws_mbuf) + sizeof(((struct ws_mbuf*)0)->mbuf[0]) * targ->nb_txrings;
}
else if (targ->nb_txports != 0) {
memsize += sizeof(struct port_queue) * targ->nb_txports;
memsize = RTE_ALIGN_CEIL(memsize, RTE_CACHE_LINE_SIZE);
memsize += sizeof(struct ws_mbuf) + sizeof(((struct ws_mbuf*)0)->mbuf[0]) * targ->nb_txports;
}
else {
memsize = RTE_ALIGN_CEIL(memsize, RTE_CACHE_LINE_SIZE);
memsize += sizeof(struct ws_mbuf) + sizeof(((struct ws_mbuf*)0)->mbuf[0]);
}
return memsize;
}
static void *flush_function(struct task_args *targ)
{
if (targ->flags & TASK_ARG_DROP) {
return targ->nb_txrings ? flush_queues_sw : flush_queues_hw;
}
else {
return targ->nb_txrings ? flush_queues_no_drop_sw : flush_queues_no_drop_hw;
}
}
static size_t init_rx_tx_rings_ports(struct task_args *targ, struct task_base *tbase, size_t offset)
{
if (targ->tx_opt_ring_task) {
tbase->rx_pkt = rx_pkt_self;
}
else if (targ->nb_rxrings != 0) {
if (targ->nb_rxrings == 1) {
tbase->rx_pkt = rx_pkt_sw1;
tbase->rx_params_sw1.rx_ring = targ->rx_rings[0];
}
else {
tbase->rx_pkt = rx_pkt_sw;
tbase->rx_params_sw.nb_rxrings = targ->nb_rxrings;
tbase->rx_params_sw.rx_rings = (struct rte_ring **)(((uint8_t *)tbase) + offset);
offset += sizeof(struct rte_ring *)*tbase->rx_params_sw.nb_rxrings;
for (uint8_t i = 0; i < tbase->rx_params_sw.nb_rxrings; ++i) {
tbase->rx_params_sw.rx_rings[i] = targ->rx_rings[i];
}
if (rte_is_power_of_2(targ->nb_rxrings)) {
tbase->rx_pkt = rx_pkt_sw_pow2;
tbase->rx_params_sw.rxrings_mask = targ->nb_rxrings - 1;
}
}
}
else {
if (targ->nb_rxports == 1) {
if (targ->task_init->flag_features & TASK_FEATURE_L3)
tbase->rx_pkt = (targ->task_init->flag_features & TASK_FEATURE_MULTI_RX)? rx_pkt_hw1_multi_l3 : rx_pkt_hw1_l3;
else
tbase->rx_pkt = (targ->task_init->flag_features & TASK_FEATURE_MULTI_RX)? rx_pkt_hw1_multi : rx_pkt_hw1;
tbase->rx_params_hw1.rx_pq.port = targ->rx_port_queue[0].port;
tbase->rx_params_hw1.rx_pq.queue = targ->rx_port_queue[0].queue;
}
else {
PROX_ASSERT((targ->nb_rxports != 0) || (targ->task_init->flag_features & TASK_FEATURE_NO_RX));
if (targ->task_init->flag_features & TASK_FEATURE_L3)
tbase->rx_pkt = (targ->task_init->flag_features & TASK_FEATURE_MULTI_RX)? rx_pkt_hw_multi_l3 : rx_pkt_hw_l3;
else
tbase->rx_pkt = (targ->task_init->flag_features & TASK_FEATURE_MULTI_RX)? rx_pkt_hw_multi : rx_pkt_hw;
tbase->rx_params_hw.nb_rxports = targ->nb_rxports;
tbase->rx_params_hw.rx_pq = (struct port_queue *)(((uint8_t *)tbase) + offset);
offset += sizeof(struct port_queue) * tbase->rx_params_hw.nb_rxports;
for (int i = 0; i< targ->nb_rxports; i++) {
tbase->rx_params_hw.rx_pq[i].port = targ->rx_port_queue[i].port;
tbase->rx_params_hw.rx_pq[i].queue = targ->rx_port_queue[i].queue;
}
if (rte_is_power_of_2(targ->nb_rxports)) {
if (targ->task_init->flag_features & TASK_FEATURE_L3)
tbase->rx_pkt = (targ->task_init->flag_features & TASK_FEATURE_MULTI_RX)? rx_pkt_hw_pow2_multi_l3 : rx_pkt_hw_pow2_l3;
else
tbase->rx_pkt = (targ->task_init->flag_features & TASK_FEATURE_MULTI_RX)? rx_pkt_hw_pow2_multi : rx_pkt_hw_pow2;
tbase->rx_params_hw.rxport_mask = targ->nb_rxports - 1;
}
}
}
if ((targ->nb_txrings != 0) && (!targ->tx_opt_ring) && (!(targ->flags & TASK_ARG_DROP))) {
// Transmitting to a ring in NO DROP. We need to make sure the receiving task in not running on the same core.
// Otherwise we might end up in a dead lock: trying in a loop to transmit to a task which cannot receive anymore
// (as npt being scheduled).
struct core_task ct;
struct task_args *dtarg;
for (unsigned int j = 0; j < targ->nb_txrings; j++) {
ct = targ->core_task_set[0].core_task[j];
PROX_PANIC(ct.core == targ->lconf->id, "Core %d, task %d: NO_DROP task transmitting to another task (core %d, task %d) running on on same core => potential deadlock\n", targ->lconf->id, targ->id, ct.core, ct.task);
//plog_info("Core %d, task %d: NO_DROP task transmitting to another task (core %d, task %d) running on on same core => potential deadlock\n", targ->lconf->id, targ->id, ct.core, ct.task);
}
}
if ((targ->nb_txrings != 0) && (targ->nb_txports == 1)) {
/* Transmitting to multiple rings and one port */
plog_info("Initializing with 1 port %d queue %d nb_rings=%d\n", targ->tx_port_queue[0].port, targ->tx_port_queue[0].queue, targ->nb_txrings);
tbase->tx_params_hw_sw.tx_port_queue.port = targ->tx_port_queue[0].port;
tbase->tx_params_hw_sw.tx_port_queue.queue = targ->tx_port_queue[0].queue;
if (!targ->tx_opt_ring) {
tbase->tx_params_hw_sw.nb_txrings = targ->nb_txrings;
tbase->tx_params_hw_sw.tx_rings = (struct rte_ring **)(((uint8_t *)tbase) + offset);
offset += sizeof(struct rte_ring *)*tbase->tx_params_hw_sw.nb_txrings;
for (uint8_t i = 0; i < tbase->tx_params_hw_sw.nb_txrings; ++i) {
tbase->tx_params_hw_sw.tx_rings[i] = targ->tx_rings[i];
}
offset = RTE_ALIGN_CEIL(offset, RTE_CACHE_LINE_SIZE);
tbase->ws_mbuf = (struct ws_mbuf *)(((uint8_t *)tbase) + offset);
offset += sizeof(struct ws_mbuf) + sizeof(((struct ws_mbuf*)0)->mbuf[0]) * tbase->tx_params_hw_sw.nb_txrings;
}
}
else if (!targ->tx_opt_ring) {
if (targ->nb_txrings != 0) {
tbase->tx_params_sw.nb_txrings = targ->nb_txrings;
tbase->tx_params_sw.tx_rings = (struct rte_ring **)(((uint8_t *)tbase) + offset);
offset += sizeof(struct rte_ring *)*tbase->tx_params_sw.nb_txrings;
for (uint8_t i = 0; i < tbase->tx_params_sw.nb_txrings; ++i) {
tbase->tx_params_sw.tx_rings[i] = targ->tx_rings[i];
}
offset = RTE_ALIGN_CEIL(offset, RTE_CACHE_LINE_SIZE);
tbase->ws_mbuf = (struct ws_mbuf *)(((uint8_t *)tbase) + offset);
offset += sizeof(struct ws_mbuf) + sizeof(((struct ws_mbuf*)0)->mbuf[0]) * tbase->tx_params_sw.nb_txrings;
}
else if (targ->nb_txports != 0) {
tbase->tx_params_hw.nb_txports = targ->nb_txports;
tbase->tx_params_hw.tx_port_queue = (struct port_queue *)(((uint8_t *)tbase) + offset);
offset += sizeof(struct port_queue) * tbase->tx_params_hw.nb_txports;
for (uint8_t i = 0; i < tbase->tx_params_hw.nb_txports; ++i) {
tbase->tx_params_hw.tx_port_queue[i].port = targ->tx_port_queue[i].port;
tbase->tx_params_hw.tx_port_queue[i].queue = targ->tx_port_queue[i].queue;
}
offset = RTE_ALIGN_CEIL(offset, RTE_CACHE_LINE_SIZE);
tbase->ws_mbuf = (struct ws_mbuf *)(((uint8_t *)tbase) + offset);
offset += sizeof(struct ws_mbuf) + sizeof(((struct ws_mbuf*)0)->mbuf[0]) * tbase->tx_params_hw.nb_txports;
}
else {
offset = RTE_ALIGN_CEIL(offset, RTE_CACHE_LINE_SIZE);
tbase->ws_mbuf = (struct ws_mbuf *)(((uint8_t *)tbase) + offset);
offset += sizeof(struct ws_mbuf) + sizeof(((struct ws_mbuf*)0)->mbuf[0]);
}
struct ws_mbuf* w = tbase->ws_mbuf;
struct task_args *prev = targ->tx_opt_ring_task;
while (prev) {
prev->tbase->ws_mbuf = w;
prev = prev->tx_opt_ring_task;
}
}
if (targ->nb_txrings == 1 || targ->nb_txports == 1 || targ->tx_opt_ring) {
if (targ->task_init->flag_features & TASK_FEATURE_NEVER_DISCARDS) {
if (targ->tx_opt_ring) {
tbase->tx_pkt = tx_pkt_never_discard_self;
}
else if (targ->flags & TASK_ARG_DROP) {
if (targ->task_init->flag_features & TASK_FEATURE_THROUGHPUT_OPT)
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_never_discard_sw1 : tx_pkt_never_discard_hw1_thrpt_opt;
else
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_never_discard_sw1 : tx_pkt_never_discard_hw1_lat_opt;
}
else {
if (targ->task_init->flag_features & TASK_FEATURE_THROUGHPUT_OPT)
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_no_drop_never_discard_sw1 : tx_pkt_no_drop_never_discard_hw1_thrpt_opt;
else
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_no_drop_never_discard_sw1 : tx_pkt_no_drop_never_discard_hw1_lat_opt;
}
if ((targ->nb_txrings) || ((targ->task_init->flag_features & TASK_FEATURE_THROUGHPUT_OPT) == 0))
tbase->flags |= FLAG_NEVER_FLUSH;
else
targ->lconf->flush_queues[targ->task] = flush_function(targ);
}
else {
if (targ->tx_opt_ring) {
tbase->tx_pkt = tx_pkt_self;
}
else if (targ->flags & TASK_ARG_DROP) {
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_sw1 : tx_pkt_hw1;
}
else {
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_no_drop_sw1 : tx_pkt_no_drop_hw1;
}
tbase->flags |= FLAG_NEVER_FLUSH;
}
}
else {
if (targ->flags & TASK_ARG_DROP) {
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_sw : tx_pkt_hw;
}
else {
tbase->tx_pkt = targ->nb_txrings ? tx_pkt_no_drop_sw : tx_pkt_no_drop_hw;
}
targ->lconf->flush_queues[targ->task] = flush_function(targ);
}
if (targ->task_init->flag_features & TASK_FEATURE_NO_RX) {
tbase->rx_pkt = rx_pkt_dummy;
}
if (targ->nb_txrings == 0 && targ->nb_txports == 0) {
tbase->tx_pkt = tx_pkt_drop_all;
}
return offset;
}
struct task_base *init_task_struct(struct task_args *targ)
{
struct task_init* t = targ->task_init;
size_t offset = 0;
size_t memsize = calc_memsize(targ, t->size);
uint8_t task_socket = rte_lcore_to_socket_id(targ->lconf->id);
struct task_base *tbase = prox_zmalloc(memsize, task_socket);
PROX_PANIC(tbase == NULL, "Failed to allocate memory for task (%zu bytes)", memsize);
offset += t->size;
if (targ->nb_txrings == 0 && targ->nb_txports == 0)
tbase->flags |= FLAG_NEVER_FLUSH;
offset = init_rx_tx_rings_ports(targ, tbase, offset);
tbase->aux = (struct task_base_aux *)(((uint8_t *)tbase) + offset);
if ((targ->nb_txrings != 0) || (targ->nb_txports != 0)) {
if (targ->task_init->flag_features & TASK_FEATURE_L3) {
tbase->aux->tx_pkt_l2 = tbase->tx_pkt;
tbase->tx_pkt = tx_pkt_l3;
}
}
if (targ->task_init->flag_features & TASK_FEATURE_RX_ALL) {
task_base_add_rx_pkt_function(tbase, rx_pkt_all);
tbase->aux->all_mbufs = prox_zmalloc(MAX_RX_PKT_ALL * sizeof(* tbase->aux->all_mbufs), task_socket);
}
if (targ->task_init->flag_features & TASK_FEATURE_TSC_RX) {
task_base_add_rx_pkt_function(tbase, rx_pkt_tsc);
}
offset += sizeof(struct task_base_aux);
tbase->handle_bulk = t->handle;
if (targ->task_init->flag_features & TASK_FEATURE_L3) {
plog_info("\tTask configured in L3 mode\n");
tbase->l3.ctrl_plane_ring = targ->ctrl_plane_ring;
}
if ((targ->nb_txrings != 0) || (targ->nb_txports != 0)) {
if (targ->task_init->flag_features & TASK_FEATURE_L3)
task_init_l3(tbase, targ);
}
targ->tbase = tbase;
if (t->init) {
t->init(tbase, targ);
}
tbase->aux->start = t->start;
tbase->aux->stop = t->stop;
tbase->aux->start_first = t->start_first;
tbase->aux->stop_last = t->stop_last;
if ((targ->nb_txrings != 0) && (targ->nb_txports == 1)) {
tbase->aux->tx_pkt_hw = tx_pkt_no_drop_never_discard_hw1_no_pointer;
}
if (targ->tx_opt_ring) {
tbase->aux->tx_pkt_try = tx_try_self;
} else if (targ->nb_txrings == 1) {
tbase->aux->tx_pkt_try = tx_try_sw1;
} else if (targ->nb_txports) {
tbase->aux->tx_pkt_try = tx_try_hw1;
}
return tbase;
}
struct task_args *find_reachable_task_sending_to_port(struct task_args *from)
{
if (!from->nb_txrings) {
if (from->tx_port_queue[0].port != OUT_DISCARD)
return from;
else
return NULL;
}
struct core_task ct;
struct task_args *dtarg, *ret;
for (uint32_t i = 0; i < from->nb_txrings; ++i) {
ct = from->core_task_set[0].core_task[i];
dtarg = core_targ_get(ct.core, ct.task);
ret = find_reachable_task_sending_to_port(dtarg);
if (ret)
return ret;
}
return NULL;
}
struct prox_port_cfg *find_reachable_port(struct task_args *from)
{
struct task_args *dst = find_reachable_task_sending_to_port(from);
if (dst) {
int port_id = dst->tx_port_queue[0].port;
return &prox_port_cfg[port_id];
}
return NULL;
}
|