1 /*- 2 * SPDX-License-Identifier: BSD-2-Clause-FreeBSD 3 * 4 * Copyright (c) 2011 Chelsio Communications, Inc. 5 * All rights reserved. 6 * Written by: Navdeep Parhar <np@FreeBSD.org> 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 27 * SUCH DAMAGE. 28 */ 29 30 #include <sys/cdefs.h> 31 __FBSDID("$FreeBSD$"); 32 33 #include "opt_ddb.h" 34 #include "opt_inet.h" 35 #include "opt_inet6.h" 36 #include "opt_kern_tls.h" 37 #include "opt_ratelimit.h" 38 #include "opt_rss.h" 39 40 #include <sys/param.h> 41 #include <sys/conf.h> 42 #include <sys/priv.h> 43 #include <sys/kernel.h> 44 #include <sys/bus.h> 45 #include <sys/eventhandler.h> 46 #include <sys/module.h> 47 #include <sys/malloc.h> 48 #include <sys/queue.h> 49 #include <sys/taskqueue.h> 50 #include <sys/pciio.h> 51 #include <dev/pci/pcireg.h> 52 #include <dev/pci/pcivar.h> 53 #include <dev/pci/pci_private.h> 54 #include <sys/firmware.h> 55 #include <sys/sbuf.h> 56 #include <sys/smp.h> 57 #include <sys/socket.h> 58 #include <sys/sockio.h> 59 #include <sys/sysctl.h> 60 #include <net/ethernet.h> 61 #include <net/if.h> 62 #include <net/if_types.h> 63 #include <net/if_dl.h> 64 #include <net/if_vlan_var.h> 65 #ifdef RSS 66 #include <net/rss_config.h> 67 #endif 68 #include <netinet/in.h> 69 #include <netinet/ip.h> 70 #ifdef KERN_TLS 71 #include <netinet/tcp_seq.h> 72 #endif 73 #if defined(__i386__) || defined(__amd64__) 74 #include <machine/md_var.h> 75 #include <machine/cputypes.h> 76 #include <vm/vm.h> 77 #include <vm/pmap.h> 78 #endif 79 #ifdef DDB 80 #include <ddb/ddb.h> 81 #include <ddb/db_lex.h> 82 #endif 83 84 #include "common/common.h" 85 #include "common/t4_msg.h" 86 #include "common/t4_regs.h" 87 #include "common/t4_regs_values.h" 88 #include "cudbg/cudbg.h" 89 #include "t4_clip.h" 90 #include "t4_ioctl.h" 91 #include "t4_l2t.h" 92 #include "t4_mp_ring.h" 93 #include "t4_if.h" 94 #include "t4_smt.h" 95 96 /* T4 bus driver interface */ 97 static int t4_probe(device_t); 98 static int t4_attach(device_t); 99 static int t4_detach(device_t); 100 static int t4_child_location_str(device_t, device_t, char *, size_t); 101 static int t4_ready(device_t); 102 static int t4_read_port_device(device_t, int, device_t *); 103 static device_method_t t4_methods[] = { 104 DEVMETHOD(device_probe, t4_probe), 105 DEVMETHOD(device_attach, t4_attach), 106 DEVMETHOD(device_detach, t4_detach), 107 108 DEVMETHOD(bus_child_location_str, t4_child_location_str), 109 110 DEVMETHOD(t4_is_main_ready, t4_ready), 111 DEVMETHOD(t4_read_port_device, t4_read_port_device), 112 113 DEVMETHOD_END 114 }; 115 static driver_t t4_driver = { 116 "t4nex", 117 t4_methods, 118 sizeof(struct adapter) 119 }; 120 121 122 /* T4 port (cxgbe) interface */ 123 static int cxgbe_probe(device_t); 124 static int cxgbe_attach(device_t); 125 static int cxgbe_detach(device_t); 126 device_method_t cxgbe_methods[] = { 127 DEVMETHOD(device_probe, cxgbe_probe), 128 DEVMETHOD(device_attach, cxgbe_attach), 129 DEVMETHOD(device_detach, cxgbe_detach), 130 { 0, 0 } 131 }; 132 static driver_t cxgbe_driver = { 133 "cxgbe", 134 cxgbe_methods, 135 sizeof(struct port_info) 136 }; 137 138 /* T4 VI (vcxgbe) interface */ 139 static int vcxgbe_probe(device_t); 140 static int vcxgbe_attach(device_t); 141 static int vcxgbe_detach(device_t); 142 static device_method_t vcxgbe_methods[] = { 143 DEVMETHOD(device_probe, vcxgbe_probe), 144 DEVMETHOD(device_attach, vcxgbe_attach), 145 DEVMETHOD(device_detach, vcxgbe_detach), 146 { 0, 0 } 147 }; 148 static driver_t vcxgbe_driver = { 149 "vcxgbe", 150 vcxgbe_methods, 151 sizeof(struct vi_info) 152 }; 153 154 static d_ioctl_t t4_ioctl; 155 156 static struct cdevsw t4_cdevsw = { 157 .d_version = D_VERSION, 158 .d_ioctl = t4_ioctl, 159 .d_name = "t4nex", 160 }; 161 162 /* T5 bus driver interface */ 163 static int t5_probe(device_t); 164 static device_method_t t5_methods[] = { 165 DEVMETHOD(device_probe, t5_probe), 166 DEVMETHOD(device_attach, t4_attach), 167 DEVMETHOD(device_detach, t4_detach), 168 169 DEVMETHOD(bus_child_location_str, t4_child_location_str), 170 171 DEVMETHOD(t4_is_main_ready, t4_ready), 172 DEVMETHOD(t4_read_port_device, t4_read_port_device), 173 174 DEVMETHOD_END 175 }; 176 static driver_t t5_driver = { 177 "t5nex", 178 t5_methods, 179 sizeof(struct adapter) 180 }; 181 182 183 /* T5 port (cxl) interface */ 184 static driver_t cxl_driver = { 185 "cxl", 186 cxgbe_methods, 187 sizeof(struct port_info) 188 }; 189 190 /* T5 VI (vcxl) interface */ 191 static driver_t vcxl_driver = { 192 "vcxl", 193 vcxgbe_methods, 194 sizeof(struct vi_info) 195 }; 196 197 /* T6 bus driver interface */ 198 static int t6_probe(device_t); 199 static device_method_t t6_methods[] = { 200 DEVMETHOD(device_probe, t6_probe), 201 DEVMETHOD(device_attach, t4_attach), 202 DEVMETHOD(device_detach, t4_detach), 203 204 DEVMETHOD(bus_child_location_str, t4_child_location_str), 205 206 DEVMETHOD(t4_is_main_ready, t4_ready), 207 DEVMETHOD(t4_read_port_device, t4_read_port_device), 208 209 DEVMETHOD_END 210 }; 211 static driver_t t6_driver = { 212 "t6nex", 213 t6_methods, 214 sizeof(struct adapter) 215 }; 216 217 218 /* T6 port (cc) interface */ 219 static driver_t cc_driver = { 220 "cc", 221 cxgbe_methods, 222 sizeof(struct port_info) 223 }; 224 225 /* T6 VI (vcc) interface */ 226 static driver_t vcc_driver = { 227 "vcc", 228 vcxgbe_methods, 229 sizeof(struct vi_info) 230 }; 231 232 /* ifnet interface */ 233 static void cxgbe_init(void *); 234 static int cxgbe_ioctl(struct ifnet *, unsigned long, caddr_t); 235 static int cxgbe_transmit(struct ifnet *, struct mbuf *); 236 static void cxgbe_qflush(struct ifnet *); 237 #if defined(KERN_TLS) || defined(RATELIMIT) 238 static int cxgbe_snd_tag_alloc(struct ifnet *, union if_snd_tag_alloc_params *, 239 struct m_snd_tag **); 240 static int cxgbe_snd_tag_modify(struct m_snd_tag *, 241 union if_snd_tag_modify_params *); 242 static int cxgbe_snd_tag_query(struct m_snd_tag *, 243 union if_snd_tag_query_params *); 244 static void cxgbe_snd_tag_free(struct m_snd_tag *); 245 #endif 246 247 MALLOC_DEFINE(M_CXGBE, "cxgbe", "Chelsio T4/T5 Ethernet driver and services"); 248 249 /* 250 * Correct lock order when you need to acquire multiple locks is t4_list_lock, 251 * then ADAPTER_LOCK, then t4_uld_list_lock. 252 */ 253 static struct sx t4_list_lock; 254 SLIST_HEAD(, adapter) t4_list; 255 #ifdef TCP_OFFLOAD 256 static struct sx t4_uld_list_lock; 257 SLIST_HEAD(, uld_info) t4_uld_list; 258 #endif 259 260 /* 261 * Tunables. See tweak_tunables() too. 262 * 263 * Each tunable is set to a default value here if it's known at compile-time. 264 * Otherwise it is set to -n as an indication to tweak_tunables() that it should 265 * provide a reasonable default (upto n) when the driver is loaded. 266 * 267 * Tunables applicable to both T4 and T5 are under hw.cxgbe. Those specific to 268 * T5 are under hw.cxl. 269 */ 270 SYSCTL_NODE(_hw, OID_AUTO, cxgbe, CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 271 "cxgbe(4) parameters"); 272 SYSCTL_NODE(_hw, OID_AUTO, cxl, CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 273 "cxgbe(4) T5+ parameters"); 274 SYSCTL_NODE(_hw_cxgbe, OID_AUTO, toe, CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 275 "cxgbe(4) TOE parameters"); 276 277 /* 278 * Number of queues for tx and rx, NIC and offload. 279 */ 280 #define NTXQ 16 281 int t4_ntxq = -NTXQ; 282 SYSCTL_INT(_hw_cxgbe, OID_AUTO, ntxq, CTLFLAG_RDTUN, &t4_ntxq, 0, 283 "Number of TX queues per port"); 284 TUNABLE_INT("hw.cxgbe.ntxq10g", &t4_ntxq); /* Old name, undocumented */ 285 286 #define NRXQ 8 287 int t4_nrxq = -NRXQ; 288 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nrxq, CTLFLAG_RDTUN, &t4_nrxq, 0, 289 "Number of RX queues per port"); 290 TUNABLE_INT("hw.cxgbe.nrxq10g", &t4_nrxq); /* Old name, undocumented */ 291 292 #define NTXQ_VI 1 293 static int t4_ntxq_vi = -NTXQ_VI; 294 SYSCTL_INT(_hw_cxgbe, OID_AUTO, ntxq_vi, CTLFLAG_RDTUN, &t4_ntxq_vi, 0, 295 "Number of TX queues per VI"); 296 297 #define NRXQ_VI 1 298 static int t4_nrxq_vi = -NRXQ_VI; 299 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nrxq_vi, CTLFLAG_RDTUN, &t4_nrxq_vi, 0, 300 "Number of RX queues per VI"); 301 302 static int t4_rsrv_noflowq = 0; 303 SYSCTL_INT(_hw_cxgbe, OID_AUTO, rsrv_noflowq, CTLFLAG_RDTUN, &t4_rsrv_noflowq, 304 0, "Reserve TX queue 0 of each VI for non-flowid packets"); 305 306 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 307 #define NOFLDTXQ 8 308 static int t4_nofldtxq = -NOFLDTXQ; 309 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nofldtxq, CTLFLAG_RDTUN, &t4_nofldtxq, 0, 310 "Number of offload TX queues per port"); 311 312 #define NOFLDRXQ 2 313 static int t4_nofldrxq = -NOFLDRXQ; 314 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nofldrxq, CTLFLAG_RDTUN, &t4_nofldrxq, 0, 315 "Number of offload RX queues per port"); 316 317 #define NOFLDTXQ_VI 1 318 static int t4_nofldtxq_vi = -NOFLDTXQ_VI; 319 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nofldtxq_vi, CTLFLAG_RDTUN, &t4_nofldtxq_vi, 0, 320 "Number of offload TX queues per VI"); 321 322 #define NOFLDRXQ_VI 1 323 static int t4_nofldrxq_vi = -NOFLDRXQ_VI; 324 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nofldrxq_vi, CTLFLAG_RDTUN, &t4_nofldrxq_vi, 0, 325 "Number of offload RX queues per VI"); 326 327 #define TMR_IDX_OFLD 1 328 int t4_tmr_idx_ofld = TMR_IDX_OFLD; 329 SYSCTL_INT(_hw_cxgbe, OID_AUTO, holdoff_timer_idx_ofld, CTLFLAG_RDTUN, 330 &t4_tmr_idx_ofld, 0, "Holdoff timer index for offload queues"); 331 332 #define PKTC_IDX_OFLD (-1) 333 int t4_pktc_idx_ofld = PKTC_IDX_OFLD; 334 SYSCTL_INT(_hw_cxgbe, OID_AUTO, holdoff_pktc_idx_ofld, CTLFLAG_RDTUN, 335 &t4_pktc_idx_ofld, 0, "holdoff packet counter index for offload queues"); 336 337 /* 0 means chip/fw default, non-zero number is value in microseconds */ 338 static u_long t4_toe_keepalive_idle = 0; 339 SYSCTL_ULONG(_hw_cxgbe_toe, OID_AUTO, keepalive_idle, CTLFLAG_RDTUN, 340 &t4_toe_keepalive_idle, 0, "TOE keepalive idle timer (us)"); 341 342 /* 0 means chip/fw default, non-zero number is value in microseconds */ 343 static u_long t4_toe_keepalive_interval = 0; 344 SYSCTL_ULONG(_hw_cxgbe_toe, OID_AUTO, keepalive_interval, CTLFLAG_RDTUN, 345 &t4_toe_keepalive_interval, 0, "TOE keepalive interval timer (us)"); 346 347 /* 0 means chip/fw default, non-zero number is # of keepalives before abort */ 348 static int t4_toe_keepalive_count = 0; 349 SYSCTL_INT(_hw_cxgbe_toe, OID_AUTO, keepalive_count, CTLFLAG_RDTUN, 350 &t4_toe_keepalive_count, 0, "Number of TOE keepalive probes before abort"); 351 352 /* 0 means chip/fw default, non-zero number is value in microseconds */ 353 static u_long t4_toe_rexmt_min = 0; 354 SYSCTL_ULONG(_hw_cxgbe_toe, OID_AUTO, rexmt_min, CTLFLAG_RDTUN, 355 &t4_toe_rexmt_min, 0, "Minimum TOE retransmit interval (us)"); 356 357 /* 0 means chip/fw default, non-zero number is value in microseconds */ 358 static u_long t4_toe_rexmt_max = 0; 359 SYSCTL_ULONG(_hw_cxgbe_toe, OID_AUTO, rexmt_max, CTLFLAG_RDTUN, 360 &t4_toe_rexmt_max, 0, "Maximum TOE retransmit interval (us)"); 361 362 /* 0 means chip/fw default, non-zero number is # of rexmt before abort */ 363 static int t4_toe_rexmt_count = 0; 364 SYSCTL_INT(_hw_cxgbe_toe, OID_AUTO, rexmt_count, CTLFLAG_RDTUN, 365 &t4_toe_rexmt_count, 0, "Number of TOE retransmissions before abort"); 366 367 /* -1 means chip/fw default, other values are raw backoff values to use */ 368 static int t4_toe_rexmt_backoff[16] = { 369 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1 370 }; 371 SYSCTL_NODE(_hw_cxgbe_toe, OID_AUTO, rexmt_backoff, 372 CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 373 "cxgbe(4) TOE retransmit backoff values"); 374 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 0, CTLFLAG_RDTUN, 375 &t4_toe_rexmt_backoff[0], 0, ""); 376 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 1, CTLFLAG_RDTUN, 377 &t4_toe_rexmt_backoff[1], 0, ""); 378 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 2, CTLFLAG_RDTUN, 379 &t4_toe_rexmt_backoff[2], 0, ""); 380 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 3, CTLFLAG_RDTUN, 381 &t4_toe_rexmt_backoff[3], 0, ""); 382 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 4, CTLFLAG_RDTUN, 383 &t4_toe_rexmt_backoff[4], 0, ""); 384 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 5, CTLFLAG_RDTUN, 385 &t4_toe_rexmt_backoff[5], 0, ""); 386 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 6, CTLFLAG_RDTUN, 387 &t4_toe_rexmt_backoff[6], 0, ""); 388 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 7, CTLFLAG_RDTUN, 389 &t4_toe_rexmt_backoff[7], 0, ""); 390 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 8, CTLFLAG_RDTUN, 391 &t4_toe_rexmt_backoff[8], 0, ""); 392 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 9, CTLFLAG_RDTUN, 393 &t4_toe_rexmt_backoff[9], 0, ""); 394 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 10, CTLFLAG_RDTUN, 395 &t4_toe_rexmt_backoff[10], 0, ""); 396 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 11, CTLFLAG_RDTUN, 397 &t4_toe_rexmt_backoff[11], 0, ""); 398 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 12, CTLFLAG_RDTUN, 399 &t4_toe_rexmt_backoff[12], 0, ""); 400 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 13, CTLFLAG_RDTUN, 401 &t4_toe_rexmt_backoff[13], 0, ""); 402 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 14, CTLFLAG_RDTUN, 403 &t4_toe_rexmt_backoff[14], 0, ""); 404 SYSCTL_INT(_hw_cxgbe_toe_rexmt_backoff, OID_AUTO, 15, CTLFLAG_RDTUN, 405 &t4_toe_rexmt_backoff[15], 0, ""); 406 407 static int t4_toe_tls_rx_timeout = 5; 408 SYSCTL_INT(_hw_cxgbe_toe, OID_AUTO, tls_rx_timeout, CTLFLAG_RDTUN, 409 &t4_toe_tls_rx_timeout, 0, 410 "Timeout in seconds to downgrade TLS sockets to plain TOE"); 411 #endif 412 413 #ifdef DEV_NETMAP 414 #define NN_MAIN_VI (1 << 0) /* Native netmap on the main VI */ 415 #define NN_EXTRA_VI (1 << 1) /* Native netmap on the extra VI(s) */ 416 static int t4_native_netmap = NN_EXTRA_VI; 417 SYSCTL_INT(_hw_cxgbe, OID_AUTO, native_netmap, CTLFLAG_RDTUN, &t4_native_netmap, 418 0, "Native netmap support. bit 0 = main VI, bit 1 = extra VIs"); 419 420 #define NNMTXQ 8 421 static int t4_nnmtxq = -NNMTXQ; 422 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nnmtxq, CTLFLAG_RDTUN, &t4_nnmtxq, 0, 423 "Number of netmap TX queues"); 424 425 #define NNMRXQ 8 426 static int t4_nnmrxq = -NNMRXQ; 427 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nnmrxq, CTLFLAG_RDTUN, &t4_nnmrxq, 0, 428 "Number of netmap RX queues"); 429 430 #define NNMTXQ_VI 2 431 static int t4_nnmtxq_vi = -NNMTXQ_VI; 432 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nnmtxq_vi, CTLFLAG_RDTUN, &t4_nnmtxq_vi, 0, 433 "Number of netmap TX queues per VI"); 434 435 #define NNMRXQ_VI 2 436 static int t4_nnmrxq_vi = -NNMRXQ_VI; 437 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nnmrxq_vi, CTLFLAG_RDTUN, &t4_nnmrxq_vi, 0, 438 "Number of netmap RX queues per VI"); 439 #endif 440 441 /* 442 * Holdoff parameters for ports. 443 */ 444 #define TMR_IDX 1 445 int t4_tmr_idx = TMR_IDX; 446 SYSCTL_INT(_hw_cxgbe, OID_AUTO, holdoff_timer_idx, CTLFLAG_RDTUN, &t4_tmr_idx, 447 0, "Holdoff timer index"); 448 TUNABLE_INT("hw.cxgbe.holdoff_timer_idx_10G", &t4_tmr_idx); /* Old name */ 449 450 #define PKTC_IDX (-1) 451 int t4_pktc_idx = PKTC_IDX; 452 SYSCTL_INT(_hw_cxgbe, OID_AUTO, holdoff_pktc_idx, CTLFLAG_RDTUN, &t4_pktc_idx, 453 0, "Holdoff packet counter index"); 454 TUNABLE_INT("hw.cxgbe.holdoff_pktc_idx_10G", &t4_pktc_idx); /* Old name */ 455 456 /* 457 * Size (# of entries) of each tx and rx queue. 458 */ 459 unsigned int t4_qsize_txq = TX_EQ_QSIZE; 460 SYSCTL_INT(_hw_cxgbe, OID_AUTO, qsize_txq, CTLFLAG_RDTUN, &t4_qsize_txq, 0, 461 "Number of descriptors in each TX queue"); 462 463 unsigned int t4_qsize_rxq = RX_IQ_QSIZE; 464 SYSCTL_INT(_hw_cxgbe, OID_AUTO, qsize_rxq, CTLFLAG_RDTUN, &t4_qsize_rxq, 0, 465 "Number of descriptors in each RX queue"); 466 467 /* 468 * Interrupt types allowed (bits 0, 1, 2 = INTx, MSI, MSI-X respectively). 469 */ 470 int t4_intr_types = INTR_MSIX | INTR_MSI | INTR_INTX; 471 SYSCTL_INT(_hw_cxgbe, OID_AUTO, interrupt_types, CTLFLAG_RDTUN, &t4_intr_types, 472 0, "Interrupt types allowed (bit 0 = INTx, 1 = MSI, 2 = MSI-X)"); 473 474 /* 475 * Configuration file. All the _CF names here are special. 476 */ 477 #define DEFAULT_CF "default" 478 #define BUILTIN_CF "built-in" 479 #define FLASH_CF "flash" 480 #define UWIRE_CF "uwire" 481 #define FPGA_CF "fpga" 482 static char t4_cfg_file[32] = DEFAULT_CF; 483 SYSCTL_STRING(_hw_cxgbe, OID_AUTO, config_file, CTLFLAG_RDTUN, t4_cfg_file, 484 sizeof(t4_cfg_file), "Firmware configuration file"); 485 486 /* 487 * PAUSE settings (bit 0, 1, 2 = rx_pause, tx_pause, pause_autoneg respectively). 488 * rx_pause = 1 to heed incoming PAUSE frames, 0 to ignore them. 489 * tx_pause = 1 to emit PAUSE frames when the rx FIFO reaches its high water 490 * mark or when signalled to do so, 0 to never emit PAUSE. 491 * pause_autoneg = 1 means PAUSE will be negotiated if possible and the 492 * negotiated settings will override rx_pause/tx_pause. 493 * Otherwise rx_pause/tx_pause are applied forcibly. 494 */ 495 static int t4_pause_settings = PAUSE_RX | PAUSE_TX | PAUSE_AUTONEG; 496 SYSCTL_INT(_hw_cxgbe, OID_AUTO, pause_settings, CTLFLAG_RDTUN, 497 &t4_pause_settings, 0, 498 "PAUSE settings (bit 0 = rx_pause, 1 = tx_pause, 2 = pause_autoneg)"); 499 500 /* 501 * Forward Error Correction settings (bit 0, 1 = RS, BASER respectively). 502 * -1 to run with the firmware default. Same as FEC_AUTO (bit 5) 503 * 0 to disable FEC. 504 */ 505 static int t4_fec = -1; 506 SYSCTL_INT(_hw_cxgbe, OID_AUTO, fec, CTLFLAG_RDTUN, &t4_fec, 0, 507 "Forward Error Correction (bit 0 = RS, bit 1 = BASER_RS)"); 508 509 /* 510 * Link autonegotiation. 511 * -1 to run with the firmware default. 512 * 0 to disable. 513 * 1 to enable. 514 */ 515 static int t4_autoneg = -1; 516 SYSCTL_INT(_hw_cxgbe, OID_AUTO, autoneg, CTLFLAG_RDTUN, &t4_autoneg, 0, 517 "Link autonegotiation"); 518 519 /* 520 * Firmware auto-install by driver during attach (0, 1, 2 = prohibited, allowed, 521 * encouraged respectively). '-n' is the same as 'n' except the firmware 522 * version used in the checks is read from the firmware bundled with the driver. 523 */ 524 static int t4_fw_install = 1; 525 SYSCTL_INT(_hw_cxgbe, OID_AUTO, fw_install, CTLFLAG_RDTUN, &t4_fw_install, 0, 526 "Firmware auto-install (0 = prohibited, 1 = allowed, 2 = encouraged)"); 527 528 /* 529 * ASIC features that will be used. Disable the ones you don't want so that the 530 * chip resources aren't wasted on features that will not be used. 531 */ 532 static int t4_nbmcaps_allowed = 0; 533 SYSCTL_INT(_hw_cxgbe, OID_AUTO, nbmcaps_allowed, CTLFLAG_RDTUN, 534 &t4_nbmcaps_allowed, 0, "Default NBM capabilities"); 535 536 static int t4_linkcaps_allowed = 0; /* No DCBX, PPP, etc. by default */ 537 SYSCTL_INT(_hw_cxgbe, OID_AUTO, linkcaps_allowed, CTLFLAG_RDTUN, 538 &t4_linkcaps_allowed, 0, "Default link capabilities"); 539 540 static int t4_switchcaps_allowed = FW_CAPS_CONFIG_SWITCH_INGRESS | 541 FW_CAPS_CONFIG_SWITCH_EGRESS; 542 SYSCTL_INT(_hw_cxgbe, OID_AUTO, switchcaps_allowed, CTLFLAG_RDTUN, 543 &t4_switchcaps_allowed, 0, "Default switch capabilities"); 544 545 #ifdef RATELIMIT 546 static int t4_niccaps_allowed = FW_CAPS_CONFIG_NIC | 547 FW_CAPS_CONFIG_NIC_HASHFILTER | FW_CAPS_CONFIG_NIC_ETHOFLD; 548 #else 549 static int t4_niccaps_allowed = FW_CAPS_CONFIG_NIC | 550 FW_CAPS_CONFIG_NIC_HASHFILTER; 551 #endif 552 SYSCTL_INT(_hw_cxgbe, OID_AUTO, niccaps_allowed, CTLFLAG_RDTUN, 553 &t4_niccaps_allowed, 0, "Default NIC capabilities"); 554 555 static int t4_toecaps_allowed = -1; 556 SYSCTL_INT(_hw_cxgbe, OID_AUTO, toecaps_allowed, CTLFLAG_RDTUN, 557 &t4_toecaps_allowed, 0, "Default TCP offload capabilities"); 558 559 static int t4_rdmacaps_allowed = -1; 560 SYSCTL_INT(_hw_cxgbe, OID_AUTO, rdmacaps_allowed, CTLFLAG_RDTUN, 561 &t4_rdmacaps_allowed, 0, "Default RDMA capabilities"); 562 563 static int t4_cryptocaps_allowed = -1; 564 SYSCTL_INT(_hw_cxgbe, OID_AUTO, cryptocaps_allowed, CTLFLAG_RDTUN, 565 &t4_cryptocaps_allowed, 0, "Default crypto capabilities"); 566 567 static int t4_iscsicaps_allowed = -1; 568 SYSCTL_INT(_hw_cxgbe, OID_AUTO, iscsicaps_allowed, CTLFLAG_RDTUN, 569 &t4_iscsicaps_allowed, 0, "Default iSCSI capabilities"); 570 571 static int t4_fcoecaps_allowed = 0; 572 SYSCTL_INT(_hw_cxgbe, OID_AUTO, fcoecaps_allowed, CTLFLAG_RDTUN, 573 &t4_fcoecaps_allowed, 0, "Default FCoE capabilities"); 574 575 static int t5_write_combine = 0; 576 SYSCTL_INT(_hw_cxl, OID_AUTO, write_combine, CTLFLAG_RDTUN, &t5_write_combine, 577 0, "Use WC instead of UC for BAR2"); 578 579 static int t4_num_vis = 1; 580 SYSCTL_INT(_hw_cxgbe, OID_AUTO, num_vis, CTLFLAG_RDTUN, &t4_num_vis, 0, 581 "Number of VIs per port"); 582 583 /* 584 * PCIe Relaxed Ordering. 585 * -1: driver should figure out a good value. 586 * 0: disable RO. 587 * 1: enable RO. 588 * 2: leave RO alone. 589 */ 590 static int pcie_relaxed_ordering = -1; 591 SYSCTL_INT(_hw_cxgbe, OID_AUTO, pcie_relaxed_ordering, CTLFLAG_RDTUN, 592 &pcie_relaxed_ordering, 0, 593 "PCIe Relaxed Ordering: 0 = disable, 1 = enable, 2 = leave alone"); 594 595 static int t4_panic_on_fatal_err = 0; 596 SYSCTL_INT(_hw_cxgbe, OID_AUTO, panic_on_fatal_err, CTLFLAG_RDTUN, 597 &t4_panic_on_fatal_err, 0, "panic on fatal errors"); 598 599 static int t4_tx_vm_wr = 0; 600 SYSCTL_INT(_hw_cxgbe, OID_AUTO, tx_vm_wr, CTLFLAG_RWTUN, &t4_tx_vm_wr, 0, 601 "Use VM work requests to transmit packets."); 602 603 /* 604 * Set to non-zero to enable the attack filter. A packet that matches any of 605 * these conditions will get dropped on ingress: 606 * 1) IP && source address == destination address. 607 * 2) TCP/IP && source address is not a unicast address. 608 * 3) TCP/IP && destination address is not a unicast address. 609 * 4) IP && source address is loopback (127.x.y.z). 610 * 5) IP && destination address is loopback (127.x.y.z). 611 * 6) IPv6 && source address == destination address. 612 * 7) IPv6 && source address is not a unicast address. 613 * 8) IPv6 && source address is loopback (::1/128). 614 * 9) IPv6 && destination address is loopback (::1/128). 615 * 10) IPv6 && source address is unspecified (::/128). 616 * 11) IPv6 && destination address is unspecified (::/128). 617 * 12) TCP/IPv6 && source address is multicast (ff00::/8). 618 * 13) TCP/IPv6 && destination address is multicast (ff00::/8). 619 */ 620 static int t4_attack_filter = 0; 621 SYSCTL_INT(_hw_cxgbe, OID_AUTO, attack_filter, CTLFLAG_RDTUN, 622 &t4_attack_filter, 0, "Drop suspicious traffic"); 623 624 static int t4_drop_ip_fragments = 0; 625 SYSCTL_INT(_hw_cxgbe, OID_AUTO, drop_ip_fragments, CTLFLAG_RDTUN, 626 &t4_drop_ip_fragments, 0, "Drop IP fragments"); 627 628 static int t4_drop_pkts_with_l2_errors = 1; 629 SYSCTL_INT(_hw_cxgbe, OID_AUTO, drop_pkts_with_l2_errors, CTLFLAG_RDTUN, 630 &t4_drop_pkts_with_l2_errors, 0, 631 "Drop all frames with Layer 2 length or checksum errors"); 632 633 static int t4_drop_pkts_with_l3_errors = 0; 634 SYSCTL_INT(_hw_cxgbe, OID_AUTO, drop_pkts_with_l3_errors, CTLFLAG_RDTUN, 635 &t4_drop_pkts_with_l3_errors, 0, 636 "Drop all frames with IP version, length, or checksum errors"); 637 638 static int t4_drop_pkts_with_l4_errors = 0; 639 SYSCTL_INT(_hw_cxgbe, OID_AUTO, drop_pkts_with_l4_errors, CTLFLAG_RDTUN, 640 &t4_drop_pkts_with_l4_errors, 0, 641 "Drop all frames with Layer 4 length, checksum, or other errors"); 642 643 #ifdef TCP_OFFLOAD 644 /* 645 * TOE tunables. 646 */ 647 static int t4_cop_managed_offloading = 0; 648 SYSCTL_INT(_hw_cxgbe, OID_AUTO, cop_managed_offloading, CTLFLAG_RDTUN, 649 &t4_cop_managed_offloading, 0, 650 "COP (Connection Offload Policy) controls all TOE offload"); 651 #endif 652 653 #ifdef KERN_TLS 654 /* 655 * This enables KERN_TLS for all adapters if set. 656 */ 657 static int t4_kern_tls = 0; 658 SYSCTL_INT(_hw_cxgbe, OID_AUTO, kern_tls, CTLFLAG_RDTUN, &t4_kern_tls, 0, 659 "Enable KERN_TLS mode for all supported adapters"); 660 661 SYSCTL_NODE(_hw_cxgbe, OID_AUTO, tls, CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 662 "cxgbe(4) KERN_TLS parameters"); 663 664 static int t4_tls_inline_keys = 0; 665 SYSCTL_INT(_hw_cxgbe_tls, OID_AUTO, inline_keys, CTLFLAG_RDTUN, 666 &t4_tls_inline_keys, 0, 667 "Always pass TLS keys in work requests (1) or attempt to store TLS keys " 668 "in card memory."); 669 670 static int t4_tls_combo_wrs = 0; 671 SYSCTL_INT(_hw_cxgbe_tls, OID_AUTO, combo_wrs, CTLFLAG_RDTUN, &t4_tls_combo_wrs, 672 0, "Attempt to combine TCB field updates with TLS record work requests."); 673 #endif 674 675 /* Functions used by VIs to obtain unique MAC addresses for each VI. */ 676 static int vi_mac_funcs[] = { 677 FW_VI_FUNC_ETH, 678 FW_VI_FUNC_OFLD, 679 FW_VI_FUNC_IWARP, 680 FW_VI_FUNC_OPENISCSI, 681 FW_VI_FUNC_OPENFCOE, 682 FW_VI_FUNC_FOISCSI, 683 FW_VI_FUNC_FOFCOE, 684 }; 685 686 struct intrs_and_queues { 687 uint16_t intr_type; /* INTx, MSI, or MSI-X */ 688 uint16_t num_vis; /* number of VIs for each port */ 689 uint16_t nirq; /* Total # of vectors */ 690 uint16_t ntxq; /* # of NIC txq's for each port */ 691 uint16_t nrxq; /* # of NIC rxq's for each port */ 692 uint16_t nofldtxq; /* # of TOE/ETHOFLD txq's for each port */ 693 uint16_t nofldrxq; /* # of TOE rxq's for each port */ 694 uint16_t nnmtxq; /* # of netmap txq's */ 695 uint16_t nnmrxq; /* # of netmap rxq's */ 696 697 /* The vcxgbe/vcxl interfaces use these and not the ones above. */ 698 uint16_t ntxq_vi; /* # of NIC txq's */ 699 uint16_t nrxq_vi; /* # of NIC rxq's */ 700 uint16_t nofldtxq_vi; /* # of TOE txq's */ 701 uint16_t nofldrxq_vi; /* # of TOE rxq's */ 702 uint16_t nnmtxq_vi; /* # of netmap txq's */ 703 uint16_t nnmrxq_vi; /* # of netmap rxq's */ 704 }; 705 706 static void setup_memwin(struct adapter *); 707 static void position_memwin(struct adapter *, int, uint32_t); 708 static int validate_mem_range(struct adapter *, uint32_t, uint32_t); 709 static int fwmtype_to_hwmtype(int); 710 static int validate_mt_off_len(struct adapter *, int, uint32_t, uint32_t, 711 uint32_t *); 712 static int fixup_devlog_params(struct adapter *); 713 static int cfg_itype_and_nqueues(struct adapter *, struct intrs_and_queues *); 714 static int contact_firmware(struct adapter *); 715 static int partition_resources(struct adapter *); 716 static int get_params__pre_init(struct adapter *); 717 static int set_params__pre_init(struct adapter *); 718 static int get_params__post_init(struct adapter *); 719 static int set_params__post_init(struct adapter *); 720 static void t4_set_desc(struct adapter *); 721 static bool fixed_ifmedia(struct port_info *); 722 static void build_medialist(struct port_info *); 723 static void init_link_config(struct port_info *); 724 static int fixup_link_config(struct port_info *); 725 static int apply_link_config(struct port_info *); 726 static int cxgbe_init_synchronized(struct vi_info *); 727 static int cxgbe_uninit_synchronized(struct vi_info *); 728 static void quiesce_txq(struct adapter *, struct sge_txq *); 729 static void quiesce_wrq(struct adapter *, struct sge_wrq *); 730 static void quiesce_iq(struct adapter *, struct sge_iq *); 731 static void quiesce_fl(struct adapter *, struct sge_fl *); 732 static int t4_alloc_irq(struct adapter *, struct irq *, int rid, 733 driver_intr_t *, void *, char *); 734 static int t4_free_irq(struct adapter *, struct irq *); 735 static void t4_init_atid_table(struct adapter *); 736 static void t4_free_atid_table(struct adapter *); 737 static void get_regs(struct adapter *, struct t4_regdump *, uint8_t *); 738 static void vi_refresh_stats(struct adapter *, struct vi_info *); 739 static void cxgbe_refresh_stats(struct adapter *, struct port_info *); 740 static void cxgbe_tick(void *); 741 static void cxgbe_sysctls(struct port_info *); 742 static int sysctl_int_array(SYSCTL_HANDLER_ARGS); 743 static int sysctl_bitfield_8b(SYSCTL_HANDLER_ARGS); 744 static int sysctl_bitfield_16b(SYSCTL_HANDLER_ARGS); 745 static int sysctl_btphy(SYSCTL_HANDLER_ARGS); 746 static int sysctl_noflowq(SYSCTL_HANDLER_ARGS); 747 static int sysctl_tx_vm_wr(SYSCTL_HANDLER_ARGS); 748 static int sysctl_holdoff_tmr_idx(SYSCTL_HANDLER_ARGS); 749 static int sysctl_holdoff_pktc_idx(SYSCTL_HANDLER_ARGS); 750 static int sysctl_qsize_rxq(SYSCTL_HANDLER_ARGS); 751 static int sysctl_qsize_txq(SYSCTL_HANDLER_ARGS); 752 static int sysctl_pause_settings(SYSCTL_HANDLER_ARGS); 753 static int sysctl_fec(SYSCTL_HANDLER_ARGS); 754 static int sysctl_module_fec(SYSCTL_HANDLER_ARGS); 755 static int sysctl_autoneg(SYSCTL_HANDLER_ARGS); 756 static int sysctl_handle_t4_reg64(SYSCTL_HANDLER_ARGS); 757 static int sysctl_temperature(SYSCTL_HANDLER_ARGS); 758 static int sysctl_vdd(SYSCTL_HANDLER_ARGS); 759 static int sysctl_reset_sensor(SYSCTL_HANDLER_ARGS); 760 static int sysctl_loadavg(SYSCTL_HANDLER_ARGS); 761 static int sysctl_cctrl(SYSCTL_HANDLER_ARGS); 762 static int sysctl_cim_ibq_obq(SYSCTL_HANDLER_ARGS); 763 static int sysctl_cim_la(SYSCTL_HANDLER_ARGS); 764 static int sysctl_cim_ma_la(SYSCTL_HANDLER_ARGS); 765 static int sysctl_cim_pif_la(SYSCTL_HANDLER_ARGS); 766 static int sysctl_cim_qcfg(SYSCTL_HANDLER_ARGS); 767 static int sysctl_cpl_stats(SYSCTL_HANDLER_ARGS); 768 static int sysctl_ddp_stats(SYSCTL_HANDLER_ARGS); 769 static int sysctl_tid_stats(SYSCTL_HANDLER_ARGS); 770 static int sysctl_devlog(SYSCTL_HANDLER_ARGS); 771 static int sysctl_fcoe_stats(SYSCTL_HANDLER_ARGS); 772 static int sysctl_hw_sched(SYSCTL_HANDLER_ARGS); 773 static int sysctl_lb_stats(SYSCTL_HANDLER_ARGS); 774 static int sysctl_linkdnrc(SYSCTL_HANDLER_ARGS); 775 static int sysctl_meminfo(SYSCTL_HANDLER_ARGS); 776 static int sysctl_mps_tcam(SYSCTL_HANDLER_ARGS); 777 static int sysctl_mps_tcam_t6(SYSCTL_HANDLER_ARGS); 778 static int sysctl_path_mtus(SYSCTL_HANDLER_ARGS); 779 static int sysctl_pm_stats(SYSCTL_HANDLER_ARGS); 780 static int sysctl_rdma_stats(SYSCTL_HANDLER_ARGS); 781 static int sysctl_tcp_stats(SYSCTL_HANDLER_ARGS); 782 static int sysctl_tids(SYSCTL_HANDLER_ARGS); 783 static int sysctl_tp_err_stats(SYSCTL_HANDLER_ARGS); 784 static int sysctl_tnl_stats(SYSCTL_HANDLER_ARGS); 785 static int sysctl_tp_la_mask(SYSCTL_HANDLER_ARGS); 786 static int sysctl_tp_la(SYSCTL_HANDLER_ARGS); 787 static int sysctl_tx_rate(SYSCTL_HANDLER_ARGS); 788 static int sysctl_ulprx_la(SYSCTL_HANDLER_ARGS); 789 static int sysctl_wcwr_stats(SYSCTL_HANDLER_ARGS); 790 static int sysctl_cpus(SYSCTL_HANDLER_ARGS); 791 #ifdef TCP_OFFLOAD 792 static int sysctl_tls(SYSCTL_HANDLER_ARGS); 793 static int sysctl_tls_rx_ports(SYSCTL_HANDLER_ARGS); 794 static int sysctl_tls_rx_timeout(SYSCTL_HANDLER_ARGS); 795 static int sysctl_tp_tick(SYSCTL_HANDLER_ARGS); 796 static int sysctl_tp_dack_timer(SYSCTL_HANDLER_ARGS); 797 static int sysctl_tp_timer(SYSCTL_HANDLER_ARGS); 798 static int sysctl_tp_shift_cnt(SYSCTL_HANDLER_ARGS); 799 static int sysctl_tp_backoff(SYSCTL_HANDLER_ARGS); 800 static int sysctl_holdoff_tmr_idx_ofld(SYSCTL_HANDLER_ARGS); 801 static int sysctl_holdoff_pktc_idx_ofld(SYSCTL_HANDLER_ARGS); 802 #endif 803 static int get_sge_context(struct adapter *, struct t4_sge_context *); 804 static int load_fw(struct adapter *, struct t4_data *); 805 static int load_cfg(struct adapter *, struct t4_data *); 806 static int load_boot(struct adapter *, struct t4_bootrom *); 807 static int load_bootcfg(struct adapter *, struct t4_data *); 808 static int cudbg_dump(struct adapter *, struct t4_cudbg_dump *); 809 static void free_offload_policy(struct t4_offload_policy *); 810 static int set_offload_policy(struct adapter *, struct t4_offload_policy *); 811 static int read_card_mem(struct adapter *, int, struct t4_mem_range *); 812 static int read_i2c(struct adapter *, struct t4_i2c_data *); 813 static int clear_stats(struct adapter *, u_int); 814 #ifdef TCP_OFFLOAD 815 static int toe_capability(struct vi_info *, int); 816 static void t4_async_event(void *, int); 817 #endif 818 static int mod_event(module_t, int, void *); 819 static int notify_siblings(device_t, int); 820 821 struct { 822 uint16_t device; 823 char *desc; 824 } t4_pciids[] = { 825 {0xa000, "Chelsio Terminator 4 FPGA"}, 826 {0x4400, "Chelsio T440-dbg"}, 827 {0x4401, "Chelsio T420-CR"}, 828 {0x4402, "Chelsio T422-CR"}, 829 {0x4403, "Chelsio T440-CR"}, 830 {0x4404, "Chelsio T420-BCH"}, 831 {0x4405, "Chelsio T440-BCH"}, 832 {0x4406, "Chelsio T440-CH"}, 833 {0x4407, "Chelsio T420-SO"}, 834 {0x4408, "Chelsio T420-CX"}, 835 {0x4409, "Chelsio T420-BT"}, 836 {0x440a, "Chelsio T404-BT"}, 837 {0x440e, "Chelsio T440-LP-CR"}, 838 }, t5_pciids[] = { 839 {0xb000, "Chelsio Terminator 5 FPGA"}, 840 {0x5400, "Chelsio T580-dbg"}, 841 {0x5401, "Chelsio T520-CR"}, /* 2 x 10G */ 842 {0x5402, "Chelsio T522-CR"}, /* 2 x 10G, 2 X 1G */ 843 {0x5403, "Chelsio T540-CR"}, /* 4 x 10G */ 844 {0x5407, "Chelsio T520-SO"}, /* 2 x 10G, nomem */ 845 {0x5409, "Chelsio T520-BT"}, /* 2 x 10GBaseT */ 846 {0x540a, "Chelsio T504-BT"}, /* 4 x 1G */ 847 {0x540d, "Chelsio T580-CR"}, /* 2 x 40G */ 848 {0x540e, "Chelsio T540-LP-CR"}, /* 4 x 10G */ 849 {0x5410, "Chelsio T580-LP-CR"}, /* 2 x 40G */ 850 {0x5411, "Chelsio T520-LL-CR"}, /* 2 x 10G */ 851 {0x5412, "Chelsio T560-CR"}, /* 1 x 40G, 2 x 10G */ 852 {0x5414, "Chelsio T580-LP-SO-CR"}, /* 2 x 40G, nomem */ 853 {0x5415, "Chelsio T502-BT"}, /* 2 x 1G */ 854 {0x5418, "Chelsio T540-BT"}, /* 4 x 10GBaseT */ 855 {0x5419, "Chelsio T540-LP-BT"}, /* 4 x 10GBaseT */ 856 {0x541a, "Chelsio T540-SO-BT"}, /* 4 x 10GBaseT, nomem */ 857 {0x541b, "Chelsio T540-SO-CR"}, /* 4 x 10G, nomem */ 858 859 /* Custom */ 860 {0x5483, "Custom T540-CR"}, 861 {0x5484, "Custom T540-BT"}, 862 }, t6_pciids[] = { 863 {0xc006, "Chelsio Terminator 6 FPGA"}, /* T6 PE10K6 FPGA (PF0) */ 864 {0x6400, "Chelsio T6-DBG-25"}, /* 2 x 10/25G, debug */ 865 {0x6401, "Chelsio T6225-CR"}, /* 2 x 10/25G */ 866 {0x6402, "Chelsio T6225-SO-CR"}, /* 2 x 10/25G, nomem */ 867 {0x6403, "Chelsio T6425-CR"}, /* 4 x 10/25G */ 868 {0x6404, "Chelsio T6425-SO-CR"}, /* 4 x 10/25G, nomem */ 869 {0x6405, "Chelsio T6225-OCP-SO"}, /* 2 x 10/25G, nomem */ 870 {0x6406, "Chelsio T62100-OCP-SO"}, /* 2 x 40/50/100G, nomem */ 871 {0x6407, "Chelsio T62100-LP-CR"}, /* 2 x 40/50/100G */ 872 {0x6408, "Chelsio T62100-SO-CR"}, /* 2 x 40/50/100G, nomem */ 873 {0x6409, "Chelsio T6210-BT"}, /* 2 x 10GBASE-T */ 874 {0x640d, "Chelsio T62100-CR"}, /* 2 x 40/50/100G */ 875 {0x6410, "Chelsio T6-DBG-100"}, /* 2 x 40/50/100G, debug */ 876 {0x6411, "Chelsio T6225-LL-CR"}, /* 2 x 10/25G */ 877 {0x6414, "Chelsio T61100-OCP-SO"}, /* 1 x 40/50/100G, nomem */ 878 {0x6415, "Chelsio T6201-BT"}, /* 2 x 1000BASE-T */ 879 880 /* Custom */ 881 {0x6480, "Custom T6225-CR"}, 882 {0x6481, "Custom T62100-CR"}, 883 {0x6482, "Custom T6225-CR"}, 884 {0x6483, "Custom T62100-CR"}, 885 {0x6484, "Custom T64100-CR"}, 886 {0x6485, "Custom T6240-SO"}, 887 {0x6486, "Custom T6225-SO-CR"}, 888 {0x6487, "Custom T6225-CR"}, 889 }; 890 891 #ifdef TCP_OFFLOAD 892 /* 893 * service_iq_fl() has an iq and needs the fl. Offset of fl from the iq should 894 * be exactly the same for both rxq and ofld_rxq. 895 */ 896 CTASSERT(offsetof(struct sge_ofld_rxq, iq) == offsetof(struct sge_rxq, iq)); 897 CTASSERT(offsetof(struct sge_ofld_rxq, fl) == offsetof(struct sge_rxq, fl)); 898 #endif 899 CTASSERT(sizeof(struct cluster_metadata) <= CL_METADATA_SIZE); 900 901 static int 902 t4_probe(device_t dev) 903 { 904 int i; 905 uint16_t v = pci_get_vendor(dev); 906 uint16_t d = pci_get_device(dev); 907 uint8_t f = pci_get_function(dev); 908 909 if (v != PCI_VENDOR_ID_CHELSIO) 910 return (ENXIO); 911 912 /* Attach only to PF0 of the FPGA */ 913 if (d == 0xa000 && f != 0) 914 return (ENXIO); 915 916 for (i = 0; i < nitems(t4_pciids); i++) { 917 if (d == t4_pciids[i].device) { 918 device_set_desc(dev, t4_pciids[i].desc); 919 return (BUS_PROBE_DEFAULT); 920 } 921 } 922 923 return (ENXIO); 924 } 925 926 static int 927 t5_probe(device_t dev) 928 { 929 int i; 930 uint16_t v = pci_get_vendor(dev); 931 uint16_t d = pci_get_device(dev); 932 uint8_t f = pci_get_function(dev); 933 934 if (v != PCI_VENDOR_ID_CHELSIO) 935 return (ENXIO); 936 937 /* Attach only to PF0 of the FPGA */ 938 if (d == 0xb000 && f != 0) 939 return (ENXIO); 940 941 for (i = 0; i < nitems(t5_pciids); i++) { 942 if (d == t5_pciids[i].device) { 943 device_set_desc(dev, t5_pciids[i].desc); 944 return (BUS_PROBE_DEFAULT); 945 } 946 } 947 948 return (ENXIO); 949 } 950 951 static int 952 t6_probe(device_t dev) 953 { 954 int i; 955 uint16_t v = pci_get_vendor(dev); 956 uint16_t d = pci_get_device(dev); 957 958 if (v != PCI_VENDOR_ID_CHELSIO) 959 return (ENXIO); 960 961 for (i = 0; i < nitems(t6_pciids); i++) { 962 if (d == t6_pciids[i].device) { 963 device_set_desc(dev, t6_pciids[i].desc); 964 return (BUS_PROBE_DEFAULT); 965 } 966 } 967 968 return (ENXIO); 969 } 970 971 static void 972 t5_attribute_workaround(device_t dev) 973 { 974 device_t root_port; 975 uint32_t v; 976 977 /* 978 * The T5 chips do not properly echo the No Snoop and Relaxed 979 * Ordering attributes when replying to a TLP from a Root 980 * Port. As a workaround, find the parent Root Port and 981 * disable No Snoop and Relaxed Ordering. Note that this 982 * affects all devices under this root port. 983 */ 984 root_port = pci_find_pcie_root_port(dev); 985 if (root_port == NULL) { 986 device_printf(dev, "Unable to find parent root port\n"); 987 return; 988 } 989 990 v = pcie_adjust_config(root_port, PCIER_DEVICE_CTL, 991 PCIEM_CTL_RELAXED_ORD_ENABLE | PCIEM_CTL_NOSNOOP_ENABLE, 0, 2); 992 if ((v & (PCIEM_CTL_RELAXED_ORD_ENABLE | PCIEM_CTL_NOSNOOP_ENABLE)) != 993 0) 994 device_printf(dev, "Disabled No Snoop/Relaxed Ordering on %s\n", 995 device_get_nameunit(root_port)); 996 } 997 998 static const struct devnames devnames[] = { 999 { 1000 .nexus_name = "t4nex", 1001 .ifnet_name = "cxgbe", 1002 .vi_ifnet_name = "vcxgbe", 1003 .pf03_drv_name = "t4iov", 1004 .vf_nexus_name = "t4vf", 1005 .vf_ifnet_name = "cxgbev" 1006 }, { 1007 .nexus_name = "t5nex", 1008 .ifnet_name = "cxl", 1009 .vi_ifnet_name = "vcxl", 1010 .pf03_drv_name = "t5iov", 1011 .vf_nexus_name = "t5vf", 1012 .vf_ifnet_name = "cxlv" 1013 }, { 1014 .nexus_name = "t6nex", 1015 .ifnet_name = "cc", 1016 .vi_ifnet_name = "vcc", 1017 .pf03_drv_name = "t6iov", 1018 .vf_nexus_name = "t6vf", 1019 .vf_ifnet_name = "ccv" 1020 } 1021 }; 1022 1023 void 1024 t4_init_devnames(struct adapter *sc) 1025 { 1026 int id; 1027 1028 id = chip_id(sc); 1029 if (id >= CHELSIO_T4 && id - CHELSIO_T4 < nitems(devnames)) 1030 sc->names = &devnames[id - CHELSIO_T4]; 1031 else { 1032 device_printf(sc->dev, "chip id %d is not supported.\n", id); 1033 sc->names = NULL; 1034 } 1035 } 1036 1037 static int 1038 t4_ifnet_unit(struct adapter *sc, struct port_info *pi) 1039 { 1040 const char *parent, *name; 1041 long value; 1042 int line, unit; 1043 1044 line = 0; 1045 parent = device_get_nameunit(sc->dev); 1046 name = sc->names->ifnet_name; 1047 while (resource_find_dev(&line, name, &unit, "at", parent) == 0) { 1048 if (resource_long_value(name, unit, "port", &value) == 0 && 1049 value == pi->port_id) 1050 return (unit); 1051 } 1052 return (-1); 1053 } 1054 1055 static int 1056 t4_attach(device_t dev) 1057 { 1058 struct adapter *sc; 1059 int rc = 0, i, j, rqidx, tqidx, nports; 1060 struct make_dev_args mda; 1061 struct intrs_and_queues iaq; 1062 struct sge *s; 1063 uint32_t *buf; 1064 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 1065 int ofld_tqidx; 1066 #endif 1067 #ifdef TCP_OFFLOAD 1068 int ofld_rqidx; 1069 #endif 1070 #ifdef DEV_NETMAP 1071 int nm_rqidx, nm_tqidx; 1072 #endif 1073 int num_vis; 1074 1075 sc = device_get_softc(dev); 1076 sc->dev = dev; 1077 TUNABLE_INT_FETCH("hw.cxgbe.dflags", &sc->debug_flags); 1078 1079 if ((pci_get_device(dev) & 0xff00) == 0x5400) 1080 t5_attribute_workaround(dev); 1081 pci_enable_busmaster(dev); 1082 if (pci_find_cap(dev, PCIY_EXPRESS, &i) == 0) { 1083 uint32_t v; 1084 1085 pci_set_max_read_req(dev, 4096); 1086 v = pci_read_config(dev, i + PCIER_DEVICE_CTL, 2); 1087 sc->params.pci.mps = 128 << ((v & PCIEM_CTL_MAX_PAYLOAD) >> 5); 1088 if (pcie_relaxed_ordering == 0 && 1089 (v & PCIEM_CTL_RELAXED_ORD_ENABLE) != 0) { 1090 v &= ~PCIEM_CTL_RELAXED_ORD_ENABLE; 1091 pci_write_config(dev, i + PCIER_DEVICE_CTL, v, 2); 1092 } else if (pcie_relaxed_ordering == 1 && 1093 (v & PCIEM_CTL_RELAXED_ORD_ENABLE) == 0) { 1094 v |= PCIEM_CTL_RELAXED_ORD_ENABLE; 1095 pci_write_config(dev, i + PCIER_DEVICE_CTL, v, 2); 1096 } 1097 } 1098 1099 sc->sge_gts_reg = MYPF_REG(A_SGE_PF_GTS); 1100 sc->sge_kdoorbell_reg = MYPF_REG(A_SGE_PF_KDOORBELL); 1101 sc->traceq = -1; 1102 mtx_init(&sc->ifp_lock, sc->ifp_lockname, 0, MTX_DEF); 1103 snprintf(sc->ifp_lockname, sizeof(sc->ifp_lockname), "%s tracer", 1104 device_get_nameunit(dev)); 1105 1106 snprintf(sc->lockname, sizeof(sc->lockname), "%s", 1107 device_get_nameunit(dev)); 1108 mtx_init(&sc->sc_lock, sc->lockname, 0, MTX_DEF); 1109 t4_add_adapter(sc); 1110 1111 mtx_init(&sc->sfl_lock, "starving freelists", 0, MTX_DEF); 1112 TAILQ_INIT(&sc->sfl); 1113 callout_init_mtx(&sc->sfl_callout, &sc->sfl_lock, 0); 1114 1115 mtx_init(&sc->reg_lock, "indirect register access", 0, MTX_DEF); 1116 1117 sc->policy = NULL; 1118 rw_init(&sc->policy_lock, "connection offload policy"); 1119 1120 callout_init(&sc->ktls_tick, 1); 1121 1122 #ifdef TCP_OFFLOAD 1123 TASK_INIT(&sc->async_event_task, 0, t4_async_event, sc); 1124 #endif 1125 1126 refcount_init(&sc->vxlan_refcount, 0); 1127 1128 rc = t4_map_bars_0_and_4(sc); 1129 if (rc != 0) 1130 goto done; /* error message displayed already */ 1131 1132 memset(sc->chan_map, 0xff, sizeof(sc->chan_map)); 1133 1134 /* Prepare the adapter for operation. */ 1135 buf = malloc(PAGE_SIZE, M_CXGBE, M_ZERO | M_WAITOK); 1136 rc = -t4_prep_adapter(sc, buf); 1137 free(buf, M_CXGBE); 1138 if (rc != 0) { 1139 device_printf(dev, "failed to prepare adapter: %d.\n", rc); 1140 goto done; 1141 } 1142 1143 /* 1144 * This is the real PF# to which we're attaching. Works from within PCI 1145 * passthrough environments too, where pci_get_function() could return a 1146 * different PF# depending on the passthrough configuration. We need to 1147 * use the real PF# in all our communication with the firmware. 1148 */ 1149 j = t4_read_reg(sc, A_PL_WHOAMI); 1150 sc->pf = chip_id(sc) <= CHELSIO_T5 ? G_SOURCEPF(j) : G_T6_SOURCEPF(j); 1151 sc->mbox = sc->pf; 1152 1153 t4_init_devnames(sc); 1154 if (sc->names == NULL) { 1155 rc = ENOTSUP; 1156 goto done; /* error message displayed already */ 1157 } 1158 1159 /* 1160 * Do this really early, with the memory windows set up even before the 1161 * character device. The userland tool's register i/o and mem read 1162 * will work even in "recovery mode". 1163 */ 1164 setup_memwin(sc); 1165 if (t4_init_devlog_params(sc, 0) == 0) 1166 fixup_devlog_params(sc); 1167 make_dev_args_init(&mda); 1168 mda.mda_devsw = &t4_cdevsw; 1169 mda.mda_uid = UID_ROOT; 1170 mda.mda_gid = GID_WHEEL; 1171 mda.mda_mode = 0600; 1172 mda.mda_si_drv1 = sc; 1173 rc = make_dev_s(&mda, &sc->cdev, "%s", device_get_nameunit(dev)); 1174 if (rc != 0) 1175 device_printf(dev, "failed to create nexus char device: %d.\n", 1176 rc); 1177 1178 /* Go no further if recovery mode has been requested. */ 1179 if (TUNABLE_INT_FETCH("hw.cxgbe.sos", &i) && i != 0) { 1180 device_printf(dev, "recovery mode.\n"); 1181 goto done; 1182 } 1183 1184 #if defined(__i386__) 1185 if ((cpu_feature & CPUID_CX8) == 0) { 1186 device_printf(dev, "64 bit atomics not available.\n"); 1187 rc = ENOTSUP; 1188 goto done; 1189 } 1190 #endif 1191 1192 /* Contact the firmware and try to become the master driver. */ 1193 rc = contact_firmware(sc); 1194 if (rc != 0) 1195 goto done; /* error message displayed already */ 1196 MPASS(sc->flags & FW_OK); 1197 1198 rc = get_params__pre_init(sc); 1199 if (rc != 0) 1200 goto done; /* error message displayed already */ 1201 1202 if (sc->flags & MASTER_PF) { 1203 rc = partition_resources(sc); 1204 if (rc != 0) 1205 goto done; /* error message displayed already */ 1206 t4_intr_clear(sc); 1207 } 1208 1209 rc = get_params__post_init(sc); 1210 if (rc != 0) 1211 goto done; /* error message displayed already */ 1212 1213 rc = set_params__post_init(sc); 1214 if (rc != 0) 1215 goto done; /* error message displayed already */ 1216 1217 rc = t4_map_bar_2(sc); 1218 if (rc != 0) 1219 goto done; /* error message displayed already */ 1220 1221 rc = t4_create_dma_tag(sc); 1222 if (rc != 0) 1223 goto done; /* error message displayed already */ 1224 1225 /* 1226 * First pass over all the ports - allocate VIs and initialize some 1227 * basic parameters like mac address, port type, etc. 1228 */ 1229 for_each_port(sc, i) { 1230 struct port_info *pi; 1231 1232 pi = malloc(sizeof(*pi), M_CXGBE, M_ZERO | M_WAITOK); 1233 sc->port[i] = pi; 1234 1235 /* These must be set before t4_port_init */ 1236 pi->adapter = sc; 1237 pi->port_id = i; 1238 /* 1239 * XXX: vi[0] is special so we can't delay this allocation until 1240 * pi->nvi's final value is known. 1241 */ 1242 pi->vi = malloc(sizeof(struct vi_info) * t4_num_vis, M_CXGBE, 1243 M_ZERO | M_WAITOK); 1244 1245 /* 1246 * Allocate the "main" VI and initialize parameters 1247 * like mac addr. 1248 */ 1249 rc = -t4_port_init(sc, sc->mbox, sc->pf, 0, i); 1250 if (rc != 0) { 1251 device_printf(dev, "unable to initialize port %d: %d\n", 1252 i, rc); 1253 free(pi->vi, M_CXGBE); 1254 free(pi, M_CXGBE); 1255 sc->port[i] = NULL; 1256 goto done; 1257 } 1258 1259 snprintf(pi->lockname, sizeof(pi->lockname), "%sp%d", 1260 device_get_nameunit(dev), i); 1261 mtx_init(&pi->pi_lock, pi->lockname, 0, MTX_DEF); 1262 sc->chan_map[pi->tx_chan] = i; 1263 1264 /* 1265 * The MPS counter for FCS errors doesn't work correctly on the 1266 * T6 so we use the MAC counter here. Which MAC is in use 1267 * depends on the link settings which will be known when the 1268 * link comes up. 1269 */ 1270 if (is_t6(sc)) { 1271 pi->fcs_reg = -1; 1272 } else if (is_t4(sc)) { 1273 pi->fcs_reg = PORT_REG(pi->tx_chan, 1274 A_MPS_PORT_STAT_RX_PORT_CRC_ERROR_L); 1275 } else { 1276 pi->fcs_reg = T5_PORT_REG(pi->tx_chan, 1277 A_MPS_PORT_STAT_RX_PORT_CRC_ERROR_L); 1278 } 1279 pi->fcs_base = 0; 1280 1281 /* All VIs on this port share this media. */ 1282 ifmedia_init(&pi->media, IFM_IMASK, cxgbe_media_change, 1283 cxgbe_media_status); 1284 1285 PORT_LOCK(pi); 1286 init_link_config(pi); 1287 fixup_link_config(pi); 1288 build_medialist(pi); 1289 if (fixed_ifmedia(pi)) 1290 pi->flags |= FIXED_IFMEDIA; 1291 PORT_UNLOCK(pi); 1292 1293 pi->dev = device_add_child(dev, sc->names->ifnet_name, 1294 t4_ifnet_unit(sc, pi)); 1295 if (pi->dev == NULL) { 1296 device_printf(dev, 1297 "failed to add device for port %d.\n", i); 1298 rc = ENXIO; 1299 goto done; 1300 } 1301 pi->vi[0].dev = pi->dev; 1302 device_set_softc(pi->dev, pi); 1303 } 1304 1305 /* 1306 * Interrupt type, # of interrupts, # of rx/tx queues, etc. 1307 */ 1308 nports = sc->params.nports; 1309 rc = cfg_itype_and_nqueues(sc, &iaq); 1310 if (rc != 0) 1311 goto done; /* error message displayed already */ 1312 1313 num_vis = iaq.num_vis; 1314 sc->intr_type = iaq.intr_type; 1315 sc->intr_count = iaq.nirq; 1316 1317 s = &sc->sge; 1318 s->nrxq = nports * iaq.nrxq; 1319 s->ntxq = nports * iaq.ntxq; 1320 if (num_vis > 1) { 1321 s->nrxq += nports * (num_vis - 1) * iaq.nrxq_vi; 1322 s->ntxq += nports * (num_vis - 1) * iaq.ntxq_vi; 1323 } 1324 s->neq = s->ntxq + s->nrxq; /* the free list in an rxq is an eq */ 1325 s->neq += nports; /* ctrl queues: 1 per port */ 1326 s->niq = s->nrxq + 1; /* 1 extra for firmware event queue */ 1327 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 1328 if (is_offload(sc) || is_ethoffload(sc)) { 1329 s->nofldtxq = nports * iaq.nofldtxq; 1330 if (num_vis > 1) 1331 s->nofldtxq += nports * (num_vis - 1) * iaq.nofldtxq_vi; 1332 s->neq += s->nofldtxq; 1333 1334 s->ofld_txq = malloc(s->nofldtxq * sizeof(struct sge_wrq), 1335 M_CXGBE, M_ZERO | M_WAITOK); 1336 } 1337 #endif 1338 #ifdef TCP_OFFLOAD 1339 if (is_offload(sc)) { 1340 s->nofldrxq = nports * iaq.nofldrxq; 1341 if (num_vis > 1) 1342 s->nofldrxq += nports * (num_vis - 1) * iaq.nofldrxq_vi; 1343 s->neq += s->nofldrxq; /* free list */ 1344 s->niq += s->nofldrxq; 1345 1346 s->ofld_rxq = malloc(s->nofldrxq * sizeof(struct sge_ofld_rxq), 1347 M_CXGBE, M_ZERO | M_WAITOK); 1348 } 1349 #endif 1350 #ifdef DEV_NETMAP 1351 s->nnmrxq = 0; 1352 s->nnmtxq = 0; 1353 if (t4_native_netmap & NN_MAIN_VI) { 1354 s->nnmrxq += nports * iaq.nnmrxq; 1355 s->nnmtxq += nports * iaq.nnmtxq; 1356 } 1357 if (num_vis > 1 && t4_native_netmap & NN_EXTRA_VI) { 1358 s->nnmrxq += nports * (num_vis - 1) * iaq.nnmrxq_vi; 1359 s->nnmtxq += nports * (num_vis - 1) * iaq.nnmtxq_vi; 1360 } 1361 s->neq += s->nnmtxq + s->nnmrxq; 1362 s->niq += s->nnmrxq; 1363 1364 s->nm_rxq = malloc(s->nnmrxq * sizeof(struct sge_nm_rxq), 1365 M_CXGBE, M_ZERO | M_WAITOK); 1366 s->nm_txq = malloc(s->nnmtxq * sizeof(struct sge_nm_txq), 1367 M_CXGBE, M_ZERO | M_WAITOK); 1368 #endif 1369 MPASS(s->niq <= s->iqmap_sz); 1370 MPASS(s->neq <= s->eqmap_sz); 1371 1372 s->ctrlq = malloc(nports * sizeof(struct sge_wrq), M_CXGBE, 1373 M_ZERO | M_WAITOK); 1374 s->rxq = malloc(s->nrxq * sizeof(struct sge_rxq), M_CXGBE, 1375 M_ZERO | M_WAITOK); 1376 s->txq = malloc(s->ntxq * sizeof(struct sge_txq), M_CXGBE, 1377 M_ZERO | M_WAITOK); 1378 s->iqmap = malloc(s->iqmap_sz * sizeof(struct sge_iq *), M_CXGBE, 1379 M_ZERO | M_WAITOK); 1380 s->eqmap = malloc(s->eqmap_sz * sizeof(struct sge_eq *), M_CXGBE, 1381 M_ZERO | M_WAITOK); 1382 1383 sc->irq = malloc(sc->intr_count * sizeof(struct irq), M_CXGBE, 1384 M_ZERO | M_WAITOK); 1385 1386 t4_init_l2t(sc, M_WAITOK); 1387 t4_init_smt(sc, M_WAITOK); 1388 t4_init_tx_sched(sc); 1389 t4_init_atid_table(sc); 1390 #ifdef RATELIMIT 1391 t4_init_etid_table(sc); 1392 #endif 1393 #ifdef INET6 1394 t4_init_clip_table(sc); 1395 #endif 1396 if (sc->vres.key.size != 0) 1397 sc->key_map = vmem_create("T4TLS key map", sc->vres.key.start, 1398 sc->vres.key.size, 32, 0, M_FIRSTFIT | M_WAITOK); 1399 1400 /* 1401 * Second pass over the ports. This time we know the number of rx and 1402 * tx queues that each port should get. 1403 */ 1404 rqidx = tqidx = 0; 1405 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 1406 ofld_tqidx = 0; 1407 #endif 1408 #ifdef TCP_OFFLOAD 1409 ofld_rqidx = 0; 1410 #endif 1411 #ifdef DEV_NETMAP 1412 nm_rqidx = nm_tqidx = 0; 1413 #endif 1414 for_each_port(sc, i) { 1415 struct port_info *pi = sc->port[i]; 1416 struct vi_info *vi; 1417 1418 if (pi == NULL) 1419 continue; 1420 1421 pi->nvi = num_vis; 1422 for_each_vi(pi, j, vi) { 1423 vi->pi = pi; 1424 vi->adapter = sc; 1425 vi->qsize_rxq = t4_qsize_rxq; 1426 vi->qsize_txq = t4_qsize_txq; 1427 1428 vi->first_rxq = rqidx; 1429 vi->first_txq = tqidx; 1430 vi->tmr_idx = t4_tmr_idx; 1431 vi->pktc_idx = t4_pktc_idx; 1432 vi->nrxq = j == 0 ? iaq.nrxq : iaq.nrxq_vi; 1433 vi->ntxq = j == 0 ? iaq.ntxq : iaq.ntxq_vi; 1434 1435 rqidx += vi->nrxq; 1436 tqidx += vi->ntxq; 1437 1438 if (j == 0 && vi->ntxq > 1) 1439 vi->rsrv_noflowq = t4_rsrv_noflowq ? 1 : 0; 1440 else 1441 vi->rsrv_noflowq = 0; 1442 1443 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 1444 vi->first_ofld_txq = ofld_tqidx; 1445 vi->nofldtxq = j == 0 ? iaq.nofldtxq : iaq.nofldtxq_vi; 1446 ofld_tqidx += vi->nofldtxq; 1447 #endif 1448 #ifdef TCP_OFFLOAD 1449 vi->ofld_tmr_idx = t4_tmr_idx_ofld; 1450 vi->ofld_pktc_idx = t4_pktc_idx_ofld; 1451 vi->first_ofld_rxq = ofld_rqidx; 1452 vi->nofldrxq = j == 0 ? iaq.nofldrxq : iaq.nofldrxq_vi; 1453 1454 ofld_rqidx += vi->nofldrxq; 1455 #endif 1456 #ifdef DEV_NETMAP 1457 vi->first_nm_rxq = nm_rqidx; 1458 vi->first_nm_txq = nm_tqidx; 1459 if (j == 0) { 1460 vi->nnmrxq = iaq.nnmrxq; 1461 vi->nnmtxq = iaq.nnmtxq; 1462 } else { 1463 vi->nnmrxq = iaq.nnmrxq_vi; 1464 vi->nnmtxq = iaq.nnmtxq_vi; 1465 } 1466 nm_rqidx += vi->nnmrxq; 1467 nm_tqidx += vi->nnmtxq; 1468 #endif 1469 } 1470 } 1471 1472 rc = t4_setup_intr_handlers(sc); 1473 if (rc != 0) { 1474 device_printf(dev, 1475 "failed to setup interrupt handlers: %d\n", rc); 1476 goto done; 1477 } 1478 1479 rc = bus_generic_probe(dev); 1480 if (rc != 0) { 1481 device_printf(dev, "failed to probe child drivers: %d\n", rc); 1482 goto done; 1483 } 1484 1485 /* 1486 * Ensure thread-safe mailbox access (in debug builds). 1487 * 1488 * So far this was the only thread accessing the mailbox but various 1489 * ifnets and sysctls are about to be created and their handlers/ioctls 1490 * will access the mailbox from different threads. 1491 */ 1492 sc->flags |= CHK_MBOX_ACCESS; 1493 1494 rc = bus_generic_attach(dev); 1495 if (rc != 0) { 1496 device_printf(dev, 1497 "failed to attach all child ports: %d\n", rc); 1498 goto done; 1499 } 1500 1501 device_printf(dev, 1502 "PCIe gen%d x%d, %d ports, %d %s interrupt%s, %d eq, %d iq\n", 1503 sc->params.pci.speed, sc->params.pci.width, sc->params.nports, 1504 sc->intr_count, sc->intr_type == INTR_MSIX ? "MSI-X" : 1505 (sc->intr_type == INTR_MSI ? "MSI" : "INTx"), 1506 sc->intr_count > 1 ? "s" : "", sc->sge.neq, sc->sge.niq); 1507 1508 t4_set_desc(sc); 1509 1510 notify_siblings(dev, 0); 1511 1512 done: 1513 if (rc != 0 && sc->cdev) { 1514 /* cdev was created and so cxgbetool works; recover that way. */ 1515 device_printf(dev, 1516 "error during attach, adapter is now in recovery mode.\n"); 1517 rc = 0; 1518 } 1519 1520 if (rc != 0) 1521 t4_detach_common(dev); 1522 else 1523 t4_sysctls(sc); 1524 1525 return (rc); 1526 } 1527 1528 static int 1529 t4_child_location_str(device_t bus, device_t dev, char *buf, size_t buflen) 1530 { 1531 struct adapter *sc; 1532 struct port_info *pi; 1533 int i; 1534 1535 sc = device_get_softc(bus); 1536 buf[0] = '\0'; 1537 for_each_port(sc, i) { 1538 pi = sc->port[i]; 1539 if (pi != NULL && pi->dev == dev) { 1540 snprintf(buf, buflen, "port=%d", pi->port_id); 1541 break; 1542 } 1543 } 1544 return (0); 1545 } 1546 1547 static int 1548 t4_ready(device_t dev) 1549 { 1550 struct adapter *sc; 1551 1552 sc = device_get_softc(dev); 1553 if (sc->flags & FW_OK) 1554 return (0); 1555 return (ENXIO); 1556 } 1557 1558 static int 1559 t4_read_port_device(device_t dev, int port, device_t *child) 1560 { 1561 struct adapter *sc; 1562 struct port_info *pi; 1563 1564 sc = device_get_softc(dev); 1565 if (port < 0 || port >= MAX_NPORTS) 1566 return (EINVAL); 1567 pi = sc->port[port]; 1568 if (pi == NULL || pi->dev == NULL) 1569 return (ENXIO); 1570 *child = pi->dev; 1571 return (0); 1572 } 1573 1574 static int 1575 notify_siblings(device_t dev, int detaching) 1576 { 1577 device_t sibling; 1578 int error, i; 1579 1580 error = 0; 1581 for (i = 0; i < PCI_FUNCMAX; i++) { 1582 if (i == pci_get_function(dev)) 1583 continue; 1584 sibling = pci_find_dbsf(pci_get_domain(dev), pci_get_bus(dev), 1585 pci_get_slot(dev), i); 1586 if (sibling == NULL || !device_is_attached(sibling)) 1587 continue; 1588 if (detaching) 1589 error = T4_DETACH_CHILD(sibling); 1590 else 1591 (void)T4_ATTACH_CHILD(sibling); 1592 if (error) 1593 break; 1594 } 1595 return (error); 1596 } 1597 1598 /* 1599 * Idempotent 1600 */ 1601 static int 1602 t4_detach(device_t dev) 1603 { 1604 struct adapter *sc; 1605 int rc; 1606 1607 sc = device_get_softc(dev); 1608 1609 rc = notify_siblings(dev, 1); 1610 if (rc) { 1611 device_printf(dev, 1612 "failed to detach sibling devices: %d\n", rc); 1613 return (rc); 1614 } 1615 1616 return (t4_detach_common(dev)); 1617 } 1618 1619 int 1620 t4_detach_common(device_t dev) 1621 { 1622 struct adapter *sc; 1623 struct port_info *pi; 1624 int i, rc; 1625 1626 sc = device_get_softc(dev); 1627 1628 if (sc->cdev) { 1629 destroy_dev(sc->cdev); 1630 sc->cdev = NULL; 1631 } 1632 1633 sx_xlock(&t4_list_lock); 1634 SLIST_REMOVE(&t4_list, sc, adapter, link); 1635 sx_xunlock(&t4_list_lock); 1636 1637 sc->flags &= ~CHK_MBOX_ACCESS; 1638 if (sc->flags & FULL_INIT_DONE) { 1639 if (!(sc->flags & IS_VF)) 1640 t4_intr_disable(sc); 1641 } 1642 1643 if (device_is_attached(dev)) { 1644 rc = bus_generic_detach(dev); 1645 if (rc) { 1646 device_printf(dev, 1647 "failed to detach child devices: %d\n", rc); 1648 return (rc); 1649 } 1650 } 1651 1652 #ifdef TCP_OFFLOAD 1653 taskqueue_drain(taskqueue_thread, &sc->async_event_task); 1654 #endif 1655 1656 for (i = 0; i < sc->intr_count; i++) 1657 t4_free_irq(sc, &sc->irq[i]); 1658 1659 if ((sc->flags & (IS_VF | FW_OK)) == FW_OK) 1660 t4_free_tx_sched(sc); 1661 1662 for (i = 0; i < MAX_NPORTS; i++) { 1663 pi = sc->port[i]; 1664 if (pi) { 1665 t4_free_vi(sc, sc->mbox, sc->pf, 0, pi->vi[0].viid); 1666 if (pi->dev) 1667 device_delete_child(dev, pi->dev); 1668 1669 mtx_destroy(&pi->pi_lock); 1670 free(pi->vi, M_CXGBE); 1671 free(pi, M_CXGBE); 1672 } 1673 } 1674 1675 device_delete_children(dev); 1676 1677 if (sc->flags & FULL_INIT_DONE) 1678 adapter_full_uninit(sc); 1679 1680 if ((sc->flags & (IS_VF | FW_OK)) == FW_OK) 1681 t4_fw_bye(sc, sc->mbox); 1682 1683 if (sc->intr_type == INTR_MSI || sc->intr_type == INTR_MSIX) 1684 pci_release_msi(dev); 1685 1686 if (sc->regs_res) 1687 bus_release_resource(dev, SYS_RES_MEMORY, sc->regs_rid, 1688 sc->regs_res); 1689 1690 if (sc->udbs_res) 1691 bus_release_resource(dev, SYS_RES_MEMORY, sc->udbs_rid, 1692 sc->udbs_res); 1693 1694 if (sc->msix_res) 1695 bus_release_resource(dev, SYS_RES_MEMORY, sc->msix_rid, 1696 sc->msix_res); 1697 1698 if (sc->l2t) 1699 t4_free_l2t(sc->l2t); 1700 if (sc->smt) 1701 t4_free_smt(sc->smt); 1702 t4_free_atid_table(sc); 1703 #ifdef RATELIMIT 1704 t4_free_etid_table(sc); 1705 #endif 1706 if (sc->key_map) 1707 vmem_destroy(sc->key_map); 1708 #ifdef INET6 1709 t4_destroy_clip_table(sc); 1710 #endif 1711 1712 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 1713 free(sc->sge.ofld_txq, M_CXGBE); 1714 #endif 1715 #ifdef TCP_OFFLOAD 1716 free(sc->sge.ofld_rxq, M_CXGBE); 1717 #endif 1718 #ifdef DEV_NETMAP 1719 free(sc->sge.nm_rxq, M_CXGBE); 1720 free(sc->sge.nm_txq, M_CXGBE); 1721 #endif 1722 free(sc->irq, M_CXGBE); 1723 free(sc->sge.rxq, M_CXGBE); 1724 free(sc->sge.txq, M_CXGBE); 1725 free(sc->sge.ctrlq, M_CXGBE); 1726 free(sc->sge.iqmap, M_CXGBE); 1727 free(sc->sge.eqmap, M_CXGBE); 1728 free(sc->tids.ftid_tab, M_CXGBE); 1729 free(sc->tids.hpftid_tab, M_CXGBE); 1730 free_hftid_hash(&sc->tids); 1731 free(sc->tids.tid_tab, M_CXGBE); 1732 free(sc->tt.tls_rx_ports, M_CXGBE); 1733 t4_destroy_dma_tag(sc); 1734 1735 callout_drain(&sc->ktls_tick); 1736 callout_drain(&sc->sfl_callout); 1737 if (mtx_initialized(&sc->tids.ftid_lock)) { 1738 mtx_destroy(&sc->tids.ftid_lock); 1739 cv_destroy(&sc->tids.ftid_cv); 1740 } 1741 if (mtx_initialized(&sc->tids.atid_lock)) 1742 mtx_destroy(&sc->tids.atid_lock); 1743 if (mtx_initialized(&sc->ifp_lock)) 1744 mtx_destroy(&sc->ifp_lock); 1745 1746 if (rw_initialized(&sc->policy_lock)) { 1747 rw_destroy(&sc->policy_lock); 1748 #ifdef TCP_OFFLOAD 1749 if (sc->policy != NULL) 1750 free_offload_policy(sc->policy); 1751 #endif 1752 } 1753 1754 for (i = 0; i < NUM_MEMWIN; i++) { 1755 struct memwin *mw = &sc->memwin[i]; 1756 1757 if (rw_initialized(&mw->mw_lock)) 1758 rw_destroy(&mw->mw_lock); 1759 } 1760 1761 mtx_destroy(&sc->sfl_lock); 1762 mtx_destroy(&sc->reg_lock); 1763 mtx_destroy(&sc->sc_lock); 1764 1765 bzero(sc, sizeof(*sc)); 1766 1767 return (0); 1768 } 1769 1770 static int 1771 cxgbe_probe(device_t dev) 1772 { 1773 char buf[128]; 1774 struct port_info *pi = device_get_softc(dev); 1775 1776 snprintf(buf, sizeof(buf), "port %d", pi->port_id); 1777 device_set_desc_copy(dev, buf); 1778 1779 return (BUS_PROBE_DEFAULT); 1780 } 1781 1782 #define T4_CAP (IFCAP_VLAN_HWTAGGING | IFCAP_VLAN_MTU | IFCAP_HWCSUM | \ 1783 IFCAP_VLAN_HWCSUM | IFCAP_TSO | IFCAP_JUMBO_MTU | IFCAP_LRO | \ 1784 IFCAP_VLAN_HWTSO | IFCAP_LINKSTATE | IFCAP_HWCSUM_IPV6 | IFCAP_HWSTATS | \ 1785 IFCAP_HWRXTSTMP | IFCAP_MEXTPG) 1786 #define T4_CAP_ENABLE (T4_CAP) 1787 1788 static int 1789 cxgbe_vi_attach(device_t dev, struct vi_info *vi) 1790 { 1791 struct ifnet *ifp; 1792 struct sbuf *sb; 1793 struct pfil_head_args pa; 1794 struct adapter *sc = vi->adapter; 1795 1796 vi->xact_addr_filt = -1; 1797 callout_init(&vi->tick, 1); 1798 if (sc->flags & IS_VF || t4_tx_vm_wr != 0) 1799 vi->flags |= TX_USES_VM_WR; 1800 1801 /* Allocate an ifnet and set it up */ 1802 ifp = if_alloc_dev(IFT_ETHER, dev); 1803 if (ifp == NULL) { 1804 device_printf(dev, "Cannot allocate ifnet\n"); 1805 return (ENOMEM); 1806 } 1807 vi->ifp = ifp; 1808 ifp->if_softc = vi; 1809 1810 if_initname(ifp, device_get_name(dev), device_get_unit(dev)); 1811 ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST; 1812 1813 ifp->if_init = cxgbe_init; 1814 ifp->if_ioctl = cxgbe_ioctl; 1815 ifp->if_transmit = cxgbe_transmit; 1816 ifp->if_qflush = cxgbe_qflush; 1817 ifp->if_get_counter = cxgbe_get_counter; 1818 #if defined(KERN_TLS) || defined(RATELIMIT) 1819 ifp->if_snd_tag_alloc = cxgbe_snd_tag_alloc; 1820 ifp->if_snd_tag_modify = cxgbe_snd_tag_modify; 1821 ifp->if_snd_tag_query = cxgbe_snd_tag_query; 1822 ifp->if_snd_tag_free = cxgbe_snd_tag_free; 1823 #endif 1824 #ifdef RATELIMIT 1825 ifp->if_ratelimit_query = cxgbe_ratelimit_query; 1826 #endif 1827 1828 ifp->if_capabilities = T4_CAP; 1829 ifp->if_capenable = T4_CAP_ENABLE; 1830 ifp->if_hwassist = CSUM_TCP | CSUM_UDP | CSUM_IP | CSUM_TSO | 1831 CSUM_UDP_IPV6 | CSUM_TCP_IPV6; 1832 if (chip_id(sc) >= CHELSIO_T6) { 1833 ifp->if_capabilities |= IFCAP_VXLAN_HWCSUM | IFCAP_VXLAN_HWTSO; 1834 ifp->if_capenable |= IFCAP_VXLAN_HWCSUM | IFCAP_VXLAN_HWTSO; 1835 ifp->if_hwassist |= CSUM_INNER_IP6_UDP | CSUM_INNER_IP6_TCP | 1836 CSUM_INNER_IP6_TSO | CSUM_INNER_IP | CSUM_INNER_IP_UDP | 1837 CSUM_INNER_IP_TCP | CSUM_INNER_IP_TSO | CSUM_ENCAP_VXLAN; 1838 } 1839 1840 #ifdef TCP_OFFLOAD 1841 if (vi->nofldrxq != 0 && (sc->flags & KERN_TLS_OK) == 0) 1842 ifp->if_capabilities |= IFCAP_TOE; 1843 #endif 1844 #ifdef RATELIMIT 1845 if (is_ethoffload(sc) && vi->nofldtxq != 0) { 1846 ifp->if_capabilities |= IFCAP_TXRTLMT; 1847 ifp->if_capenable |= IFCAP_TXRTLMT; 1848 } 1849 #endif 1850 1851 ifp->if_hw_tsomax = IP_MAXPACKET; 1852 if (vi->flags & TX_USES_VM_WR) 1853 ifp->if_hw_tsomaxsegcount = TX_SGL_SEGS_VM_TSO; 1854 else 1855 ifp->if_hw_tsomaxsegcount = TX_SGL_SEGS_TSO; 1856 #ifdef RATELIMIT 1857 if (is_ethoffload(sc) && vi->nofldtxq != 0) 1858 ifp->if_hw_tsomaxsegcount = TX_SGL_SEGS_EO_TSO; 1859 #endif 1860 ifp->if_hw_tsomaxsegsize = 65536; 1861 #ifdef KERN_TLS 1862 if (sc->flags & KERN_TLS_OK) { 1863 ifp->if_capabilities |= IFCAP_TXTLS; 1864 ifp->if_capenable |= IFCAP_TXTLS; 1865 } 1866 #endif 1867 1868 ether_ifattach(ifp, vi->hw_addr); 1869 #ifdef DEV_NETMAP 1870 if (vi->nnmrxq != 0) 1871 cxgbe_nm_attach(vi); 1872 #endif 1873 sb = sbuf_new_auto(); 1874 sbuf_printf(sb, "%d txq, %d rxq (NIC)", vi->ntxq, vi->nrxq); 1875 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 1876 switch (ifp->if_capabilities & (IFCAP_TOE | IFCAP_TXRTLMT)) { 1877 case IFCAP_TOE: 1878 sbuf_printf(sb, "; %d txq (TOE)", vi->nofldtxq); 1879 break; 1880 case IFCAP_TOE | IFCAP_TXRTLMT: 1881 sbuf_printf(sb, "; %d txq (TOE/ETHOFLD)", vi->nofldtxq); 1882 break; 1883 case IFCAP_TXRTLMT: 1884 sbuf_printf(sb, "; %d txq (ETHOFLD)", vi->nofldtxq); 1885 break; 1886 } 1887 #endif 1888 #ifdef TCP_OFFLOAD 1889 if (ifp->if_capabilities & IFCAP_TOE) 1890 sbuf_printf(sb, ", %d rxq (TOE)", vi->nofldrxq); 1891 #endif 1892 #ifdef DEV_NETMAP 1893 if (ifp->if_capabilities & IFCAP_NETMAP) 1894 sbuf_printf(sb, "; %d txq, %d rxq (netmap)", 1895 vi->nnmtxq, vi->nnmrxq); 1896 #endif 1897 sbuf_finish(sb); 1898 device_printf(dev, "%s\n", sbuf_data(sb)); 1899 sbuf_delete(sb); 1900 1901 vi_sysctls(vi); 1902 1903 pa.pa_version = PFIL_VERSION; 1904 pa.pa_flags = PFIL_IN; 1905 pa.pa_type = PFIL_TYPE_ETHERNET; 1906 pa.pa_headname = ifp->if_xname; 1907 vi->pfil = pfil_head_register(&pa); 1908 1909 return (0); 1910 } 1911 1912 static int 1913 cxgbe_attach(device_t dev) 1914 { 1915 struct port_info *pi = device_get_softc(dev); 1916 struct adapter *sc = pi->adapter; 1917 struct vi_info *vi; 1918 int i, rc; 1919 1920 callout_init_mtx(&pi->tick, &pi->pi_lock, 0); 1921 1922 rc = cxgbe_vi_attach(dev, &pi->vi[0]); 1923 if (rc) 1924 return (rc); 1925 1926 for_each_vi(pi, i, vi) { 1927 if (i == 0) 1928 continue; 1929 vi->dev = device_add_child(dev, sc->names->vi_ifnet_name, -1); 1930 if (vi->dev == NULL) { 1931 device_printf(dev, "failed to add VI %d\n", i); 1932 continue; 1933 } 1934 device_set_softc(vi->dev, vi); 1935 } 1936 1937 cxgbe_sysctls(pi); 1938 1939 bus_generic_attach(dev); 1940 1941 return (0); 1942 } 1943 1944 static void 1945 cxgbe_vi_detach(struct vi_info *vi) 1946 { 1947 struct ifnet *ifp = vi->ifp; 1948 1949 if (vi->pfil != NULL) { 1950 pfil_head_unregister(vi->pfil); 1951 vi->pfil = NULL; 1952 } 1953 1954 ether_ifdetach(ifp); 1955 1956 /* Let detach proceed even if these fail. */ 1957 #ifdef DEV_NETMAP 1958 if (ifp->if_capabilities & IFCAP_NETMAP) 1959 cxgbe_nm_detach(vi); 1960 #endif 1961 cxgbe_uninit_synchronized(vi); 1962 callout_drain(&vi->tick); 1963 vi_full_uninit(vi); 1964 1965 if_free(vi->ifp); 1966 vi->ifp = NULL; 1967 } 1968 1969 static int 1970 cxgbe_detach(device_t dev) 1971 { 1972 struct port_info *pi = device_get_softc(dev); 1973 struct adapter *sc = pi->adapter; 1974 int rc; 1975 1976 /* Detach the extra VIs first. */ 1977 rc = bus_generic_detach(dev); 1978 if (rc) 1979 return (rc); 1980 device_delete_children(dev); 1981 1982 doom_vi(sc, &pi->vi[0]); 1983 1984 if (pi->flags & HAS_TRACEQ) { 1985 sc->traceq = -1; /* cloner should not create ifnet */ 1986 t4_tracer_port_detach(sc); 1987 } 1988 1989 cxgbe_vi_detach(&pi->vi[0]); 1990 callout_drain(&pi->tick); 1991 ifmedia_removeall(&pi->media); 1992 1993 end_synchronized_op(sc, 0); 1994 1995 return (0); 1996 } 1997 1998 static void 1999 cxgbe_init(void *arg) 2000 { 2001 struct vi_info *vi = arg; 2002 struct adapter *sc = vi->adapter; 2003 2004 if (begin_synchronized_op(sc, vi, SLEEP_OK | INTR_OK, "t4init") != 0) 2005 return; 2006 cxgbe_init_synchronized(vi); 2007 end_synchronized_op(sc, 0); 2008 } 2009 2010 static int 2011 cxgbe_ioctl(struct ifnet *ifp, unsigned long cmd, caddr_t data) 2012 { 2013 int rc = 0, mtu, flags; 2014 struct vi_info *vi = ifp->if_softc; 2015 struct port_info *pi = vi->pi; 2016 struct adapter *sc = pi->adapter; 2017 struct ifreq *ifr = (struct ifreq *)data; 2018 uint32_t mask; 2019 2020 switch (cmd) { 2021 case SIOCSIFMTU: 2022 mtu = ifr->ifr_mtu; 2023 if (mtu < ETHERMIN || mtu > MAX_MTU) 2024 return (EINVAL); 2025 2026 rc = begin_synchronized_op(sc, vi, SLEEP_OK | INTR_OK, "t4mtu"); 2027 if (rc) 2028 return (rc); 2029 ifp->if_mtu = mtu; 2030 if (vi->flags & VI_INIT_DONE) { 2031 t4_update_fl_bufsize(ifp); 2032 if (ifp->if_drv_flags & IFF_DRV_RUNNING) 2033 rc = update_mac_settings(ifp, XGMAC_MTU); 2034 } 2035 end_synchronized_op(sc, 0); 2036 break; 2037 2038 case SIOCSIFFLAGS: 2039 rc = begin_synchronized_op(sc, vi, SLEEP_OK | INTR_OK, "t4flg"); 2040 if (rc) 2041 return (rc); 2042 2043 if (ifp->if_flags & IFF_UP) { 2044 if (ifp->if_drv_flags & IFF_DRV_RUNNING) { 2045 flags = vi->if_flags; 2046 if ((ifp->if_flags ^ flags) & 2047 (IFF_PROMISC | IFF_ALLMULTI)) { 2048 rc = update_mac_settings(ifp, 2049 XGMAC_PROMISC | XGMAC_ALLMULTI); 2050 } 2051 } else { 2052 rc = cxgbe_init_synchronized(vi); 2053 } 2054 vi->if_flags = ifp->if_flags; 2055 } else if (ifp->if_drv_flags & IFF_DRV_RUNNING) { 2056 rc = cxgbe_uninit_synchronized(vi); 2057 } 2058 end_synchronized_op(sc, 0); 2059 break; 2060 2061 case SIOCADDMULTI: 2062 case SIOCDELMULTI: 2063 rc = begin_synchronized_op(sc, vi, SLEEP_OK | INTR_OK, "t4multi"); 2064 if (rc) 2065 return (rc); 2066 if (ifp->if_drv_flags & IFF_DRV_RUNNING) 2067 rc = update_mac_settings(ifp, XGMAC_MCADDRS); 2068 end_synchronized_op(sc, 0); 2069 break; 2070 2071 case SIOCSIFCAP: 2072 rc = begin_synchronized_op(sc, vi, SLEEP_OK | INTR_OK, "t4cap"); 2073 if (rc) 2074 return (rc); 2075 2076 mask = ifr->ifr_reqcap ^ ifp->if_capenable; 2077 if (mask & IFCAP_TXCSUM) { 2078 ifp->if_capenable ^= IFCAP_TXCSUM; 2079 ifp->if_hwassist ^= (CSUM_TCP | CSUM_UDP | CSUM_IP); 2080 2081 if (IFCAP_TSO4 & ifp->if_capenable && 2082 !(IFCAP_TXCSUM & ifp->if_capenable)) { 2083 mask &= ~IFCAP_TSO4; 2084 ifp->if_capenable &= ~IFCAP_TSO4; 2085 if_printf(ifp, 2086 "tso4 disabled due to -txcsum.\n"); 2087 } 2088 } 2089 if (mask & IFCAP_TXCSUM_IPV6) { 2090 ifp->if_capenable ^= IFCAP_TXCSUM_IPV6; 2091 ifp->if_hwassist ^= (CSUM_UDP_IPV6 | CSUM_TCP_IPV6); 2092 2093 if (IFCAP_TSO6 & ifp->if_capenable && 2094 !(IFCAP_TXCSUM_IPV6 & ifp->if_capenable)) { 2095 mask &= ~IFCAP_TSO6; 2096 ifp->if_capenable &= ~IFCAP_TSO6; 2097 if_printf(ifp, 2098 "tso6 disabled due to -txcsum6.\n"); 2099 } 2100 } 2101 if (mask & IFCAP_RXCSUM) 2102 ifp->if_capenable ^= IFCAP_RXCSUM; 2103 if (mask & IFCAP_RXCSUM_IPV6) 2104 ifp->if_capenable ^= IFCAP_RXCSUM_IPV6; 2105 2106 /* 2107 * Note that we leave CSUM_TSO alone (it is always set). The 2108 * kernel takes both IFCAP_TSOx and CSUM_TSO into account before 2109 * sending a TSO request our way, so it's sufficient to toggle 2110 * IFCAP_TSOx only. 2111 */ 2112 if (mask & IFCAP_TSO4) { 2113 if (!(IFCAP_TSO4 & ifp->if_capenable) && 2114 !(IFCAP_TXCSUM & ifp->if_capenable)) { 2115 if_printf(ifp, "enable txcsum first.\n"); 2116 rc = EAGAIN; 2117 goto fail; 2118 } 2119 ifp->if_capenable ^= IFCAP_TSO4; 2120 } 2121 if (mask & IFCAP_TSO6) { 2122 if (!(IFCAP_TSO6 & ifp->if_capenable) && 2123 !(IFCAP_TXCSUM_IPV6 & ifp->if_capenable)) { 2124 if_printf(ifp, "enable txcsum6 first.\n"); 2125 rc = EAGAIN; 2126 goto fail; 2127 } 2128 ifp->if_capenable ^= IFCAP_TSO6; 2129 } 2130 if (mask & IFCAP_LRO) { 2131 #if defined(INET) || defined(INET6) 2132 int i; 2133 struct sge_rxq *rxq; 2134 2135 ifp->if_capenable ^= IFCAP_LRO; 2136 for_each_rxq(vi, i, rxq) { 2137 if (ifp->if_capenable & IFCAP_LRO) 2138 rxq->iq.flags |= IQ_LRO_ENABLED; 2139 else 2140 rxq->iq.flags &= ~IQ_LRO_ENABLED; 2141 } 2142 #endif 2143 } 2144 #ifdef TCP_OFFLOAD 2145 if (mask & IFCAP_TOE) { 2146 int enable = (ifp->if_capenable ^ mask) & IFCAP_TOE; 2147 2148 rc = toe_capability(vi, enable); 2149 if (rc != 0) 2150 goto fail; 2151 2152 ifp->if_capenable ^= mask; 2153 } 2154 #endif 2155 if (mask & IFCAP_VLAN_HWTAGGING) { 2156 ifp->if_capenable ^= IFCAP_VLAN_HWTAGGING; 2157 if (ifp->if_drv_flags & IFF_DRV_RUNNING) 2158 rc = update_mac_settings(ifp, XGMAC_VLANEX); 2159 } 2160 if (mask & IFCAP_VLAN_MTU) { 2161 ifp->if_capenable ^= IFCAP_VLAN_MTU; 2162 2163 /* Need to find out how to disable auto-mtu-inflation */ 2164 } 2165 if (mask & IFCAP_VLAN_HWTSO) 2166 ifp->if_capenable ^= IFCAP_VLAN_HWTSO; 2167 if (mask & IFCAP_VLAN_HWCSUM) 2168 ifp->if_capenable ^= IFCAP_VLAN_HWCSUM; 2169 #ifdef RATELIMIT 2170 if (mask & IFCAP_TXRTLMT) 2171 ifp->if_capenable ^= IFCAP_TXRTLMT; 2172 #endif 2173 if (mask & IFCAP_HWRXTSTMP) { 2174 int i; 2175 struct sge_rxq *rxq; 2176 2177 ifp->if_capenable ^= IFCAP_HWRXTSTMP; 2178 for_each_rxq(vi, i, rxq) { 2179 if (ifp->if_capenable & IFCAP_HWRXTSTMP) 2180 rxq->iq.flags |= IQ_RX_TIMESTAMP; 2181 else 2182 rxq->iq.flags &= ~IQ_RX_TIMESTAMP; 2183 } 2184 } 2185 if (mask & IFCAP_MEXTPG) 2186 ifp->if_capenable ^= IFCAP_MEXTPG; 2187 2188 #ifdef KERN_TLS 2189 if (mask & IFCAP_TXTLS) 2190 ifp->if_capenable ^= (mask & IFCAP_TXTLS); 2191 #endif 2192 if (mask & IFCAP_VXLAN_HWCSUM) { 2193 ifp->if_capenable ^= IFCAP_VXLAN_HWCSUM; 2194 ifp->if_hwassist ^= CSUM_INNER_IP6_UDP | 2195 CSUM_INNER_IP6_TCP | CSUM_INNER_IP | 2196 CSUM_INNER_IP_UDP | CSUM_INNER_IP_TCP; 2197 } 2198 if (mask & IFCAP_VXLAN_HWTSO) { 2199 ifp->if_capenable ^= IFCAP_VXLAN_HWTSO; 2200 ifp->if_hwassist ^= CSUM_INNER_IP6_TSO | 2201 CSUM_INNER_IP_TSO; 2202 } 2203 2204 #ifdef VLAN_CAPABILITIES 2205 VLAN_CAPABILITIES(ifp); 2206 #endif 2207 fail: 2208 end_synchronized_op(sc, 0); 2209 break; 2210 2211 case SIOCSIFMEDIA: 2212 case SIOCGIFMEDIA: 2213 case SIOCGIFXMEDIA: 2214 ifmedia_ioctl(ifp, ifr, &pi->media, cmd); 2215 break; 2216 2217 case SIOCGI2C: { 2218 struct ifi2creq i2c; 2219 2220 rc = copyin(ifr_data_get_ptr(ifr), &i2c, sizeof(i2c)); 2221 if (rc != 0) 2222 break; 2223 if (i2c.dev_addr != 0xA0 && i2c.dev_addr != 0xA2) { 2224 rc = EPERM; 2225 break; 2226 } 2227 if (i2c.len > sizeof(i2c.data)) { 2228 rc = EINVAL; 2229 break; 2230 } 2231 rc = begin_synchronized_op(sc, vi, SLEEP_OK | INTR_OK, "t4i2c"); 2232 if (rc) 2233 return (rc); 2234 rc = -t4_i2c_rd(sc, sc->mbox, pi->port_id, i2c.dev_addr, 2235 i2c.offset, i2c.len, &i2c.data[0]); 2236 end_synchronized_op(sc, 0); 2237 if (rc == 0) 2238 rc = copyout(&i2c, ifr_data_get_ptr(ifr), sizeof(i2c)); 2239 break; 2240 } 2241 2242 default: 2243 rc = ether_ioctl(ifp, cmd, data); 2244 } 2245 2246 return (rc); 2247 } 2248 2249 static int 2250 cxgbe_transmit(struct ifnet *ifp, struct mbuf *m) 2251 { 2252 struct vi_info *vi = ifp->if_softc; 2253 struct port_info *pi = vi->pi; 2254 struct adapter *sc; 2255 struct sge_txq *txq; 2256 void *items[1]; 2257 int rc; 2258 2259 M_ASSERTPKTHDR(m); 2260 MPASS(m->m_nextpkt == NULL); /* not quite ready for this yet */ 2261 #if defined(KERN_TLS) || defined(RATELIMIT) 2262 if (m->m_pkthdr.csum_flags & CSUM_SND_TAG) 2263 MPASS(m->m_pkthdr.snd_tag->ifp == ifp); 2264 #endif 2265 2266 if (__predict_false(pi->link_cfg.link_ok == false)) { 2267 m_freem(m); 2268 return (ENETDOWN); 2269 } 2270 2271 rc = parse_pkt(&m, vi->flags & TX_USES_VM_WR); 2272 if (__predict_false(rc != 0)) { 2273 MPASS(m == NULL); /* was freed already */ 2274 atomic_add_int(&pi->tx_parse_error, 1); /* rare, atomic is ok */ 2275 return (rc); 2276 } 2277 #ifdef RATELIMIT 2278 if (m->m_pkthdr.csum_flags & CSUM_SND_TAG) { 2279 if (m->m_pkthdr.snd_tag->type == IF_SND_TAG_TYPE_RATE_LIMIT) 2280 return (ethofld_transmit(ifp, m)); 2281 } 2282 #endif 2283 2284 /* Select a txq. */ 2285 sc = vi->adapter; 2286 txq = &sc->sge.txq[vi->first_txq]; 2287 if (M_HASHTYPE_GET(m) != M_HASHTYPE_NONE) 2288 txq += ((m->m_pkthdr.flowid % (vi->ntxq - vi->rsrv_noflowq)) + 2289 vi->rsrv_noflowq); 2290 2291 items[0] = m; 2292 rc = mp_ring_enqueue(txq->r, items, 1, 256); 2293 if (__predict_false(rc != 0)) 2294 m_freem(m); 2295 2296 return (rc); 2297 } 2298 2299 static void 2300 cxgbe_qflush(struct ifnet *ifp) 2301 { 2302 struct vi_info *vi = ifp->if_softc; 2303 struct sge_txq *txq; 2304 int i; 2305 2306 /* queues do not exist if !VI_INIT_DONE. */ 2307 if (vi->flags & VI_INIT_DONE) { 2308 for_each_txq(vi, i, txq) { 2309 TXQ_LOCK(txq); 2310 txq->eq.flags |= EQ_QFLUSH; 2311 TXQ_UNLOCK(txq); 2312 while (!mp_ring_is_idle(txq->r)) { 2313 mp_ring_check_drainage(txq->r, 4096); 2314 pause("qflush", 1); 2315 } 2316 TXQ_LOCK(txq); 2317 txq->eq.flags &= ~EQ_QFLUSH; 2318 TXQ_UNLOCK(txq); 2319 } 2320 } 2321 if_qflush(ifp); 2322 } 2323 2324 static uint64_t 2325 vi_get_counter(struct ifnet *ifp, ift_counter c) 2326 { 2327 struct vi_info *vi = ifp->if_softc; 2328 struct fw_vi_stats_vf *s = &vi->stats; 2329 2330 vi_refresh_stats(vi->adapter, vi); 2331 2332 switch (c) { 2333 case IFCOUNTER_IPACKETS: 2334 return (s->rx_bcast_frames + s->rx_mcast_frames + 2335 s->rx_ucast_frames); 2336 case IFCOUNTER_IERRORS: 2337 return (s->rx_err_frames); 2338 case IFCOUNTER_OPACKETS: 2339 return (s->tx_bcast_frames + s->tx_mcast_frames + 2340 s->tx_ucast_frames + s->tx_offload_frames); 2341 case IFCOUNTER_OERRORS: 2342 return (s->tx_drop_frames); 2343 case IFCOUNTER_IBYTES: 2344 return (s->rx_bcast_bytes + s->rx_mcast_bytes + 2345 s->rx_ucast_bytes); 2346 case IFCOUNTER_OBYTES: 2347 return (s->tx_bcast_bytes + s->tx_mcast_bytes + 2348 s->tx_ucast_bytes + s->tx_offload_bytes); 2349 case IFCOUNTER_IMCASTS: 2350 return (s->rx_mcast_frames); 2351 case IFCOUNTER_OMCASTS: 2352 return (s->tx_mcast_frames); 2353 case IFCOUNTER_OQDROPS: { 2354 uint64_t drops; 2355 2356 drops = 0; 2357 if (vi->flags & VI_INIT_DONE) { 2358 int i; 2359 struct sge_txq *txq; 2360 2361 for_each_txq(vi, i, txq) 2362 drops += counter_u64_fetch(txq->r->dropped); 2363 } 2364 2365 return (drops); 2366 2367 } 2368 2369 default: 2370 return (if_get_counter_default(ifp, c)); 2371 } 2372 } 2373 2374 uint64_t 2375 cxgbe_get_counter(struct ifnet *ifp, ift_counter c) 2376 { 2377 struct vi_info *vi = ifp->if_softc; 2378 struct port_info *pi = vi->pi; 2379 struct adapter *sc = pi->adapter; 2380 struct port_stats *s = &pi->stats; 2381 2382 if (pi->nvi > 1 || sc->flags & IS_VF) 2383 return (vi_get_counter(ifp, c)); 2384 2385 cxgbe_refresh_stats(sc, pi); 2386 2387 switch (c) { 2388 case IFCOUNTER_IPACKETS: 2389 return (s->rx_frames); 2390 2391 case IFCOUNTER_IERRORS: 2392 return (s->rx_jabber + s->rx_runt + s->rx_too_long + 2393 s->rx_fcs_err + s->rx_len_err); 2394 2395 case IFCOUNTER_OPACKETS: 2396 return (s->tx_frames); 2397 2398 case IFCOUNTER_OERRORS: 2399 return (s->tx_error_frames); 2400 2401 case IFCOUNTER_IBYTES: 2402 return (s->rx_octets); 2403 2404 case IFCOUNTER_OBYTES: 2405 return (s->tx_octets); 2406 2407 case IFCOUNTER_IMCASTS: 2408 return (s->rx_mcast_frames); 2409 2410 case IFCOUNTER_OMCASTS: 2411 return (s->tx_mcast_frames); 2412 2413 case IFCOUNTER_IQDROPS: 2414 return (s->rx_ovflow0 + s->rx_ovflow1 + s->rx_ovflow2 + 2415 s->rx_ovflow3 + s->rx_trunc0 + s->rx_trunc1 + s->rx_trunc2 + 2416 s->rx_trunc3 + pi->tnl_cong_drops); 2417 2418 case IFCOUNTER_OQDROPS: { 2419 uint64_t drops; 2420 2421 drops = s->tx_drop; 2422 if (vi->flags & VI_INIT_DONE) { 2423 int i; 2424 struct sge_txq *txq; 2425 2426 for_each_txq(vi, i, txq) 2427 drops += counter_u64_fetch(txq->r->dropped); 2428 } 2429 2430 return (drops); 2431 2432 } 2433 2434 default: 2435 return (if_get_counter_default(ifp, c)); 2436 } 2437 } 2438 2439 #if defined(KERN_TLS) || defined(RATELIMIT) 2440 static int 2441 cxgbe_snd_tag_alloc(struct ifnet *ifp, union if_snd_tag_alloc_params *params, 2442 struct m_snd_tag **pt) 2443 { 2444 int error; 2445 2446 switch (params->hdr.type) { 2447 #ifdef RATELIMIT 2448 case IF_SND_TAG_TYPE_RATE_LIMIT: 2449 error = cxgbe_rate_tag_alloc(ifp, params, pt); 2450 break; 2451 #endif 2452 #ifdef KERN_TLS 2453 case IF_SND_TAG_TYPE_TLS: 2454 error = cxgbe_tls_tag_alloc(ifp, params, pt); 2455 break; 2456 #endif 2457 default: 2458 error = EOPNOTSUPP; 2459 } 2460 return (error); 2461 } 2462 2463 static int 2464 cxgbe_snd_tag_modify(struct m_snd_tag *mst, 2465 union if_snd_tag_modify_params *params) 2466 { 2467 2468 switch (mst->type) { 2469 #ifdef RATELIMIT 2470 case IF_SND_TAG_TYPE_RATE_LIMIT: 2471 return (cxgbe_rate_tag_modify(mst, params)); 2472 #endif 2473 default: 2474 return (EOPNOTSUPP); 2475 } 2476 } 2477 2478 static int 2479 cxgbe_snd_tag_query(struct m_snd_tag *mst, 2480 union if_snd_tag_query_params *params) 2481 { 2482 2483 switch (mst->type) { 2484 #ifdef RATELIMIT 2485 case IF_SND_TAG_TYPE_RATE_LIMIT: 2486 return (cxgbe_rate_tag_query(mst, params)); 2487 #endif 2488 default: 2489 return (EOPNOTSUPP); 2490 } 2491 } 2492 2493 static void 2494 cxgbe_snd_tag_free(struct m_snd_tag *mst) 2495 { 2496 2497 switch (mst->type) { 2498 #ifdef RATELIMIT 2499 case IF_SND_TAG_TYPE_RATE_LIMIT: 2500 cxgbe_rate_tag_free(mst); 2501 return; 2502 #endif 2503 #ifdef KERN_TLS 2504 case IF_SND_TAG_TYPE_TLS: 2505 cxgbe_tls_tag_free(mst); 2506 return; 2507 #endif 2508 default: 2509 panic("shouldn't get here"); 2510 } 2511 } 2512 #endif 2513 2514 /* 2515 * The kernel picks a media from the list we had provided but we still validate 2516 * the requeste. 2517 */ 2518 int 2519 cxgbe_media_change(struct ifnet *ifp) 2520 { 2521 struct vi_info *vi = ifp->if_softc; 2522 struct port_info *pi = vi->pi; 2523 struct ifmedia *ifm = &pi->media; 2524 struct link_config *lc = &pi->link_cfg; 2525 struct adapter *sc = pi->adapter; 2526 int rc; 2527 2528 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4mec"); 2529 if (rc != 0) 2530 return (rc); 2531 PORT_LOCK(pi); 2532 if (IFM_SUBTYPE(ifm->ifm_media) == IFM_AUTO) { 2533 /* ifconfig .. media autoselect */ 2534 if (!(lc->pcaps & FW_PORT_CAP32_ANEG)) { 2535 rc = ENOTSUP; /* AN not supported by transceiver */ 2536 goto done; 2537 } 2538 lc->requested_aneg = AUTONEG_ENABLE; 2539 lc->requested_speed = 0; 2540 lc->requested_fc |= PAUSE_AUTONEG; 2541 } else { 2542 lc->requested_aneg = AUTONEG_DISABLE; 2543 lc->requested_speed = 2544 ifmedia_baudrate(ifm->ifm_media) / 1000000; 2545 lc->requested_fc = 0; 2546 if (IFM_OPTIONS(ifm->ifm_media) & IFM_ETH_RXPAUSE) 2547 lc->requested_fc |= PAUSE_RX; 2548 if (IFM_OPTIONS(ifm->ifm_media) & IFM_ETH_TXPAUSE) 2549 lc->requested_fc |= PAUSE_TX; 2550 } 2551 if (pi->up_vis > 0) { 2552 fixup_link_config(pi); 2553 rc = apply_link_config(pi); 2554 } 2555 done: 2556 PORT_UNLOCK(pi); 2557 end_synchronized_op(sc, 0); 2558 return (rc); 2559 } 2560 2561 /* 2562 * Base media word (without ETHER, pause, link active, etc.) for the port at the 2563 * given speed. 2564 */ 2565 static int 2566 port_mword(struct port_info *pi, uint32_t speed) 2567 { 2568 2569 MPASS(speed & M_FW_PORT_CAP32_SPEED); 2570 MPASS(powerof2(speed)); 2571 2572 switch(pi->port_type) { 2573 case FW_PORT_TYPE_BT_SGMII: 2574 case FW_PORT_TYPE_BT_XFI: 2575 case FW_PORT_TYPE_BT_XAUI: 2576 /* BaseT */ 2577 switch (speed) { 2578 case FW_PORT_CAP32_SPEED_100M: 2579 return (IFM_100_T); 2580 case FW_PORT_CAP32_SPEED_1G: 2581 return (IFM_1000_T); 2582 case FW_PORT_CAP32_SPEED_10G: 2583 return (IFM_10G_T); 2584 } 2585 break; 2586 case FW_PORT_TYPE_KX4: 2587 if (speed == FW_PORT_CAP32_SPEED_10G) 2588 return (IFM_10G_KX4); 2589 break; 2590 case FW_PORT_TYPE_CX4: 2591 if (speed == FW_PORT_CAP32_SPEED_10G) 2592 return (IFM_10G_CX4); 2593 break; 2594 case FW_PORT_TYPE_KX: 2595 if (speed == FW_PORT_CAP32_SPEED_1G) 2596 return (IFM_1000_KX); 2597 break; 2598 case FW_PORT_TYPE_KR: 2599 case FW_PORT_TYPE_BP_AP: 2600 case FW_PORT_TYPE_BP4_AP: 2601 case FW_PORT_TYPE_BP40_BA: 2602 case FW_PORT_TYPE_KR4_100G: 2603 case FW_PORT_TYPE_KR_SFP28: 2604 case FW_PORT_TYPE_KR_XLAUI: 2605 switch (speed) { 2606 case FW_PORT_CAP32_SPEED_1G: 2607 return (IFM_1000_KX); 2608 case FW_PORT_CAP32_SPEED_10G: 2609 return (IFM_10G_KR); 2610 case FW_PORT_CAP32_SPEED_25G: 2611 return (IFM_25G_KR); 2612 case FW_PORT_CAP32_SPEED_40G: 2613 return (IFM_40G_KR4); 2614 case FW_PORT_CAP32_SPEED_50G: 2615 return (IFM_50G_KR2); 2616 case FW_PORT_CAP32_SPEED_100G: 2617 return (IFM_100G_KR4); 2618 } 2619 break; 2620 case FW_PORT_TYPE_FIBER_XFI: 2621 case FW_PORT_TYPE_FIBER_XAUI: 2622 case FW_PORT_TYPE_SFP: 2623 case FW_PORT_TYPE_QSFP_10G: 2624 case FW_PORT_TYPE_QSA: 2625 case FW_PORT_TYPE_QSFP: 2626 case FW_PORT_TYPE_CR4_QSFP: 2627 case FW_PORT_TYPE_CR_QSFP: 2628 case FW_PORT_TYPE_CR2_QSFP: 2629 case FW_PORT_TYPE_SFP28: 2630 /* Pluggable transceiver */ 2631 switch (pi->mod_type) { 2632 case FW_PORT_MOD_TYPE_LR: 2633 switch (speed) { 2634 case FW_PORT_CAP32_SPEED_1G: 2635 return (IFM_1000_LX); 2636 case FW_PORT_CAP32_SPEED_10G: 2637 return (IFM_10G_LR); 2638 case FW_PORT_CAP32_SPEED_25G: 2639 return (IFM_25G_LR); 2640 case FW_PORT_CAP32_SPEED_40G: 2641 return (IFM_40G_LR4); 2642 case FW_PORT_CAP32_SPEED_50G: 2643 return (IFM_50G_LR2); 2644 case FW_PORT_CAP32_SPEED_100G: 2645 return (IFM_100G_LR4); 2646 } 2647 break; 2648 case FW_PORT_MOD_TYPE_SR: 2649 switch (speed) { 2650 case FW_PORT_CAP32_SPEED_1G: 2651 return (IFM_1000_SX); 2652 case FW_PORT_CAP32_SPEED_10G: 2653 return (IFM_10G_SR); 2654 case FW_PORT_CAP32_SPEED_25G: 2655 return (IFM_25G_SR); 2656 case FW_PORT_CAP32_SPEED_40G: 2657 return (IFM_40G_SR4); 2658 case FW_PORT_CAP32_SPEED_50G: 2659 return (IFM_50G_SR2); 2660 case FW_PORT_CAP32_SPEED_100G: 2661 return (IFM_100G_SR4); 2662 } 2663 break; 2664 case FW_PORT_MOD_TYPE_ER: 2665 if (speed == FW_PORT_CAP32_SPEED_10G) 2666 return (IFM_10G_ER); 2667 break; 2668 case FW_PORT_MOD_TYPE_TWINAX_PASSIVE: 2669 case FW_PORT_MOD_TYPE_TWINAX_ACTIVE: 2670 switch (speed) { 2671 case FW_PORT_CAP32_SPEED_1G: 2672 return (IFM_1000_CX); 2673 case FW_PORT_CAP32_SPEED_10G: 2674 return (IFM_10G_TWINAX); 2675 case FW_PORT_CAP32_SPEED_25G: 2676 return (IFM_25G_CR); 2677 case FW_PORT_CAP32_SPEED_40G: 2678 return (IFM_40G_CR4); 2679 case FW_PORT_CAP32_SPEED_50G: 2680 return (IFM_50G_CR2); 2681 case FW_PORT_CAP32_SPEED_100G: 2682 return (IFM_100G_CR4); 2683 } 2684 break; 2685 case FW_PORT_MOD_TYPE_LRM: 2686 if (speed == FW_PORT_CAP32_SPEED_10G) 2687 return (IFM_10G_LRM); 2688 break; 2689 case FW_PORT_MOD_TYPE_NA: 2690 MPASS(0); /* Not pluggable? */ 2691 /* fall throough */ 2692 case FW_PORT_MOD_TYPE_ERROR: 2693 case FW_PORT_MOD_TYPE_UNKNOWN: 2694 case FW_PORT_MOD_TYPE_NOTSUPPORTED: 2695 break; 2696 case FW_PORT_MOD_TYPE_NONE: 2697 return (IFM_NONE); 2698 } 2699 break; 2700 case FW_PORT_TYPE_NONE: 2701 return (IFM_NONE); 2702 } 2703 2704 return (IFM_UNKNOWN); 2705 } 2706 2707 void 2708 cxgbe_media_status(struct ifnet *ifp, struct ifmediareq *ifmr) 2709 { 2710 struct vi_info *vi = ifp->if_softc; 2711 struct port_info *pi = vi->pi; 2712 struct adapter *sc = pi->adapter; 2713 struct link_config *lc = &pi->link_cfg; 2714 2715 if (begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4med") != 0) 2716 return; 2717 PORT_LOCK(pi); 2718 2719 if (pi->up_vis == 0) { 2720 /* 2721 * If all the interfaces are administratively down the firmware 2722 * does not report transceiver changes. Refresh port info here 2723 * so that ifconfig displays accurate ifmedia at all times. 2724 * This is the only reason we have a synchronized op in this 2725 * function. Just PORT_LOCK would have been enough otherwise. 2726 */ 2727 t4_update_port_info(pi); 2728 build_medialist(pi); 2729 } 2730 2731 /* ifm_status */ 2732 ifmr->ifm_status = IFM_AVALID; 2733 if (lc->link_ok == false) 2734 goto done; 2735 ifmr->ifm_status |= IFM_ACTIVE; 2736 2737 /* ifm_active */ 2738 ifmr->ifm_active = IFM_ETHER | IFM_FDX; 2739 ifmr->ifm_active &= ~(IFM_ETH_TXPAUSE | IFM_ETH_RXPAUSE); 2740 if (lc->fc & PAUSE_RX) 2741 ifmr->ifm_active |= IFM_ETH_RXPAUSE; 2742 if (lc->fc & PAUSE_TX) 2743 ifmr->ifm_active |= IFM_ETH_TXPAUSE; 2744 ifmr->ifm_active |= port_mword(pi, speed_to_fwcap(lc->speed)); 2745 done: 2746 PORT_UNLOCK(pi); 2747 end_synchronized_op(sc, 0); 2748 } 2749 2750 static int 2751 vcxgbe_probe(device_t dev) 2752 { 2753 char buf[128]; 2754 struct vi_info *vi = device_get_softc(dev); 2755 2756 snprintf(buf, sizeof(buf), "port %d vi %td", vi->pi->port_id, 2757 vi - vi->pi->vi); 2758 device_set_desc_copy(dev, buf); 2759 2760 return (BUS_PROBE_DEFAULT); 2761 } 2762 2763 static int 2764 alloc_extra_vi(struct adapter *sc, struct port_info *pi, struct vi_info *vi) 2765 { 2766 int func, index, rc; 2767 uint32_t param, val; 2768 2769 ASSERT_SYNCHRONIZED_OP(sc); 2770 2771 index = vi - pi->vi; 2772 MPASS(index > 0); /* This function deals with _extra_ VIs only */ 2773 KASSERT(index < nitems(vi_mac_funcs), 2774 ("%s: VI %s doesn't have a MAC func", __func__, 2775 device_get_nameunit(vi->dev))); 2776 func = vi_mac_funcs[index]; 2777 rc = t4_alloc_vi_func(sc, sc->mbox, pi->tx_chan, sc->pf, 0, 1, 2778 vi->hw_addr, &vi->rss_size, &vi->vfvld, &vi->vin, func, 0); 2779 if (rc < 0) { 2780 device_printf(vi->dev, "failed to allocate virtual interface %d" 2781 "for port %d: %d\n", index, pi->port_id, -rc); 2782 return (-rc); 2783 } 2784 vi->viid = rc; 2785 2786 if (vi->rss_size == 1) { 2787 /* 2788 * This VI didn't get a slice of the RSS table. Reduce the 2789 * number of VIs being created (hw.cxgbe.num_vis) or modify the 2790 * configuration file (nvi, rssnvi for this PF) if this is a 2791 * problem. 2792 */ 2793 device_printf(vi->dev, "RSS table not available.\n"); 2794 vi->rss_base = 0xffff; 2795 2796 return (0); 2797 } 2798 2799 param = V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | 2800 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_RSSINFO) | 2801 V_FW_PARAMS_PARAM_YZ(vi->viid); 2802 rc = t4_query_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 2803 if (rc) 2804 vi->rss_base = 0xffff; 2805 else { 2806 MPASS((val >> 16) == vi->rss_size); 2807 vi->rss_base = val & 0xffff; 2808 } 2809 2810 return (0); 2811 } 2812 2813 static int 2814 vcxgbe_attach(device_t dev) 2815 { 2816 struct vi_info *vi; 2817 struct port_info *pi; 2818 struct adapter *sc; 2819 int rc; 2820 2821 vi = device_get_softc(dev); 2822 pi = vi->pi; 2823 sc = pi->adapter; 2824 2825 rc = begin_synchronized_op(sc, vi, SLEEP_OK | INTR_OK, "t4via"); 2826 if (rc) 2827 return (rc); 2828 rc = alloc_extra_vi(sc, pi, vi); 2829 end_synchronized_op(sc, 0); 2830 if (rc) 2831 return (rc); 2832 2833 rc = cxgbe_vi_attach(dev, vi); 2834 if (rc) { 2835 t4_free_vi(sc, sc->mbox, sc->pf, 0, vi->viid); 2836 return (rc); 2837 } 2838 return (0); 2839 } 2840 2841 static int 2842 vcxgbe_detach(device_t dev) 2843 { 2844 struct vi_info *vi; 2845 struct adapter *sc; 2846 2847 vi = device_get_softc(dev); 2848 sc = vi->adapter; 2849 2850 doom_vi(sc, vi); 2851 2852 cxgbe_vi_detach(vi); 2853 t4_free_vi(sc, sc->mbox, sc->pf, 0, vi->viid); 2854 2855 end_synchronized_op(sc, 0); 2856 2857 return (0); 2858 } 2859 2860 static struct callout fatal_callout; 2861 2862 static void 2863 delayed_panic(void *arg) 2864 { 2865 struct adapter *sc = arg; 2866 2867 panic("%s: panic on fatal error", device_get_nameunit(sc->dev)); 2868 } 2869 2870 void 2871 t4_fatal_err(struct adapter *sc, bool fw_error) 2872 { 2873 2874 t4_shutdown_adapter(sc); 2875 log(LOG_ALERT, "%s: encountered fatal error, adapter stopped.\n", 2876 device_get_nameunit(sc->dev)); 2877 if (fw_error) { 2878 ASSERT_SYNCHRONIZED_OP(sc); 2879 sc->flags |= ADAP_ERR; 2880 } else { 2881 ADAPTER_LOCK(sc); 2882 sc->flags |= ADAP_ERR; 2883 ADAPTER_UNLOCK(sc); 2884 } 2885 #ifdef TCP_OFFLOAD 2886 taskqueue_enqueue(taskqueue_thread, &sc->async_event_task); 2887 #endif 2888 2889 if (t4_panic_on_fatal_err) { 2890 log(LOG_ALERT, "%s: panic on fatal error after 30s", 2891 device_get_nameunit(sc->dev)); 2892 callout_reset(&fatal_callout, hz * 30, delayed_panic, sc); 2893 } 2894 } 2895 2896 void 2897 t4_add_adapter(struct adapter *sc) 2898 { 2899 sx_xlock(&t4_list_lock); 2900 SLIST_INSERT_HEAD(&t4_list, sc, link); 2901 sx_xunlock(&t4_list_lock); 2902 } 2903 2904 int 2905 t4_map_bars_0_and_4(struct adapter *sc) 2906 { 2907 sc->regs_rid = PCIR_BAR(0); 2908 sc->regs_res = bus_alloc_resource_any(sc->dev, SYS_RES_MEMORY, 2909 &sc->regs_rid, RF_ACTIVE); 2910 if (sc->regs_res == NULL) { 2911 device_printf(sc->dev, "cannot map registers.\n"); 2912 return (ENXIO); 2913 } 2914 sc->bt = rman_get_bustag(sc->regs_res); 2915 sc->bh = rman_get_bushandle(sc->regs_res); 2916 sc->mmio_len = rman_get_size(sc->regs_res); 2917 setbit(&sc->doorbells, DOORBELL_KDB); 2918 2919 sc->msix_rid = PCIR_BAR(4); 2920 sc->msix_res = bus_alloc_resource_any(sc->dev, SYS_RES_MEMORY, 2921 &sc->msix_rid, RF_ACTIVE); 2922 if (sc->msix_res == NULL) { 2923 device_printf(sc->dev, "cannot map MSI-X BAR.\n"); 2924 return (ENXIO); 2925 } 2926 2927 return (0); 2928 } 2929 2930 int 2931 t4_map_bar_2(struct adapter *sc) 2932 { 2933 2934 /* 2935 * T4: only iWARP driver uses the userspace doorbells. There is no need 2936 * to map it if RDMA is disabled. 2937 */ 2938 if (is_t4(sc) && sc->rdmacaps == 0) 2939 return (0); 2940 2941 sc->udbs_rid = PCIR_BAR(2); 2942 sc->udbs_res = bus_alloc_resource_any(sc->dev, SYS_RES_MEMORY, 2943 &sc->udbs_rid, RF_ACTIVE); 2944 if (sc->udbs_res == NULL) { 2945 device_printf(sc->dev, "cannot map doorbell BAR.\n"); 2946 return (ENXIO); 2947 } 2948 sc->udbs_base = rman_get_virtual(sc->udbs_res); 2949 2950 if (chip_id(sc) >= CHELSIO_T5) { 2951 setbit(&sc->doorbells, DOORBELL_UDB); 2952 #if defined(__i386__) || defined(__amd64__) 2953 if (t5_write_combine) { 2954 int rc, mode; 2955 2956 /* 2957 * Enable write combining on BAR2. This is the 2958 * userspace doorbell BAR and is split into 128B 2959 * (UDBS_SEG_SIZE) doorbell regions, each associated 2960 * with an egress queue. The first 64B has the doorbell 2961 * and the second 64B can be used to submit a tx work 2962 * request with an implicit doorbell. 2963 */ 2964 2965 rc = pmap_change_attr((vm_offset_t)sc->udbs_base, 2966 rman_get_size(sc->udbs_res), PAT_WRITE_COMBINING); 2967 if (rc == 0) { 2968 clrbit(&sc->doorbells, DOORBELL_UDB); 2969 setbit(&sc->doorbells, DOORBELL_WCWR); 2970 setbit(&sc->doorbells, DOORBELL_UDBWC); 2971 } else { 2972 device_printf(sc->dev, 2973 "couldn't enable write combining: %d\n", 2974 rc); 2975 } 2976 2977 mode = is_t5(sc) ? V_STATMODE(0) : V_T6_STATMODE(0); 2978 t4_write_reg(sc, A_SGE_STAT_CFG, 2979 V_STATSOURCE_T5(7) | mode); 2980 } 2981 #endif 2982 } 2983 sc->iwt.wc_en = isset(&sc->doorbells, DOORBELL_UDBWC) ? 1 : 0; 2984 2985 return (0); 2986 } 2987 2988 struct memwin_init { 2989 uint32_t base; 2990 uint32_t aperture; 2991 }; 2992 2993 static const struct memwin_init t4_memwin[NUM_MEMWIN] = { 2994 { MEMWIN0_BASE, MEMWIN0_APERTURE }, 2995 { MEMWIN1_BASE, MEMWIN1_APERTURE }, 2996 { MEMWIN2_BASE_T4, MEMWIN2_APERTURE_T4 } 2997 }; 2998 2999 static const struct memwin_init t5_memwin[NUM_MEMWIN] = { 3000 { MEMWIN0_BASE, MEMWIN0_APERTURE }, 3001 { MEMWIN1_BASE, MEMWIN1_APERTURE }, 3002 { MEMWIN2_BASE_T5, MEMWIN2_APERTURE_T5 }, 3003 }; 3004 3005 static void 3006 setup_memwin(struct adapter *sc) 3007 { 3008 const struct memwin_init *mw_init; 3009 struct memwin *mw; 3010 int i; 3011 uint32_t bar0; 3012 3013 if (is_t4(sc)) { 3014 /* 3015 * Read low 32b of bar0 indirectly via the hardware backdoor 3016 * mechanism. Works from within PCI passthrough environments 3017 * too, where rman_get_start() can return a different value. We 3018 * need to program the T4 memory window decoders with the actual 3019 * addresses that will be coming across the PCIe link. 3020 */ 3021 bar0 = t4_hw_pci_read_cfg4(sc, PCIR_BAR(0)); 3022 bar0 &= (uint32_t) PCIM_BAR_MEM_BASE; 3023 3024 mw_init = &t4_memwin[0]; 3025 } else { 3026 /* T5+ use the relative offset inside the PCIe BAR */ 3027 bar0 = 0; 3028 3029 mw_init = &t5_memwin[0]; 3030 } 3031 3032 for (i = 0, mw = &sc->memwin[0]; i < NUM_MEMWIN; i++, mw_init++, mw++) { 3033 rw_init(&mw->mw_lock, "memory window access"); 3034 mw->mw_base = mw_init->base; 3035 mw->mw_aperture = mw_init->aperture; 3036 mw->mw_curpos = 0; 3037 t4_write_reg(sc, 3038 PCIE_MEM_ACCESS_REG(A_PCIE_MEM_ACCESS_BASE_WIN, i), 3039 (mw->mw_base + bar0) | V_BIR(0) | 3040 V_WINDOW(ilog2(mw->mw_aperture) - 10)); 3041 rw_wlock(&mw->mw_lock); 3042 position_memwin(sc, i, 0); 3043 rw_wunlock(&mw->mw_lock); 3044 } 3045 3046 /* flush */ 3047 t4_read_reg(sc, PCIE_MEM_ACCESS_REG(A_PCIE_MEM_ACCESS_BASE_WIN, 2)); 3048 } 3049 3050 /* 3051 * Positions the memory window at the given address in the card's address space. 3052 * There are some alignment requirements and the actual position may be at an 3053 * address prior to the requested address. mw->mw_curpos always has the actual 3054 * position of the window. 3055 */ 3056 static void 3057 position_memwin(struct adapter *sc, int idx, uint32_t addr) 3058 { 3059 struct memwin *mw; 3060 uint32_t pf; 3061 uint32_t reg; 3062 3063 MPASS(idx >= 0 && idx < NUM_MEMWIN); 3064 mw = &sc->memwin[idx]; 3065 rw_assert(&mw->mw_lock, RA_WLOCKED); 3066 3067 if (is_t4(sc)) { 3068 pf = 0; 3069 mw->mw_curpos = addr & ~0xf; /* start must be 16B aligned */ 3070 } else { 3071 pf = V_PFNUM(sc->pf); 3072 mw->mw_curpos = addr & ~0x7f; /* start must be 128B aligned */ 3073 } 3074 reg = PCIE_MEM_ACCESS_REG(A_PCIE_MEM_ACCESS_OFFSET, idx); 3075 t4_write_reg(sc, reg, mw->mw_curpos | pf); 3076 t4_read_reg(sc, reg); /* flush */ 3077 } 3078 3079 int 3080 rw_via_memwin(struct adapter *sc, int idx, uint32_t addr, uint32_t *val, 3081 int len, int rw) 3082 { 3083 struct memwin *mw; 3084 uint32_t mw_end, v; 3085 3086 MPASS(idx >= 0 && idx < NUM_MEMWIN); 3087 3088 /* Memory can only be accessed in naturally aligned 4 byte units */ 3089 if (addr & 3 || len & 3 || len <= 0) 3090 return (EINVAL); 3091 3092 mw = &sc->memwin[idx]; 3093 while (len > 0) { 3094 rw_rlock(&mw->mw_lock); 3095 mw_end = mw->mw_curpos + mw->mw_aperture; 3096 if (addr >= mw_end || addr < mw->mw_curpos) { 3097 /* Will need to reposition the window */ 3098 if (!rw_try_upgrade(&mw->mw_lock)) { 3099 rw_runlock(&mw->mw_lock); 3100 rw_wlock(&mw->mw_lock); 3101 } 3102 rw_assert(&mw->mw_lock, RA_WLOCKED); 3103 position_memwin(sc, idx, addr); 3104 rw_downgrade(&mw->mw_lock); 3105 mw_end = mw->mw_curpos + mw->mw_aperture; 3106 } 3107 rw_assert(&mw->mw_lock, RA_RLOCKED); 3108 while (addr < mw_end && len > 0) { 3109 if (rw == 0) { 3110 v = t4_read_reg(sc, mw->mw_base + addr - 3111 mw->mw_curpos); 3112 *val++ = le32toh(v); 3113 } else { 3114 v = *val++; 3115 t4_write_reg(sc, mw->mw_base + addr - 3116 mw->mw_curpos, htole32(v)); 3117 } 3118 addr += 4; 3119 len -= 4; 3120 } 3121 rw_runlock(&mw->mw_lock); 3122 } 3123 3124 return (0); 3125 } 3126 3127 static void 3128 t4_init_atid_table(struct adapter *sc) 3129 { 3130 struct tid_info *t; 3131 int i; 3132 3133 t = &sc->tids; 3134 if (t->natids == 0) 3135 return; 3136 3137 MPASS(t->atid_tab == NULL); 3138 3139 t->atid_tab = malloc(t->natids * sizeof(*t->atid_tab), M_CXGBE, 3140 M_ZERO | M_WAITOK); 3141 mtx_init(&t->atid_lock, "atid lock", NULL, MTX_DEF); 3142 t->afree = t->atid_tab; 3143 t->atids_in_use = 0; 3144 for (i = 1; i < t->natids; i++) 3145 t->atid_tab[i - 1].next = &t->atid_tab[i]; 3146 t->atid_tab[t->natids - 1].next = NULL; 3147 } 3148 3149 static void 3150 t4_free_atid_table(struct adapter *sc) 3151 { 3152 struct tid_info *t; 3153 3154 t = &sc->tids; 3155 3156 KASSERT(t->atids_in_use == 0, 3157 ("%s: %d atids still in use.", __func__, t->atids_in_use)); 3158 3159 if (mtx_initialized(&t->atid_lock)) 3160 mtx_destroy(&t->atid_lock); 3161 free(t->atid_tab, M_CXGBE); 3162 t->atid_tab = NULL; 3163 } 3164 3165 int 3166 alloc_atid(struct adapter *sc, void *ctx) 3167 { 3168 struct tid_info *t = &sc->tids; 3169 int atid = -1; 3170 3171 mtx_lock(&t->atid_lock); 3172 if (t->afree) { 3173 union aopen_entry *p = t->afree; 3174 3175 atid = p - t->atid_tab; 3176 MPASS(atid <= M_TID_TID); 3177 t->afree = p->next; 3178 p->data = ctx; 3179 t->atids_in_use++; 3180 } 3181 mtx_unlock(&t->atid_lock); 3182 return (atid); 3183 } 3184 3185 void * 3186 lookup_atid(struct adapter *sc, int atid) 3187 { 3188 struct tid_info *t = &sc->tids; 3189 3190 return (t->atid_tab[atid].data); 3191 } 3192 3193 void 3194 free_atid(struct adapter *sc, int atid) 3195 { 3196 struct tid_info *t = &sc->tids; 3197 union aopen_entry *p = &t->atid_tab[atid]; 3198 3199 mtx_lock(&t->atid_lock); 3200 p->next = t->afree; 3201 t->afree = p; 3202 t->atids_in_use--; 3203 mtx_unlock(&t->atid_lock); 3204 } 3205 3206 static void 3207 queue_tid_release(struct adapter *sc, int tid) 3208 { 3209 3210 CXGBE_UNIMPLEMENTED("deferred tid release"); 3211 } 3212 3213 void 3214 release_tid(struct adapter *sc, int tid, struct sge_wrq *ctrlq) 3215 { 3216 struct wrqe *wr; 3217 struct cpl_tid_release *req; 3218 3219 wr = alloc_wrqe(sizeof(*req), ctrlq); 3220 if (wr == NULL) { 3221 queue_tid_release(sc, tid); /* defer */ 3222 return; 3223 } 3224 req = wrtod(wr); 3225 3226 INIT_TP_WR_MIT_CPL(req, CPL_TID_RELEASE, tid); 3227 3228 t4_wrq_tx(sc, wr); 3229 } 3230 3231 static int 3232 t4_range_cmp(const void *a, const void *b) 3233 { 3234 return ((const struct t4_range *)a)->start - 3235 ((const struct t4_range *)b)->start; 3236 } 3237 3238 /* 3239 * Verify that the memory range specified by the addr/len pair is valid within 3240 * the card's address space. 3241 */ 3242 static int 3243 validate_mem_range(struct adapter *sc, uint32_t addr, uint32_t len) 3244 { 3245 struct t4_range mem_ranges[4], *r, *next; 3246 uint32_t em, addr_len; 3247 int i, n, remaining; 3248 3249 /* Memory can only be accessed in naturally aligned 4 byte units */ 3250 if (addr & 3 || len & 3 || len == 0) 3251 return (EINVAL); 3252 3253 /* Enabled memories */ 3254 em = t4_read_reg(sc, A_MA_TARGET_MEM_ENABLE); 3255 3256 r = &mem_ranges[0]; 3257 n = 0; 3258 bzero(r, sizeof(mem_ranges)); 3259 if (em & F_EDRAM0_ENABLE) { 3260 addr_len = t4_read_reg(sc, A_MA_EDRAM0_BAR); 3261 r->size = G_EDRAM0_SIZE(addr_len) << 20; 3262 if (r->size > 0) { 3263 r->start = G_EDRAM0_BASE(addr_len) << 20; 3264 if (addr >= r->start && 3265 addr + len <= r->start + r->size) 3266 return (0); 3267 r++; 3268 n++; 3269 } 3270 } 3271 if (em & F_EDRAM1_ENABLE) { 3272 addr_len = t4_read_reg(sc, A_MA_EDRAM1_BAR); 3273 r->size = G_EDRAM1_SIZE(addr_len) << 20; 3274 if (r->size > 0) { 3275 r->start = G_EDRAM1_BASE(addr_len) << 20; 3276 if (addr >= r->start && 3277 addr + len <= r->start + r->size) 3278 return (0); 3279 r++; 3280 n++; 3281 } 3282 } 3283 if (em & F_EXT_MEM_ENABLE) { 3284 addr_len = t4_read_reg(sc, A_MA_EXT_MEMORY_BAR); 3285 r->size = G_EXT_MEM_SIZE(addr_len) << 20; 3286 if (r->size > 0) { 3287 r->start = G_EXT_MEM_BASE(addr_len) << 20; 3288 if (addr >= r->start && 3289 addr + len <= r->start + r->size) 3290 return (0); 3291 r++; 3292 n++; 3293 } 3294 } 3295 if (is_t5(sc) && em & F_EXT_MEM1_ENABLE) { 3296 addr_len = t4_read_reg(sc, A_MA_EXT_MEMORY1_BAR); 3297 r->size = G_EXT_MEM1_SIZE(addr_len) << 20; 3298 if (r->size > 0) { 3299 r->start = G_EXT_MEM1_BASE(addr_len) << 20; 3300 if (addr >= r->start && 3301 addr + len <= r->start + r->size) 3302 return (0); 3303 r++; 3304 n++; 3305 } 3306 } 3307 MPASS(n <= nitems(mem_ranges)); 3308 3309 if (n > 1) { 3310 /* Sort and merge the ranges. */ 3311 qsort(mem_ranges, n, sizeof(struct t4_range), t4_range_cmp); 3312 3313 /* Start from index 0 and examine the next n - 1 entries. */ 3314 r = &mem_ranges[0]; 3315 for (remaining = n - 1; remaining > 0; remaining--, r++) { 3316 3317 MPASS(r->size > 0); /* r is a valid entry. */ 3318 next = r + 1; 3319 MPASS(next->size > 0); /* and so is the next one. */ 3320 3321 while (r->start + r->size >= next->start) { 3322 /* Merge the next one into the current entry. */ 3323 r->size = max(r->start + r->size, 3324 next->start + next->size) - r->start; 3325 n--; /* One fewer entry in total. */ 3326 if (--remaining == 0) 3327 goto done; /* short circuit */ 3328 next++; 3329 } 3330 if (next != r + 1) { 3331 /* 3332 * Some entries were merged into r and next 3333 * points to the first valid entry that couldn't 3334 * be merged. 3335 */ 3336 MPASS(next->size > 0); /* must be valid */ 3337 memcpy(r + 1, next, remaining * sizeof(*r)); 3338 #ifdef INVARIANTS 3339 /* 3340 * This so that the foo->size assertion in the 3341 * next iteration of the loop do the right 3342 * thing for entries that were pulled up and are 3343 * no longer valid. 3344 */ 3345 MPASS(n < nitems(mem_ranges)); 3346 bzero(&mem_ranges[n], (nitems(mem_ranges) - n) * 3347 sizeof(struct t4_range)); 3348 #endif 3349 } 3350 } 3351 done: 3352 /* Done merging the ranges. */ 3353 MPASS(n > 0); 3354 r = &mem_ranges[0]; 3355 for (i = 0; i < n; i++, r++) { 3356 if (addr >= r->start && 3357 addr + len <= r->start + r->size) 3358 return (0); 3359 } 3360 } 3361 3362 return (EFAULT); 3363 } 3364 3365 static int 3366 fwmtype_to_hwmtype(int mtype) 3367 { 3368 3369 switch (mtype) { 3370 case FW_MEMTYPE_EDC0: 3371 return (MEM_EDC0); 3372 case FW_MEMTYPE_EDC1: 3373 return (MEM_EDC1); 3374 case FW_MEMTYPE_EXTMEM: 3375 return (MEM_MC0); 3376 case FW_MEMTYPE_EXTMEM1: 3377 return (MEM_MC1); 3378 default: 3379 panic("%s: cannot translate fw mtype %d.", __func__, mtype); 3380 } 3381 } 3382 3383 /* 3384 * Verify that the memory range specified by the memtype/offset/len pair is 3385 * valid and lies entirely within the memtype specified. The global address of 3386 * the start of the range is returned in addr. 3387 */ 3388 static int 3389 validate_mt_off_len(struct adapter *sc, int mtype, uint32_t off, uint32_t len, 3390 uint32_t *addr) 3391 { 3392 uint32_t em, addr_len, maddr; 3393 3394 /* Memory can only be accessed in naturally aligned 4 byte units */ 3395 if (off & 3 || len & 3 || len == 0) 3396 return (EINVAL); 3397 3398 em = t4_read_reg(sc, A_MA_TARGET_MEM_ENABLE); 3399 switch (fwmtype_to_hwmtype(mtype)) { 3400 case MEM_EDC0: 3401 if (!(em & F_EDRAM0_ENABLE)) 3402 return (EINVAL); 3403 addr_len = t4_read_reg(sc, A_MA_EDRAM0_BAR); 3404 maddr = G_EDRAM0_BASE(addr_len) << 20; 3405 break; 3406 case MEM_EDC1: 3407 if (!(em & F_EDRAM1_ENABLE)) 3408 return (EINVAL); 3409 addr_len = t4_read_reg(sc, A_MA_EDRAM1_BAR); 3410 maddr = G_EDRAM1_BASE(addr_len) << 20; 3411 break; 3412 case MEM_MC: 3413 if (!(em & F_EXT_MEM_ENABLE)) 3414 return (EINVAL); 3415 addr_len = t4_read_reg(sc, A_MA_EXT_MEMORY_BAR); 3416 maddr = G_EXT_MEM_BASE(addr_len) << 20; 3417 break; 3418 case MEM_MC1: 3419 if (!is_t5(sc) || !(em & F_EXT_MEM1_ENABLE)) 3420 return (EINVAL); 3421 addr_len = t4_read_reg(sc, A_MA_EXT_MEMORY1_BAR); 3422 maddr = G_EXT_MEM1_BASE(addr_len) << 20; 3423 break; 3424 default: 3425 return (EINVAL); 3426 } 3427 3428 *addr = maddr + off; /* global address */ 3429 return (validate_mem_range(sc, *addr, len)); 3430 } 3431 3432 static int 3433 fixup_devlog_params(struct adapter *sc) 3434 { 3435 struct devlog_params *dparams = &sc->params.devlog; 3436 int rc; 3437 3438 rc = validate_mt_off_len(sc, dparams->memtype, dparams->start, 3439 dparams->size, &dparams->addr); 3440 3441 return (rc); 3442 } 3443 3444 static void 3445 update_nirq(struct intrs_and_queues *iaq, int nports) 3446 { 3447 3448 iaq->nirq = T4_EXTRA_INTR; 3449 iaq->nirq += nports * max(iaq->nrxq, iaq->nnmrxq); 3450 iaq->nirq += nports * iaq->nofldrxq; 3451 iaq->nirq += nports * (iaq->num_vis - 1) * 3452 max(iaq->nrxq_vi, iaq->nnmrxq_vi); 3453 iaq->nirq += nports * (iaq->num_vis - 1) * iaq->nofldrxq_vi; 3454 } 3455 3456 /* 3457 * Adjust requirements to fit the number of interrupts available. 3458 */ 3459 static void 3460 calculate_iaq(struct adapter *sc, struct intrs_and_queues *iaq, int itype, 3461 int navail) 3462 { 3463 int old_nirq; 3464 const int nports = sc->params.nports; 3465 3466 MPASS(nports > 0); 3467 MPASS(navail > 0); 3468 3469 bzero(iaq, sizeof(*iaq)); 3470 iaq->intr_type = itype; 3471 iaq->num_vis = t4_num_vis; 3472 iaq->ntxq = t4_ntxq; 3473 iaq->ntxq_vi = t4_ntxq_vi; 3474 iaq->nrxq = t4_nrxq; 3475 iaq->nrxq_vi = t4_nrxq_vi; 3476 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 3477 if (is_offload(sc) || is_ethoffload(sc)) { 3478 iaq->nofldtxq = t4_nofldtxq; 3479 iaq->nofldtxq_vi = t4_nofldtxq_vi; 3480 } 3481 #endif 3482 #ifdef TCP_OFFLOAD 3483 if (is_offload(sc)) { 3484 iaq->nofldrxq = t4_nofldrxq; 3485 iaq->nofldrxq_vi = t4_nofldrxq_vi; 3486 } 3487 #endif 3488 #ifdef DEV_NETMAP 3489 if (t4_native_netmap & NN_MAIN_VI) { 3490 iaq->nnmtxq = t4_nnmtxq; 3491 iaq->nnmrxq = t4_nnmrxq; 3492 } 3493 if (t4_native_netmap & NN_EXTRA_VI) { 3494 iaq->nnmtxq_vi = t4_nnmtxq_vi; 3495 iaq->nnmrxq_vi = t4_nnmrxq_vi; 3496 } 3497 #endif 3498 3499 update_nirq(iaq, nports); 3500 if (iaq->nirq <= navail && 3501 (itype != INTR_MSI || powerof2(iaq->nirq))) { 3502 /* 3503 * This is the normal case -- there are enough interrupts for 3504 * everything. 3505 */ 3506 goto done; 3507 } 3508 3509 /* 3510 * If extra VIs have been configured try reducing their count and see if 3511 * that works. 3512 */ 3513 while (iaq->num_vis > 1) { 3514 iaq->num_vis--; 3515 update_nirq(iaq, nports); 3516 if (iaq->nirq <= navail && 3517 (itype != INTR_MSI || powerof2(iaq->nirq))) { 3518 device_printf(sc->dev, "virtual interfaces per port " 3519 "reduced to %d from %d. nrxq=%u, nofldrxq=%u, " 3520 "nrxq_vi=%u nofldrxq_vi=%u, nnmrxq_vi=%u. " 3521 "itype %d, navail %u, nirq %d.\n", 3522 iaq->num_vis, t4_num_vis, iaq->nrxq, iaq->nofldrxq, 3523 iaq->nrxq_vi, iaq->nofldrxq_vi, iaq->nnmrxq_vi, 3524 itype, navail, iaq->nirq); 3525 goto done; 3526 } 3527 } 3528 3529 /* 3530 * Extra VIs will not be created. Log a message if they were requested. 3531 */ 3532 MPASS(iaq->num_vis == 1); 3533 iaq->ntxq_vi = iaq->nrxq_vi = 0; 3534 iaq->nofldtxq_vi = iaq->nofldrxq_vi = 0; 3535 iaq->nnmtxq_vi = iaq->nnmrxq_vi = 0; 3536 if (iaq->num_vis != t4_num_vis) { 3537 device_printf(sc->dev, "extra virtual interfaces disabled. " 3538 "nrxq=%u, nofldrxq=%u, nrxq_vi=%u nofldrxq_vi=%u, " 3539 "nnmrxq_vi=%u. itype %d, navail %u, nirq %d.\n", 3540 iaq->nrxq, iaq->nofldrxq, iaq->nrxq_vi, iaq->nofldrxq_vi, 3541 iaq->nnmrxq_vi, itype, navail, iaq->nirq); 3542 } 3543 3544 /* 3545 * Keep reducing the number of NIC rx queues to the next lower power of 3546 * 2 (for even RSS distribution) and halving the TOE rx queues and see 3547 * if that works. 3548 */ 3549 do { 3550 if (iaq->nrxq > 1) { 3551 do { 3552 iaq->nrxq--; 3553 } while (!powerof2(iaq->nrxq)); 3554 if (iaq->nnmrxq > iaq->nrxq) 3555 iaq->nnmrxq = iaq->nrxq; 3556 } 3557 if (iaq->nofldrxq > 1) 3558 iaq->nofldrxq >>= 1; 3559 3560 old_nirq = iaq->nirq; 3561 update_nirq(iaq, nports); 3562 if (iaq->nirq <= navail && 3563 (itype != INTR_MSI || powerof2(iaq->nirq))) { 3564 device_printf(sc->dev, "running with reduced number of " 3565 "rx queues because of shortage of interrupts. " 3566 "nrxq=%u, nofldrxq=%u. " 3567 "itype %d, navail %u, nirq %d.\n", iaq->nrxq, 3568 iaq->nofldrxq, itype, navail, iaq->nirq); 3569 goto done; 3570 } 3571 } while (old_nirq != iaq->nirq); 3572 3573 /* One interrupt for everything. Ugh. */ 3574 device_printf(sc->dev, "running with minimal number of queues. " 3575 "itype %d, navail %u.\n", itype, navail); 3576 iaq->nirq = 1; 3577 iaq->nrxq = 1; 3578 iaq->ntxq = 1; 3579 if (iaq->nofldrxq > 0) { 3580 iaq->nofldrxq = 1; 3581 iaq->nofldtxq = 1; 3582 } 3583 iaq->nnmtxq = 0; 3584 iaq->nnmrxq = 0; 3585 done: 3586 MPASS(iaq->num_vis > 0); 3587 if (iaq->num_vis > 1) { 3588 MPASS(iaq->nrxq_vi > 0); 3589 MPASS(iaq->ntxq_vi > 0); 3590 } 3591 MPASS(iaq->nirq > 0); 3592 MPASS(iaq->nrxq > 0); 3593 MPASS(iaq->ntxq > 0); 3594 if (itype == INTR_MSI) { 3595 MPASS(powerof2(iaq->nirq)); 3596 } 3597 } 3598 3599 static int 3600 cfg_itype_and_nqueues(struct adapter *sc, struct intrs_and_queues *iaq) 3601 { 3602 int rc, itype, navail, nalloc; 3603 3604 for (itype = INTR_MSIX; itype; itype >>= 1) { 3605 3606 if ((itype & t4_intr_types) == 0) 3607 continue; /* not allowed */ 3608 3609 if (itype == INTR_MSIX) 3610 navail = pci_msix_count(sc->dev); 3611 else if (itype == INTR_MSI) 3612 navail = pci_msi_count(sc->dev); 3613 else 3614 navail = 1; 3615 restart: 3616 if (navail == 0) 3617 continue; 3618 3619 calculate_iaq(sc, iaq, itype, navail); 3620 nalloc = iaq->nirq; 3621 rc = 0; 3622 if (itype == INTR_MSIX) 3623 rc = pci_alloc_msix(sc->dev, &nalloc); 3624 else if (itype == INTR_MSI) 3625 rc = pci_alloc_msi(sc->dev, &nalloc); 3626 3627 if (rc == 0 && nalloc > 0) { 3628 if (nalloc == iaq->nirq) 3629 return (0); 3630 3631 /* 3632 * Didn't get the number requested. Use whatever number 3633 * the kernel is willing to allocate. 3634 */ 3635 device_printf(sc->dev, "fewer vectors than requested, " 3636 "type=%d, req=%d, rcvd=%d; will downshift req.\n", 3637 itype, iaq->nirq, nalloc); 3638 pci_release_msi(sc->dev); 3639 navail = nalloc; 3640 goto restart; 3641 } 3642 3643 device_printf(sc->dev, 3644 "failed to allocate vectors:%d, type=%d, req=%d, rcvd=%d\n", 3645 itype, rc, iaq->nirq, nalloc); 3646 } 3647 3648 device_printf(sc->dev, 3649 "failed to find a usable interrupt type. " 3650 "allowed=%d, msi-x=%d, msi=%d, intx=1", t4_intr_types, 3651 pci_msix_count(sc->dev), pci_msi_count(sc->dev)); 3652 3653 return (ENXIO); 3654 } 3655 3656 #define FW_VERSION(chip) ( \ 3657 V_FW_HDR_FW_VER_MAJOR(chip##FW_VERSION_MAJOR) | \ 3658 V_FW_HDR_FW_VER_MINOR(chip##FW_VERSION_MINOR) | \ 3659 V_FW_HDR_FW_VER_MICRO(chip##FW_VERSION_MICRO) | \ 3660 V_FW_HDR_FW_VER_BUILD(chip##FW_VERSION_BUILD)) 3661 #define FW_INTFVER(chip, intf) (chip##FW_HDR_INTFVER_##intf) 3662 3663 /* Just enough of fw_hdr to cover all version info. */ 3664 struct fw_h { 3665 __u8 ver; 3666 __u8 chip; 3667 __be16 len512; 3668 __be32 fw_ver; 3669 __be32 tp_microcode_ver; 3670 __u8 intfver_nic; 3671 __u8 intfver_vnic; 3672 __u8 intfver_ofld; 3673 __u8 intfver_ri; 3674 __u8 intfver_iscsipdu; 3675 __u8 intfver_iscsi; 3676 __u8 intfver_fcoepdu; 3677 __u8 intfver_fcoe; 3678 }; 3679 /* Spot check a couple of fields. */ 3680 CTASSERT(offsetof(struct fw_h, fw_ver) == offsetof(struct fw_hdr, fw_ver)); 3681 CTASSERT(offsetof(struct fw_h, intfver_nic) == offsetof(struct fw_hdr, intfver_nic)); 3682 CTASSERT(offsetof(struct fw_h, intfver_fcoe) == offsetof(struct fw_hdr, intfver_fcoe)); 3683 3684 struct fw_info { 3685 uint8_t chip; 3686 char *kld_name; 3687 char *fw_mod_name; 3688 struct fw_h fw_h; 3689 } fw_info[] = { 3690 { 3691 .chip = CHELSIO_T4, 3692 .kld_name = "t4fw_cfg", 3693 .fw_mod_name = "t4fw", 3694 .fw_h = { 3695 .chip = FW_HDR_CHIP_T4, 3696 .fw_ver = htobe32(FW_VERSION(T4)), 3697 .intfver_nic = FW_INTFVER(T4, NIC), 3698 .intfver_vnic = FW_INTFVER(T4, VNIC), 3699 .intfver_ofld = FW_INTFVER(T4, OFLD), 3700 .intfver_ri = FW_INTFVER(T4, RI), 3701 .intfver_iscsipdu = FW_INTFVER(T4, ISCSIPDU), 3702 .intfver_iscsi = FW_INTFVER(T4, ISCSI), 3703 .intfver_fcoepdu = FW_INTFVER(T4, FCOEPDU), 3704 .intfver_fcoe = FW_INTFVER(T4, FCOE), 3705 }, 3706 }, { 3707 .chip = CHELSIO_T5, 3708 .kld_name = "t5fw_cfg", 3709 .fw_mod_name = "t5fw", 3710 .fw_h = { 3711 .chip = FW_HDR_CHIP_T5, 3712 .fw_ver = htobe32(FW_VERSION(T5)), 3713 .intfver_nic = FW_INTFVER(T5, NIC), 3714 .intfver_vnic = FW_INTFVER(T5, VNIC), 3715 .intfver_ofld = FW_INTFVER(T5, OFLD), 3716 .intfver_ri = FW_INTFVER(T5, RI), 3717 .intfver_iscsipdu = FW_INTFVER(T5, ISCSIPDU), 3718 .intfver_iscsi = FW_INTFVER(T5, ISCSI), 3719 .intfver_fcoepdu = FW_INTFVER(T5, FCOEPDU), 3720 .intfver_fcoe = FW_INTFVER(T5, FCOE), 3721 }, 3722 }, { 3723 .chip = CHELSIO_T6, 3724 .kld_name = "t6fw_cfg", 3725 .fw_mod_name = "t6fw", 3726 .fw_h = { 3727 .chip = FW_HDR_CHIP_T6, 3728 .fw_ver = htobe32(FW_VERSION(T6)), 3729 .intfver_nic = FW_INTFVER(T6, NIC), 3730 .intfver_vnic = FW_INTFVER(T6, VNIC), 3731 .intfver_ofld = FW_INTFVER(T6, OFLD), 3732 .intfver_ri = FW_INTFVER(T6, RI), 3733 .intfver_iscsipdu = FW_INTFVER(T6, ISCSIPDU), 3734 .intfver_iscsi = FW_INTFVER(T6, ISCSI), 3735 .intfver_fcoepdu = FW_INTFVER(T6, FCOEPDU), 3736 .intfver_fcoe = FW_INTFVER(T6, FCOE), 3737 }, 3738 } 3739 }; 3740 3741 static struct fw_info * 3742 find_fw_info(int chip) 3743 { 3744 int i; 3745 3746 for (i = 0; i < nitems(fw_info); i++) { 3747 if (fw_info[i].chip == chip) 3748 return (&fw_info[i]); 3749 } 3750 return (NULL); 3751 } 3752 3753 /* 3754 * Is the given firmware API compatible with the one the driver was compiled 3755 * with? 3756 */ 3757 static int 3758 fw_compatible(const struct fw_h *hdr1, const struct fw_h *hdr2) 3759 { 3760 3761 /* short circuit if it's the exact same firmware version */ 3762 if (hdr1->chip == hdr2->chip && hdr1->fw_ver == hdr2->fw_ver) 3763 return (1); 3764 3765 /* 3766 * XXX: Is this too conservative? Perhaps I should limit this to the 3767 * features that are supported in the driver. 3768 */ 3769 #define SAME_INTF(x) (hdr1->intfver_##x == hdr2->intfver_##x) 3770 if (hdr1->chip == hdr2->chip && SAME_INTF(nic) && SAME_INTF(vnic) && 3771 SAME_INTF(ofld) && SAME_INTF(ri) && SAME_INTF(iscsipdu) && 3772 SAME_INTF(iscsi) && SAME_INTF(fcoepdu) && SAME_INTF(fcoe)) 3773 return (1); 3774 #undef SAME_INTF 3775 3776 return (0); 3777 } 3778 3779 static int 3780 load_fw_module(struct adapter *sc, const struct firmware **dcfg, 3781 const struct firmware **fw) 3782 { 3783 struct fw_info *fw_info; 3784 3785 *dcfg = NULL; 3786 if (fw != NULL) 3787 *fw = NULL; 3788 3789 fw_info = find_fw_info(chip_id(sc)); 3790 if (fw_info == NULL) { 3791 device_printf(sc->dev, 3792 "unable to look up firmware information for chip %d.\n", 3793 chip_id(sc)); 3794 return (EINVAL); 3795 } 3796 3797 *dcfg = firmware_get(fw_info->kld_name); 3798 if (*dcfg != NULL) { 3799 if (fw != NULL) 3800 *fw = firmware_get(fw_info->fw_mod_name); 3801 return (0); 3802 } 3803 3804 return (ENOENT); 3805 } 3806 3807 static void 3808 unload_fw_module(struct adapter *sc, const struct firmware *dcfg, 3809 const struct firmware *fw) 3810 { 3811 3812 if (fw != NULL) 3813 firmware_put(fw, FIRMWARE_UNLOAD); 3814 if (dcfg != NULL) 3815 firmware_put(dcfg, FIRMWARE_UNLOAD); 3816 } 3817 3818 /* 3819 * Return values: 3820 * 0 means no firmware install attempted. 3821 * ERESTART means a firmware install was attempted and was successful. 3822 * +ve errno means a firmware install was attempted but failed. 3823 */ 3824 static int 3825 install_kld_firmware(struct adapter *sc, struct fw_h *card_fw, 3826 const struct fw_h *drv_fw, const char *reason, int *already) 3827 { 3828 const struct firmware *cfg, *fw; 3829 const uint32_t c = be32toh(card_fw->fw_ver); 3830 uint32_t d, k; 3831 int rc, fw_install; 3832 struct fw_h bundled_fw; 3833 bool load_attempted; 3834 3835 cfg = fw = NULL; 3836 load_attempted = false; 3837 fw_install = t4_fw_install < 0 ? -t4_fw_install : t4_fw_install; 3838 3839 memcpy(&bundled_fw, drv_fw, sizeof(bundled_fw)); 3840 if (t4_fw_install < 0) { 3841 rc = load_fw_module(sc, &cfg, &fw); 3842 if (rc != 0 || fw == NULL) { 3843 device_printf(sc->dev, 3844 "failed to load firmware module: %d. cfg %p, fw %p;" 3845 " will use compiled-in firmware version for" 3846 "hw.cxgbe.fw_install checks.\n", 3847 rc, cfg, fw); 3848 } else { 3849 memcpy(&bundled_fw, fw->data, sizeof(bundled_fw)); 3850 } 3851 load_attempted = true; 3852 } 3853 d = be32toh(bundled_fw.fw_ver); 3854 3855 if (reason != NULL) 3856 goto install; 3857 3858 if ((sc->flags & FW_OK) == 0) { 3859 3860 if (c == 0xffffffff) { 3861 reason = "missing"; 3862 goto install; 3863 } 3864 3865 rc = 0; 3866 goto done; 3867 } 3868 3869 if (!fw_compatible(card_fw, &bundled_fw)) { 3870 reason = "incompatible or unusable"; 3871 goto install; 3872 } 3873 3874 if (d > c) { 3875 reason = "older than the version bundled with this driver"; 3876 goto install; 3877 } 3878 3879 if (fw_install == 2 && d != c) { 3880 reason = "different than the version bundled with this driver"; 3881 goto install; 3882 } 3883 3884 /* No reason to do anything to the firmware already on the card. */ 3885 rc = 0; 3886 goto done; 3887 3888 install: 3889 rc = 0; 3890 if ((*already)++) 3891 goto done; 3892 3893 if (fw_install == 0) { 3894 device_printf(sc->dev, "firmware on card (%u.%u.%u.%u) is %s, " 3895 "but the driver is prohibited from installing a firmware " 3896 "on the card.\n", 3897 G_FW_HDR_FW_VER_MAJOR(c), G_FW_HDR_FW_VER_MINOR(c), 3898 G_FW_HDR_FW_VER_MICRO(c), G_FW_HDR_FW_VER_BUILD(c), reason); 3899 3900 goto done; 3901 } 3902 3903 /* 3904 * We'll attempt to install a firmware. Load the module first (if it 3905 * hasn't been loaded already). 3906 */ 3907 if (!load_attempted) { 3908 rc = load_fw_module(sc, &cfg, &fw); 3909 if (rc != 0 || fw == NULL) { 3910 device_printf(sc->dev, 3911 "failed to load firmware module: %d. cfg %p, fw %p\n", 3912 rc, cfg, fw); 3913 /* carry on */ 3914 } 3915 } 3916 if (fw == NULL) { 3917 device_printf(sc->dev, "firmware on card (%u.%u.%u.%u) is %s, " 3918 "but the driver cannot take corrective action because it " 3919 "is unable to load the firmware module.\n", 3920 G_FW_HDR_FW_VER_MAJOR(c), G_FW_HDR_FW_VER_MINOR(c), 3921 G_FW_HDR_FW_VER_MICRO(c), G_FW_HDR_FW_VER_BUILD(c), reason); 3922 rc = sc->flags & FW_OK ? 0 : ENOENT; 3923 goto done; 3924 } 3925 k = be32toh(((const struct fw_hdr *)fw->data)->fw_ver); 3926 if (k != d) { 3927 MPASS(t4_fw_install > 0); 3928 device_printf(sc->dev, 3929 "firmware in KLD (%u.%u.%u.%u) is not what the driver was " 3930 "expecting (%u.%u.%u.%u) and will not be used.\n", 3931 G_FW_HDR_FW_VER_MAJOR(k), G_FW_HDR_FW_VER_MINOR(k), 3932 G_FW_HDR_FW_VER_MICRO(k), G_FW_HDR_FW_VER_BUILD(k), 3933 G_FW_HDR_FW_VER_MAJOR(d), G_FW_HDR_FW_VER_MINOR(d), 3934 G_FW_HDR_FW_VER_MICRO(d), G_FW_HDR_FW_VER_BUILD(d)); 3935 rc = sc->flags & FW_OK ? 0 : EINVAL; 3936 goto done; 3937 } 3938 3939 device_printf(sc->dev, "firmware on card (%u.%u.%u.%u) is %s, " 3940 "installing firmware %u.%u.%u.%u on card.\n", 3941 G_FW_HDR_FW_VER_MAJOR(c), G_FW_HDR_FW_VER_MINOR(c), 3942 G_FW_HDR_FW_VER_MICRO(c), G_FW_HDR_FW_VER_BUILD(c), reason, 3943 G_FW_HDR_FW_VER_MAJOR(d), G_FW_HDR_FW_VER_MINOR(d), 3944 G_FW_HDR_FW_VER_MICRO(d), G_FW_HDR_FW_VER_BUILD(d)); 3945 3946 rc = -t4_fw_upgrade(sc, sc->mbox, fw->data, fw->datasize, 0); 3947 if (rc != 0) { 3948 device_printf(sc->dev, "failed to install firmware: %d\n", rc); 3949 } else { 3950 /* Installed successfully, update the cached header too. */ 3951 rc = ERESTART; 3952 memcpy(card_fw, fw->data, sizeof(*card_fw)); 3953 } 3954 done: 3955 unload_fw_module(sc, cfg, fw); 3956 3957 return (rc); 3958 } 3959 3960 /* 3961 * Establish contact with the firmware and attempt to become the master driver. 3962 * 3963 * A firmware will be installed to the card if needed (if the driver is allowed 3964 * to do so). 3965 */ 3966 static int 3967 contact_firmware(struct adapter *sc) 3968 { 3969 int rc, already = 0; 3970 enum dev_state state; 3971 struct fw_info *fw_info; 3972 struct fw_hdr *card_fw; /* fw on the card */ 3973 const struct fw_h *drv_fw; 3974 3975 fw_info = find_fw_info(chip_id(sc)); 3976 if (fw_info == NULL) { 3977 device_printf(sc->dev, 3978 "unable to look up firmware information for chip %d.\n", 3979 chip_id(sc)); 3980 return (EINVAL); 3981 } 3982 drv_fw = &fw_info->fw_h; 3983 3984 /* Read the header of the firmware on the card */ 3985 card_fw = malloc(sizeof(*card_fw), M_CXGBE, M_ZERO | M_WAITOK); 3986 restart: 3987 rc = -t4_get_fw_hdr(sc, card_fw); 3988 if (rc != 0) { 3989 device_printf(sc->dev, 3990 "unable to read firmware header from card's flash: %d\n", 3991 rc); 3992 goto done; 3993 } 3994 3995 rc = install_kld_firmware(sc, (struct fw_h *)card_fw, drv_fw, NULL, 3996 &already); 3997 if (rc == ERESTART) 3998 goto restart; 3999 if (rc != 0) 4000 goto done; 4001 4002 rc = t4_fw_hello(sc, sc->mbox, sc->mbox, MASTER_MAY, &state); 4003 if (rc < 0 || state == DEV_STATE_ERR) { 4004 rc = -rc; 4005 device_printf(sc->dev, 4006 "failed to connect to the firmware: %d, %d. " 4007 "PCIE_FW 0x%08x\n", rc, state, t4_read_reg(sc, A_PCIE_FW)); 4008 #if 0 4009 if (install_kld_firmware(sc, (struct fw_h *)card_fw, drv_fw, 4010 "not responding properly to HELLO", &already) == ERESTART) 4011 goto restart; 4012 #endif 4013 goto done; 4014 } 4015 MPASS(be32toh(card_fw->flags) & FW_HDR_FLAGS_RESET_HALT); 4016 sc->flags |= FW_OK; /* The firmware responded to the FW_HELLO. */ 4017 4018 if (rc == sc->pf) { 4019 sc->flags |= MASTER_PF; 4020 rc = install_kld_firmware(sc, (struct fw_h *)card_fw, drv_fw, 4021 NULL, &already); 4022 if (rc == ERESTART) 4023 rc = 0; 4024 else if (rc != 0) 4025 goto done; 4026 } else if (state == DEV_STATE_UNINIT) { 4027 /* 4028 * We didn't get to be the master so we definitely won't be 4029 * configuring the chip. It's a bug if someone else hasn't 4030 * configured it already. 4031 */ 4032 device_printf(sc->dev, "couldn't be master(%d), " 4033 "device not already initialized either(%d). " 4034 "PCIE_FW 0x%08x\n", rc, state, t4_read_reg(sc, A_PCIE_FW)); 4035 rc = EPROTO; 4036 goto done; 4037 } else { 4038 /* 4039 * Some other PF is the master and has configured the chip. 4040 * This is allowed but untested. 4041 */ 4042 device_printf(sc->dev, "PF%d is master, device state %d. " 4043 "PCIE_FW 0x%08x\n", rc, state, t4_read_reg(sc, A_PCIE_FW)); 4044 snprintf(sc->cfg_file, sizeof(sc->cfg_file), "pf%d", rc); 4045 sc->cfcsum = 0; 4046 rc = 0; 4047 } 4048 done: 4049 if (rc != 0 && sc->flags & FW_OK) { 4050 t4_fw_bye(sc, sc->mbox); 4051 sc->flags &= ~FW_OK; 4052 } 4053 free(card_fw, M_CXGBE); 4054 return (rc); 4055 } 4056 4057 static int 4058 copy_cfg_file_to_card(struct adapter *sc, char *cfg_file, 4059 uint32_t mtype, uint32_t moff) 4060 { 4061 struct fw_info *fw_info; 4062 const struct firmware *dcfg, *rcfg = NULL; 4063 const uint32_t *cfdata; 4064 uint32_t cflen, addr; 4065 int rc; 4066 4067 load_fw_module(sc, &dcfg, NULL); 4068 4069 /* Card specific interpretation of "default". */ 4070 if (strncmp(cfg_file, DEFAULT_CF, sizeof(t4_cfg_file)) == 0) { 4071 if (pci_get_device(sc->dev) == 0x440a) 4072 snprintf(cfg_file, sizeof(t4_cfg_file), UWIRE_CF); 4073 if (is_fpga(sc)) 4074 snprintf(cfg_file, sizeof(t4_cfg_file), FPGA_CF); 4075 } 4076 4077 if (strncmp(cfg_file, DEFAULT_CF, sizeof(t4_cfg_file)) == 0) { 4078 if (dcfg == NULL) { 4079 device_printf(sc->dev, 4080 "KLD with default config is not available.\n"); 4081 rc = ENOENT; 4082 goto done; 4083 } 4084 cfdata = dcfg->data; 4085 cflen = dcfg->datasize & ~3; 4086 } else { 4087 char s[32]; 4088 4089 fw_info = find_fw_info(chip_id(sc)); 4090 if (fw_info == NULL) { 4091 device_printf(sc->dev, 4092 "unable to look up firmware information for chip %d.\n", 4093 chip_id(sc)); 4094 rc = EINVAL; 4095 goto done; 4096 } 4097 snprintf(s, sizeof(s), "%s_%s", fw_info->kld_name, cfg_file); 4098 4099 rcfg = firmware_get(s); 4100 if (rcfg == NULL) { 4101 device_printf(sc->dev, 4102 "unable to load module \"%s\" for configuration " 4103 "profile \"%s\".\n", s, cfg_file); 4104 rc = ENOENT; 4105 goto done; 4106 } 4107 cfdata = rcfg->data; 4108 cflen = rcfg->datasize & ~3; 4109 } 4110 4111 if (cflen > FLASH_CFG_MAX_SIZE) { 4112 device_printf(sc->dev, 4113 "config file too long (%d, max allowed is %d).\n", 4114 cflen, FLASH_CFG_MAX_SIZE); 4115 rc = EINVAL; 4116 goto done; 4117 } 4118 4119 rc = validate_mt_off_len(sc, mtype, moff, cflen, &addr); 4120 if (rc != 0) { 4121 device_printf(sc->dev, 4122 "%s: addr (%d/0x%x) or len %d is not valid: %d.\n", 4123 __func__, mtype, moff, cflen, rc); 4124 rc = EINVAL; 4125 goto done; 4126 } 4127 write_via_memwin(sc, 2, addr, cfdata, cflen); 4128 done: 4129 if (rcfg != NULL) 4130 firmware_put(rcfg, FIRMWARE_UNLOAD); 4131 unload_fw_module(sc, dcfg, NULL); 4132 return (rc); 4133 } 4134 4135 struct caps_allowed { 4136 uint16_t nbmcaps; 4137 uint16_t linkcaps; 4138 uint16_t switchcaps; 4139 uint16_t niccaps; 4140 uint16_t toecaps; 4141 uint16_t rdmacaps; 4142 uint16_t cryptocaps; 4143 uint16_t iscsicaps; 4144 uint16_t fcoecaps; 4145 }; 4146 4147 #define FW_PARAM_DEV(param) \ 4148 (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | \ 4149 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_##param)) 4150 #define FW_PARAM_PFVF(param) \ 4151 (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_PFVF) | \ 4152 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_PFVF_##param)) 4153 4154 /* 4155 * Provide a configuration profile to the firmware and have it initialize the 4156 * chip accordingly. This may involve uploading a configuration file to the 4157 * card. 4158 */ 4159 static int 4160 apply_cfg_and_initialize(struct adapter *sc, char *cfg_file, 4161 const struct caps_allowed *caps_allowed) 4162 { 4163 int rc; 4164 struct fw_caps_config_cmd caps; 4165 uint32_t mtype, moff, finicsum, cfcsum, param, val; 4166 4167 rc = -t4_fw_reset(sc, sc->mbox, F_PIORSTMODE | F_PIORST); 4168 if (rc != 0) { 4169 device_printf(sc->dev, "firmware reset failed: %d.\n", rc); 4170 return (rc); 4171 } 4172 4173 bzero(&caps, sizeof(caps)); 4174 caps.op_to_write = htobe32(V_FW_CMD_OP(FW_CAPS_CONFIG_CMD) | 4175 F_FW_CMD_REQUEST | F_FW_CMD_READ); 4176 if (strncmp(cfg_file, BUILTIN_CF, sizeof(t4_cfg_file)) == 0) { 4177 mtype = 0; 4178 moff = 0; 4179 caps.cfvalid_to_len16 = htobe32(FW_LEN16(caps)); 4180 } else if (strncmp(cfg_file, FLASH_CF, sizeof(t4_cfg_file)) == 0) { 4181 mtype = FW_MEMTYPE_FLASH; 4182 moff = t4_flash_cfg_addr(sc); 4183 caps.cfvalid_to_len16 = htobe32(F_FW_CAPS_CONFIG_CMD_CFVALID | 4184 V_FW_CAPS_CONFIG_CMD_MEMTYPE_CF(mtype) | 4185 V_FW_CAPS_CONFIG_CMD_MEMADDR64K_CF(moff >> 16) | 4186 FW_LEN16(caps)); 4187 } else { 4188 /* 4189 * Ask the firmware where it wants us to upload the config file. 4190 */ 4191 param = FW_PARAM_DEV(CF); 4192 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 4193 if (rc != 0) { 4194 /* No support for config file? Shouldn't happen. */ 4195 device_printf(sc->dev, 4196 "failed to query config file location: %d.\n", rc); 4197 goto done; 4198 } 4199 mtype = G_FW_PARAMS_PARAM_Y(val); 4200 moff = G_FW_PARAMS_PARAM_Z(val) << 16; 4201 caps.cfvalid_to_len16 = htobe32(F_FW_CAPS_CONFIG_CMD_CFVALID | 4202 V_FW_CAPS_CONFIG_CMD_MEMTYPE_CF(mtype) | 4203 V_FW_CAPS_CONFIG_CMD_MEMADDR64K_CF(moff >> 16) | 4204 FW_LEN16(caps)); 4205 4206 rc = copy_cfg_file_to_card(sc, cfg_file, mtype, moff); 4207 if (rc != 0) { 4208 device_printf(sc->dev, 4209 "failed to upload config file to card: %d.\n", rc); 4210 goto done; 4211 } 4212 } 4213 rc = -t4_wr_mbox(sc, sc->mbox, &caps, sizeof(caps), &caps); 4214 if (rc != 0) { 4215 device_printf(sc->dev, "failed to pre-process config file: %d " 4216 "(mtype %d, moff 0x%x).\n", rc, mtype, moff); 4217 goto done; 4218 } 4219 4220 finicsum = be32toh(caps.finicsum); 4221 cfcsum = be32toh(caps.cfcsum); /* actual */ 4222 if (finicsum != cfcsum) { 4223 device_printf(sc->dev, 4224 "WARNING: config file checksum mismatch: %08x %08x\n", 4225 finicsum, cfcsum); 4226 } 4227 sc->cfcsum = cfcsum; 4228 snprintf(sc->cfg_file, sizeof(sc->cfg_file), "%s", cfg_file); 4229 4230 /* 4231 * Let the firmware know what features will (not) be used so it can tune 4232 * things accordingly. 4233 */ 4234 #define LIMIT_CAPS(x) do { \ 4235 caps.x##caps &= htobe16(caps_allowed->x##caps); \ 4236 } while (0) 4237 LIMIT_CAPS(nbm); 4238 LIMIT_CAPS(link); 4239 LIMIT_CAPS(switch); 4240 LIMIT_CAPS(nic); 4241 LIMIT_CAPS(toe); 4242 LIMIT_CAPS(rdma); 4243 LIMIT_CAPS(crypto); 4244 LIMIT_CAPS(iscsi); 4245 LIMIT_CAPS(fcoe); 4246 #undef LIMIT_CAPS 4247 if (caps.niccaps & htobe16(FW_CAPS_CONFIG_NIC_HASHFILTER)) { 4248 /* 4249 * TOE and hashfilters are mutually exclusive. It is a config 4250 * file or firmware bug if both are reported as available. Try 4251 * to cope with the situation in non-debug builds by disabling 4252 * TOE. 4253 */ 4254 MPASS(caps.toecaps == 0); 4255 4256 caps.toecaps = 0; 4257 caps.rdmacaps = 0; 4258 caps.iscsicaps = 0; 4259 } 4260 4261 caps.op_to_write = htobe32(V_FW_CMD_OP(FW_CAPS_CONFIG_CMD) | 4262 F_FW_CMD_REQUEST | F_FW_CMD_WRITE); 4263 caps.cfvalid_to_len16 = htobe32(FW_LEN16(caps)); 4264 rc = -t4_wr_mbox(sc, sc->mbox, &caps, sizeof(caps), NULL); 4265 if (rc != 0) { 4266 device_printf(sc->dev, 4267 "failed to process config file: %d.\n", rc); 4268 goto done; 4269 } 4270 4271 t4_tweak_chip_settings(sc); 4272 set_params__pre_init(sc); 4273 4274 /* get basic stuff going */ 4275 rc = -t4_fw_initialize(sc, sc->mbox); 4276 if (rc != 0) { 4277 device_printf(sc->dev, "fw_initialize failed: %d.\n", rc); 4278 goto done; 4279 } 4280 done: 4281 return (rc); 4282 } 4283 4284 /* 4285 * Partition chip resources for use between various PFs, VFs, etc. 4286 */ 4287 static int 4288 partition_resources(struct adapter *sc) 4289 { 4290 char cfg_file[sizeof(t4_cfg_file)]; 4291 struct caps_allowed caps_allowed; 4292 int rc; 4293 bool fallback; 4294 4295 /* Only the master driver gets to configure the chip resources. */ 4296 MPASS(sc->flags & MASTER_PF); 4297 4298 #define COPY_CAPS(x) do { \ 4299 caps_allowed.x##caps = t4_##x##caps_allowed; \ 4300 } while (0) 4301 bzero(&caps_allowed, sizeof(caps_allowed)); 4302 COPY_CAPS(nbm); 4303 COPY_CAPS(link); 4304 COPY_CAPS(switch); 4305 COPY_CAPS(nic); 4306 COPY_CAPS(toe); 4307 COPY_CAPS(rdma); 4308 COPY_CAPS(crypto); 4309 COPY_CAPS(iscsi); 4310 COPY_CAPS(fcoe); 4311 fallback = sc->debug_flags & DF_DISABLE_CFG_RETRY ? false : true; 4312 snprintf(cfg_file, sizeof(cfg_file), "%s", t4_cfg_file); 4313 retry: 4314 rc = apply_cfg_and_initialize(sc, cfg_file, &caps_allowed); 4315 if (rc != 0 && fallback) { 4316 device_printf(sc->dev, 4317 "failed (%d) to configure card with \"%s\" profile, " 4318 "will fall back to a basic configuration and retry.\n", 4319 rc, cfg_file); 4320 snprintf(cfg_file, sizeof(cfg_file), "%s", BUILTIN_CF); 4321 bzero(&caps_allowed, sizeof(caps_allowed)); 4322 COPY_CAPS(switch); 4323 caps_allowed.niccaps = FW_CAPS_CONFIG_NIC; 4324 fallback = false; 4325 goto retry; 4326 } 4327 #undef COPY_CAPS 4328 return (rc); 4329 } 4330 4331 /* 4332 * Retrieve parameters that are needed (or nice to have) very early. 4333 */ 4334 static int 4335 get_params__pre_init(struct adapter *sc) 4336 { 4337 int rc; 4338 uint32_t param[2], val[2]; 4339 4340 t4_get_version_info(sc); 4341 4342 snprintf(sc->fw_version, sizeof(sc->fw_version), "%u.%u.%u.%u", 4343 G_FW_HDR_FW_VER_MAJOR(sc->params.fw_vers), 4344 G_FW_HDR_FW_VER_MINOR(sc->params.fw_vers), 4345 G_FW_HDR_FW_VER_MICRO(sc->params.fw_vers), 4346 G_FW_HDR_FW_VER_BUILD(sc->params.fw_vers)); 4347 4348 snprintf(sc->bs_version, sizeof(sc->bs_version), "%u.%u.%u.%u", 4349 G_FW_HDR_FW_VER_MAJOR(sc->params.bs_vers), 4350 G_FW_HDR_FW_VER_MINOR(sc->params.bs_vers), 4351 G_FW_HDR_FW_VER_MICRO(sc->params.bs_vers), 4352 G_FW_HDR_FW_VER_BUILD(sc->params.bs_vers)); 4353 4354 snprintf(sc->tp_version, sizeof(sc->tp_version), "%u.%u.%u.%u", 4355 G_FW_HDR_FW_VER_MAJOR(sc->params.tp_vers), 4356 G_FW_HDR_FW_VER_MINOR(sc->params.tp_vers), 4357 G_FW_HDR_FW_VER_MICRO(sc->params.tp_vers), 4358 G_FW_HDR_FW_VER_BUILD(sc->params.tp_vers)); 4359 4360 snprintf(sc->er_version, sizeof(sc->er_version), "%u.%u.%u.%u", 4361 G_FW_HDR_FW_VER_MAJOR(sc->params.er_vers), 4362 G_FW_HDR_FW_VER_MINOR(sc->params.er_vers), 4363 G_FW_HDR_FW_VER_MICRO(sc->params.er_vers), 4364 G_FW_HDR_FW_VER_BUILD(sc->params.er_vers)); 4365 4366 param[0] = FW_PARAM_DEV(PORTVEC); 4367 param[1] = FW_PARAM_DEV(CCLK); 4368 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 2, param, val); 4369 if (rc != 0) { 4370 device_printf(sc->dev, 4371 "failed to query parameters (pre_init): %d.\n", rc); 4372 return (rc); 4373 } 4374 4375 sc->params.portvec = val[0]; 4376 sc->params.nports = bitcount32(val[0]); 4377 sc->params.vpd.cclk = val[1]; 4378 4379 /* Read device log parameters. */ 4380 rc = -t4_init_devlog_params(sc, 1); 4381 if (rc == 0) 4382 fixup_devlog_params(sc); 4383 else { 4384 device_printf(sc->dev, 4385 "failed to get devlog parameters: %d.\n", rc); 4386 rc = 0; /* devlog isn't critical for device operation */ 4387 } 4388 4389 return (rc); 4390 } 4391 4392 /* 4393 * Any params that need to be set before FW_INITIALIZE. 4394 */ 4395 static int 4396 set_params__pre_init(struct adapter *sc) 4397 { 4398 int rc = 0; 4399 uint32_t param, val; 4400 4401 if (chip_id(sc) >= CHELSIO_T6) { 4402 param = FW_PARAM_DEV(HPFILTER_REGION_SUPPORT); 4403 val = 1; 4404 rc = -t4_set_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 4405 /* firmwares < 1.20.1.0 do not have this param. */ 4406 if (rc == FW_EINVAL && 4407 sc->params.fw_vers < FW_VERSION32(1, 20, 1, 0)) { 4408 rc = 0; 4409 } 4410 if (rc != 0) { 4411 device_printf(sc->dev, 4412 "failed to enable high priority filters :%d.\n", 4413 rc); 4414 } 4415 } 4416 4417 /* Enable opaque VIIDs with firmwares that support it. */ 4418 param = FW_PARAM_DEV(OPAQUE_VIID_SMT_EXTN); 4419 val = 1; 4420 rc = -t4_set_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 4421 if (rc == 0 && val == 1) 4422 sc->params.viid_smt_extn_support = true; 4423 else 4424 sc->params.viid_smt_extn_support = false; 4425 4426 return (rc); 4427 } 4428 4429 /* 4430 * Retrieve various parameters that are of interest to the driver. The device 4431 * has been initialized by the firmware at this point. 4432 */ 4433 static int 4434 get_params__post_init(struct adapter *sc) 4435 { 4436 int rc; 4437 uint32_t param[7], val[7]; 4438 struct fw_caps_config_cmd caps; 4439 4440 param[0] = FW_PARAM_PFVF(IQFLINT_START); 4441 param[1] = FW_PARAM_PFVF(EQ_START); 4442 param[2] = FW_PARAM_PFVF(FILTER_START); 4443 param[3] = FW_PARAM_PFVF(FILTER_END); 4444 param[4] = FW_PARAM_PFVF(L2T_START); 4445 param[5] = FW_PARAM_PFVF(L2T_END); 4446 param[6] = V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | 4447 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_DIAG) | 4448 V_FW_PARAMS_PARAM_Y(FW_PARAM_DEV_DIAG_VDD); 4449 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 7, param, val); 4450 if (rc != 0) { 4451 device_printf(sc->dev, 4452 "failed to query parameters (post_init): %d.\n", rc); 4453 return (rc); 4454 } 4455 4456 sc->sge.iq_start = val[0]; 4457 sc->sge.eq_start = val[1]; 4458 if ((int)val[3] > (int)val[2]) { 4459 sc->tids.ftid_base = val[2]; 4460 sc->tids.ftid_end = val[3]; 4461 sc->tids.nftids = val[3] - val[2] + 1; 4462 } 4463 sc->vres.l2t.start = val[4]; 4464 sc->vres.l2t.size = val[5] - val[4] + 1; 4465 KASSERT(sc->vres.l2t.size <= L2T_SIZE, 4466 ("%s: L2 table size (%u) larger than expected (%u)", 4467 __func__, sc->vres.l2t.size, L2T_SIZE)); 4468 sc->params.core_vdd = val[6]; 4469 4470 param[0] = FW_PARAM_PFVF(IQFLINT_END); 4471 param[1] = FW_PARAM_PFVF(EQ_END); 4472 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 2, param, val); 4473 if (rc != 0) { 4474 device_printf(sc->dev, 4475 "failed to query parameters (post_init2): %d.\n", rc); 4476 return (rc); 4477 } 4478 MPASS((int)val[0] >= sc->sge.iq_start); 4479 sc->sge.iqmap_sz = val[0] - sc->sge.iq_start + 1; 4480 MPASS((int)val[1] >= sc->sge.eq_start); 4481 sc->sge.eqmap_sz = val[1] - sc->sge.eq_start + 1; 4482 4483 if (chip_id(sc) >= CHELSIO_T6) { 4484 4485 sc->tids.tid_base = t4_read_reg(sc, 4486 A_LE_DB_ACTIVE_TABLE_START_INDEX); 4487 4488 param[0] = FW_PARAM_PFVF(HPFILTER_START); 4489 param[1] = FW_PARAM_PFVF(HPFILTER_END); 4490 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 2, param, val); 4491 if (rc != 0) { 4492 device_printf(sc->dev, 4493 "failed to query hpfilter parameters: %d.\n", rc); 4494 return (rc); 4495 } 4496 if ((int)val[1] > (int)val[0]) { 4497 sc->tids.hpftid_base = val[0]; 4498 sc->tids.hpftid_end = val[1]; 4499 sc->tids.nhpftids = val[1] - val[0] + 1; 4500 4501 /* 4502 * These should go off if the layout changes and the 4503 * driver needs to catch up. 4504 */ 4505 MPASS(sc->tids.hpftid_base == 0); 4506 MPASS(sc->tids.tid_base == sc->tids.nhpftids); 4507 } 4508 4509 param[0] = FW_PARAM_PFVF(RAWF_START); 4510 param[1] = FW_PARAM_PFVF(RAWF_END); 4511 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 2, param, val); 4512 if (rc != 0) { 4513 device_printf(sc->dev, 4514 "failed to query rawf parameters: %d.\n", rc); 4515 return (rc); 4516 } 4517 if ((int)val[1] > (int)val[0]) { 4518 sc->rawf_base = val[0]; 4519 sc->nrawf = val[1] - val[0] + 1; 4520 } 4521 } 4522 4523 /* 4524 * MPSBGMAP is queried separately because only recent firmwares support 4525 * it as a parameter and we don't want the compound query above to fail 4526 * on older firmwares. 4527 */ 4528 param[0] = FW_PARAM_DEV(MPSBGMAP); 4529 val[0] = 0; 4530 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, param, val); 4531 if (rc == 0) 4532 sc->params.mps_bg_map = val[0]; 4533 else 4534 sc->params.mps_bg_map = 0; 4535 4536 /* 4537 * Determine whether the firmware supports the filter2 work request. 4538 * This is queried separately for the same reason as MPSBGMAP above. 4539 */ 4540 param[0] = FW_PARAM_DEV(FILTER2_WR); 4541 val[0] = 0; 4542 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, param, val); 4543 if (rc == 0) 4544 sc->params.filter2_wr_support = val[0] != 0; 4545 else 4546 sc->params.filter2_wr_support = 0; 4547 4548 /* 4549 * Find out whether we're allowed to use the ULPTX MEMWRITE DSGL. 4550 * This is queried separately for the same reason as other params above. 4551 */ 4552 param[0] = FW_PARAM_DEV(ULPTX_MEMWRITE_DSGL); 4553 val[0] = 0; 4554 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, param, val); 4555 if (rc == 0) 4556 sc->params.ulptx_memwrite_dsgl = val[0] != 0; 4557 else 4558 sc->params.ulptx_memwrite_dsgl = false; 4559 4560 /* FW_RI_FR_NSMR_TPTE_WR support */ 4561 param[0] = FW_PARAM_DEV(RI_FR_NSMR_TPTE_WR); 4562 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, param, val); 4563 if (rc == 0) 4564 sc->params.fr_nsmr_tpte_wr_support = val[0] != 0; 4565 else 4566 sc->params.fr_nsmr_tpte_wr_support = false; 4567 4568 param[0] = FW_PARAM_PFVF(MAX_PKTS_PER_ETH_TX_PKTS_WR); 4569 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, param, val); 4570 if (rc == 0) 4571 sc->params.max_pkts_per_eth_tx_pkts_wr = val[0]; 4572 else 4573 sc->params.max_pkts_per_eth_tx_pkts_wr = 15; 4574 4575 /* get capabilites */ 4576 bzero(&caps, sizeof(caps)); 4577 caps.op_to_write = htobe32(V_FW_CMD_OP(FW_CAPS_CONFIG_CMD) | 4578 F_FW_CMD_REQUEST | F_FW_CMD_READ); 4579 caps.cfvalid_to_len16 = htobe32(FW_LEN16(caps)); 4580 rc = -t4_wr_mbox(sc, sc->mbox, &caps, sizeof(caps), &caps); 4581 if (rc != 0) { 4582 device_printf(sc->dev, 4583 "failed to get card capabilities: %d.\n", rc); 4584 return (rc); 4585 } 4586 4587 #define READ_CAPS(x) do { \ 4588 sc->x = htobe16(caps.x); \ 4589 } while (0) 4590 READ_CAPS(nbmcaps); 4591 READ_CAPS(linkcaps); 4592 READ_CAPS(switchcaps); 4593 READ_CAPS(niccaps); 4594 READ_CAPS(toecaps); 4595 READ_CAPS(rdmacaps); 4596 READ_CAPS(cryptocaps); 4597 READ_CAPS(iscsicaps); 4598 READ_CAPS(fcoecaps); 4599 4600 if (sc->niccaps & FW_CAPS_CONFIG_NIC_HASHFILTER) { 4601 MPASS(chip_id(sc) > CHELSIO_T4); 4602 MPASS(sc->toecaps == 0); 4603 sc->toecaps = 0; 4604 4605 param[0] = FW_PARAM_DEV(NTID); 4606 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, param, val); 4607 if (rc != 0) { 4608 device_printf(sc->dev, 4609 "failed to query HASHFILTER parameters: %d.\n", rc); 4610 return (rc); 4611 } 4612 sc->tids.ntids = val[0]; 4613 if (sc->params.fw_vers < FW_VERSION32(1, 20, 5, 0)) { 4614 MPASS(sc->tids.ntids >= sc->tids.nhpftids); 4615 sc->tids.ntids -= sc->tids.nhpftids; 4616 } 4617 sc->tids.natids = min(sc->tids.ntids / 2, MAX_ATIDS); 4618 sc->params.hash_filter = 1; 4619 } 4620 if (sc->niccaps & FW_CAPS_CONFIG_NIC_ETHOFLD) { 4621 param[0] = FW_PARAM_PFVF(ETHOFLD_START); 4622 param[1] = FW_PARAM_PFVF(ETHOFLD_END); 4623 param[2] = FW_PARAM_DEV(FLOWC_BUFFIFO_SZ); 4624 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 3, param, val); 4625 if (rc != 0) { 4626 device_printf(sc->dev, 4627 "failed to query NIC parameters: %d.\n", rc); 4628 return (rc); 4629 } 4630 if ((int)val[1] > (int)val[0]) { 4631 sc->tids.etid_base = val[0]; 4632 sc->tids.etid_end = val[1]; 4633 sc->tids.netids = val[1] - val[0] + 1; 4634 sc->params.eo_wr_cred = val[2]; 4635 sc->params.ethoffload = 1; 4636 } 4637 } 4638 if (sc->toecaps) { 4639 /* query offload-related parameters */ 4640 param[0] = FW_PARAM_DEV(NTID); 4641 param[1] = FW_PARAM_PFVF(SERVER_START); 4642 param[2] = FW_PARAM_PFVF(SERVER_END); 4643 param[3] = FW_PARAM_PFVF(TDDP_START); 4644 param[4] = FW_PARAM_PFVF(TDDP_END); 4645 param[5] = FW_PARAM_DEV(FLOWC_BUFFIFO_SZ); 4646 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 6, param, val); 4647 if (rc != 0) { 4648 device_printf(sc->dev, 4649 "failed to query TOE parameters: %d.\n", rc); 4650 return (rc); 4651 } 4652 sc->tids.ntids = val[0]; 4653 if (sc->params.fw_vers < FW_VERSION32(1, 20, 5, 0)) { 4654 MPASS(sc->tids.ntids >= sc->tids.nhpftids); 4655 sc->tids.ntids -= sc->tids.nhpftids; 4656 } 4657 sc->tids.natids = min(sc->tids.ntids / 2, MAX_ATIDS); 4658 if ((int)val[2] > (int)val[1]) { 4659 sc->tids.stid_base = val[1]; 4660 sc->tids.nstids = val[2] - val[1] + 1; 4661 } 4662 sc->vres.ddp.start = val[3]; 4663 sc->vres.ddp.size = val[4] - val[3] + 1; 4664 sc->params.ofldq_wr_cred = val[5]; 4665 sc->params.offload = 1; 4666 } else { 4667 /* 4668 * The firmware attempts memfree TOE configuration for -SO cards 4669 * and will report toecaps=0 if it runs out of resources (this 4670 * depends on the config file). It may not report 0 for other 4671 * capabilities dependent on the TOE in this case. Set them to 4672 * 0 here so that the driver doesn't bother tracking resources 4673 * that will never be used. 4674 */ 4675 sc->iscsicaps = 0; 4676 sc->rdmacaps = 0; 4677 } 4678 if (sc->rdmacaps) { 4679 param[0] = FW_PARAM_PFVF(STAG_START); 4680 param[1] = FW_PARAM_PFVF(STAG_END); 4681 param[2] = FW_PARAM_PFVF(RQ_START); 4682 param[3] = FW_PARAM_PFVF(RQ_END); 4683 param[4] = FW_PARAM_PFVF(PBL_START); 4684 param[5] = FW_PARAM_PFVF(PBL_END); 4685 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 6, param, val); 4686 if (rc != 0) { 4687 device_printf(sc->dev, 4688 "failed to query RDMA parameters(1): %d.\n", rc); 4689 return (rc); 4690 } 4691 sc->vres.stag.start = val[0]; 4692 sc->vres.stag.size = val[1] - val[0] + 1; 4693 sc->vres.rq.start = val[2]; 4694 sc->vres.rq.size = val[3] - val[2] + 1; 4695 sc->vres.pbl.start = val[4]; 4696 sc->vres.pbl.size = val[5] - val[4] + 1; 4697 4698 param[0] = FW_PARAM_PFVF(SQRQ_START); 4699 param[1] = FW_PARAM_PFVF(SQRQ_END); 4700 param[2] = FW_PARAM_PFVF(CQ_START); 4701 param[3] = FW_PARAM_PFVF(CQ_END); 4702 param[4] = FW_PARAM_PFVF(OCQ_START); 4703 param[5] = FW_PARAM_PFVF(OCQ_END); 4704 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 6, param, val); 4705 if (rc != 0) { 4706 device_printf(sc->dev, 4707 "failed to query RDMA parameters(2): %d.\n", rc); 4708 return (rc); 4709 } 4710 sc->vres.qp.start = val[0]; 4711 sc->vres.qp.size = val[1] - val[0] + 1; 4712 sc->vres.cq.start = val[2]; 4713 sc->vres.cq.size = val[3] - val[2] + 1; 4714 sc->vres.ocq.start = val[4]; 4715 sc->vres.ocq.size = val[5] - val[4] + 1; 4716 4717 param[0] = FW_PARAM_PFVF(SRQ_START); 4718 param[1] = FW_PARAM_PFVF(SRQ_END); 4719 param[2] = FW_PARAM_DEV(MAXORDIRD_QP); 4720 param[3] = FW_PARAM_DEV(MAXIRD_ADAPTER); 4721 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 4, param, val); 4722 if (rc != 0) { 4723 device_printf(sc->dev, 4724 "failed to query RDMA parameters(3): %d.\n", rc); 4725 return (rc); 4726 } 4727 sc->vres.srq.start = val[0]; 4728 sc->vres.srq.size = val[1] - val[0] + 1; 4729 sc->params.max_ordird_qp = val[2]; 4730 sc->params.max_ird_adapter = val[3]; 4731 } 4732 if (sc->iscsicaps) { 4733 param[0] = FW_PARAM_PFVF(ISCSI_START); 4734 param[1] = FW_PARAM_PFVF(ISCSI_END); 4735 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 2, param, val); 4736 if (rc != 0) { 4737 device_printf(sc->dev, 4738 "failed to query iSCSI parameters: %d.\n", rc); 4739 return (rc); 4740 } 4741 sc->vres.iscsi.start = val[0]; 4742 sc->vres.iscsi.size = val[1] - val[0] + 1; 4743 } 4744 if (sc->cryptocaps & FW_CAPS_CONFIG_TLSKEYS) { 4745 param[0] = FW_PARAM_PFVF(TLS_START); 4746 param[1] = FW_PARAM_PFVF(TLS_END); 4747 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 2, param, val); 4748 if (rc != 0) { 4749 device_printf(sc->dev, 4750 "failed to query TLS parameters: %d.\n", rc); 4751 return (rc); 4752 } 4753 sc->vres.key.start = val[0]; 4754 sc->vres.key.size = val[1] - val[0] + 1; 4755 } 4756 4757 /* 4758 * We've got the params we wanted to query directly from the firmware. 4759 * Grab some others via other means. 4760 */ 4761 t4_init_sge_params(sc); 4762 t4_init_tp_params(sc); 4763 t4_read_mtu_tbl(sc, sc->params.mtus, NULL); 4764 t4_load_mtus(sc, sc->params.mtus, sc->params.a_wnd, sc->params.b_wnd); 4765 4766 rc = t4_verify_chip_settings(sc); 4767 if (rc != 0) 4768 return (rc); 4769 t4_init_rx_buf_info(sc); 4770 4771 return (rc); 4772 } 4773 4774 #ifdef KERN_TLS 4775 static void 4776 ktls_tick(void *arg) 4777 { 4778 struct adapter *sc; 4779 uint32_t tstamp; 4780 4781 sc = arg; 4782 4783 tstamp = tcp_ts_getticks(); 4784 t4_write_reg(sc, A_TP_SYNC_TIME_HI, tstamp >> 1); 4785 t4_write_reg(sc, A_TP_SYNC_TIME_LO, tstamp << 31); 4786 4787 callout_schedule_sbt(&sc->ktls_tick, SBT_1MS, 0, C_HARDCLOCK); 4788 } 4789 4790 static void 4791 t4_enable_kern_tls(struct adapter *sc) 4792 { 4793 uint32_t m, v; 4794 4795 m = F_ENABLECBYP; 4796 v = F_ENABLECBYP; 4797 t4_set_reg_field(sc, A_TP_PARA_REG6, m, v); 4798 4799 m = F_CPL_FLAGS_UPDATE_EN | F_SEQ_UPDATE_EN; 4800 v = F_CPL_FLAGS_UPDATE_EN | F_SEQ_UPDATE_EN; 4801 t4_set_reg_field(sc, A_ULP_TX_CONFIG, m, v); 4802 4803 m = F_NICMODE; 4804 v = F_NICMODE; 4805 t4_set_reg_field(sc, A_TP_IN_CONFIG, m, v); 4806 4807 m = F_LOOKUPEVERYPKT; 4808 v = 0; 4809 t4_set_reg_field(sc, A_TP_INGRESS_CONFIG, m, v); 4810 4811 m = F_TXDEFERENABLE | F_DISABLEWINDOWPSH | F_DISABLESEPPSHFLAG; 4812 v = F_DISABLEWINDOWPSH; 4813 t4_set_reg_field(sc, A_TP_PC_CONFIG, m, v); 4814 4815 m = V_TIMESTAMPRESOLUTION(M_TIMESTAMPRESOLUTION); 4816 v = V_TIMESTAMPRESOLUTION(0x1f); 4817 t4_set_reg_field(sc, A_TP_TIMER_RESOLUTION, m, v); 4818 4819 sc->flags |= KERN_TLS_OK; 4820 4821 sc->tlst.inline_keys = t4_tls_inline_keys; 4822 sc->tlst.combo_wrs = t4_tls_combo_wrs; 4823 } 4824 #endif 4825 4826 static int 4827 set_params__post_init(struct adapter *sc) 4828 { 4829 uint32_t mask, param, val; 4830 #ifdef TCP_OFFLOAD 4831 int i, v, shift; 4832 #endif 4833 4834 /* ask for encapsulated CPLs */ 4835 param = FW_PARAM_PFVF(CPLFW4MSG_ENCAP); 4836 val = 1; 4837 (void)t4_set_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 4838 4839 /* Enable 32b port caps if the firmware supports it. */ 4840 param = FW_PARAM_PFVF(PORT_CAPS32); 4841 val = 1; 4842 if (t4_set_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val) == 0) 4843 sc->params.port_caps32 = 1; 4844 4845 /* Let filter + maskhash steer to a part of the VI's RSS region. */ 4846 val = 1 << (G_MASKSIZE(t4_read_reg(sc, A_TP_RSS_CONFIG_TNL)) - 1); 4847 t4_set_reg_field(sc, A_TP_RSS_CONFIG_TNL, V_MASKFILTER(M_MASKFILTER), 4848 V_MASKFILTER(val - 1)); 4849 4850 mask = F_DROPERRORANY | F_DROPERRORMAC | F_DROPERRORIPVER | 4851 F_DROPERRORFRAG | F_DROPERRORATTACK | F_DROPERRORETHHDRLEN | 4852 F_DROPERRORIPHDRLEN | F_DROPERRORTCPHDRLEN | F_DROPERRORPKTLEN | 4853 F_DROPERRORTCPOPT | F_DROPERRORCSUMIP | F_DROPERRORCSUM; 4854 val = 0; 4855 if (chip_id(sc) < CHELSIO_T6 && t4_attack_filter != 0) { 4856 t4_set_reg_field(sc, A_TP_GLOBAL_CONFIG, F_ATTACKFILTERENABLE, 4857 F_ATTACKFILTERENABLE); 4858 val |= F_DROPERRORATTACK; 4859 } 4860 if (t4_drop_ip_fragments != 0) { 4861 t4_set_reg_field(sc, A_TP_GLOBAL_CONFIG, F_FRAGMENTDROP, 4862 F_FRAGMENTDROP); 4863 val |= F_DROPERRORFRAG; 4864 } 4865 if (t4_drop_pkts_with_l2_errors != 0) 4866 val |= F_DROPERRORMAC | F_DROPERRORETHHDRLEN; 4867 if (t4_drop_pkts_with_l3_errors != 0) { 4868 val |= F_DROPERRORIPVER | F_DROPERRORIPHDRLEN | 4869 F_DROPERRORCSUMIP; 4870 } 4871 if (t4_drop_pkts_with_l4_errors != 0) { 4872 val |= F_DROPERRORTCPHDRLEN | F_DROPERRORPKTLEN | 4873 F_DROPERRORTCPOPT | F_DROPERRORCSUM; 4874 } 4875 t4_set_reg_field(sc, A_TP_ERR_CONFIG, mask, val); 4876 4877 #ifdef TCP_OFFLOAD 4878 /* 4879 * Override the TOE timers with user provided tunables. This is not the 4880 * recommended way to change the timers (the firmware config file is) so 4881 * these tunables are not documented. 4882 * 4883 * All the timer tunables are in microseconds. 4884 */ 4885 if (t4_toe_keepalive_idle != 0) { 4886 v = us_to_tcp_ticks(sc, t4_toe_keepalive_idle); 4887 v &= M_KEEPALIVEIDLE; 4888 t4_set_reg_field(sc, A_TP_KEEP_IDLE, 4889 V_KEEPALIVEIDLE(M_KEEPALIVEIDLE), V_KEEPALIVEIDLE(v)); 4890 } 4891 if (t4_toe_keepalive_interval != 0) { 4892 v = us_to_tcp_ticks(sc, t4_toe_keepalive_interval); 4893 v &= M_KEEPALIVEINTVL; 4894 t4_set_reg_field(sc, A_TP_KEEP_INTVL, 4895 V_KEEPALIVEINTVL(M_KEEPALIVEINTVL), V_KEEPALIVEINTVL(v)); 4896 } 4897 if (t4_toe_keepalive_count != 0) { 4898 v = t4_toe_keepalive_count & M_KEEPALIVEMAXR2; 4899 t4_set_reg_field(sc, A_TP_SHIFT_CNT, 4900 V_KEEPALIVEMAXR1(M_KEEPALIVEMAXR1) | 4901 V_KEEPALIVEMAXR2(M_KEEPALIVEMAXR2), 4902 V_KEEPALIVEMAXR1(1) | V_KEEPALIVEMAXR2(v)); 4903 } 4904 if (t4_toe_rexmt_min != 0) { 4905 v = us_to_tcp_ticks(sc, t4_toe_rexmt_min); 4906 v &= M_RXTMIN; 4907 t4_set_reg_field(sc, A_TP_RXT_MIN, 4908 V_RXTMIN(M_RXTMIN), V_RXTMIN(v)); 4909 } 4910 if (t4_toe_rexmt_max != 0) { 4911 v = us_to_tcp_ticks(sc, t4_toe_rexmt_max); 4912 v &= M_RXTMAX; 4913 t4_set_reg_field(sc, A_TP_RXT_MAX, 4914 V_RXTMAX(M_RXTMAX), V_RXTMAX(v)); 4915 } 4916 if (t4_toe_rexmt_count != 0) { 4917 v = t4_toe_rexmt_count & M_RXTSHIFTMAXR2; 4918 t4_set_reg_field(sc, A_TP_SHIFT_CNT, 4919 V_RXTSHIFTMAXR1(M_RXTSHIFTMAXR1) | 4920 V_RXTSHIFTMAXR2(M_RXTSHIFTMAXR2), 4921 V_RXTSHIFTMAXR1(1) | V_RXTSHIFTMAXR2(v)); 4922 } 4923 for (i = 0; i < nitems(t4_toe_rexmt_backoff); i++) { 4924 if (t4_toe_rexmt_backoff[i] != -1) { 4925 v = t4_toe_rexmt_backoff[i] & M_TIMERBACKOFFINDEX0; 4926 shift = (i & 3) << 3; 4927 t4_set_reg_field(sc, A_TP_TCP_BACKOFF_REG0 + (i & ~3), 4928 M_TIMERBACKOFFINDEX0 << shift, v << shift); 4929 } 4930 } 4931 #endif 4932 4933 #ifdef KERN_TLS 4934 if (sc->cryptocaps & FW_CAPS_CONFIG_TLSKEYS && 4935 sc->toecaps & FW_CAPS_CONFIG_TOE) { 4936 if (t4_kern_tls != 0) 4937 t4_enable_kern_tls(sc); 4938 else { 4939 /* 4940 * Limit TOE connections to 2 reassembly 4941 * "islands". This is required for TOE TLS 4942 * connections to downgrade to plain TOE 4943 * connections if an unsupported TLS version 4944 * or ciphersuite is used. 4945 */ 4946 t4_tp_wr_bits_indirect(sc, A_TP_FRAG_CONFIG, 4947 V_PASSMODE(M_PASSMODE), V_PASSMODE(2)); 4948 } 4949 } 4950 #endif 4951 return (0); 4952 } 4953 4954 #undef FW_PARAM_PFVF 4955 #undef FW_PARAM_DEV 4956 4957 static void 4958 t4_set_desc(struct adapter *sc) 4959 { 4960 char buf[128]; 4961 struct adapter_params *p = &sc->params; 4962 4963 snprintf(buf, sizeof(buf), "Chelsio %s", p->vpd.id); 4964 4965 device_set_desc_copy(sc->dev, buf); 4966 } 4967 4968 static inline void 4969 ifmedia_add4(struct ifmedia *ifm, int m) 4970 { 4971 4972 ifmedia_add(ifm, m, 0, NULL); 4973 ifmedia_add(ifm, m | IFM_ETH_TXPAUSE, 0, NULL); 4974 ifmedia_add(ifm, m | IFM_ETH_RXPAUSE, 0, NULL); 4975 ifmedia_add(ifm, m | IFM_ETH_TXPAUSE | IFM_ETH_RXPAUSE, 0, NULL); 4976 } 4977 4978 /* 4979 * This is the selected media, which is not quite the same as the active media. 4980 * The media line in ifconfig is "media: Ethernet selected (active)" if selected 4981 * and active are not the same, and "media: Ethernet selected" otherwise. 4982 */ 4983 static void 4984 set_current_media(struct port_info *pi) 4985 { 4986 struct link_config *lc; 4987 struct ifmedia *ifm; 4988 int mword; 4989 u_int speed; 4990 4991 PORT_LOCK_ASSERT_OWNED(pi); 4992 4993 /* Leave current media alone if it's already set to IFM_NONE. */ 4994 ifm = &pi->media; 4995 if (ifm->ifm_cur != NULL && 4996 IFM_SUBTYPE(ifm->ifm_cur->ifm_media) == IFM_NONE) 4997 return; 4998 4999 lc = &pi->link_cfg; 5000 if (lc->requested_aneg != AUTONEG_DISABLE && 5001 lc->pcaps & FW_PORT_CAP32_ANEG) { 5002 ifmedia_set(ifm, IFM_ETHER | IFM_AUTO); 5003 return; 5004 } 5005 mword = IFM_ETHER | IFM_FDX; 5006 if (lc->requested_fc & PAUSE_TX) 5007 mword |= IFM_ETH_TXPAUSE; 5008 if (lc->requested_fc & PAUSE_RX) 5009 mword |= IFM_ETH_RXPAUSE; 5010 if (lc->requested_speed == 0) 5011 speed = port_top_speed(pi) * 1000; /* Gbps -> Mbps */ 5012 else 5013 speed = lc->requested_speed; 5014 mword |= port_mword(pi, speed_to_fwcap(speed)); 5015 ifmedia_set(ifm, mword); 5016 } 5017 5018 /* 5019 * Returns true if the ifmedia list for the port cannot change. 5020 */ 5021 static bool 5022 fixed_ifmedia(struct port_info *pi) 5023 { 5024 5025 return (pi->port_type == FW_PORT_TYPE_BT_SGMII || 5026 pi->port_type == FW_PORT_TYPE_BT_XFI || 5027 pi->port_type == FW_PORT_TYPE_BT_XAUI || 5028 pi->port_type == FW_PORT_TYPE_KX4 || 5029 pi->port_type == FW_PORT_TYPE_KX || 5030 pi->port_type == FW_PORT_TYPE_KR || 5031 pi->port_type == FW_PORT_TYPE_BP_AP || 5032 pi->port_type == FW_PORT_TYPE_BP4_AP || 5033 pi->port_type == FW_PORT_TYPE_BP40_BA || 5034 pi->port_type == FW_PORT_TYPE_KR4_100G || 5035 pi->port_type == FW_PORT_TYPE_KR_SFP28 || 5036 pi->port_type == FW_PORT_TYPE_KR_XLAUI); 5037 } 5038 5039 static void 5040 build_medialist(struct port_info *pi) 5041 { 5042 uint32_t ss, speed; 5043 int unknown, mword, bit; 5044 struct link_config *lc; 5045 struct ifmedia *ifm; 5046 5047 PORT_LOCK_ASSERT_OWNED(pi); 5048 5049 if (pi->flags & FIXED_IFMEDIA) 5050 return; 5051 5052 /* 5053 * Rebuild the ifmedia list. 5054 */ 5055 ifm = &pi->media; 5056 ifmedia_removeall(ifm); 5057 lc = &pi->link_cfg; 5058 ss = G_FW_PORT_CAP32_SPEED(lc->pcaps); /* Supported Speeds */ 5059 if (__predict_false(ss == 0)) { /* not supposed to happen. */ 5060 MPASS(ss != 0); 5061 no_media: 5062 MPASS(LIST_EMPTY(&ifm->ifm_list)); 5063 ifmedia_add(ifm, IFM_ETHER | IFM_NONE, 0, NULL); 5064 ifmedia_set(ifm, IFM_ETHER | IFM_NONE); 5065 return; 5066 } 5067 5068 unknown = 0; 5069 for (bit = S_FW_PORT_CAP32_SPEED; bit < fls(ss); bit++) { 5070 speed = 1 << bit; 5071 MPASS(speed & M_FW_PORT_CAP32_SPEED); 5072 if (ss & speed) { 5073 mword = port_mword(pi, speed); 5074 if (mword == IFM_NONE) { 5075 goto no_media; 5076 } else if (mword == IFM_UNKNOWN) 5077 unknown++; 5078 else 5079 ifmedia_add4(ifm, IFM_ETHER | IFM_FDX | mword); 5080 } 5081 } 5082 if (unknown > 0) /* Add one unknown for all unknown media types. */ 5083 ifmedia_add4(ifm, IFM_ETHER | IFM_FDX | IFM_UNKNOWN); 5084 if (lc->pcaps & FW_PORT_CAP32_ANEG) 5085 ifmedia_add(ifm, IFM_ETHER | IFM_AUTO, 0, NULL); 5086 5087 set_current_media(pi); 5088 } 5089 5090 /* 5091 * Initialize the requested fields in the link config based on driver tunables. 5092 */ 5093 static void 5094 init_link_config(struct port_info *pi) 5095 { 5096 struct link_config *lc = &pi->link_cfg; 5097 5098 PORT_LOCK_ASSERT_OWNED(pi); 5099 5100 lc->requested_speed = 0; 5101 5102 if (t4_autoneg == 0) 5103 lc->requested_aneg = AUTONEG_DISABLE; 5104 else if (t4_autoneg == 1) 5105 lc->requested_aneg = AUTONEG_ENABLE; 5106 else 5107 lc->requested_aneg = AUTONEG_AUTO; 5108 5109 lc->requested_fc = t4_pause_settings & (PAUSE_TX | PAUSE_RX | 5110 PAUSE_AUTONEG); 5111 5112 if (t4_fec & FEC_AUTO) 5113 lc->requested_fec = FEC_AUTO; 5114 else if (t4_fec == 0) 5115 lc->requested_fec = FEC_NONE; 5116 else { 5117 /* -1 is handled by the FEC_AUTO block above and not here. */ 5118 lc->requested_fec = t4_fec & 5119 (FEC_RS | FEC_BASER_RS | FEC_NONE | FEC_MODULE); 5120 if (lc->requested_fec == 0) 5121 lc->requested_fec = FEC_AUTO; 5122 } 5123 } 5124 5125 /* 5126 * Makes sure that all requested settings comply with what's supported by the 5127 * port. Returns the number of settings that were invalid and had to be fixed. 5128 */ 5129 static int 5130 fixup_link_config(struct port_info *pi) 5131 { 5132 int n = 0; 5133 struct link_config *lc = &pi->link_cfg; 5134 uint32_t fwspeed; 5135 5136 PORT_LOCK_ASSERT_OWNED(pi); 5137 5138 /* Speed (when not autonegotiating) */ 5139 if (lc->requested_speed != 0) { 5140 fwspeed = speed_to_fwcap(lc->requested_speed); 5141 if ((fwspeed & lc->pcaps) == 0) { 5142 n++; 5143 lc->requested_speed = 0; 5144 } 5145 } 5146 5147 /* Link autonegotiation */ 5148 MPASS(lc->requested_aneg == AUTONEG_ENABLE || 5149 lc->requested_aneg == AUTONEG_DISABLE || 5150 lc->requested_aneg == AUTONEG_AUTO); 5151 if (lc->requested_aneg == AUTONEG_ENABLE && 5152 !(lc->pcaps & FW_PORT_CAP32_ANEG)) { 5153 n++; 5154 lc->requested_aneg = AUTONEG_AUTO; 5155 } 5156 5157 /* Flow control */ 5158 MPASS((lc->requested_fc & ~(PAUSE_TX | PAUSE_RX | PAUSE_AUTONEG)) == 0); 5159 if (lc->requested_fc & PAUSE_TX && 5160 !(lc->pcaps & FW_PORT_CAP32_FC_TX)) { 5161 n++; 5162 lc->requested_fc &= ~PAUSE_TX; 5163 } 5164 if (lc->requested_fc & PAUSE_RX && 5165 !(lc->pcaps & FW_PORT_CAP32_FC_RX)) { 5166 n++; 5167 lc->requested_fc &= ~PAUSE_RX; 5168 } 5169 if (!(lc->requested_fc & PAUSE_AUTONEG) && 5170 !(lc->pcaps & FW_PORT_CAP32_FORCE_PAUSE)) { 5171 n++; 5172 lc->requested_fc |= PAUSE_AUTONEG; 5173 } 5174 5175 /* FEC */ 5176 if ((lc->requested_fec & FEC_RS && 5177 !(lc->pcaps & FW_PORT_CAP32_FEC_RS)) || 5178 (lc->requested_fec & FEC_BASER_RS && 5179 !(lc->pcaps & FW_PORT_CAP32_FEC_BASER_RS))) { 5180 n++; 5181 lc->requested_fec = FEC_AUTO; 5182 } 5183 5184 return (n); 5185 } 5186 5187 /* 5188 * Apply the requested L1 settings, which are expected to be valid, to the 5189 * hardware. 5190 */ 5191 static int 5192 apply_link_config(struct port_info *pi) 5193 { 5194 struct adapter *sc = pi->adapter; 5195 struct link_config *lc = &pi->link_cfg; 5196 int rc; 5197 5198 #ifdef INVARIANTS 5199 ASSERT_SYNCHRONIZED_OP(sc); 5200 PORT_LOCK_ASSERT_OWNED(pi); 5201 5202 if (lc->requested_aneg == AUTONEG_ENABLE) 5203 MPASS(lc->pcaps & FW_PORT_CAP32_ANEG); 5204 if (!(lc->requested_fc & PAUSE_AUTONEG)) 5205 MPASS(lc->pcaps & FW_PORT_CAP32_FORCE_PAUSE); 5206 if (lc->requested_fc & PAUSE_TX) 5207 MPASS(lc->pcaps & FW_PORT_CAP32_FC_TX); 5208 if (lc->requested_fc & PAUSE_RX) 5209 MPASS(lc->pcaps & FW_PORT_CAP32_FC_RX); 5210 if (lc->requested_fec & FEC_RS) 5211 MPASS(lc->pcaps & FW_PORT_CAP32_FEC_RS); 5212 if (lc->requested_fec & FEC_BASER_RS) 5213 MPASS(lc->pcaps & FW_PORT_CAP32_FEC_BASER_RS); 5214 #endif 5215 rc = -t4_link_l1cfg(sc, sc->mbox, pi->tx_chan, lc); 5216 if (rc != 0) { 5217 /* Don't complain if the VF driver gets back an EPERM. */ 5218 if (!(sc->flags & IS_VF) || rc != FW_EPERM) 5219 device_printf(pi->dev, "l1cfg failed: %d\n", rc); 5220 } else { 5221 /* 5222 * An L1_CFG will almost always result in a link-change event if 5223 * the link is up, and the driver will refresh the actual 5224 * fec/fc/etc. when the notification is processed. If the link 5225 * is down then the actual settings are meaningless. 5226 * 5227 * This takes care of the case where a change in the L1 settings 5228 * may not result in a notification. 5229 */ 5230 if (lc->link_ok && !(lc->requested_fc & PAUSE_AUTONEG)) 5231 lc->fc = lc->requested_fc & (PAUSE_TX | PAUSE_RX); 5232 } 5233 return (rc); 5234 } 5235 5236 #define FW_MAC_EXACT_CHUNK 7 5237 struct mcaddr_ctx { 5238 struct ifnet *ifp; 5239 const uint8_t *mcaddr[FW_MAC_EXACT_CHUNK]; 5240 uint64_t hash; 5241 int i; 5242 int del; 5243 int rc; 5244 }; 5245 5246 static u_int 5247 add_maddr(void *arg, struct sockaddr_dl *sdl, u_int cnt) 5248 { 5249 struct mcaddr_ctx *ctx = arg; 5250 struct vi_info *vi = ctx->ifp->if_softc; 5251 struct port_info *pi = vi->pi; 5252 struct adapter *sc = pi->adapter; 5253 5254 if (ctx->rc < 0) 5255 return (0); 5256 5257 ctx->mcaddr[ctx->i] = LLADDR(sdl); 5258 MPASS(ETHER_IS_MULTICAST(ctx->mcaddr[ctx->i])); 5259 ctx->i++; 5260 5261 if (ctx->i == FW_MAC_EXACT_CHUNK) { 5262 ctx->rc = t4_alloc_mac_filt(sc, sc->mbox, vi->viid, ctx->del, 5263 ctx->i, ctx->mcaddr, NULL, &ctx->hash, 0); 5264 if (ctx->rc < 0) { 5265 int j; 5266 5267 for (j = 0; j < ctx->i; j++) { 5268 if_printf(ctx->ifp, 5269 "failed to add mc address" 5270 " %02x:%02x:%02x:" 5271 "%02x:%02x:%02x rc=%d\n", 5272 ctx->mcaddr[j][0], ctx->mcaddr[j][1], 5273 ctx->mcaddr[j][2], ctx->mcaddr[j][3], 5274 ctx->mcaddr[j][4], ctx->mcaddr[j][5], 5275 -ctx->rc); 5276 } 5277 return (0); 5278 } 5279 ctx->del = 0; 5280 ctx->i = 0; 5281 } 5282 5283 return (1); 5284 } 5285 5286 /* 5287 * Program the port's XGMAC based on parameters in ifnet. The caller also 5288 * indicates which parameters should be programmed (the rest are left alone). 5289 */ 5290 int 5291 update_mac_settings(struct ifnet *ifp, int flags) 5292 { 5293 int rc = 0; 5294 struct vi_info *vi = ifp->if_softc; 5295 struct port_info *pi = vi->pi; 5296 struct adapter *sc = pi->adapter; 5297 int mtu = -1, promisc = -1, allmulti = -1, vlanex = -1; 5298 uint8_t match_all_mac[ETHER_ADDR_LEN] = {0}; 5299 5300 ASSERT_SYNCHRONIZED_OP(sc); 5301 KASSERT(flags, ("%s: not told what to update.", __func__)); 5302 5303 if (flags & XGMAC_MTU) 5304 mtu = ifp->if_mtu; 5305 5306 if (flags & XGMAC_PROMISC) 5307 promisc = ifp->if_flags & IFF_PROMISC ? 1 : 0; 5308 5309 if (flags & XGMAC_ALLMULTI) 5310 allmulti = ifp->if_flags & IFF_ALLMULTI ? 1 : 0; 5311 5312 if (flags & XGMAC_VLANEX) 5313 vlanex = ifp->if_capenable & IFCAP_VLAN_HWTAGGING ? 1 : 0; 5314 5315 if (flags & (XGMAC_MTU|XGMAC_PROMISC|XGMAC_ALLMULTI|XGMAC_VLANEX)) { 5316 rc = -t4_set_rxmode(sc, sc->mbox, vi->viid, mtu, promisc, 5317 allmulti, 1, vlanex, false); 5318 if (rc) { 5319 if_printf(ifp, "set_rxmode (%x) failed: %d\n", flags, 5320 rc); 5321 return (rc); 5322 } 5323 } 5324 5325 if (flags & XGMAC_UCADDR) { 5326 uint8_t ucaddr[ETHER_ADDR_LEN]; 5327 5328 bcopy(IF_LLADDR(ifp), ucaddr, sizeof(ucaddr)); 5329 rc = t4_change_mac(sc, sc->mbox, vi->viid, vi->xact_addr_filt, 5330 ucaddr, true, &vi->smt_idx); 5331 if (rc < 0) { 5332 rc = -rc; 5333 if_printf(ifp, "change_mac failed: %d\n", rc); 5334 return (rc); 5335 } else { 5336 vi->xact_addr_filt = rc; 5337 rc = 0; 5338 } 5339 } 5340 5341 if (flags & XGMAC_MCADDRS) { 5342 struct epoch_tracker et; 5343 struct mcaddr_ctx ctx; 5344 int j; 5345 5346 ctx.ifp = ifp; 5347 ctx.hash = 0; 5348 ctx.i = 0; 5349 ctx.del = 1; 5350 ctx.rc = 0; 5351 /* 5352 * Unlike other drivers, we accumulate list of pointers into 5353 * interface address lists and we need to keep it safe even 5354 * after if_foreach_llmaddr() returns, thus we must enter the 5355 * network epoch. 5356 */ 5357 NET_EPOCH_ENTER(et); 5358 if_foreach_llmaddr(ifp, add_maddr, &ctx); 5359 if (ctx.rc < 0) { 5360 NET_EPOCH_EXIT(et); 5361 rc = -ctx.rc; 5362 return (rc); 5363 } 5364 if (ctx.i > 0) { 5365 rc = t4_alloc_mac_filt(sc, sc->mbox, vi->viid, 5366 ctx.del, ctx.i, ctx.mcaddr, NULL, &ctx.hash, 0); 5367 NET_EPOCH_EXIT(et); 5368 if (rc < 0) { 5369 rc = -rc; 5370 for (j = 0; j < ctx.i; j++) { 5371 if_printf(ifp, 5372 "failed to add mcast address" 5373 " %02x:%02x:%02x:" 5374 "%02x:%02x:%02x rc=%d\n", 5375 ctx.mcaddr[j][0], ctx.mcaddr[j][1], 5376 ctx.mcaddr[j][2], ctx.mcaddr[j][3], 5377 ctx.mcaddr[j][4], ctx.mcaddr[j][5], 5378 rc); 5379 } 5380 return (rc); 5381 } 5382 ctx.del = 0; 5383 } else 5384 NET_EPOCH_EXIT(et); 5385 5386 rc = -t4_set_addr_hash(sc, sc->mbox, vi->viid, 0, ctx.hash, 0); 5387 if (rc != 0) 5388 if_printf(ifp, "failed to set mcast address hash: %d\n", 5389 rc); 5390 if (ctx.del == 0) { 5391 /* We clobbered the VXLAN entry if there was one. */ 5392 pi->vxlan_tcam_entry = false; 5393 } 5394 } 5395 5396 if (IS_MAIN_VI(vi) && sc->vxlan_refcount > 0 && 5397 pi->vxlan_tcam_entry == false) { 5398 rc = t4_alloc_raw_mac_filt(sc, vi->viid, match_all_mac, 5399 match_all_mac, sc->rawf_base + pi->port_id, 1, pi->port_id, 5400 true); 5401 if (rc < 0) { 5402 rc = -rc; 5403 if_printf(ifp, "failed to add VXLAN TCAM entry: %d.\n", 5404 rc); 5405 } else { 5406 MPASS(rc == sc->rawf_base + pi->port_id); 5407 rc = 0; 5408 pi->vxlan_tcam_entry = true; 5409 } 5410 } 5411 5412 return (rc); 5413 } 5414 5415 /* 5416 * {begin|end}_synchronized_op must be called from the same thread. 5417 */ 5418 int 5419 begin_synchronized_op(struct adapter *sc, struct vi_info *vi, int flags, 5420 char *wmesg) 5421 { 5422 int rc, pri; 5423 5424 #ifdef WITNESS 5425 /* the caller thinks it's ok to sleep, but is it really? */ 5426 if (flags & SLEEP_OK) 5427 WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL, 5428 "begin_synchronized_op"); 5429 #endif 5430 5431 if (INTR_OK) 5432 pri = PCATCH; 5433 else 5434 pri = 0; 5435 5436 ADAPTER_LOCK(sc); 5437 for (;;) { 5438 5439 if (vi && IS_DOOMED(vi)) { 5440 rc = ENXIO; 5441 goto done; 5442 } 5443 5444 if (!IS_BUSY(sc)) { 5445 rc = 0; 5446 break; 5447 } 5448 5449 if (!(flags & SLEEP_OK)) { 5450 rc = EBUSY; 5451 goto done; 5452 } 5453 5454 if (mtx_sleep(&sc->flags, &sc->sc_lock, pri, wmesg, 0)) { 5455 rc = EINTR; 5456 goto done; 5457 } 5458 } 5459 5460 KASSERT(!IS_BUSY(sc), ("%s: controller busy.", __func__)); 5461 SET_BUSY(sc); 5462 #ifdef INVARIANTS 5463 sc->last_op = wmesg; 5464 sc->last_op_thr = curthread; 5465 sc->last_op_flags = flags; 5466 #endif 5467 5468 done: 5469 if (!(flags & HOLD_LOCK) || rc) 5470 ADAPTER_UNLOCK(sc); 5471 5472 return (rc); 5473 } 5474 5475 /* 5476 * Tell if_ioctl and if_init that the VI is going away. This is 5477 * special variant of begin_synchronized_op and must be paired with a 5478 * call to end_synchronized_op. 5479 */ 5480 void 5481 doom_vi(struct adapter *sc, struct vi_info *vi) 5482 { 5483 5484 ADAPTER_LOCK(sc); 5485 SET_DOOMED(vi); 5486 wakeup(&sc->flags); 5487 while (IS_BUSY(sc)) 5488 mtx_sleep(&sc->flags, &sc->sc_lock, 0, "t4detach", 0); 5489 SET_BUSY(sc); 5490 #ifdef INVARIANTS 5491 sc->last_op = "t4detach"; 5492 sc->last_op_thr = curthread; 5493 sc->last_op_flags = 0; 5494 #endif 5495 ADAPTER_UNLOCK(sc); 5496 } 5497 5498 /* 5499 * {begin|end}_synchronized_op must be called from the same thread. 5500 */ 5501 void 5502 end_synchronized_op(struct adapter *sc, int flags) 5503 { 5504 5505 if (flags & LOCK_HELD) 5506 ADAPTER_LOCK_ASSERT_OWNED(sc); 5507 else 5508 ADAPTER_LOCK(sc); 5509 5510 KASSERT(IS_BUSY(sc), ("%s: controller not busy.", __func__)); 5511 CLR_BUSY(sc); 5512 wakeup(&sc->flags); 5513 ADAPTER_UNLOCK(sc); 5514 } 5515 5516 static int 5517 cxgbe_init_synchronized(struct vi_info *vi) 5518 { 5519 struct port_info *pi = vi->pi; 5520 struct adapter *sc = pi->adapter; 5521 struct ifnet *ifp = vi->ifp; 5522 int rc = 0, i; 5523 struct sge_txq *txq; 5524 5525 ASSERT_SYNCHRONIZED_OP(sc); 5526 5527 if (ifp->if_drv_flags & IFF_DRV_RUNNING) 5528 return (0); /* already running */ 5529 5530 if (!(sc->flags & FULL_INIT_DONE) && 5531 ((rc = adapter_full_init(sc)) != 0)) 5532 return (rc); /* error message displayed already */ 5533 5534 if (!(vi->flags & VI_INIT_DONE) && 5535 ((rc = vi_full_init(vi)) != 0)) 5536 return (rc); /* error message displayed already */ 5537 5538 rc = update_mac_settings(ifp, XGMAC_ALL); 5539 if (rc) 5540 goto done; /* error message displayed already */ 5541 5542 PORT_LOCK(pi); 5543 if (pi->up_vis == 0) { 5544 t4_update_port_info(pi); 5545 fixup_link_config(pi); 5546 build_medialist(pi); 5547 apply_link_config(pi); 5548 } 5549 5550 rc = -t4_enable_vi(sc, sc->mbox, vi->viid, true, true); 5551 if (rc != 0) { 5552 if_printf(ifp, "enable_vi failed: %d\n", rc); 5553 PORT_UNLOCK(pi); 5554 goto done; 5555 } 5556 5557 /* 5558 * Can't fail from this point onwards. Review cxgbe_uninit_synchronized 5559 * if this changes. 5560 */ 5561 5562 for_each_txq(vi, i, txq) { 5563 TXQ_LOCK(txq); 5564 txq->eq.flags |= EQ_ENABLED; 5565 TXQ_UNLOCK(txq); 5566 } 5567 5568 /* 5569 * The first iq of the first port to come up is used for tracing. 5570 */ 5571 if (sc->traceq < 0 && IS_MAIN_VI(vi)) { 5572 sc->traceq = sc->sge.rxq[vi->first_rxq].iq.abs_id; 5573 t4_write_reg(sc, is_t4(sc) ? A_MPS_TRC_RSS_CONTROL : 5574 A_MPS_T5_TRC_RSS_CONTROL, V_RSSCONTROL(pi->tx_chan) | 5575 V_QUEUENUMBER(sc->traceq)); 5576 pi->flags |= HAS_TRACEQ; 5577 } 5578 5579 /* all ok */ 5580 pi->up_vis++; 5581 ifp->if_drv_flags |= IFF_DRV_RUNNING; 5582 5583 if (pi->nvi > 1 || sc->flags & IS_VF) 5584 callout_reset(&vi->tick, hz, vi_tick, vi); 5585 else 5586 callout_reset(&pi->tick, hz, cxgbe_tick, pi); 5587 if (pi->link_cfg.link_ok) 5588 t4_os_link_changed(pi); 5589 PORT_UNLOCK(pi); 5590 done: 5591 if (rc != 0) 5592 cxgbe_uninit_synchronized(vi); 5593 5594 return (rc); 5595 } 5596 5597 /* 5598 * Idempotent. 5599 */ 5600 static int 5601 cxgbe_uninit_synchronized(struct vi_info *vi) 5602 { 5603 struct port_info *pi = vi->pi; 5604 struct adapter *sc = pi->adapter; 5605 struct ifnet *ifp = vi->ifp; 5606 int rc, i; 5607 struct sge_txq *txq; 5608 5609 ASSERT_SYNCHRONIZED_OP(sc); 5610 5611 if (!(vi->flags & VI_INIT_DONE)) { 5612 if (__predict_false(ifp->if_drv_flags & IFF_DRV_RUNNING)) { 5613 KASSERT(0, ("uninited VI is running")); 5614 if_printf(ifp, "uninited VI with running ifnet. " 5615 "vi->flags 0x%016lx, if_flags 0x%08x, " 5616 "if_drv_flags 0x%08x\n", vi->flags, ifp->if_flags, 5617 ifp->if_drv_flags); 5618 } 5619 return (0); 5620 } 5621 5622 /* 5623 * Disable the VI so that all its data in either direction is discarded 5624 * by the MPS. Leave everything else (the queues, interrupts, and 1Hz 5625 * tick) intact as the TP can deliver negative advice or data that it's 5626 * holding in its RAM (for an offloaded connection) even after the VI is 5627 * disabled. 5628 */ 5629 rc = -t4_enable_vi(sc, sc->mbox, vi->viid, false, false); 5630 if (rc) { 5631 if_printf(ifp, "disable_vi failed: %d\n", rc); 5632 return (rc); 5633 } 5634 5635 for_each_txq(vi, i, txq) { 5636 TXQ_LOCK(txq); 5637 txq->eq.flags &= ~EQ_ENABLED; 5638 TXQ_UNLOCK(txq); 5639 } 5640 5641 PORT_LOCK(pi); 5642 if (pi->nvi > 1 || sc->flags & IS_VF) 5643 callout_stop(&vi->tick); 5644 else 5645 callout_stop(&pi->tick); 5646 if (!(ifp->if_drv_flags & IFF_DRV_RUNNING)) { 5647 PORT_UNLOCK(pi); 5648 return (0); 5649 } 5650 ifp->if_drv_flags &= ~IFF_DRV_RUNNING; 5651 pi->up_vis--; 5652 if (pi->up_vis > 0) { 5653 PORT_UNLOCK(pi); 5654 return (0); 5655 } 5656 5657 pi->link_cfg.link_ok = false; 5658 pi->link_cfg.speed = 0; 5659 pi->link_cfg.link_down_rc = 255; 5660 t4_os_link_changed(pi); 5661 PORT_UNLOCK(pi); 5662 5663 return (0); 5664 } 5665 5666 /* 5667 * It is ok for this function to fail midway and return right away. t4_detach 5668 * will walk the entire sc->irq list and clean up whatever is valid. 5669 */ 5670 int 5671 t4_setup_intr_handlers(struct adapter *sc) 5672 { 5673 int rc, rid, p, q, v; 5674 char s[8]; 5675 struct irq *irq; 5676 struct port_info *pi; 5677 struct vi_info *vi; 5678 struct sge *sge = &sc->sge; 5679 struct sge_rxq *rxq; 5680 #ifdef TCP_OFFLOAD 5681 struct sge_ofld_rxq *ofld_rxq; 5682 #endif 5683 #ifdef DEV_NETMAP 5684 struct sge_nm_rxq *nm_rxq; 5685 #endif 5686 #ifdef RSS 5687 int nbuckets = rss_getnumbuckets(); 5688 #endif 5689 5690 /* 5691 * Setup interrupts. 5692 */ 5693 irq = &sc->irq[0]; 5694 rid = sc->intr_type == INTR_INTX ? 0 : 1; 5695 if (forwarding_intr_to_fwq(sc)) 5696 return (t4_alloc_irq(sc, irq, rid, t4_intr_all, sc, "all")); 5697 5698 /* Multiple interrupts. */ 5699 if (sc->flags & IS_VF) 5700 KASSERT(sc->intr_count >= T4VF_EXTRA_INTR + sc->params.nports, 5701 ("%s: too few intr.", __func__)); 5702 else 5703 KASSERT(sc->intr_count >= T4_EXTRA_INTR + sc->params.nports, 5704 ("%s: too few intr.", __func__)); 5705 5706 /* The first one is always error intr on PFs */ 5707 if (!(sc->flags & IS_VF)) { 5708 rc = t4_alloc_irq(sc, irq, rid, t4_intr_err, sc, "err"); 5709 if (rc != 0) 5710 return (rc); 5711 irq++; 5712 rid++; 5713 } 5714 5715 /* The second one is always the firmware event queue (first on VFs) */ 5716 rc = t4_alloc_irq(sc, irq, rid, t4_intr_evt, &sge->fwq, "evt"); 5717 if (rc != 0) 5718 return (rc); 5719 irq++; 5720 rid++; 5721 5722 for_each_port(sc, p) { 5723 pi = sc->port[p]; 5724 for_each_vi(pi, v, vi) { 5725 vi->first_intr = rid - 1; 5726 5727 if (vi->nnmrxq > 0) { 5728 int n = max(vi->nrxq, vi->nnmrxq); 5729 5730 rxq = &sge->rxq[vi->first_rxq]; 5731 #ifdef DEV_NETMAP 5732 nm_rxq = &sge->nm_rxq[vi->first_nm_rxq]; 5733 #endif 5734 for (q = 0; q < n; q++) { 5735 snprintf(s, sizeof(s), "%x%c%x", p, 5736 'a' + v, q); 5737 if (q < vi->nrxq) 5738 irq->rxq = rxq++; 5739 #ifdef DEV_NETMAP 5740 if (q < vi->nnmrxq) 5741 irq->nm_rxq = nm_rxq++; 5742 5743 if (irq->nm_rxq != NULL && 5744 irq->rxq == NULL) { 5745 /* Netmap rx only */ 5746 rc = t4_alloc_irq(sc, irq, rid, 5747 t4_nm_intr, irq->nm_rxq, s); 5748 } 5749 if (irq->nm_rxq != NULL && 5750 irq->rxq != NULL) { 5751 /* NIC and Netmap rx */ 5752 rc = t4_alloc_irq(sc, irq, rid, 5753 t4_vi_intr, irq, s); 5754 } 5755 #endif 5756 if (irq->rxq != NULL && 5757 irq->nm_rxq == NULL) { 5758 /* NIC rx only */ 5759 rc = t4_alloc_irq(sc, irq, rid, 5760 t4_intr, irq->rxq, s); 5761 } 5762 if (rc != 0) 5763 return (rc); 5764 #ifdef RSS 5765 if (q < vi->nrxq) { 5766 bus_bind_intr(sc->dev, irq->res, 5767 rss_getcpu(q % nbuckets)); 5768 } 5769 #endif 5770 irq++; 5771 rid++; 5772 vi->nintr++; 5773 } 5774 } else { 5775 for_each_rxq(vi, q, rxq) { 5776 snprintf(s, sizeof(s), "%x%c%x", p, 5777 'a' + v, q); 5778 rc = t4_alloc_irq(sc, irq, rid, 5779 t4_intr, rxq, s); 5780 if (rc != 0) 5781 return (rc); 5782 #ifdef RSS 5783 bus_bind_intr(sc->dev, irq->res, 5784 rss_getcpu(q % nbuckets)); 5785 #endif 5786 irq++; 5787 rid++; 5788 vi->nintr++; 5789 } 5790 } 5791 #ifdef TCP_OFFLOAD 5792 for_each_ofld_rxq(vi, q, ofld_rxq) { 5793 snprintf(s, sizeof(s), "%x%c%x", p, 'A' + v, q); 5794 rc = t4_alloc_irq(sc, irq, rid, t4_intr, 5795 ofld_rxq, s); 5796 if (rc != 0) 5797 return (rc); 5798 irq++; 5799 rid++; 5800 vi->nintr++; 5801 } 5802 #endif 5803 } 5804 } 5805 MPASS(irq == &sc->irq[sc->intr_count]); 5806 5807 return (0); 5808 } 5809 5810 int 5811 adapter_full_init(struct adapter *sc) 5812 { 5813 int rc, i; 5814 #ifdef RSS 5815 uint32_t raw_rss_key[RSS_KEYSIZE / sizeof(uint32_t)]; 5816 uint32_t rss_key[RSS_KEYSIZE / sizeof(uint32_t)]; 5817 #endif 5818 5819 ASSERT_SYNCHRONIZED_OP(sc); 5820 ADAPTER_LOCK_ASSERT_NOTOWNED(sc); 5821 KASSERT((sc->flags & FULL_INIT_DONE) == 0, 5822 ("%s: FULL_INIT_DONE already", __func__)); 5823 5824 /* 5825 * queues that belong to the adapter (not any particular port). 5826 */ 5827 rc = t4_setup_adapter_queues(sc); 5828 if (rc != 0) 5829 goto done; 5830 5831 for (i = 0; i < nitems(sc->tq); i++) { 5832 sc->tq[i] = taskqueue_create("t4 taskq", M_NOWAIT, 5833 taskqueue_thread_enqueue, &sc->tq[i]); 5834 if (sc->tq[i] == NULL) { 5835 device_printf(sc->dev, 5836 "failed to allocate task queue %d\n", i); 5837 rc = ENOMEM; 5838 goto done; 5839 } 5840 taskqueue_start_threads(&sc->tq[i], 1, PI_NET, "%s tq%d", 5841 device_get_nameunit(sc->dev), i); 5842 } 5843 #ifdef RSS 5844 MPASS(RSS_KEYSIZE == 40); 5845 rss_getkey((void *)&raw_rss_key[0]); 5846 for (i = 0; i < nitems(rss_key); i++) { 5847 rss_key[i] = htobe32(raw_rss_key[nitems(rss_key) - 1 - i]); 5848 } 5849 t4_write_rss_key(sc, &rss_key[0], -1, 1); 5850 #endif 5851 5852 if (!(sc->flags & IS_VF)) 5853 t4_intr_enable(sc); 5854 #ifdef KERN_TLS 5855 if (sc->flags & KERN_TLS_OK) 5856 callout_reset_sbt(&sc->ktls_tick, SBT_1MS, 0, ktls_tick, sc, 5857 C_HARDCLOCK); 5858 #endif 5859 sc->flags |= FULL_INIT_DONE; 5860 done: 5861 if (rc != 0) 5862 adapter_full_uninit(sc); 5863 5864 return (rc); 5865 } 5866 5867 int 5868 adapter_full_uninit(struct adapter *sc) 5869 { 5870 int i; 5871 5872 ADAPTER_LOCK_ASSERT_NOTOWNED(sc); 5873 5874 t4_teardown_adapter_queues(sc); 5875 5876 for (i = 0; i < nitems(sc->tq) && sc->tq[i]; i++) { 5877 taskqueue_free(sc->tq[i]); 5878 sc->tq[i] = NULL; 5879 } 5880 5881 sc->flags &= ~FULL_INIT_DONE; 5882 5883 return (0); 5884 } 5885 5886 #ifdef RSS 5887 #define SUPPORTED_RSS_HASHTYPES (RSS_HASHTYPE_RSS_IPV4 | \ 5888 RSS_HASHTYPE_RSS_TCP_IPV4 | RSS_HASHTYPE_RSS_IPV6 | \ 5889 RSS_HASHTYPE_RSS_TCP_IPV6 | RSS_HASHTYPE_RSS_UDP_IPV4 | \ 5890 RSS_HASHTYPE_RSS_UDP_IPV6) 5891 5892 /* Translates kernel hash types to hardware. */ 5893 static int 5894 hashconfig_to_hashen(int hashconfig) 5895 { 5896 int hashen = 0; 5897 5898 if (hashconfig & RSS_HASHTYPE_RSS_IPV4) 5899 hashen |= F_FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN; 5900 if (hashconfig & RSS_HASHTYPE_RSS_IPV6) 5901 hashen |= F_FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN; 5902 if (hashconfig & RSS_HASHTYPE_RSS_UDP_IPV4) { 5903 hashen |= F_FW_RSS_VI_CONFIG_CMD_UDPEN | 5904 F_FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN; 5905 } 5906 if (hashconfig & RSS_HASHTYPE_RSS_UDP_IPV6) { 5907 hashen |= F_FW_RSS_VI_CONFIG_CMD_UDPEN | 5908 F_FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN; 5909 } 5910 if (hashconfig & RSS_HASHTYPE_RSS_TCP_IPV4) 5911 hashen |= F_FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN; 5912 if (hashconfig & RSS_HASHTYPE_RSS_TCP_IPV6) 5913 hashen |= F_FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN; 5914 5915 return (hashen); 5916 } 5917 5918 /* Translates hardware hash types to kernel. */ 5919 static int 5920 hashen_to_hashconfig(int hashen) 5921 { 5922 int hashconfig = 0; 5923 5924 if (hashen & F_FW_RSS_VI_CONFIG_CMD_UDPEN) { 5925 /* 5926 * If UDP hashing was enabled it must have been enabled for 5927 * either IPv4 or IPv6 (inclusive or). Enabling UDP without 5928 * enabling any 4-tuple hash is nonsense configuration. 5929 */ 5930 MPASS(hashen & (F_FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN | 5931 F_FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN)); 5932 5933 if (hashen & F_FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN) 5934 hashconfig |= RSS_HASHTYPE_RSS_UDP_IPV4; 5935 if (hashen & F_FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN) 5936 hashconfig |= RSS_HASHTYPE_RSS_UDP_IPV6; 5937 } 5938 if (hashen & F_FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN) 5939 hashconfig |= RSS_HASHTYPE_RSS_TCP_IPV4; 5940 if (hashen & F_FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN) 5941 hashconfig |= RSS_HASHTYPE_RSS_TCP_IPV6; 5942 if (hashen & F_FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN) 5943 hashconfig |= RSS_HASHTYPE_RSS_IPV4; 5944 if (hashen & F_FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN) 5945 hashconfig |= RSS_HASHTYPE_RSS_IPV6; 5946 5947 return (hashconfig); 5948 } 5949 #endif 5950 5951 int 5952 vi_full_init(struct vi_info *vi) 5953 { 5954 struct adapter *sc = vi->adapter; 5955 struct ifnet *ifp = vi->ifp; 5956 uint16_t *rss; 5957 struct sge_rxq *rxq; 5958 int rc, i, j; 5959 #ifdef RSS 5960 int nbuckets = rss_getnumbuckets(); 5961 int hashconfig = rss_gethashconfig(); 5962 int extra; 5963 #endif 5964 5965 ASSERT_SYNCHRONIZED_OP(sc); 5966 KASSERT((vi->flags & VI_INIT_DONE) == 0, 5967 ("%s: VI_INIT_DONE already", __func__)); 5968 5969 sysctl_ctx_init(&vi->ctx); 5970 vi->flags |= VI_SYSCTL_CTX; 5971 5972 /* 5973 * Allocate tx/rx/fl queues for this VI. 5974 */ 5975 rc = t4_setup_vi_queues(vi); 5976 if (rc != 0) 5977 goto done; /* error message displayed already */ 5978 5979 /* 5980 * Setup RSS for this VI. Save a copy of the RSS table for later use. 5981 */ 5982 if (vi->nrxq > vi->rss_size) { 5983 if_printf(ifp, "nrxq (%d) > hw RSS table size (%d); " 5984 "some queues will never receive traffic.\n", vi->nrxq, 5985 vi->rss_size); 5986 } else if (vi->rss_size % vi->nrxq) { 5987 if_printf(ifp, "nrxq (%d), hw RSS table size (%d); " 5988 "expect uneven traffic distribution.\n", vi->nrxq, 5989 vi->rss_size); 5990 } 5991 #ifdef RSS 5992 if (vi->nrxq != nbuckets) { 5993 if_printf(ifp, "nrxq (%d) != kernel RSS buckets (%d);" 5994 "performance will be impacted.\n", vi->nrxq, nbuckets); 5995 } 5996 #endif 5997 rss = malloc(vi->rss_size * sizeof (*rss), M_CXGBE, M_ZERO | M_WAITOK); 5998 for (i = 0; i < vi->rss_size;) { 5999 #ifdef RSS 6000 j = rss_get_indirection_to_bucket(i); 6001 j %= vi->nrxq; 6002 rxq = &sc->sge.rxq[vi->first_rxq + j]; 6003 rss[i++] = rxq->iq.abs_id; 6004 #else 6005 for_each_rxq(vi, j, rxq) { 6006 rss[i++] = rxq->iq.abs_id; 6007 if (i == vi->rss_size) 6008 break; 6009 } 6010 #endif 6011 } 6012 6013 rc = -t4_config_rss_range(sc, sc->mbox, vi->viid, 0, vi->rss_size, rss, 6014 vi->rss_size); 6015 if (rc != 0) { 6016 free(rss, M_CXGBE); 6017 if_printf(ifp, "rss_config failed: %d\n", rc); 6018 goto done; 6019 } 6020 6021 #ifdef RSS 6022 vi->hashen = hashconfig_to_hashen(hashconfig); 6023 6024 /* 6025 * We may have had to enable some hashes even though the global config 6026 * wants them disabled. This is a potential problem that must be 6027 * reported to the user. 6028 */ 6029 extra = hashen_to_hashconfig(vi->hashen) ^ hashconfig; 6030 6031 /* 6032 * If we consider only the supported hash types, then the enabled hashes 6033 * are a superset of the requested hashes. In other words, there cannot 6034 * be any supported hash that was requested but not enabled, but there 6035 * can be hashes that were not requested but had to be enabled. 6036 */ 6037 extra &= SUPPORTED_RSS_HASHTYPES; 6038 MPASS((extra & hashconfig) == 0); 6039 6040 if (extra) { 6041 if_printf(ifp, 6042 "global RSS config (0x%x) cannot be accommodated.\n", 6043 hashconfig); 6044 } 6045 if (extra & RSS_HASHTYPE_RSS_IPV4) 6046 if_printf(ifp, "IPv4 2-tuple hashing forced on.\n"); 6047 if (extra & RSS_HASHTYPE_RSS_TCP_IPV4) 6048 if_printf(ifp, "TCP/IPv4 4-tuple hashing forced on.\n"); 6049 if (extra & RSS_HASHTYPE_RSS_IPV6) 6050 if_printf(ifp, "IPv6 2-tuple hashing forced on.\n"); 6051 if (extra & RSS_HASHTYPE_RSS_TCP_IPV6) 6052 if_printf(ifp, "TCP/IPv6 4-tuple hashing forced on.\n"); 6053 if (extra & RSS_HASHTYPE_RSS_UDP_IPV4) 6054 if_printf(ifp, "UDP/IPv4 4-tuple hashing forced on.\n"); 6055 if (extra & RSS_HASHTYPE_RSS_UDP_IPV6) 6056 if_printf(ifp, "UDP/IPv6 4-tuple hashing forced on.\n"); 6057 #else 6058 vi->hashen = F_FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN | 6059 F_FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN | 6060 F_FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN | 6061 F_FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN | F_FW_RSS_VI_CONFIG_CMD_UDPEN; 6062 #endif 6063 rc = -t4_config_vi_rss(sc, sc->mbox, vi->viid, vi->hashen, rss[0], 0, 0); 6064 if (rc != 0) { 6065 free(rss, M_CXGBE); 6066 if_printf(ifp, "rss hash/defaultq config failed: %d\n", rc); 6067 goto done; 6068 } 6069 6070 vi->rss = rss; 6071 vi->flags |= VI_INIT_DONE; 6072 done: 6073 if (rc != 0) 6074 vi_full_uninit(vi); 6075 6076 return (rc); 6077 } 6078 6079 /* 6080 * Idempotent. 6081 */ 6082 int 6083 vi_full_uninit(struct vi_info *vi) 6084 { 6085 struct port_info *pi = vi->pi; 6086 struct adapter *sc = pi->adapter; 6087 int i; 6088 struct sge_rxq *rxq; 6089 struct sge_txq *txq; 6090 #ifdef TCP_OFFLOAD 6091 struct sge_ofld_rxq *ofld_rxq; 6092 #endif 6093 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 6094 struct sge_wrq *ofld_txq; 6095 #endif 6096 6097 if (vi->flags & VI_INIT_DONE) { 6098 6099 /* Need to quiesce queues. */ 6100 6101 /* XXX: Only for the first VI? */ 6102 if (IS_MAIN_VI(vi) && !(sc->flags & IS_VF)) 6103 quiesce_wrq(sc, &sc->sge.ctrlq[pi->port_id]); 6104 6105 for_each_txq(vi, i, txq) { 6106 quiesce_txq(sc, txq); 6107 } 6108 6109 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 6110 for_each_ofld_txq(vi, i, ofld_txq) { 6111 quiesce_wrq(sc, ofld_txq); 6112 } 6113 #endif 6114 6115 for_each_rxq(vi, i, rxq) { 6116 quiesce_iq(sc, &rxq->iq); 6117 quiesce_fl(sc, &rxq->fl); 6118 } 6119 6120 #ifdef TCP_OFFLOAD 6121 for_each_ofld_rxq(vi, i, ofld_rxq) { 6122 quiesce_iq(sc, &ofld_rxq->iq); 6123 quiesce_fl(sc, &ofld_rxq->fl); 6124 } 6125 #endif 6126 free(vi->rss, M_CXGBE); 6127 free(vi->nm_rss, M_CXGBE); 6128 } 6129 6130 t4_teardown_vi_queues(vi); 6131 vi->flags &= ~VI_INIT_DONE; 6132 6133 return (0); 6134 } 6135 6136 static void 6137 quiesce_txq(struct adapter *sc, struct sge_txq *txq) 6138 { 6139 struct sge_eq *eq = &txq->eq; 6140 struct sge_qstat *spg = (void *)&eq->desc[eq->sidx]; 6141 6142 (void) sc; /* unused */ 6143 6144 #ifdef INVARIANTS 6145 TXQ_LOCK(txq); 6146 MPASS((eq->flags & EQ_ENABLED) == 0); 6147 TXQ_UNLOCK(txq); 6148 #endif 6149 6150 /* Wait for the mp_ring to empty. */ 6151 while (!mp_ring_is_idle(txq->r)) { 6152 mp_ring_check_drainage(txq->r, 4096); 6153 pause("rquiesce", 1); 6154 } 6155 6156 /* Then wait for the hardware to finish. */ 6157 while (spg->cidx != htobe16(eq->pidx)) 6158 pause("equiesce", 1); 6159 6160 /* Finally, wait for the driver to reclaim all descriptors. */ 6161 while (eq->cidx != eq->pidx) 6162 pause("dquiesce", 1); 6163 } 6164 6165 static void 6166 quiesce_wrq(struct adapter *sc, struct sge_wrq *wrq) 6167 { 6168 6169 /* XXXTX */ 6170 } 6171 6172 static void 6173 quiesce_iq(struct adapter *sc, struct sge_iq *iq) 6174 { 6175 (void) sc; /* unused */ 6176 6177 /* Synchronize with the interrupt handler */ 6178 while (!atomic_cmpset_int(&iq->state, IQS_IDLE, IQS_DISABLED)) 6179 pause("iqfree", 1); 6180 } 6181 6182 static void 6183 quiesce_fl(struct adapter *sc, struct sge_fl *fl) 6184 { 6185 mtx_lock(&sc->sfl_lock); 6186 FL_LOCK(fl); 6187 fl->flags |= FL_DOOMED; 6188 FL_UNLOCK(fl); 6189 callout_stop(&sc->sfl_callout); 6190 mtx_unlock(&sc->sfl_lock); 6191 6192 KASSERT((fl->flags & FL_STARVING) == 0, 6193 ("%s: still starving", __func__)); 6194 } 6195 6196 static int 6197 t4_alloc_irq(struct adapter *sc, struct irq *irq, int rid, 6198 driver_intr_t *handler, void *arg, char *name) 6199 { 6200 int rc; 6201 6202 irq->rid = rid; 6203 irq->res = bus_alloc_resource_any(sc->dev, SYS_RES_IRQ, &irq->rid, 6204 RF_SHAREABLE | RF_ACTIVE); 6205 if (irq->res == NULL) { 6206 device_printf(sc->dev, 6207 "failed to allocate IRQ for rid %d, name %s.\n", rid, name); 6208 return (ENOMEM); 6209 } 6210 6211 rc = bus_setup_intr(sc->dev, irq->res, INTR_MPSAFE | INTR_TYPE_NET, 6212 NULL, handler, arg, &irq->tag); 6213 if (rc != 0) { 6214 device_printf(sc->dev, 6215 "failed to setup interrupt for rid %d, name %s: %d\n", 6216 rid, name, rc); 6217 } else if (name) 6218 bus_describe_intr(sc->dev, irq->res, irq->tag, "%s", name); 6219 6220 return (rc); 6221 } 6222 6223 static int 6224 t4_free_irq(struct adapter *sc, struct irq *irq) 6225 { 6226 if (irq->tag) 6227 bus_teardown_intr(sc->dev, irq->res, irq->tag); 6228 if (irq->res) 6229 bus_release_resource(sc->dev, SYS_RES_IRQ, irq->rid, irq->res); 6230 6231 bzero(irq, sizeof(*irq)); 6232 6233 return (0); 6234 } 6235 6236 static void 6237 get_regs(struct adapter *sc, struct t4_regdump *regs, uint8_t *buf) 6238 { 6239 6240 regs->version = chip_id(sc) | chip_rev(sc) << 10; 6241 t4_get_regs(sc, buf, regs->len); 6242 } 6243 6244 #define A_PL_INDIR_CMD 0x1f8 6245 6246 #define S_PL_AUTOINC 31 6247 #define M_PL_AUTOINC 0x1U 6248 #define V_PL_AUTOINC(x) ((x) << S_PL_AUTOINC) 6249 #define G_PL_AUTOINC(x) (((x) >> S_PL_AUTOINC) & M_PL_AUTOINC) 6250 6251 #define S_PL_VFID 20 6252 #define M_PL_VFID 0xffU 6253 #define V_PL_VFID(x) ((x) << S_PL_VFID) 6254 #define G_PL_VFID(x) (((x) >> S_PL_VFID) & M_PL_VFID) 6255 6256 #define S_PL_ADDR 0 6257 #define M_PL_ADDR 0xfffffU 6258 #define V_PL_ADDR(x) ((x) << S_PL_ADDR) 6259 #define G_PL_ADDR(x) (((x) >> S_PL_ADDR) & M_PL_ADDR) 6260 6261 #define A_PL_INDIR_DATA 0x1fc 6262 6263 static uint64_t 6264 read_vf_stat(struct adapter *sc, u_int vin, int reg) 6265 { 6266 u32 stats[2]; 6267 6268 mtx_assert(&sc->reg_lock, MA_OWNED); 6269 if (sc->flags & IS_VF) { 6270 stats[0] = t4_read_reg(sc, VF_MPS_REG(reg)); 6271 stats[1] = t4_read_reg(sc, VF_MPS_REG(reg + 4)); 6272 } else { 6273 t4_write_reg(sc, A_PL_INDIR_CMD, V_PL_AUTOINC(1) | 6274 V_PL_VFID(vin) | V_PL_ADDR(VF_MPS_REG(reg))); 6275 stats[0] = t4_read_reg(sc, A_PL_INDIR_DATA); 6276 stats[1] = t4_read_reg(sc, A_PL_INDIR_DATA); 6277 } 6278 return (((uint64_t)stats[1]) << 32 | stats[0]); 6279 } 6280 6281 static void 6282 t4_get_vi_stats(struct adapter *sc, u_int vin, struct fw_vi_stats_vf *stats) 6283 { 6284 6285 #define GET_STAT(name) \ 6286 read_vf_stat(sc, vin, A_MPS_VF_STAT_##name##_L) 6287 6288 stats->tx_bcast_bytes = GET_STAT(TX_VF_BCAST_BYTES); 6289 stats->tx_bcast_frames = GET_STAT(TX_VF_BCAST_FRAMES); 6290 stats->tx_mcast_bytes = GET_STAT(TX_VF_MCAST_BYTES); 6291 stats->tx_mcast_frames = GET_STAT(TX_VF_MCAST_FRAMES); 6292 stats->tx_ucast_bytes = GET_STAT(TX_VF_UCAST_BYTES); 6293 stats->tx_ucast_frames = GET_STAT(TX_VF_UCAST_FRAMES); 6294 stats->tx_drop_frames = GET_STAT(TX_VF_DROP_FRAMES); 6295 stats->tx_offload_bytes = GET_STAT(TX_VF_OFFLOAD_BYTES); 6296 stats->tx_offload_frames = GET_STAT(TX_VF_OFFLOAD_FRAMES); 6297 stats->rx_bcast_bytes = GET_STAT(RX_VF_BCAST_BYTES); 6298 stats->rx_bcast_frames = GET_STAT(RX_VF_BCAST_FRAMES); 6299 stats->rx_mcast_bytes = GET_STAT(RX_VF_MCAST_BYTES); 6300 stats->rx_mcast_frames = GET_STAT(RX_VF_MCAST_FRAMES); 6301 stats->rx_ucast_bytes = GET_STAT(RX_VF_UCAST_BYTES); 6302 stats->rx_ucast_frames = GET_STAT(RX_VF_UCAST_FRAMES); 6303 stats->rx_err_frames = GET_STAT(RX_VF_ERR_FRAMES); 6304 6305 #undef GET_STAT 6306 } 6307 6308 static void 6309 t4_clr_vi_stats(struct adapter *sc, u_int vin) 6310 { 6311 int reg; 6312 6313 t4_write_reg(sc, A_PL_INDIR_CMD, V_PL_AUTOINC(1) | V_PL_VFID(vin) | 6314 V_PL_ADDR(VF_MPS_REG(A_MPS_VF_STAT_TX_VF_BCAST_BYTES_L))); 6315 for (reg = A_MPS_VF_STAT_TX_VF_BCAST_BYTES_L; 6316 reg <= A_MPS_VF_STAT_RX_VF_ERR_FRAMES_H; reg += 4) 6317 t4_write_reg(sc, A_PL_INDIR_DATA, 0); 6318 } 6319 6320 static void 6321 vi_refresh_stats(struct adapter *sc, struct vi_info *vi) 6322 { 6323 struct timeval tv; 6324 const struct timeval interval = {0, 250000}; /* 250ms */ 6325 6326 if (!(vi->flags & VI_INIT_DONE)) 6327 return; 6328 6329 getmicrotime(&tv); 6330 timevalsub(&tv, &interval); 6331 if (timevalcmp(&tv, &vi->last_refreshed, <)) 6332 return; 6333 6334 mtx_lock(&sc->reg_lock); 6335 t4_get_vi_stats(sc, vi->vin, &vi->stats); 6336 getmicrotime(&vi->last_refreshed); 6337 mtx_unlock(&sc->reg_lock); 6338 } 6339 6340 static void 6341 cxgbe_refresh_stats(struct adapter *sc, struct port_info *pi) 6342 { 6343 u_int i, v, tnl_cong_drops, chan_map; 6344 struct timeval tv; 6345 const struct timeval interval = {0, 250000}; /* 250ms */ 6346 6347 getmicrotime(&tv); 6348 timevalsub(&tv, &interval); 6349 if (timevalcmp(&tv, &pi->last_refreshed, <)) 6350 return; 6351 6352 tnl_cong_drops = 0; 6353 t4_get_port_stats(sc, pi->tx_chan, &pi->stats); 6354 chan_map = pi->rx_e_chan_map; 6355 while (chan_map) { 6356 i = ffs(chan_map) - 1; 6357 mtx_lock(&sc->reg_lock); 6358 t4_read_indirect(sc, A_TP_MIB_INDEX, A_TP_MIB_DATA, &v, 1, 6359 A_TP_MIB_TNL_CNG_DROP_0 + i); 6360 mtx_unlock(&sc->reg_lock); 6361 tnl_cong_drops += v; 6362 chan_map &= ~(1 << i); 6363 } 6364 pi->tnl_cong_drops = tnl_cong_drops; 6365 getmicrotime(&pi->last_refreshed); 6366 } 6367 6368 static void 6369 cxgbe_tick(void *arg) 6370 { 6371 struct port_info *pi = arg; 6372 struct adapter *sc = pi->adapter; 6373 6374 PORT_LOCK_ASSERT_OWNED(pi); 6375 cxgbe_refresh_stats(sc, pi); 6376 6377 callout_schedule(&pi->tick, hz); 6378 } 6379 6380 void 6381 vi_tick(void *arg) 6382 { 6383 struct vi_info *vi = arg; 6384 struct adapter *sc = vi->adapter; 6385 6386 vi_refresh_stats(sc, vi); 6387 6388 callout_schedule(&vi->tick, hz); 6389 } 6390 6391 /* 6392 * Should match fw_caps_config_<foo> enums in t4fw_interface.h 6393 */ 6394 static char *caps_decoder[] = { 6395 "\20\001IPMI\002NCSI", /* 0: NBM */ 6396 "\20\001PPP\002QFC\003DCBX", /* 1: link */ 6397 "\20\001INGRESS\002EGRESS", /* 2: switch */ 6398 "\20\001NIC\002VM\003IDS\004UM\005UM_ISGL" /* 3: NIC */ 6399 "\006HASHFILTER\007ETHOFLD", 6400 "\20\001TOE", /* 4: TOE */ 6401 "\20\001RDDP\002RDMAC", /* 5: RDMA */ 6402 "\20\001INITIATOR_PDU\002TARGET_PDU" /* 6: iSCSI */ 6403 "\003INITIATOR_CNXOFLD\004TARGET_CNXOFLD" 6404 "\005INITIATOR_SSNOFLD\006TARGET_SSNOFLD" 6405 "\007T10DIF" 6406 "\010INITIATOR_CMDOFLD\011TARGET_CMDOFLD", 6407 "\20\001LOOKASIDE\002TLSKEYS", /* 7: Crypto */ 6408 "\20\001INITIATOR\002TARGET\003CTRL_OFLD" /* 8: FCoE */ 6409 "\004PO_INITIATOR\005PO_TARGET", 6410 }; 6411 6412 void 6413 t4_sysctls(struct adapter *sc) 6414 { 6415 struct sysctl_ctx_list *ctx; 6416 struct sysctl_oid *oid; 6417 struct sysctl_oid_list *children, *c0; 6418 static char *doorbells = {"\20\1UDB\2WCWR\3UDBWC\4KDB"}; 6419 6420 ctx = device_get_sysctl_ctx(sc->dev); 6421 6422 /* 6423 * dev.t4nex.X. 6424 */ 6425 oid = device_get_sysctl_tree(sc->dev); 6426 c0 = children = SYSCTL_CHILDREN(oid); 6427 6428 sc->sc_do_rxcopy = 1; 6429 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "do_rx_copy", CTLFLAG_RW, 6430 &sc->sc_do_rxcopy, 1, "Do RX copy of small frames"); 6431 6432 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "nports", CTLFLAG_RD, NULL, 6433 sc->params.nports, "# of ports"); 6434 6435 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "doorbells", 6436 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, doorbells, 6437 (uintptr_t)&sc->doorbells, sysctl_bitfield_8b, "A", 6438 "available doorbells"); 6439 6440 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "core_clock", CTLFLAG_RD, NULL, 6441 sc->params.vpd.cclk, "core clock frequency (in KHz)"); 6442 6443 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "holdoff_timers", 6444 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, 6445 sc->params.sge.timer_val, sizeof(sc->params.sge.timer_val), 6446 sysctl_int_array, "A", "interrupt holdoff timer values (us)"); 6447 6448 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "holdoff_pkt_counts", 6449 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, 6450 sc->params.sge.counter_val, sizeof(sc->params.sge.counter_val), 6451 sysctl_int_array, "A", "interrupt holdoff packet counter values"); 6452 6453 t4_sge_sysctls(sc, ctx, children); 6454 6455 sc->lro_timeout = 100; 6456 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "lro_timeout", CTLFLAG_RW, 6457 &sc->lro_timeout, 0, "lro inactive-flush timeout (in us)"); 6458 6459 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "dflags", CTLFLAG_RW, 6460 &sc->debug_flags, 0, "flags to enable runtime debugging"); 6461 6462 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "tp_version", 6463 CTLFLAG_RD, sc->tp_version, 0, "TP microcode version"); 6464 6465 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "firmware_version", 6466 CTLFLAG_RD, sc->fw_version, 0, "firmware version"); 6467 6468 if (sc->flags & IS_VF) 6469 return; 6470 6471 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "hw_revision", CTLFLAG_RD, 6472 NULL, chip_rev(sc), "chip hardware revision"); 6473 6474 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "sn", 6475 CTLFLAG_RD, sc->params.vpd.sn, 0, "serial number"); 6476 6477 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "pn", 6478 CTLFLAG_RD, sc->params.vpd.pn, 0, "part number"); 6479 6480 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "ec", 6481 CTLFLAG_RD, sc->params.vpd.ec, 0, "engineering change"); 6482 6483 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "md_version", 6484 CTLFLAG_RD, sc->params.vpd.md, 0, "manufacturing diags version"); 6485 6486 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "na", 6487 CTLFLAG_RD, sc->params.vpd.na, 0, "network address"); 6488 6489 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "er_version", CTLFLAG_RD, 6490 sc->er_version, 0, "expansion ROM version"); 6491 6492 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "bs_version", CTLFLAG_RD, 6493 sc->bs_version, 0, "bootstrap firmware version"); 6494 6495 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "scfg_version", CTLFLAG_RD, 6496 NULL, sc->params.scfg_vers, "serial config version"); 6497 6498 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "vpd_version", CTLFLAG_RD, 6499 NULL, sc->params.vpd_vers, "VPD version"); 6500 6501 SYSCTL_ADD_STRING(ctx, children, OID_AUTO, "cf", 6502 CTLFLAG_RD, sc->cfg_file, 0, "configuration file"); 6503 6504 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "cfcsum", CTLFLAG_RD, NULL, 6505 sc->cfcsum, "config file checksum"); 6506 6507 #define SYSCTL_CAP(name, n, text) \ 6508 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, #name, \ 6509 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, caps_decoder[n], \ 6510 (uintptr_t)&sc->name, sysctl_bitfield_16b, "A", \ 6511 "available " text " capabilities") 6512 6513 SYSCTL_CAP(nbmcaps, 0, "NBM"); 6514 SYSCTL_CAP(linkcaps, 1, "link"); 6515 SYSCTL_CAP(switchcaps, 2, "switch"); 6516 SYSCTL_CAP(niccaps, 3, "NIC"); 6517 SYSCTL_CAP(toecaps, 4, "TCP offload"); 6518 SYSCTL_CAP(rdmacaps, 5, "RDMA"); 6519 SYSCTL_CAP(iscsicaps, 6, "iSCSI"); 6520 SYSCTL_CAP(cryptocaps, 7, "crypto"); 6521 SYSCTL_CAP(fcoecaps, 8, "FCoE"); 6522 #undef SYSCTL_CAP 6523 6524 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "nfilters", CTLFLAG_RD, 6525 NULL, sc->tids.nftids, "number of filters"); 6526 6527 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "temperature", 6528 CTLTYPE_INT | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6529 sysctl_temperature, "I", "chip temperature (in Celsius)"); 6530 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "reset_sensor", 6531 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, sc, 0, 6532 sysctl_reset_sensor, "I", "reset the chip's temperature sensor."); 6533 6534 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "loadavg", 6535 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6536 sysctl_loadavg, "A", 6537 "microprocessor load averages (debug firmwares only)"); 6538 6539 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "core_vdd", 6540 CTLTYPE_INT | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, sysctl_vdd, 6541 "I", "core Vdd (in mV)"); 6542 6543 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "local_cpus", 6544 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, LOCAL_CPUS, 6545 sysctl_cpus, "A", "local CPUs"); 6546 6547 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "intr_cpus", 6548 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, INTR_CPUS, 6549 sysctl_cpus, "A", "preferred CPUs for interrupts"); 6550 6551 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "swintr", CTLFLAG_RW, 6552 &sc->swintr, 0, "software triggered interrupts"); 6553 6554 /* 6555 * dev.t4nex.X.misc. Marked CTLFLAG_SKIP to avoid information overload. 6556 */ 6557 oid = SYSCTL_ADD_NODE(ctx, c0, OID_AUTO, "misc", 6558 CTLFLAG_RD | CTLFLAG_SKIP | CTLFLAG_MPSAFE, NULL, 6559 "logs and miscellaneous information"); 6560 children = SYSCTL_CHILDREN(oid); 6561 6562 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cctrl", 6563 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6564 sysctl_cctrl, "A", "congestion control"); 6565 6566 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_ibq_tp0", 6567 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6568 sysctl_cim_ibq_obq, "A", "CIM IBQ 0 (TP0)"); 6569 6570 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_ibq_tp1", 6571 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 1, 6572 sysctl_cim_ibq_obq, "A", "CIM IBQ 1 (TP1)"); 6573 6574 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_ibq_ulp", 6575 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 2, 6576 sysctl_cim_ibq_obq, "A", "CIM IBQ 2 (ULP)"); 6577 6578 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_ibq_sge0", 6579 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 3, 6580 sysctl_cim_ibq_obq, "A", "CIM IBQ 3 (SGE0)"); 6581 6582 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_ibq_sge1", 6583 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 4, 6584 sysctl_cim_ibq_obq, "A", "CIM IBQ 4 (SGE1)"); 6585 6586 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_ibq_ncsi", 6587 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 5, 6588 sysctl_cim_ibq_obq, "A", "CIM IBQ 5 (NCSI)"); 6589 6590 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_la", 6591 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6592 sysctl_cim_la, "A", "CIM logic analyzer"); 6593 6594 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_ma_la", 6595 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6596 sysctl_cim_ma_la, "A", "CIM MA logic analyzer"); 6597 6598 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_ulp0", 6599 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6600 0 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", "CIM OBQ 0 (ULP0)"); 6601 6602 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_ulp1", 6603 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6604 1 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", "CIM OBQ 1 (ULP1)"); 6605 6606 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_ulp2", 6607 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6608 2 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", "CIM OBQ 2 (ULP2)"); 6609 6610 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_ulp3", 6611 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6612 3 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", "CIM OBQ 3 (ULP3)"); 6613 6614 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_sge", 6615 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6616 4 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", "CIM OBQ 4 (SGE)"); 6617 6618 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_ncsi", 6619 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6620 5 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", "CIM OBQ 5 (NCSI)"); 6621 6622 if (chip_id(sc) > CHELSIO_T4) { 6623 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_sge0_rx", 6624 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6625 6 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", 6626 "CIM OBQ 6 (SGE0-RX)"); 6627 6628 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_obq_sge1_rx", 6629 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6630 7 + CIM_NUM_IBQ, sysctl_cim_ibq_obq, "A", 6631 "CIM OBQ 7 (SGE1-RX)"); 6632 } 6633 6634 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_pif_la", 6635 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6636 sysctl_cim_pif_la, "A", "CIM PIF logic analyzer"); 6637 6638 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cim_qcfg", 6639 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6640 sysctl_cim_qcfg, "A", "CIM queue configuration"); 6641 6642 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "cpl_stats", 6643 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6644 sysctl_cpl_stats, "A", "CPL statistics"); 6645 6646 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "ddp_stats", 6647 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6648 sysctl_ddp_stats, "A", "non-TCP DDP statistics"); 6649 6650 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tid_stats", 6651 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6652 sysctl_tid_stats, "A", "tid stats"); 6653 6654 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "devlog", 6655 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6656 sysctl_devlog, "A", "firmware's device log"); 6657 6658 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "fcoe_stats", 6659 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6660 sysctl_fcoe_stats, "A", "FCoE statistics"); 6661 6662 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "hw_sched", 6663 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6664 sysctl_hw_sched, "A", "hardware scheduler "); 6665 6666 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "l2t", 6667 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6668 sysctl_l2t, "A", "hardware L2 table"); 6669 6670 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "smt", 6671 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6672 sysctl_smt, "A", "hardware source MAC table"); 6673 6674 #ifdef INET6 6675 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "clip", 6676 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6677 sysctl_clip, "A", "active CLIP table entries"); 6678 #endif 6679 6680 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "lb_stats", 6681 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6682 sysctl_lb_stats, "A", "loopback statistics"); 6683 6684 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "meminfo", 6685 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6686 sysctl_meminfo, "A", "memory regions"); 6687 6688 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "mps_tcam", 6689 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6690 chip_id(sc) <= CHELSIO_T5 ? sysctl_mps_tcam : sysctl_mps_tcam_t6, 6691 "A", "MPS TCAM entries"); 6692 6693 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "path_mtus", 6694 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6695 sysctl_path_mtus, "A", "path MTUs"); 6696 6697 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "pm_stats", 6698 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6699 sysctl_pm_stats, "A", "PM statistics"); 6700 6701 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "rdma_stats", 6702 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6703 sysctl_rdma_stats, "A", "RDMA statistics"); 6704 6705 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tcp_stats", 6706 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6707 sysctl_tcp_stats, "A", "TCP statistics"); 6708 6709 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tids", 6710 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6711 sysctl_tids, "A", "TID information"); 6712 6713 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tp_err_stats", 6714 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6715 sysctl_tp_err_stats, "A", "TP error statistics"); 6716 6717 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tnl_stats", 6718 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6719 sysctl_tnl_stats, "A", "TP tunnel statistics"); 6720 6721 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tp_la_mask", 6722 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, sc, 0, 6723 sysctl_tp_la_mask, "I", "TP logic analyzer event capture mask"); 6724 6725 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tp_la", 6726 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6727 sysctl_tp_la, "A", "TP logic analyzer"); 6728 6729 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tx_rate", 6730 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6731 sysctl_tx_rate, "A", "Tx rate"); 6732 6733 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "ulprx_la", 6734 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6735 sysctl_ulprx_la, "A", "ULPRX logic analyzer"); 6736 6737 if (chip_id(sc) >= CHELSIO_T5) { 6738 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "wcwr_stats", 6739 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6740 sysctl_wcwr_stats, "A", "write combined work requests"); 6741 } 6742 6743 #ifdef KERN_TLS 6744 if (sc->flags & KERN_TLS_OK) { 6745 /* 6746 * dev.t4nex.0.tls. 6747 */ 6748 oid = SYSCTL_ADD_NODE(ctx, c0, OID_AUTO, "tls", 6749 CTLFLAG_RD | CTLFLAG_MPSAFE, NULL, "KERN_TLS parameters"); 6750 children = SYSCTL_CHILDREN(oid); 6751 6752 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "inline_keys", 6753 CTLFLAG_RW, &sc->tlst.inline_keys, 0, "Always pass TLS " 6754 "keys in work requests (1) or attempt to store TLS keys " 6755 "in card memory."); 6756 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "combo_wrs", 6757 CTLFLAG_RW, &sc->tlst.combo_wrs, 0, "Attempt to combine " 6758 "TCB field updates with TLS record work requests."); 6759 } 6760 #endif 6761 6762 #ifdef TCP_OFFLOAD 6763 if (is_offload(sc)) { 6764 int i; 6765 char s[4]; 6766 6767 /* 6768 * dev.t4nex.X.toe. 6769 */ 6770 oid = SYSCTL_ADD_NODE(ctx, c0, OID_AUTO, "toe", 6771 CTLFLAG_RD | CTLFLAG_MPSAFE, NULL, "TOE parameters"); 6772 children = SYSCTL_CHILDREN(oid); 6773 6774 sc->tt.cong_algorithm = -1; 6775 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "cong_algorithm", 6776 CTLFLAG_RW, &sc->tt.cong_algorithm, 0, "congestion control " 6777 "(-1 = default, 0 = reno, 1 = tahoe, 2 = newreno, " 6778 "3 = highspeed)"); 6779 6780 sc->tt.sndbuf = -1; 6781 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "sndbuf", CTLFLAG_RW, 6782 &sc->tt.sndbuf, 0, "hardware send buffer"); 6783 6784 sc->tt.ddp = 0; 6785 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "ddp", 6786 CTLFLAG_RW | CTLFLAG_SKIP, &sc->tt.ddp, 0, ""); 6787 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "rx_zcopy", CTLFLAG_RW, 6788 &sc->tt.ddp, 0, "Enable zero-copy aio_read(2)"); 6789 6790 sc->tt.rx_coalesce = -1; 6791 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "rx_coalesce", 6792 CTLFLAG_RW, &sc->tt.rx_coalesce, 0, "receive coalescing"); 6793 6794 sc->tt.tls = 0; 6795 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tls", CTLTYPE_INT | 6796 CTLFLAG_RW | CTLFLAG_MPSAFE, sc, 0, sysctl_tls, "I", 6797 "Inline TLS allowed"); 6798 6799 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tls_rx_ports", 6800 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, sc, 0, 6801 sysctl_tls_rx_ports, "I", 6802 "TCP ports that use inline TLS+TOE RX"); 6803 6804 sc->tt.tls_rx_timeout = t4_toe_tls_rx_timeout; 6805 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tls_rx_timeout", 6806 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, sc, 0, 6807 sysctl_tls_rx_timeout, "I", 6808 "Timeout in seconds to downgrade TLS sockets to plain TOE"); 6809 6810 sc->tt.tx_align = -1; 6811 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "tx_align", 6812 CTLFLAG_RW, &sc->tt.tx_align, 0, "chop and align payload"); 6813 6814 sc->tt.tx_zcopy = 0; 6815 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "tx_zcopy", 6816 CTLFLAG_RW, &sc->tt.tx_zcopy, 0, 6817 "Enable zero-copy aio_write(2)"); 6818 6819 sc->tt.cop_managed_offloading = !!t4_cop_managed_offloading; 6820 SYSCTL_ADD_INT(ctx, children, OID_AUTO, 6821 "cop_managed_offloading", CTLFLAG_RW, 6822 &sc->tt.cop_managed_offloading, 0, 6823 "COP (Connection Offload Policy) controls all TOE offload"); 6824 6825 sc->tt.autorcvbuf_inc = 16 * 1024; 6826 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "autorcvbuf_inc", 6827 CTLFLAG_RW, &sc->tt.autorcvbuf_inc, 0, 6828 "autorcvbuf increment"); 6829 6830 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "timer_tick", 6831 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6832 sysctl_tp_tick, "A", "TP timer tick (us)"); 6833 6834 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "timestamp_tick", 6835 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 1, 6836 sysctl_tp_tick, "A", "TCP timestamp tick (us)"); 6837 6838 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "dack_tick", 6839 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 2, 6840 sysctl_tp_tick, "A", "DACK tick (us)"); 6841 6842 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "dack_timer", 6843 CTLTYPE_UINT | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 0, 6844 sysctl_tp_dack_timer, "IU", "DACK timer (us)"); 6845 6846 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "rexmt_min", 6847 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6848 A_TP_RXT_MIN, sysctl_tp_timer, "LU", 6849 "Minimum retransmit interval (us)"); 6850 6851 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "rexmt_max", 6852 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6853 A_TP_RXT_MAX, sysctl_tp_timer, "LU", 6854 "Maximum retransmit interval (us)"); 6855 6856 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "persist_min", 6857 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6858 A_TP_PERS_MIN, sysctl_tp_timer, "LU", 6859 "Persist timer min (us)"); 6860 6861 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "persist_max", 6862 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6863 A_TP_PERS_MAX, sysctl_tp_timer, "LU", 6864 "Persist timer max (us)"); 6865 6866 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "keepalive_idle", 6867 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6868 A_TP_KEEP_IDLE, sysctl_tp_timer, "LU", 6869 "Keepalive idle timer (us)"); 6870 6871 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "keepalive_interval", 6872 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6873 A_TP_KEEP_INTVL, sysctl_tp_timer, "LU", 6874 "Keepalive interval timer (us)"); 6875 6876 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "initial_srtt", 6877 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6878 A_TP_INIT_SRTT, sysctl_tp_timer, "LU", "Initial SRTT (us)"); 6879 6880 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "finwait2_timer", 6881 CTLTYPE_ULONG | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6882 A_TP_FINWAIT2_TIMER, sysctl_tp_timer, "LU", 6883 "FINWAIT2 timer (us)"); 6884 6885 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "syn_rexmt_count", 6886 CTLTYPE_UINT | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6887 S_SYNSHIFTMAX, sysctl_tp_shift_cnt, "IU", 6888 "Number of SYN retransmissions before abort"); 6889 6890 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "rexmt_count", 6891 CTLTYPE_UINT | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6892 S_RXTSHIFTMAXR2, sysctl_tp_shift_cnt, "IU", 6893 "Number of retransmissions before abort"); 6894 6895 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "keepalive_count", 6896 CTLTYPE_UINT | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6897 S_KEEPALIVEMAXR2, sysctl_tp_shift_cnt, "IU", 6898 "Number of keepalive probes before abort"); 6899 6900 oid = SYSCTL_ADD_NODE(ctx, children, OID_AUTO, "rexmt_backoff", 6901 CTLFLAG_RD | CTLFLAG_MPSAFE, NULL, 6902 "TOE retransmit backoffs"); 6903 children = SYSCTL_CHILDREN(oid); 6904 for (i = 0; i < 16; i++) { 6905 snprintf(s, sizeof(s), "%u", i); 6906 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, s, 6907 CTLTYPE_UINT | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 6908 i, sysctl_tp_backoff, "IU", 6909 "TOE retransmit backoff"); 6910 } 6911 } 6912 #endif 6913 } 6914 6915 void 6916 vi_sysctls(struct vi_info *vi) 6917 { 6918 struct sysctl_ctx_list *ctx; 6919 struct sysctl_oid *oid; 6920 struct sysctl_oid_list *children; 6921 6922 ctx = device_get_sysctl_ctx(vi->dev); 6923 6924 /* 6925 * dev.v?(cxgbe|cxl).X. 6926 */ 6927 oid = device_get_sysctl_tree(vi->dev); 6928 children = SYSCTL_CHILDREN(oid); 6929 6930 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "viid", CTLFLAG_RD, NULL, 6931 vi->viid, "VI identifer"); 6932 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "nrxq", CTLFLAG_RD, 6933 &vi->nrxq, 0, "# of rx queues"); 6934 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "ntxq", CTLFLAG_RD, 6935 &vi->ntxq, 0, "# of tx queues"); 6936 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "first_rxq", CTLFLAG_RD, 6937 &vi->first_rxq, 0, "index of first rx queue"); 6938 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "first_txq", CTLFLAG_RD, 6939 &vi->first_txq, 0, "index of first tx queue"); 6940 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "rss_base", CTLFLAG_RD, NULL, 6941 vi->rss_base, "start of RSS indirection table"); 6942 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "rss_size", CTLFLAG_RD, NULL, 6943 vi->rss_size, "size of RSS indirection table"); 6944 6945 if (IS_MAIN_VI(vi)) { 6946 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "rsrv_noflowq", 6947 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 6948 sysctl_noflowq, "IU", 6949 "Reserve queue 0 for non-flowid packets"); 6950 } 6951 6952 if (vi->adapter->flags & IS_VF) { 6953 MPASS(vi->flags & TX_USES_VM_WR); 6954 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "tx_vm_wr", CTLFLAG_RD, 6955 NULL, 1, "use VM work requests for transmit"); 6956 } else { 6957 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "tx_vm_wr", 6958 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 6959 sysctl_tx_vm_wr, "I", "use VM work requestes for transmit"); 6960 } 6961 6962 #ifdef TCP_OFFLOAD 6963 if (vi->nofldrxq != 0) { 6964 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "nofldrxq", CTLFLAG_RD, 6965 &vi->nofldrxq, 0, 6966 "# of rx queues for offloaded TCP connections"); 6967 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "first_ofld_rxq", 6968 CTLFLAG_RD, &vi->first_ofld_rxq, 0, 6969 "index of first TOE rx queue"); 6970 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "holdoff_tmr_idx_ofld", 6971 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 6972 sysctl_holdoff_tmr_idx_ofld, "I", 6973 "holdoff timer index for TOE queues"); 6974 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "holdoff_pktc_idx_ofld", 6975 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 6976 sysctl_holdoff_pktc_idx_ofld, "I", 6977 "holdoff packet counter index for TOE queues"); 6978 } 6979 #endif 6980 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 6981 if (vi->nofldtxq != 0) { 6982 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "nofldtxq", CTLFLAG_RD, 6983 &vi->nofldtxq, 0, 6984 "# of tx queues for TOE/ETHOFLD"); 6985 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "first_ofld_txq", 6986 CTLFLAG_RD, &vi->first_ofld_txq, 0, 6987 "index of first TOE/ETHOFLD tx queue"); 6988 } 6989 #endif 6990 #ifdef DEV_NETMAP 6991 if (vi->nnmrxq != 0) { 6992 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "nnmrxq", CTLFLAG_RD, 6993 &vi->nnmrxq, 0, "# of netmap rx queues"); 6994 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "nnmtxq", CTLFLAG_RD, 6995 &vi->nnmtxq, 0, "# of netmap tx queues"); 6996 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "first_nm_rxq", 6997 CTLFLAG_RD, &vi->first_nm_rxq, 0, 6998 "index of first netmap rx queue"); 6999 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "first_nm_txq", 7000 CTLFLAG_RD, &vi->first_nm_txq, 0, 7001 "index of first netmap tx queue"); 7002 } 7003 #endif 7004 7005 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "holdoff_tmr_idx", 7006 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 7007 sysctl_holdoff_tmr_idx, "I", "holdoff timer index"); 7008 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "holdoff_pktc_idx", 7009 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 7010 sysctl_holdoff_pktc_idx, "I", "holdoff packet counter index"); 7011 7012 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "qsize_rxq", 7013 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 7014 sysctl_qsize_rxq, "I", "rx queue size"); 7015 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "qsize_txq", 7016 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, vi, 0, 7017 sysctl_qsize_txq, "I", "tx queue size"); 7018 } 7019 7020 static void 7021 cxgbe_sysctls(struct port_info *pi) 7022 { 7023 struct sysctl_ctx_list *ctx; 7024 struct sysctl_oid *oid; 7025 struct sysctl_oid_list *children, *children2; 7026 struct adapter *sc = pi->adapter; 7027 int i; 7028 char name[16]; 7029 static char *tc_flags = {"\20\1USER\2SYNC\3ASYNC\4ERR"}; 7030 7031 ctx = device_get_sysctl_ctx(pi->dev); 7032 7033 /* 7034 * dev.cxgbe.X. 7035 */ 7036 oid = device_get_sysctl_tree(pi->dev); 7037 children = SYSCTL_CHILDREN(oid); 7038 7039 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "linkdnrc", 7040 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, pi, 0, 7041 sysctl_linkdnrc, "A", "reason why link is down"); 7042 if (pi->port_type == FW_PORT_TYPE_BT_XAUI) { 7043 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "temperature", 7044 CTLTYPE_INT | CTLFLAG_RD | CTLFLAG_MPSAFE, pi, 0, 7045 sysctl_btphy, "I", "PHY temperature (in Celsius)"); 7046 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "fw_version", 7047 CTLTYPE_INT | CTLFLAG_RD | CTLFLAG_MPSAFE, pi, 1, 7048 sysctl_btphy, "I", "PHY firmware version"); 7049 } 7050 7051 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "pause_settings", 7052 CTLTYPE_STRING | CTLFLAG_RW | CTLFLAG_MPSAFE, pi, 0, 7053 sysctl_pause_settings, "A", 7054 "PAUSE settings (bit 0 = rx_pause, 1 = tx_pause, 2 = pause_autoneg)"); 7055 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "fec", 7056 CTLTYPE_STRING | CTLFLAG_RW | CTLFLAG_MPSAFE, pi, 0, 7057 sysctl_fec, "A", 7058 "FECs to use (bit 0 = RS, 1 = FC, 2 = none, 5 = auto, 6 = module)"); 7059 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "module_fec", 7060 CTLTYPE_STRING | CTLFLAG_MPSAFE, pi, 0, sysctl_module_fec, "A", 7061 "FEC recommended by the cable/transceiver"); 7062 SYSCTL_ADD_PROC(ctx, children, OID_AUTO, "autoneg", 7063 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, pi, 0, 7064 sysctl_autoneg, "I", 7065 "autonegotiation (-1 = not supported)"); 7066 7067 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "pcaps", CTLFLAG_RD, 7068 &pi->link_cfg.pcaps, 0, "port capabilities"); 7069 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "acaps", CTLFLAG_RD, 7070 &pi->link_cfg.acaps, 0, "advertised capabilities"); 7071 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "lpacaps", CTLFLAG_RD, 7072 &pi->link_cfg.lpacaps, 0, "link partner advertised capabilities"); 7073 7074 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "max_speed", CTLFLAG_RD, NULL, 7075 port_top_speed(pi), "max speed (in Gbps)"); 7076 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "mps_bg_map", CTLFLAG_RD, NULL, 7077 pi->mps_bg_map, "MPS buffer group map"); 7078 SYSCTL_ADD_INT(ctx, children, OID_AUTO, "rx_e_chan_map", CTLFLAG_RD, 7079 NULL, pi->rx_e_chan_map, "TP rx e-channel map"); 7080 7081 if (sc->flags & IS_VF) 7082 return; 7083 7084 /* 7085 * dev.(cxgbe|cxl).X.tc. 7086 */ 7087 oid = SYSCTL_ADD_NODE(ctx, children, OID_AUTO, "tc", 7088 CTLFLAG_RD | CTLFLAG_MPSAFE, NULL, 7089 "Tx scheduler traffic classes (cl_rl)"); 7090 children2 = SYSCTL_CHILDREN(oid); 7091 SYSCTL_ADD_UINT(ctx, children2, OID_AUTO, "pktsize", 7092 CTLFLAG_RW, &pi->sched_params->pktsize, 0, 7093 "pktsize for per-flow cl-rl (0 means up to the driver )"); 7094 SYSCTL_ADD_UINT(ctx, children2, OID_AUTO, "burstsize", 7095 CTLFLAG_RW, &pi->sched_params->burstsize, 0, 7096 "burstsize for per-flow cl-rl (0 means up to the driver)"); 7097 for (i = 0; i < sc->chip_params->nsched_cls; i++) { 7098 struct tx_cl_rl_params *tc = &pi->sched_params->cl_rl[i]; 7099 7100 snprintf(name, sizeof(name), "%d", i); 7101 children2 = SYSCTL_CHILDREN(SYSCTL_ADD_NODE(ctx, 7102 SYSCTL_CHILDREN(oid), OID_AUTO, name, 7103 CTLFLAG_RD | CTLFLAG_MPSAFE, NULL, "traffic class")); 7104 SYSCTL_ADD_PROC(ctx, children2, OID_AUTO, "flags", 7105 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, tc_flags, 7106 (uintptr_t)&tc->flags, sysctl_bitfield_8b, "A", "flags"); 7107 SYSCTL_ADD_UINT(ctx, children2, OID_AUTO, "refcount", 7108 CTLFLAG_RD, &tc->refcount, 0, "references to this class"); 7109 SYSCTL_ADD_PROC(ctx, children2, OID_AUTO, "params", 7110 CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, 7111 (pi->port_id << 16) | i, sysctl_tc_params, "A", 7112 "traffic class parameters"); 7113 } 7114 7115 /* 7116 * dev.cxgbe.X.stats. 7117 */ 7118 oid = SYSCTL_ADD_NODE(ctx, children, OID_AUTO, "stats", 7119 CTLFLAG_RD | CTLFLAG_MPSAFE, NULL, "port statistics"); 7120 children = SYSCTL_CHILDREN(oid); 7121 SYSCTL_ADD_UINT(ctx, children, OID_AUTO, "tx_parse_error", CTLFLAG_RD, 7122 &pi->tx_parse_error, 0, 7123 "# of tx packets with invalid length or # of segments"); 7124 7125 #define T4_REGSTAT(name, stat, desc) \ 7126 SYSCTL_ADD_OID(ctx, children, OID_AUTO, #name, \ 7127 CTLTYPE_U64 | CTLFLAG_RD | CTLFLAG_MPSAFE, sc, \ 7128 (is_t4(sc) ? PORT_REG(pi->tx_chan, A_MPS_PORT_STAT_##stat##_L) : \ 7129 T5_PORT_REG(pi->tx_chan, A_MPS_PORT_STAT_##stat##_L)), \ 7130 sysctl_handle_t4_reg64, "QU", desc) 7131 7132 /* We get these from port_stats and they may be stale by up to 1s */ 7133 #define T4_PORTSTAT(name, desc) \ 7134 SYSCTL_ADD_UQUAD(ctx, children, OID_AUTO, #name, CTLFLAG_RD, \ 7135 &pi->stats.name, desc) 7136 7137 T4_REGSTAT(tx_octets, TX_PORT_BYTES, "# of octets in good frames"); 7138 T4_REGSTAT(tx_frames, TX_PORT_FRAMES, "total # of good frames"); 7139 T4_REGSTAT(tx_bcast_frames, TX_PORT_BCAST, "# of broadcast frames"); 7140 T4_REGSTAT(tx_mcast_frames, TX_PORT_MCAST, "# of multicast frames"); 7141 T4_REGSTAT(tx_ucast_frames, TX_PORT_UCAST, "# of unicast frames"); 7142 T4_REGSTAT(tx_error_frames, TX_PORT_ERROR, "# of error frames"); 7143 T4_REGSTAT(tx_frames_64, TX_PORT_64B, "# of tx frames in this range"); 7144 T4_REGSTAT(tx_frames_65_127, TX_PORT_65B_127B, "# of tx frames in this range"); 7145 T4_REGSTAT(tx_frames_128_255, TX_PORT_128B_255B, "# of tx frames in this range"); 7146 T4_REGSTAT(tx_frames_256_511, TX_PORT_256B_511B, "# of tx frames in this range"); 7147 T4_REGSTAT(tx_frames_512_1023, TX_PORT_512B_1023B, "# of tx frames in this range"); 7148 T4_REGSTAT(tx_frames_1024_1518, TX_PORT_1024B_1518B, "# of tx frames in this range"); 7149 T4_REGSTAT(tx_frames_1519_max, TX_PORT_1519B_MAX, "# of tx frames in this range"); 7150 T4_REGSTAT(tx_drop, TX_PORT_DROP, "# of dropped tx frames"); 7151 T4_REGSTAT(tx_pause, TX_PORT_PAUSE, "# of pause frames transmitted"); 7152 T4_REGSTAT(tx_ppp0, TX_PORT_PPP0, "# of PPP prio 0 frames transmitted"); 7153 T4_REGSTAT(tx_ppp1, TX_PORT_PPP1, "# of PPP prio 1 frames transmitted"); 7154 T4_REGSTAT(tx_ppp2, TX_PORT_PPP2, "# of PPP prio 2 frames transmitted"); 7155 T4_REGSTAT(tx_ppp3, TX_PORT_PPP3, "# of PPP prio 3 frames transmitted"); 7156 T4_REGSTAT(tx_ppp4, TX_PORT_PPP4, "# of PPP prio 4 frames transmitted"); 7157 T4_REGSTAT(tx_ppp5, TX_PORT_PPP5, "# of PPP prio 5 frames transmitted"); 7158 T4_REGSTAT(tx_ppp6, TX_PORT_PPP6, "# of PPP prio 6 frames transmitted"); 7159 T4_REGSTAT(tx_ppp7, TX_PORT_PPP7, "# of PPP prio 7 frames transmitted"); 7160 7161 T4_REGSTAT(rx_octets, RX_PORT_BYTES, "# of octets in good frames"); 7162 T4_REGSTAT(rx_frames, RX_PORT_FRAMES, "total # of good frames"); 7163 T4_REGSTAT(rx_bcast_frames, RX_PORT_BCAST, "# of broadcast frames"); 7164 T4_REGSTAT(rx_mcast_frames, RX_PORT_MCAST, "# of multicast frames"); 7165 T4_REGSTAT(rx_ucast_frames, RX_PORT_UCAST, "# of unicast frames"); 7166 T4_REGSTAT(rx_too_long, RX_PORT_MTU_ERROR, "# of frames exceeding MTU"); 7167 T4_REGSTAT(rx_jabber, RX_PORT_MTU_CRC_ERROR, "# of jabber frames"); 7168 if (is_t6(sc)) { 7169 T4_PORTSTAT(rx_fcs_err, 7170 "# of frames received with bad FCS since last link up"); 7171 } else { 7172 T4_REGSTAT(rx_fcs_err, RX_PORT_CRC_ERROR, 7173 "# of frames received with bad FCS"); 7174 } 7175 T4_REGSTAT(rx_len_err, RX_PORT_LEN_ERROR, "# of frames received with length error"); 7176 T4_REGSTAT(rx_symbol_err, RX_PORT_SYM_ERROR, "symbol errors"); 7177 T4_REGSTAT(rx_runt, RX_PORT_LESS_64B, "# of short frames received"); 7178 T4_REGSTAT(rx_frames_64, RX_PORT_64B, "# of rx frames in this range"); 7179 T4_REGSTAT(rx_frames_65_127, RX_PORT_65B_127B, "# of rx frames in this range"); 7180 T4_REGSTAT(rx_frames_128_255, RX_PORT_128B_255B, "# of rx frames in this range"); 7181 T4_REGSTAT(rx_frames_256_511, RX_PORT_256B_511B, "# of rx frames in this range"); 7182 T4_REGSTAT(rx_frames_512_1023, RX_PORT_512B_1023B, "# of rx frames in this range"); 7183 T4_REGSTAT(rx_frames_1024_1518, RX_PORT_1024B_1518B, "# of rx frames in this range"); 7184 T4_REGSTAT(rx_frames_1519_max, RX_PORT_1519B_MAX, "# of rx frames in this range"); 7185 T4_REGSTAT(rx_pause, RX_PORT_PAUSE, "# of pause frames received"); 7186 T4_REGSTAT(rx_ppp0, RX_PORT_PPP0, "# of PPP prio 0 frames received"); 7187 T4_REGSTAT(rx_ppp1, RX_PORT_PPP1, "# of PPP prio 1 frames received"); 7188 T4_REGSTAT(rx_ppp2, RX_PORT_PPP2, "# of PPP prio 2 frames received"); 7189 T4_REGSTAT(rx_ppp3, RX_PORT_PPP3, "# of PPP prio 3 frames received"); 7190 T4_REGSTAT(rx_ppp4, RX_PORT_PPP4, "# of PPP prio 4 frames received"); 7191 T4_REGSTAT(rx_ppp5, RX_PORT_PPP5, "# of PPP prio 5 frames received"); 7192 T4_REGSTAT(rx_ppp6, RX_PORT_PPP6, "# of PPP prio 6 frames received"); 7193 T4_REGSTAT(rx_ppp7, RX_PORT_PPP7, "# of PPP prio 7 frames received"); 7194 7195 T4_PORTSTAT(rx_ovflow0, "# drops due to buffer-group 0 overflows"); 7196 T4_PORTSTAT(rx_ovflow1, "# drops due to buffer-group 1 overflows"); 7197 T4_PORTSTAT(rx_ovflow2, "# drops due to buffer-group 2 overflows"); 7198 T4_PORTSTAT(rx_ovflow3, "# drops due to buffer-group 3 overflows"); 7199 T4_PORTSTAT(rx_trunc0, "# of buffer-group 0 truncated packets"); 7200 T4_PORTSTAT(rx_trunc1, "# of buffer-group 1 truncated packets"); 7201 T4_PORTSTAT(rx_trunc2, "# of buffer-group 2 truncated packets"); 7202 T4_PORTSTAT(rx_trunc3, "# of buffer-group 3 truncated packets"); 7203 7204 #undef T4_REGSTAT 7205 #undef T4_PORTSTAT 7206 7207 SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "tx_toe_tls_records", 7208 CTLFLAG_RD, &pi->tx_toe_tls_records, 7209 "# of TOE TLS records transmitted"); 7210 SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "tx_toe_tls_octets", 7211 CTLFLAG_RD, &pi->tx_toe_tls_octets, 7212 "# of payload octets in transmitted TOE TLS records"); 7213 SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "rx_toe_tls_records", 7214 CTLFLAG_RD, &pi->rx_toe_tls_records, 7215 "# of TOE TLS records received"); 7216 SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "rx_toe_tls_octets", 7217 CTLFLAG_RD, &pi->rx_toe_tls_octets, 7218 "# of payload octets in received TOE TLS records"); 7219 } 7220 7221 static int 7222 sysctl_int_array(SYSCTL_HANDLER_ARGS) 7223 { 7224 int rc, *i, space = 0; 7225 struct sbuf sb; 7226 7227 sbuf_new_for_sysctl(&sb, NULL, 64, req); 7228 for (i = arg1; arg2; arg2 -= sizeof(int), i++) { 7229 if (space) 7230 sbuf_printf(&sb, " "); 7231 sbuf_printf(&sb, "%d", *i); 7232 space = 1; 7233 } 7234 rc = sbuf_finish(&sb); 7235 sbuf_delete(&sb); 7236 return (rc); 7237 } 7238 7239 static int 7240 sysctl_bitfield_8b(SYSCTL_HANDLER_ARGS) 7241 { 7242 int rc; 7243 struct sbuf *sb; 7244 7245 rc = sysctl_wire_old_buffer(req, 0); 7246 if (rc != 0) 7247 return(rc); 7248 7249 sb = sbuf_new_for_sysctl(NULL, NULL, 128, req); 7250 if (sb == NULL) 7251 return (ENOMEM); 7252 7253 sbuf_printf(sb, "%b", *(uint8_t *)(uintptr_t)arg2, (char *)arg1); 7254 rc = sbuf_finish(sb); 7255 sbuf_delete(sb); 7256 7257 return (rc); 7258 } 7259 7260 static int 7261 sysctl_bitfield_16b(SYSCTL_HANDLER_ARGS) 7262 { 7263 int rc; 7264 struct sbuf *sb; 7265 7266 rc = sysctl_wire_old_buffer(req, 0); 7267 if (rc != 0) 7268 return(rc); 7269 7270 sb = sbuf_new_for_sysctl(NULL, NULL, 128, req); 7271 if (sb == NULL) 7272 return (ENOMEM); 7273 7274 sbuf_printf(sb, "%b", *(uint16_t *)(uintptr_t)arg2, (char *)arg1); 7275 rc = sbuf_finish(sb); 7276 sbuf_delete(sb); 7277 7278 return (rc); 7279 } 7280 7281 static int 7282 sysctl_btphy(SYSCTL_HANDLER_ARGS) 7283 { 7284 struct port_info *pi = arg1; 7285 int op = arg2; 7286 struct adapter *sc = pi->adapter; 7287 u_int v; 7288 int rc; 7289 7290 rc = begin_synchronized_op(sc, &pi->vi[0], SLEEP_OK | INTR_OK, "t4btt"); 7291 if (rc) 7292 return (rc); 7293 /* XXX: magic numbers */ 7294 rc = -t4_mdio_rd(sc, sc->mbox, pi->mdio_addr, 0x1e, op ? 0x20 : 0xc820, 7295 &v); 7296 end_synchronized_op(sc, 0); 7297 if (rc) 7298 return (rc); 7299 if (op == 0) 7300 v /= 256; 7301 7302 rc = sysctl_handle_int(oidp, &v, 0, req); 7303 return (rc); 7304 } 7305 7306 static int 7307 sysctl_noflowq(SYSCTL_HANDLER_ARGS) 7308 { 7309 struct vi_info *vi = arg1; 7310 int rc, val; 7311 7312 val = vi->rsrv_noflowq; 7313 rc = sysctl_handle_int(oidp, &val, 0, req); 7314 if (rc != 0 || req->newptr == NULL) 7315 return (rc); 7316 7317 if ((val >= 1) && (vi->ntxq > 1)) 7318 vi->rsrv_noflowq = 1; 7319 else 7320 vi->rsrv_noflowq = 0; 7321 7322 return (rc); 7323 } 7324 7325 static int 7326 sysctl_tx_vm_wr(SYSCTL_HANDLER_ARGS) 7327 { 7328 struct vi_info *vi = arg1; 7329 struct adapter *sc = vi->adapter; 7330 int rc, val, i; 7331 7332 MPASS(!(sc->flags & IS_VF)); 7333 7334 val = vi->flags & TX_USES_VM_WR ? 1 : 0; 7335 rc = sysctl_handle_int(oidp, &val, 0, req); 7336 if (rc != 0 || req->newptr == NULL) 7337 return (rc); 7338 7339 if (val != 0 && val != 1) 7340 return (EINVAL); 7341 7342 rc = begin_synchronized_op(sc, vi, HOLD_LOCK | SLEEP_OK | INTR_OK, 7343 "t4txvm"); 7344 if (rc) 7345 return (rc); 7346 if (vi->ifp->if_drv_flags & IFF_DRV_RUNNING) { 7347 /* 7348 * We don't want parse_pkt to run with one setting (VF or PF) 7349 * and then eth_tx to see a different setting but still use 7350 * stale information calculated by parse_pkt. 7351 */ 7352 rc = EBUSY; 7353 } else { 7354 struct port_info *pi = vi->pi; 7355 struct sge_txq *txq; 7356 uint32_t ctrl0; 7357 uint8_t npkt = sc->params.max_pkts_per_eth_tx_pkts_wr; 7358 7359 if (val) { 7360 vi->flags |= TX_USES_VM_WR; 7361 vi->ifp->if_hw_tsomaxsegcount = TX_SGL_SEGS_VM_TSO; 7362 ctrl0 = htobe32(V_TXPKT_OPCODE(CPL_TX_PKT_XT) | 7363 V_TXPKT_INTF(pi->tx_chan)); 7364 if (!(sc->flags & IS_VF)) 7365 npkt--; 7366 } else { 7367 vi->flags &= ~TX_USES_VM_WR; 7368 vi->ifp->if_hw_tsomaxsegcount = TX_SGL_SEGS_TSO; 7369 ctrl0 = htobe32(V_TXPKT_OPCODE(CPL_TX_PKT_XT) | 7370 V_TXPKT_INTF(pi->tx_chan) | V_TXPKT_PF(sc->pf) | 7371 V_TXPKT_VF(vi->vin) | V_TXPKT_VF_VLD(vi->vfvld)); 7372 } 7373 for_each_txq(vi, i, txq) { 7374 txq->cpl_ctrl0 = ctrl0; 7375 txq->txp.max_npkt = npkt; 7376 } 7377 } 7378 end_synchronized_op(sc, LOCK_HELD); 7379 return (rc); 7380 } 7381 7382 static int 7383 sysctl_holdoff_tmr_idx(SYSCTL_HANDLER_ARGS) 7384 { 7385 struct vi_info *vi = arg1; 7386 struct adapter *sc = vi->adapter; 7387 int idx, rc, i; 7388 struct sge_rxq *rxq; 7389 uint8_t v; 7390 7391 idx = vi->tmr_idx; 7392 7393 rc = sysctl_handle_int(oidp, &idx, 0, req); 7394 if (rc != 0 || req->newptr == NULL) 7395 return (rc); 7396 7397 if (idx < 0 || idx >= SGE_NTIMERS) 7398 return (EINVAL); 7399 7400 rc = begin_synchronized_op(sc, vi, HOLD_LOCK | SLEEP_OK | INTR_OK, 7401 "t4tmr"); 7402 if (rc) 7403 return (rc); 7404 7405 v = V_QINTR_TIMER_IDX(idx) | V_QINTR_CNT_EN(vi->pktc_idx != -1); 7406 for_each_rxq(vi, i, rxq) { 7407 #ifdef atomic_store_rel_8 7408 atomic_store_rel_8(&rxq->iq.intr_params, v); 7409 #else 7410 rxq->iq.intr_params = v; 7411 #endif 7412 } 7413 vi->tmr_idx = idx; 7414 7415 end_synchronized_op(sc, LOCK_HELD); 7416 return (0); 7417 } 7418 7419 static int 7420 sysctl_holdoff_pktc_idx(SYSCTL_HANDLER_ARGS) 7421 { 7422 struct vi_info *vi = arg1; 7423 struct adapter *sc = vi->adapter; 7424 int idx, rc; 7425 7426 idx = vi->pktc_idx; 7427 7428 rc = sysctl_handle_int(oidp, &idx, 0, req); 7429 if (rc != 0 || req->newptr == NULL) 7430 return (rc); 7431 7432 if (idx < -1 || idx >= SGE_NCOUNTERS) 7433 return (EINVAL); 7434 7435 rc = begin_synchronized_op(sc, vi, HOLD_LOCK | SLEEP_OK | INTR_OK, 7436 "t4pktc"); 7437 if (rc) 7438 return (rc); 7439 7440 if (vi->flags & VI_INIT_DONE) 7441 rc = EBUSY; /* cannot be changed once the queues are created */ 7442 else 7443 vi->pktc_idx = idx; 7444 7445 end_synchronized_op(sc, LOCK_HELD); 7446 return (rc); 7447 } 7448 7449 static int 7450 sysctl_qsize_rxq(SYSCTL_HANDLER_ARGS) 7451 { 7452 struct vi_info *vi = arg1; 7453 struct adapter *sc = vi->adapter; 7454 int qsize, rc; 7455 7456 qsize = vi->qsize_rxq; 7457 7458 rc = sysctl_handle_int(oidp, &qsize, 0, req); 7459 if (rc != 0 || req->newptr == NULL) 7460 return (rc); 7461 7462 if (qsize < 128 || (qsize & 7)) 7463 return (EINVAL); 7464 7465 rc = begin_synchronized_op(sc, vi, HOLD_LOCK | SLEEP_OK | INTR_OK, 7466 "t4rxqs"); 7467 if (rc) 7468 return (rc); 7469 7470 if (vi->flags & VI_INIT_DONE) 7471 rc = EBUSY; /* cannot be changed once the queues are created */ 7472 else 7473 vi->qsize_rxq = qsize; 7474 7475 end_synchronized_op(sc, LOCK_HELD); 7476 return (rc); 7477 } 7478 7479 static int 7480 sysctl_qsize_txq(SYSCTL_HANDLER_ARGS) 7481 { 7482 struct vi_info *vi = arg1; 7483 struct adapter *sc = vi->adapter; 7484 int qsize, rc; 7485 7486 qsize = vi->qsize_txq; 7487 7488 rc = sysctl_handle_int(oidp, &qsize, 0, req); 7489 if (rc != 0 || req->newptr == NULL) 7490 return (rc); 7491 7492 if (qsize < 128 || qsize > 65536) 7493 return (EINVAL); 7494 7495 rc = begin_synchronized_op(sc, vi, HOLD_LOCK | SLEEP_OK | INTR_OK, 7496 "t4txqs"); 7497 if (rc) 7498 return (rc); 7499 7500 if (vi->flags & VI_INIT_DONE) 7501 rc = EBUSY; /* cannot be changed once the queues are created */ 7502 else 7503 vi->qsize_txq = qsize; 7504 7505 end_synchronized_op(sc, LOCK_HELD); 7506 return (rc); 7507 } 7508 7509 static int 7510 sysctl_pause_settings(SYSCTL_HANDLER_ARGS) 7511 { 7512 struct port_info *pi = arg1; 7513 struct adapter *sc = pi->adapter; 7514 struct link_config *lc = &pi->link_cfg; 7515 int rc; 7516 7517 if (req->newptr == NULL) { 7518 struct sbuf *sb; 7519 static char *bits = "\20\1RX\2TX\3AUTO"; 7520 7521 rc = sysctl_wire_old_buffer(req, 0); 7522 if (rc != 0) 7523 return(rc); 7524 7525 sb = sbuf_new_for_sysctl(NULL, NULL, 128, req); 7526 if (sb == NULL) 7527 return (ENOMEM); 7528 7529 if (lc->link_ok) { 7530 sbuf_printf(sb, "%b", (lc->fc & (PAUSE_TX | PAUSE_RX)) | 7531 (lc->requested_fc & PAUSE_AUTONEG), bits); 7532 } else { 7533 sbuf_printf(sb, "%b", lc->requested_fc & (PAUSE_TX | 7534 PAUSE_RX | PAUSE_AUTONEG), bits); 7535 } 7536 rc = sbuf_finish(sb); 7537 sbuf_delete(sb); 7538 } else { 7539 char s[2]; 7540 int n; 7541 7542 s[0] = '0' + (lc->requested_fc & (PAUSE_TX | PAUSE_RX | 7543 PAUSE_AUTONEG)); 7544 s[1] = 0; 7545 7546 rc = sysctl_handle_string(oidp, s, sizeof(s), req); 7547 if (rc != 0) 7548 return(rc); 7549 7550 if (s[1] != 0) 7551 return (EINVAL); 7552 if (s[0] < '0' || s[0] > '9') 7553 return (EINVAL); /* not a number */ 7554 n = s[0] - '0'; 7555 if (n & ~(PAUSE_TX | PAUSE_RX | PAUSE_AUTONEG)) 7556 return (EINVAL); /* some other bit is set too */ 7557 7558 rc = begin_synchronized_op(sc, &pi->vi[0], SLEEP_OK | INTR_OK, 7559 "t4PAUSE"); 7560 if (rc) 7561 return (rc); 7562 PORT_LOCK(pi); 7563 lc->requested_fc = n; 7564 fixup_link_config(pi); 7565 if (pi->up_vis > 0) 7566 rc = apply_link_config(pi); 7567 set_current_media(pi); 7568 PORT_UNLOCK(pi); 7569 end_synchronized_op(sc, 0); 7570 } 7571 7572 return (rc); 7573 } 7574 7575 static int 7576 sysctl_fec(SYSCTL_HANDLER_ARGS) 7577 { 7578 struct port_info *pi = arg1; 7579 struct adapter *sc = pi->adapter; 7580 struct link_config *lc = &pi->link_cfg; 7581 int rc; 7582 int8_t old; 7583 7584 if (req->newptr == NULL) { 7585 struct sbuf *sb; 7586 static char *bits = "\20\1RS-FEC\2FC-FEC\3NO-FEC\4RSVD2" 7587 "\5RSVD3\6auto\7module"; 7588 7589 rc = sysctl_wire_old_buffer(req, 0); 7590 if (rc != 0) 7591 return(rc); 7592 7593 sb = sbuf_new_for_sysctl(NULL, NULL, 128, req); 7594 if (sb == NULL) 7595 return (ENOMEM); 7596 7597 /* 7598 * Display the requested_fec when the link is down -- the actual 7599 * FEC makes sense only when the link is up. 7600 */ 7601 if (lc->link_ok) { 7602 sbuf_printf(sb, "%b", (lc->fec & M_FW_PORT_CAP32_FEC) | 7603 (lc->requested_fec & (FEC_AUTO | FEC_MODULE)), 7604 bits); 7605 } else { 7606 sbuf_printf(sb, "%b", lc->requested_fec, bits); 7607 } 7608 rc = sbuf_finish(sb); 7609 sbuf_delete(sb); 7610 } else { 7611 char s[8]; 7612 int n; 7613 7614 snprintf(s, sizeof(s), "%d", 7615 lc->requested_fec == FEC_AUTO ? -1 : 7616 lc->requested_fec & (M_FW_PORT_CAP32_FEC | FEC_MODULE)); 7617 7618 rc = sysctl_handle_string(oidp, s, sizeof(s), req); 7619 if (rc != 0) 7620 return(rc); 7621 7622 n = strtol(&s[0], NULL, 0); 7623 if (n < 0 || n & FEC_AUTO) 7624 n = FEC_AUTO; 7625 else if (n & ~(M_FW_PORT_CAP32_FEC | FEC_MODULE)) 7626 return (EINVAL);/* some other bit is set too */ 7627 7628 rc = begin_synchronized_op(sc, &pi->vi[0], SLEEP_OK | INTR_OK, 7629 "t4fec"); 7630 if (rc) 7631 return (rc); 7632 PORT_LOCK(pi); 7633 old = lc->requested_fec; 7634 if (n == FEC_AUTO) 7635 lc->requested_fec = FEC_AUTO; 7636 else if (n == 0 || n == FEC_NONE) 7637 lc->requested_fec = FEC_NONE; 7638 else { 7639 if ((lc->pcaps | 7640 V_FW_PORT_CAP32_FEC(n & M_FW_PORT_CAP32_FEC)) != 7641 lc->pcaps) { 7642 rc = ENOTSUP; 7643 goto done; 7644 } 7645 lc->requested_fec = n & (M_FW_PORT_CAP32_FEC | 7646 FEC_MODULE); 7647 } 7648 fixup_link_config(pi); 7649 if (pi->up_vis > 0) { 7650 rc = apply_link_config(pi); 7651 if (rc != 0) { 7652 lc->requested_fec = old; 7653 if (rc == FW_EPROTO) 7654 rc = ENOTSUP; 7655 } 7656 } 7657 done: 7658 PORT_UNLOCK(pi); 7659 end_synchronized_op(sc, 0); 7660 } 7661 7662 return (rc); 7663 } 7664 7665 static int 7666 sysctl_module_fec(SYSCTL_HANDLER_ARGS) 7667 { 7668 struct port_info *pi = arg1; 7669 struct adapter *sc = pi->adapter; 7670 struct link_config *lc = &pi->link_cfg; 7671 int rc; 7672 int8_t fec; 7673 struct sbuf *sb; 7674 static char *bits = "\20\1RS-FEC\2FC-FEC\3NO-FEC\4RSVD2\5RSVD3"; 7675 7676 rc = sysctl_wire_old_buffer(req, 0); 7677 if (rc != 0) 7678 return (rc); 7679 7680 sb = sbuf_new_for_sysctl(NULL, NULL, 128, req); 7681 if (sb == NULL) 7682 return (ENOMEM); 7683 7684 if (begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4mfec") != 0) 7685 return (EBUSY); 7686 PORT_LOCK(pi); 7687 if (pi->up_vis == 0) { 7688 /* 7689 * If all the interfaces are administratively down the firmware 7690 * does not report transceiver changes. Refresh port info here. 7691 * This is the only reason we have a synchronized op in this 7692 * function. Just PORT_LOCK would have been enough otherwise. 7693 */ 7694 t4_update_port_info(pi); 7695 } 7696 7697 fec = lc->fec_hint; 7698 if (pi->mod_type == FW_PORT_MOD_TYPE_NONE || 7699 !fec_supported(lc->pcaps)) { 7700 sbuf_printf(sb, "n/a"); 7701 } else { 7702 if (fec == 0) 7703 fec = FEC_NONE; 7704 sbuf_printf(sb, "%b", fec & M_FW_PORT_CAP32_FEC, bits); 7705 } 7706 rc = sbuf_finish(sb); 7707 sbuf_delete(sb); 7708 7709 PORT_UNLOCK(pi); 7710 end_synchronized_op(sc, 0); 7711 7712 return (rc); 7713 } 7714 7715 static int 7716 sysctl_autoneg(SYSCTL_HANDLER_ARGS) 7717 { 7718 struct port_info *pi = arg1; 7719 struct adapter *sc = pi->adapter; 7720 struct link_config *lc = &pi->link_cfg; 7721 int rc, val; 7722 7723 if (lc->pcaps & FW_PORT_CAP32_ANEG) 7724 val = lc->requested_aneg == AUTONEG_DISABLE ? 0 : 1; 7725 else 7726 val = -1; 7727 rc = sysctl_handle_int(oidp, &val, 0, req); 7728 if (rc != 0 || req->newptr == NULL) 7729 return (rc); 7730 if (val == 0) 7731 val = AUTONEG_DISABLE; 7732 else if (val == 1) 7733 val = AUTONEG_ENABLE; 7734 else 7735 val = AUTONEG_AUTO; 7736 7737 rc = begin_synchronized_op(sc, &pi->vi[0], SLEEP_OK | INTR_OK, 7738 "t4aneg"); 7739 if (rc) 7740 return (rc); 7741 PORT_LOCK(pi); 7742 if (val == AUTONEG_ENABLE && !(lc->pcaps & FW_PORT_CAP32_ANEG)) { 7743 rc = ENOTSUP; 7744 goto done; 7745 } 7746 lc->requested_aneg = val; 7747 fixup_link_config(pi); 7748 if (pi->up_vis > 0) 7749 rc = apply_link_config(pi); 7750 set_current_media(pi); 7751 done: 7752 PORT_UNLOCK(pi); 7753 end_synchronized_op(sc, 0); 7754 return (rc); 7755 } 7756 7757 static int 7758 sysctl_handle_t4_reg64(SYSCTL_HANDLER_ARGS) 7759 { 7760 struct adapter *sc = arg1; 7761 int reg = arg2; 7762 uint64_t val; 7763 7764 val = t4_read_reg64(sc, reg); 7765 7766 return (sysctl_handle_64(oidp, &val, 0, req)); 7767 } 7768 7769 static int 7770 sysctl_temperature(SYSCTL_HANDLER_ARGS) 7771 { 7772 struct adapter *sc = arg1; 7773 int rc, t; 7774 uint32_t param, val; 7775 7776 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4temp"); 7777 if (rc) 7778 return (rc); 7779 param = V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | 7780 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_DIAG) | 7781 V_FW_PARAMS_PARAM_Y(FW_PARAM_DEV_DIAG_TMP); 7782 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 7783 end_synchronized_op(sc, 0); 7784 if (rc) 7785 return (rc); 7786 7787 /* unknown is returned as 0 but we display -1 in that case */ 7788 t = val == 0 ? -1 : val; 7789 7790 rc = sysctl_handle_int(oidp, &t, 0, req); 7791 return (rc); 7792 } 7793 7794 static int 7795 sysctl_vdd(SYSCTL_HANDLER_ARGS) 7796 { 7797 struct adapter *sc = arg1; 7798 int rc; 7799 uint32_t param, val; 7800 7801 if (sc->params.core_vdd == 0) { 7802 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, 7803 "t4vdd"); 7804 if (rc) 7805 return (rc); 7806 param = V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | 7807 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_DIAG) | 7808 V_FW_PARAMS_PARAM_Y(FW_PARAM_DEV_DIAG_VDD); 7809 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 7810 end_synchronized_op(sc, 0); 7811 if (rc) 7812 return (rc); 7813 sc->params.core_vdd = val; 7814 } 7815 7816 return (sysctl_handle_int(oidp, &sc->params.core_vdd, 0, req)); 7817 } 7818 7819 static int 7820 sysctl_reset_sensor(SYSCTL_HANDLER_ARGS) 7821 { 7822 struct adapter *sc = arg1; 7823 int rc, v; 7824 uint32_t param, val; 7825 7826 v = sc->sensor_resets; 7827 rc = sysctl_handle_int(oidp, &v, 0, req); 7828 if (rc != 0 || req->newptr == NULL || v <= 0) 7829 return (rc); 7830 7831 if (sc->params.fw_vers < FW_VERSION32(1, 24, 7, 0) || 7832 chip_id(sc) < CHELSIO_T5) 7833 return (ENOTSUP); 7834 7835 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4srst"); 7836 if (rc) 7837 return (rc); 7838 param = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | 7839 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_DIAG) | 7840 V_FW_PARAMS_PARAM_Y(FW_PARAM_DEV_DIAG_RESET_TMP_SENSOR)); 7841 val = 1; 7842 rc = -t4_set_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 7843 end_synchronized_op(sc, 0); 7844 if (rc == 0) 7845 sc->sensor_resets++; 7846 return (rc); 7847 } 7848 7849 static int 7850 sysctl_loadavg(SYSCTL_HANDLER_ARGS) 7851 { 7852 struct adapter *sc = arg1; 7853 struct sbuf *sb; 7854 int rc; 7855 uint32_t param, val; 7856 7857 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4lavg"); 7858 if (rc) 7859 return (rc); 7860 param = V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | 7861 V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_LOAD); 7862 rc = -t4_query_params(sc, sc->mbox, sc->pf, 0, 1, ¶m, &val); 7863 end_synchronized_op(sc, 0); 7864 if (rc) 7865 return (rc); 7866 7867 rc = sysctl_wire_old_buffer(req, 0); 7868 if (rc != 0) 7869 return (rc); 7870 7871 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 7872 if (sb == NULL) 7873 return (ENOMEM); 7874 7875 if (val == 0xffffffff) { 7876 /* Only debug and custom firmwares report load averages. */ 7877 sbuf_printf(sb, "not available"); 7878 } else { 7879 sbuf_printf(sb, "%d %d %d", val & 0xff, (val >> 8) & 0xff, 7880 (val >> 16) & 0xff); 7881 } 7882 rc = sbuf_finish(sb); 7883 sbuf_delete(sb); 7884 7885 return (rc); 7886 } 7887 7888 static int 7889 sysctl_cctrl(SYSCTL_HANDLER_ARGS) 7890 { 7891 struct adapter *sc = arg1; 7892 struct sbuf *sb; 7893 int rc, i; 7894 uint16_t incr[NMTUS][NCCTRL_WIN]; 7895 static const char *dec_fac[] = { 7896 "0.5", "0.5625", "0.625", "0.6875", "0.75", "0.8125", "0.875", 7897 "0.9375" 7898 }; 7899 7900 rc = sysctl_wire_old_buffer(req, 0); 7901 if (rc != 0) 7902 return (rc); 7903 7904 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 7905 if (sb == NULL) 7906 return (ENOMEM); 7907 7908 t4_read_cong_tbl(sc, incr); 7909 7910 for (i = 0; i < NCCTRL_WIN; ++i) { 7911 sbuf_printf(sb, "%2d: %4u %4u %4u %4u %4u %4u %4u %4u\n", i, 7912 incr[0][i], incr[1][i], incr[2][i], incr[3][i], incr[4][i], 7913 incr[5][i], incr[6][i], incr[7][i]); 7914 sbuf_printf(sb, "%8u %4u %4u %4u %4u %4u %4u %4u %5u %s\n", 7915 incr[8][i], incr[9][i], incr[10][i], incr[11][i], 7916 incr[12][i], incr[13][i], incr[14][i], incr[15][i], 7917 sc->params.a_wnd[i], dec_fac[sc->params.b_wnd[i]]); 7918 } 7919 7920 rc = sbuf_finish(sb); 7921 sbuf_delete(sb); 7922 7923 return (rc); 7924 } 7925 7926 static const char *qname[CIM_NUM_IBQ + CIM_NUM_OBQ_T5] = { 7927 "TP0", "TP1", "ULP", "SGE0", "SGE1", "NC-SI", /* ibq's */ 7928 "ULP0", "ULP1", "ULP2", "ULP3", "SGE", "NC-SI", /* obq's */ 7929 "SGE0-RX", "SGE1-RX" /* additional obq's (T5 onwards) */ 7930 }; 7931 7932 static int 7933 sysctl_cim_ibq_obq(SYSCTL_HANDLER_ARGS) 7934 { 7935 struct adapter *sc = arg1; 7936 struct sbuf *sb; 7937 int rc, i, n, qid = arg2; 7938 uint32_t *buf, *p; 7939 char *qtype; 7940 u_int cim_num_obq = sc->chip_params->cim_num_obq; 7941 7942 KASSERT(qid >= 0 && qid < CIM_NUM_IBQ + cim_num_obq, 7943 ("%s: bad qid %d\n", __func__, qid)); 7944 7945 if (qid < CIM_NUM_IBQ) { 7946 /* inbound queue */ 7947 qtype = "IBQ"; 7948 n = 4 * CIM_IBQ_SIZE; 7949 buf = malloc(n * sizeof(uint32_t), M_CXGBE, M_ZERO | M_WAITOK); 7950 rc = t4_read_cim_ibq(sc, qid, buf, n); 7951 } else { 7952 /* outbound queue */ 7953 qtype = "OBQ"; 7954 qid -= CIM_NUM_IBQ; 7955 n = 4 * cim_num_obq * CIM_OBQ_SIZE; 7956 buf = malloc(n * sizeof(uint32_t), M_CXGBE, M_ZERO | M_WAITOK); 7957 rc = t4_read_cim_obq(sc, qid, buf, n); 7958 } 7959 7960 if (rc < 0) { 7961 rc = -rc; 7962 goto done; 7963 } 7964 n = rc * sizeof(uint32_t); /* rc has # of words actually read */ 7965 7966 rc = sysctl_wire_old_buffer(req, 0); 7967 if (rc != 0) 7968 goto done; 7969 7970 sb = sbuf_new_for_sysctl(NULL, NULL, PAGE_SIZE, req); 7971 if (sb == NULL) { 7972 rc = ENOMEM; 7973 goto done; 7974 } 7975 7976 sbuf_printf(sb, "%s%d %s", qtype , qid, qname[arg2]); 7977 for (i = 0, p = buf; i < n; i += 16, p += 4) 7978 sbuf_printf(sb, "\n%#06x: %08x %08x %08x %08x", i, p[0], p[1], 7979 p[2], p[3]); 7980 7981 rc = sbuf_finish(sb); 7982 sbuf_delete(sb); 7983 done: 7984 free(buf, M_CXGBE); 7985 return (rc); 7986 } 7987 7988 static void 7989 sbuf_cim_la4(struct adapter *sc, struct sbuf *sb, uint32_t *buf, uint32_t cfg) 7990 { 7991 uint32_t *p; 7992 7993 sbuf_printf(sb, "Status Data PC%s", 7994 cfg & F_UPDBGLACAPTPCONLY ? "" : 7995 " LS0Stat LS0Addr LS0Data"); 7996 7997 for (p = buf; p <= &buf[sc->params.cim_la_size - 8]; p += 8) { 7998 if (cfg & F_UPDBGLACAPTPCONLY) { 7999 sbuf_printf(sb, "\n %02x %08x %08x", p[5] & 0xff, 8000 p[6], p[7]); 8001 sbuf_printf(sb, "\n %02x %02x%06x %02x%06x", 8002 (p[3] >> 8) & 0xff, p[3] & 0xff, p[4] >> 8, 8003 p[4] & 0xff, p[5] >> 8); 8004 sbuf_printf(sb, "\n %02x %x%07x %x%07x", 8005 (p[0] >> 4) & 0xff, p[0] & 0xf, p[1] >> 4, 8006 p[1] & 0xf, p[2] >> 4); 8007 } else { 8008 sbuf_printf(sb, 8009 "\n %02x %x%07x %x%07x %08x %08x " 8010 "%08x%08x%08x%08x", 8011 (p[0] >> 4) & 0xff, p[0] & 0xf, p[1] >> 4, 8012 p[1] & 0xf, p[2] >> 4, p[2] & 0xf, p[3], p[4], p[5], 8013 p[6], p[7]); 8014 } 8015 } 8016 } 8017 8018 static void 8019 sbuf_cim_la6(struct adapter *sc, struct sbuf *sb, uint32_t *buf, uint32_t cfg) 8020 { 8021 uint32_t *p; 8022 8023 sbuf_printf(sb, "Status Inst Data PC%s", 8024 cfg & F_UPDBGLACAPTPCONLY ? "" : 8025 " LS0Stat LS0Addr LS0Data LS1Stat LS1Addr LS1Data"); 8026 8027 for (p = buf; p <= &buf[sc->params.cim_la_size - 10]; p += 10) { 8028 if (cfg & F_UPDBGLACAPTPCONLY) { 8029 sbuf_printf(sb, "\n %02x %08x %08x %08x", 8030 p[3] & 0xff, p[2], p[1], p[0]); 8031 sbuf_printf(sb, "\n %02x %02x%06x %02x%06x %02x%06x", 8032 (p[6] >> 8) & 0xff, p[6] & 0xff, p[5] >> 8, 8033 p[5] & 0xff, p[4] >> 8, p[4] & 0xff, p[3] >> 8); 8034 sbuf_printf(sb, "\n %02x %04x%04x %04x%04x %04x%04x", 8035 (p[9] >> 16) & 0xff, p[9] & 0xffff, p[8] >> 16, 8036 p[8] & 0xffff, p[7] >> 16, p[7] & 0xffff, 8037 p[6] >> 16); 8038 } else { 8039 sbuf_printf(sb, "\n %02x %04x%04x %04x%04x %04x%04x " 8040 "%08x %08x %08x %08x %08x %08x", 8041 (p[9] >> 16) & 0xff, 8042 p[9] & 0xffff, p[8] >> 16, 8043 p[8] & 0xffff, p[7] >> 16, 8044 p[7] & 0xffff, p[6] >> 16, 8045 p[2], p[1], p[0], p[5], p[4], p[3]); 8046 } 8047 } 8048 } 8049 8050 static int 8051 sbuf_cim_la(struct adapter *sc, struct sbuf *sb, int flags) 8052 { 8053 uint32_t cfg, *buf; 8054 int rc; 8055 8056 rc = -t4_cim_read(sc, A_UP_UP_DBG_LA_CFG, 1, &cfg); 8057 if (rc != 0) 8058 return (rc); 8059 8060 MPASS(flags == M_WAITOK || flags == M_NOWAIT); 8061 buf = malloc(sc->params.cim_la_size * sizeof(uint32_t), M_CXGBE, 8062 M_ZERO | flags); 8063 if (buf == NULL) 8064 return (ENOMEM); 8065 8066 rc = -t4_cim_read_la(sc, buf, NULL); 8067 if (rc != 0) 8068 goto done; 8069 if (chip_id(sc) < CHELSIO_T6) 8070 sbuf_cim_la4(sc, sb, buf, cfg); 8071 else 8072 sbuf_cim_la6(sc, sb, buf, cfg); 8073 8074 done: 8075 free(buf, M_CXGBE); 8076 return (rc); 8077 } 8078 8079 static int 8080 sysctl_cim_la(SYSCTL_HANDLER_ARGS) 8081 { 8082 struct adapter *sc = arg1; 8083 struct sbuf *sb; 8084 int rc; 8085 8086 rc = sysctl_wire_old_buffer(req, 0); 8087 if (rc != 0) 8088 return (rc); 8089 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 8090 if (sb == NULL) 8091 return (ENOMEM); 8092 8093 rc = sbuf_cim_la(sc, sb, M_WAITOK); 8094 if (rc == 0) 8095 rc = sbuf_finish(sb); 8096 sbuf_delete(sb); 8097 return (rc); 8098 } 8099 8100 bool 8101 t4_os_dump_cimla(struct adapter *sc, int arg, bool verbose) 8102 { 8103 struct sbuf sb; 8104 int rc; 8105 8106 if (sbuf_new(&sb, NULL, 4096, SBUF_AUTOEXTEND) != &sb) 8107 return (false); 8108 rc = sbuf_cim_la(sc, &sb, M_NOWAIT); 8109 if (rc == 0) { 8110 rc = sbuf_finish(&sb); 8111 if (rc == 0) { 8112 log(LOG_DEBUG, "%s: CIM LA dump follows.\n%s", 8113 device_get_nameunit(sc->dev), sbuf_data(&sb)); 8114 } 8115 } 8116 sbuf_delete(&sb); 8117 return (false); 8118 } 8119 8120 static int 8121 sysctl_cim_ma_la(SYSCTL_HANDLER_ARGS) 8122 { 8123 struct adapter *sc = arg1; 8124 u_int i; 8125 struct sbuf *sb; 8126 uint32_t *buf, *p; 8127 int rc; 8128 8129 rc = sysctl_wire_old_buffer(req, 0); 8130 if (rc != 0) 8131 return (rc); 8132 8133 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 8134 if (sb == NULL) 8135 return (ENOMEM); 8136 8137 buf = malloc(2 * CIM_MALA_SIZE * 5 * sizeof(uint32_t), M_CXGBE, 8138 M_ZERO | M_WAITOK); 8139 8140 t4_cim_read_ma_la(sc, buf, buf + 5 * CIM_MALA_SIZE); 8141 p = buf; 8142 8143 for (i = 0; i < CIM_MALA_SIZE; i++, p += 5) { 8144 sbuf_printf(sb, "\n%02x%08x%08x%08x%08x", p[4], p[3], p[2], 8145 p[1], p[0]); 8146 } 8147 8148 sbuf_printf(sb, "\n\nCnt ID Tag UE Data RDY VLD"); 8149 for (i = 0; i < CIM_MALA_SIZE; i++, p += 5) { 8150 sbuf_printf(sb, "\n%3u %2u %x %u %08x%08x %u %u", 8151 (p[2] >> 10) & 0xff, (p[2] >> 7) & 7, 8152 (p[2] >> 3) & 0xf, (p[2] >> 2) & 1, 8153 (p[1] >> 2) | ((p[2] & 3) << 30), 8154 (p[0] >> 2) | ((p[1] & 3) << 30), (p[0] >> 1) & 1, 8155 p[0] & 1); 8156 } 8157 8158 rc = sbuf_finish(sb); 8159 sbuf_delete(sb); 8160 free(buf, M_CXGBE); 8161 return (rc); 8162 } 8163 8164 static int 8165 sysctl_cim_pif_la(SYSCTL_HANDLER_ARGS) 8166 { 8167 struct adapter *sc = arg1; 8168 u_int i; 8169 struct sbuf *sb; 8170 uint32_t *buf, *p; 8171 int rc; 8172 8173 rc = sysctl_wire_old_buffer(req, 0); 8174 if (rc != 0) 8175 return (rc); 8176 8177 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 8178 if (sb == NULL) 8179 return (ENOMEM); 8180 8181 buf = malloc(2 * CIM_PIFLA_SIZE * 6 * sizeof(uint32_t), M_CXGBE, 8182 M_ZERO | M_WAITOK); 8183 8184 t4_cim_read_pif_la(sc, buf, buf + 6 * CIM_PIFLA_SIZE, NULL, NULL); 8185 p = buf; 8186 8187 sbuf_printf(sb, "Cntl ID DataBE Addr Data"); 8188 for (i = 0; i < CIM_PIFLA_SIZE; i++, p += 6) { 8189 sbuf_printf(sb, "\n %02x %02x %04x %08x %08x%08x%08x%08x", 8190 (p[5] >> 22) & 0xff, (p[5] >> 16) & 0x3f, p[5] & 0xffff, 8191 p[4], p[3], p[2], p[1], p[0]); 8192 } 8193 8194 sbuf_printf(sb, "\n\nCntl ID Data"); 8195 for (i = 0; i < CIM_PIFLA_SIZE; i++, p += 6) { 8196 sbuf_printf(sb, "\n %02x %02x %08x%08x%08x%08x", 8197 (p[4] >> 6) & 0xff, p[4] & 0x3f, p[3], p[2], p[1], p[0]); 8198 } 8199 8200 rc = sbuf_finish(sb); 8201 sbuf_delete(sb); 8202 free(buf, M_CXGBE); 8203 return (rc); 8204 } 8205 8206 static int 8207 sysctl_cim_qcfg(SYSCTL_HANDLER_ARGS) 8208 { 8209 struct adapter *sc = arg1; 8210 struct sbuf *sb; 8211 int rc, i; 8212 uint16_t base[CIM_NUM_IBQ + CIM_NUM_OBQ_T5]; 8213 uint16_t size[CIM_NUM_IBQ + CIM_NUM_OBQ_T5]; 8214 uint16_t thres[CIM_NUM_IBQ]; 8215 uint32_t obq_wr[2 * CIM_NUM_OBQ_T5], *wr = obq_wr; 8216 uint32_t stat[4 * (CIM_NUM_IBQ + CIM_NUM_OBQ_T5)], *p = stat; 8217 u_int cim_num_obq, ibq_rdaddr, obq_rdaddr, nq; 8218 8219 cim_num_obq = sc->chip_params->cim_num_obq; 8220 if (is_t4(sc)) { 8221 ibq_rdaddr = A_UP_IBQ_0_RDADDR; 8222 obq_rdaddr = A_UP_OBQ_0_REALADDR; 8223 } else { 8224 ibq_rdaddr = A_UP_IBQ_0_SHADOW_RDADDR; 8225 obq_rdaddr = A_UP_OBQ_0_SHADOW_REALADDR; 8226 } 8227 nq = CIM_NUM_IBQ + cim_num_obq; 8228 8229 rc = -t4_cim_read(sc, ibq_rdaddr, 4 * nq, stat); 8230 if (rc == 0) 8231 rc = -t4_cim_read(sc, obq_rdaddr, 2 * cim_num_obq, obq_wr); 8232 if (rc != 0) 8233 return (rc); 8234 8235 t4_read_cimq_cfg(sc, base, size, thres); 8236 8237 rc = sysctl_wire_old_buffer(req, 0); 8238 if (rc != 0) 8239 return (rc); 8240 8241 sb = sbuf_new_for_sysctl(NULL, NULL, PAGE_SIZE, req); 8242 if (sb == NULL) 8243 return (ENOMEM); 8244 8245 sbuf_printf(sb, 8246 " Queue Base Size Thres RdPtr WrPtr SOP EOP Avail"); 8247 8248 for (i = 0; i < CIM_NUM_IBQ; i++, p += 4) 8249 sbuf_printf(sb, "\n%7s %5x %5u %5u %6x %4x %4u %4u %5u", 8250 qname[i], base[i], size[i], thres[i], G_IBQRDADDR(p[0]), 8251 G_IBQWRADDR(p[1]), G_QUESOPCNT(p[3]), G_QUEEOPCNT(p[3]), 8252 G_QUEREMFLITS(p[2]) * 16); 8253 for ( ; i < nq; i++, p += 4, wr += 2) 8254 sbuf_printf(sb, "\n%7s %5x %5u %12x %4x %4u %4u %5u", qname[i], 8255 base[i], size[i], G_QUERDADDR(p[0]) & 0x3fff, 8256 wr[0] - base[i], G_QUESOPCNT(p[3]), G_QUEEOPCNT(p[3]), 8257 G_QUEREMFLITS(p[2]) * 16); 8258 8259 rc = sbuf_finish(sb); 8260 sbuf_delete(sb); 8261 8262 return (rc); 8263 } 8264 8265 static int 8266 sysctl_cpl_stats(SYSCTL_HANDLER_ARGS) 8267 { 8268 struct adapter *sc = arg1; 8269 struct sbuf *sb; 8270 int rc; 8271 struct tp_cpl_stats stats; 8272 8273 rc = sysctl_wire_old_buffer(req, 0); 8274 if (rc != 0) 8275 return (rc); 8276 8277 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 8278 if (sb == NULL) 8279 return (ENOMEM); 8280 8281 mtx_lock(&sc->reg_lock); 8282 t4_tp_get_cpl_stats(sc, &stats, 0); 8283 mtx_unlock(&sc->reg_lock); 8284 8285 if (sc->chip_params->nchan > 2) { 8286 sbuf_printf(sb, " channel 0 channel 1" 8287 " channel 2 channel 3"); 8288 sbuf_printf(sb, "\nCPL requests: %10u %10u %10u %10u", 8289 stats.req[0], stats.req[1], stats.req[2], stats.req[3]); 8290 sbuf_printf(sb, "\nCPL responses: %10u %10u %10u %10u", 8291 stats.rsp[0], stats.rsp[1], stats.rsp[2], stats.rsp[3]); 8292 } else { 8293 sbuf_printf(sb, " channel 0 channel 1"); 8294 sbuf_printf(sb, "\nCPL requests: %10u %10u", 8295 stats.req[0], stats.req[1]); 8296 sbuf_printf(sb, "\nCPL responses: %10u %10u", 8297 stats.rsp[0], stats.rsp[1]); 8298 } 8299 8300 rc = sbuf_finish(sb); 8301 sbuf_delete(sb); 8302 8303 return (rc); 8304 } 8305 8306 static int 8307 sysctl_ddp_stats(SYSCTL_HANDLER_ARGS) 8308 { 8309 struct adapter *sc = arg1; 8310 struct sbuf *sb; 8311 int rc; 8312 struct tp_usm_stats stats; 8313 8314 rc = sysctl_wire_old_buffer(req, 0); 8315 if (rc != 0) 8316 return(rc); 8317 8318 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 8319 if (sb == NULL) 8320 return (ENOMEM); 8321 8322 mtx_lock(&sc->reg_lock); 8323 t4_get_usm_stats(sc, &stats, 1); 8324 mtx_unlock(&sc->reg_lock); 8325 8326 sbuf_printf(sb, "Frames: %u\n", stats.frames); 8327 sbuf_printf(sb, "Octets: %ju\n", stats.octets); 8328 sbuf_printf(sb, "Drops: %u", stats.drops); 8329 8330 rc = sbuf_finish(sb); 8331 sbuf_delete(sb); 8332 8333 return (rc); 8334 } 8335 8336 static int 8337 sysctl_tid_stats(SYSCTL_HANDLER_ARGS) 8338 { 8339 struct adapter *sc = arg1; 8340 struct sbuf *sb; 8341 int rc; 8342 struct tp_tid_stats stats; 8343 8344 rc = sysctl_wire_old_buffer(req, 0); 8345 if (rc != 0) 8346 return(rc); 8347 8348 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 8349 if (sb == NULL) 8350 return (ENOMEM); 8351 8352 mtx_lock(&sc->reg_lock); 8353 t4_tp_get_tid_stats(sc, &stats, 1); 8354 mtx_unlock(&sc->reg_lock); 8355 8356 sbuf_printf(sb, "Delete: %u\n", stats.del); 8357 sbuf_printf(sb, "Invalidate: %u\n", stats.inv); 8358 sbuf_printf(sb, "Active: %u\n", stats.act); 8359 sbuf_printf(sb, "Passive: %u", stats.pas); 8360 8361 rc = sbuf_finish(sb); 8362 sbuf_delete(sb); 8363 8364 return (rc); 8365 } 8366 8367 static const char * const devlog_level_strings[] = { 8368 [FW_DEVLOG_LEVEL_EMERG] = "EMERG", 8369 [FW_DEVLOG_LEVEL_CRIT] = "CRIT", 8370 [FW_DEVLOG_LEVEL_ERR] = "ERR", 8371 [FW_DEVLOG_LEVEL_NOTICE] = "NOTICE", 8372 [FW_DEVLOG_LEVEL_INFO] = "INFO", 8373 [FW_DEVLOG_LEVEL_DEBUG] = "DEBUG" 8374 }; 8375 8376 static const char * const devlog_facility_strings[] = { 8377 [FW_DEVLOG_FACILITY_CORE] = "CORE", 8378 [FW_DEVLOG_FACILITY_CF] = "CF", 8379 [FW_DEVLOG_FACILITY_SCHED] = "SCHED", 8380 [FW_DEVLOG_FACILITY_TIMER] = "TIMER", 8381 [FW_DEVLOG_FACILITY_RES] = "RES", 8382 [FW_DEVLOG_FACILITY_HW] = "HW", 8383 [FW_DEVLOG_FACILITY_FLR] = "FLR", 8384 [FW_DEVLOG_FACILITY_DMAQ] = "DMAQ", 8385 [FW_DEVLOG_FACILITY_PHY] = "PHY", 8386 [FW_DEVLOG_FACILITY_MAC] = "MAC", 8387 [FW_DEVLOG_FACILITY_PORT] = "PORT", 8388 [FW_DEVLOG_FACILITY_VI] = "VI", 8389 [FW_DEVLOG_FACILITY_FILTER] = "FILTER", 8390 [FW_DEVLOG_FACILITY_ACL] = "ACL", 8391 [FW_DEVLOG_FACILITY_TM] = "TM", 8392 [FW_DEVLOG_FACILITY_QFC] = "QFC", 8393 [FW_DEVLOG_FACILITY_DCB] = "DCB", 8394 [FW_DEVLOG_FACILITY_ETH] = "ETH", 8395 [FW_DEVLOG_FACILITY_OFLD] = "OFLD", 8396 [FW_DEVLOG_FACILITY_RI] = "RI", 8397 [FW_DEVLOG_FACILITY_ISCSI] = "ISCSI", 8398 [FW_DEVLOG_FACILITY_FCOE] = "FCOE", 8399 [FW_DEVLOG_FACILITY_FOISCSI] = "FOISCSI", 8400 [FW_DEVLOG_FACILITY_FOFCOE] = "FOFCOE", 8401 [FW_DEVLOG_FACILITY_CHNET] = "CHNET", 8402 }; 8403 8404 static int 8405 sbuf_devlog(struct adapter *sc, struct sbuf *sb, int flags) 8406 { 8407 int i, j, rc, nentries, first = 0; 8408 struct devlog_params *dparams = &sc->params.devlog; 8409 struct fw_devlog_e *buf, *e; 8410 uint64_t ftstamp = UINT64_MAX; 8411 8412 if (dparams->addr == 0) 8413 return (ENXIO); 8414 8415 MPASS(flags == M_WAITOK || flags == M_NOWAIT); 8416 buf = malloc(dparams->size, M_CXGBE, M_ZERO | flags); 8417 if (buf == NULL) 8418 return (ENOMEM); 8419 8420 rc = read_via_memwin(sc, 1, dparams->addr, (void *)buf, dparams->size); 8421 if (rc != 0) 8422 goto done; 8423 8424 nentries = dparams->size / sizeof(struct fw_devlog_e); 8425 for (i = 0; i < nentries; i++) { 8426 e = &buf[i]; 8427 8428 if (e->timestamp == 0) 8429 break; /* end */ 8430 8431 e->timestamp = be64toh(e->timestamp); 8432 e->seqno = be32toh(e->seqno); 8433 for (j = 0; j < 8; j++) 8434 e->params[j] = be32toh(e->params[j]); 8435 8436 if (e->timestamp < ftstamp) { 8437 ftstamp = e->timestamp; 8438 first = i; 8439 } 8440 } 8441 8442 if (buf[first].timestamp == 0) 8443 goto done; /* nothing in the log */ 8444 8445 sbuf_printf(sb, "%10s %15s %8s %8s %s\n", 8446 "Seq#", "Tstamp", "Level", "Facility", "Message"); 8447 8448 i = first; 8449 do { 8450 e = &buf[i]; 8451 if (e->timestamp == 0) 8452 break; /* end */ 8453 8454 sbuf_printf(sb, "%10d %15ju %8s %8s ", 8455 e->seqno, e->timestamp, 8456 (e->level < nitems(devlog_level_strings) ? 8457 devlog_level_strings[e->level] : "UNKNOWN"), 8458 (e->facility < nitems(devlog_facility_strings) ? 8459 devlog_facility_strings[e->facility] : "UNKNOWN")); 8460 sbuf_printf(sb, e->fmt, e->params[0], e->params[1], 8461 e->params[2], e->params[3], e->params[4], 8462 e->params[5], e->params[6], e->params[7]); 8463 8464 if (++i == nentries) 8465 i = 0; 8466 } while (i != first); 8467 done: 8468 free(buf, M_CXGBE); 8469 return (rc); 8470 } 8471 8472 static int 8473 sysctl_devlog(SYSCTL_HANDLER_ARGS) 8474 { 8475 struct adapter *sc = arg1; 8476 int rc; 8477 struct sbuf *sb; 8478 8479 rc = sysctl_wire_old_buffer(req, 0); 8480 if (rc != 0) 8481 return (rc); 8482 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 8483 if (sb == NULL) 8484 return (ENOMEM); 8485 8486 rc = sbuf_devlog(sc, sb, M_WAITOK); 8487 if (rc == 0) 8488 rc = sbuf_finish(sb); 8489 sbuf_delete(sb); 8490 return (rc); 8491 } 8492 8493 void 8494 t4_os_dump_devlog(struct adapter *sc) 8495 { 8496 int rc; 8497 struct sbuf sb; 8498 8499 if (sbuf_new(&sb, NULL, 4096, SBUF_AUTOEXTEND) != &sb) 8500 return; 8501 rc = sbuf_devlog(sc, &sb, M_NOWAIT); 8502 if (rc == 0) { 8503 rc = sbuf_finish(&sb); 8504 if (rc == 0) { 8505 log(LOG_DEBUG, "%s: device log follows.\n%s", 8506 device_get_nameunit(sc->dev), sbuf_data(&sb)); 8507 } 8508 } 8509 sbuf_delete(&sb); 8510 } 8511 8512 static int 8513 sysctl_fcoe_stats(SYSCTL_HANDLER_ARGS) 8514 { 8515 struct adapter *sc = arg1; 8516 struct sbuf *sb; 8517 int rc; 8518 struct tp_fcoe_stats stats[MAX_NCHAN]; 8519 int i, nchan = sc->chip_params->nchan; 8520 8521 rc = sysctl_wire_old_buffer(req, 0); 8522 if (rc != 0) 8523 return (rc); 8524 8525 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 8526 if (sb == NULL) 8527 return (ENOMEM); 8528 8529 mtx_lock(&sc->reg_lock); 8530 for (i = 0; i < nchan; i++) 8531 t4_get_fcoe_stats(sc, i, &stats[i], 1); 8532 mtx_unlock(&sc->reg_lock); 8533 8534 if (nchan > 2) { 8535 sbuf_printf(sb, " channel 0 channel 1" 8536 " channel 2 channel 3"); 8537 sbuf_printf(sb, "\noctetsDDP: %16ju %16ju %16ju %16ju", 8538 stats[0].octets_ddp, stats[1].octets_ddp, 8539 stats[2].octets_ddp, stats[3].octets_ddp); 8540 sbuf_printf(sb, "\nframesDDP: %16u %16u %16u %16u", 8541 stats[0].frames_ddp, stats[1].frames_ddp, 8542 stats[2].frames_ddp, stats[3].frames_ddp); 8543 sbuf_printf(sb, "\nframesDrop: %16u %16u %16u %16u", 8544 stats[0].frames_drop, stats[1].frames_drop, 8545 stats[2].frames_drop, stats[3].frames_drop); 8546 } else { 8547 sbuf_printf(sb, " channel 0 channel 1"); 8548 sbuf_printf(sb, "\noctetsDDP: %16ju %16ju", 8549 stats[0].octets_ddp, stats[1].octets_ddp); 8550 sbuf_printf(sb, "\nframesDDP: %16u %16u", 8551 stats[0].frames_ddp, stats[1].frames_ddp); 8552 sbuf_printf(sb, "\nframesDrop: %16u %16u", 8553 stats[0].frames_drop, stats[1].frames_drop); 8554 } 8555 8556 rc = sbuf_finish(sb); 8557 sbuf_delete(sb); 8558 8559 return (rc); 8560 } 8561 8562 static int 8563 sysctl_hw_sched(SYSCTL_HANDLER_ARGS) 8564 { 8565 struct adapter *sc = arg1; 8566 struct sbuf *sb; 8567 int rc, i; 8568 unsigned int map, kbps, ipg, mode; 8569 unsigned int pace_tab[NTX_SCHED]; 8570 8571 rc = sysctl_wire_old_buffer(req, 0); 8572 if (rc != 0) 8573 return (rc); 8574 8575 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 8576 if (sb == NULL) 8577 return (ENOMEM); 8578 8579 map = t4_read_reg(sc, A_TP_TX_MOD_QUEUE_REQ_MAP); 8580 mode = G_TIMERMODE(t4_read_reg(sc, A_TP_MOD_CONFIG)); 8581 t4_read_pace_tbl(sc, pace_tab); 8582 8583 sbuf_printf(sb, "Scheduler Mode Channel Rate (Kbps) " 8584 "Class IPG (0.1 ns) Flow IPG (us)"); 8585 8586 for (i = 0; i < NTX_SCHED; ++i, map >>= 2) { 8587 t4_get_tx_sched(sc, i, &kbps, &ipg, 1); 8588 sbuf_printf(sb, "\n %u %-5s %u ", i, 8589 (mode & (1 << i)) ? "flow" : "class", map & 3); 8590 if (kbps) 8591 sbuf_printf(sb, "%9u ", kbps); 8592 else 8593 sbuf_printf(sb, " disabled "); 8594 8595 if (ipg) 8596 sbuf_printf(sb, "%13u ", ipg); 8597 else 8598 sbuf_printf(sb, " disabled "); 8599 8600 if (pace_tab[i]) 8601 sbuf_printf(sb, "%10u", pace_tab[i]); 8602 else 8603 sbuf_printf(sb, " disabled"); 8604 } 8605 8606 rc = sbuf_finish(sb); 8607 sbuf_delete(sb); 8608 8609 return (rc); 8610 } 8611 8612 static int 8613 sysctl_lb_stats(SYSCTL_HANDLER_ARGS) 8614 { 8615 struct adapter *sc = arg1; 8616 struct sbuf *sb; 8617 int rc, i, j; 8618 uint64_t *p0, *p1; 8619 struct lb_port_stats s[2]; 8620 static const char *stat_name[] = { 8621 "OctetsOK:", "FramesOK:", "BcastFrames:", "McastFrames:", 8622 "UcastFrames:", "ErrorFrames:", "Frames64:", "Frames65To127:", 8623 "Frames128To255:", "Frames256To511:", "Frames512To1023:", 8624 "Frames1024To1518:", "Frames1519ToMax:", "FramesDropped:", 8625 "BG0FramesDropped:", "BG1FramesDropped:", "BG2FramesDropped:", 8626 "BG3FramesDropped:", "BG0FramesTrunc:", "BG1FramesTrunc:", 8627 "BG2FramesTrunc:", "BG3FramesTrunc:" 8628 }; 8629 8630 rc = sysctl_wire_old_buffer(req, 0); 8631 if (rc != 0) 8632 return (rc); 8633 8634 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 8635 if (sb == NULL) 8636 return (ENOMEM); 8637 8638 memset(s, 0, sizeof(s)); 8639 8640 for (i = 0; i < sc->chip_params->nchan; i += 2) { 8641 t4_get_lb_stats(sc, i, &s[0]); 8642 t4_get_lb_stats(sc, i + 1, &s[1]); 8643 8644 p0 = &s[0].octets; 8645 p1 = &s[1].octets; 8646 sbuf_printf(sb, "%s Loopback %u" 8647 " Loopback %u", i == 0 ? "" : "\n", i, i + 1); 8648 8649 for (j = 0; j < nitems(stat_name); j++) 8650 sbuf_printf(sb, "\n%-17s %20ju %20ju", stat_name[j], 8651 *p0++, *p1++); 8652 } 8653 8654 rc = sbuf_finish(sb); 8655 sbuf_delete(sb); 8656 8657 return (rc); 8658 } 8659 8660 static int 8661 sysctl_linkdnrc(SYSCTL_HANDLER_ARGS) 8662 { 8663 int rc = 0; 8664 struct port_info *pi = arg1; 8665 struct link_config *lc = &pi->link_cfg; 8666 struct sbuf *sb; 8667 8668 rc = sysctl_wire_old_buffer(req, 0); 8669 if (rc != 0) 8670 return(rc); 8671 sb = sbuf_new_for_sysctl(NULL, NULL, 64, req); 8672 if (sb == NULL) 8673 return (ENOMEM); 8674 8675 if (lc->link_ok || lc->link_down_rc == 255) 8676 sbuf_printf(sb, "n/a"); 8677 else 8678 sbuf_printf(sb, "%s", t4_link_down_rc_str(lc->link_down_rc)); 8679 8680 rc = sbuf_finish(sb); 8681 sbuf_delete(sb); 8682 8683 return (rc); 8684 } 8685 8686 struct mem_desc { 8687 unsigned int base; 8688 unsigned int limit; 8689 unsigned int idx; 8690 }; 8691 8692 static int 8693 mem_desc_cmp(const void *a, const void *b) 8694 { 8695 return ((const struct mem_desc *)a)->base - 8696 ((const struct mem_desc *)b)->base; 8697 } 8698 8699 static void 8700 mem_region_show(struct sbuf *sb, const char *name, unsigned int from, 8701 unsigned int to) 8702 { 8703 unsigned int size; 8704 8705 if (from == to) 8706 return; 8707 8708 size = to - from + 1; 8709 if (size == 0) 8710 return; 8711 8712 /* XXX: need humanize_number(3) in libkern for a more readable 'size' */ 8713 sbuf_printf(sb, "%-15s %#x-%#x [%u]\n", name, from, to, size); 8714 } 8715 8716 static int 8717 sysctl_meminfo(SYSCTL_HANDLER_ARGS) 8718 { 8719 struct adapter *sc = arg1; 8720 struct sbuf *sb; 8721 int rc, i, n; 8722 uint32_t lo, hi, used, alloc; 8723 static const char *memory[] = {"EDC0:", "EDC1:", "MC:", "MC0:", "MC1:"}; 8724 static const char *region[] = { 8725 "DBQ contexts:", "IMSG contexts:", "FLM cache:", "TCBs:", 8726 "Pstructs:", "Timers:", "Rx FL:", "Tx FL:", "Pstruct FL:", 8727 "Tx payload:", "Rx payload:", "LE hash:", "iSCSI region:", 8728 "TDDP region:", "TPT region:", "STAG region:", "RQ region:", 8729 "RQUDP region:", "PBL region:", "TXPBL region:", 8730 "DBVFIFO region:", "ULPRX state:", "ULPTX state:", 8731 "On-chip queues:", "TLS keys:", 8732 }; 8733 struct mem_desc avail[4]; 8734 struct mem_desc mem[nitems(region) + 3]; /* up to 3 holes */ 8735 struct mem_desc *md = mem; 8736 8737 rc = sysctl_wire_old_buffer(req, 0); 8738 if (rc != 0) 8739 return (rc); 8740 8741 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 8742 if (sb == NULL) 8743 return (ENOMEM); 8744 8745 for (i = 0; i < nitems(mem); i++) { 8746 mem[i].limit = 0; 8747 mem[i].idx = i; 8748 } 8749 8750 /* Find and sort the populated memory ranges */ 8751 i = 0; 8752 lo = t4_read_reg(sc, A_MA_TARGET_MEM_ENABLE); 8753 if (lo & F_EDRAM0_ENABLE) { 8754 hi = t4_read_reg(sc, A_MA_EDRAM0_BAR); 8755 avail[i].base = G_EDRAM0_BASE(hi) << 20; 8756 avail[i].limit = avail[i].base + (G_EDRAM0_SIZE(hi) << 20); 8757 avail[i].idx = 0; 8758 i++; 8759 } 8760 if (lo & F_EDRAM1_ENABLE) { 8761 hi = t4_read_reg(sc, A_MA_EDRAM1_BAR); 8762 avail[i].base = G_EDRAM1_BASE(hi) << 20; 8763 avail[i].limit = avail[i].base + (G_EDRAM1_SIZE(hi) << 20); 8764 avail[i].idx = 1; 8765 i++; 8766 } 8767 if (lo & F_EXT_MEM_ENABLE) { 8768 hi = t4_read_reg(sc, A_MA_EXT_MEMORY_BAR); 8769 avail[i].base = G_EXT_MEM_BASE(hi) << 20; 8770 avail[i].limit = avail[i].base + 8771 (G_EXT_MEM_SIZE(hi) << 20); 8772 avail[i].idx = is_t5(sc) ? 3 : 2; /* Call it MC0 for T5 */ 8773 i++; 8774 } 8775 if (is_t5(sc) && lo & F_EXT_MEM1_ENABLE) { 8776 hi = t4_read_reg(sc, A_MA_EXT_MEMORY1_BAR); 8777 avail[i].base = G_EXT_MEM1_BASE(hi) << 20; 8778 avail[i].limit = avail[i].base + 8779 (G_EXT_MEM1_SIZE(hi) << 20); 8780 avail[i].idx = 4; 8781 i++; 8782 } 8783 if (!i) /* no memory available */ 8784 return 0; 8785 qsort(avail, i, sizeof(struct mem_desc), mem_desc_cmp); 8786 8787 (md++)->base = t4_read_reg(sc, A_SGE_DBQ_CTXT_BADDR); 8788 (md++)->base = t4_read_reg(sc, A_SGE_IMSG_CTXT_BADDR); 8789 (md++)->base = t4_read_reg(sc, A_SGE_FLM_CACHE_BADDR); 8790 (md++)->base = t4_read_reg(sc, A_TP_CMM_TCB_BASE); 8791 (md++)->base = t4_read_reg(sc, A_TP_CMM_MM_BASE); 8792 (md++)->base = t4_read_reg(sc, A_TP_CMM_TIMER_BASE); 8793 (md++)->base = t4_read_reg(sc, A_TP_CMM_MM_RX_FLST_BASE); 8794 (md++)->base = t4_read_reg(sc, A_TP_CMM_MM_TX_FLST_BASE); 8795 (md++)->base = t4_read_reg(sc, A_TP_CMM_MM_PS_FLST_BASE); 8796 8797 /* the next few have explicit upper bounds */ 8798 md->base = t4_read_reg(sc, A_TP_PMM_TX_BASE); 8799 md->limit = md->base - 1 + 8800 t4_read_reg(sc, A_TP_PMM_TX_PAGE_SIZE) * 8801 G_PMTXMAXPAGE(t4_read_reg(sc, A_TP_PMM_TX_MAX_PAGE)); 8802 md++; 8803 8804 md->base = t4_read_reg(sc, A_TP_PMM_RX_BASE); 8805 md->limit = md->base - 1 + 8806 t4_read_reg(sc, A_TP_PMM_RX_PAGE_SIZE) * 8807 G_PMRXMAXPAGE(t4_read_reg(sc, A_TP_PMM_RX_MAX_PAGE)); 8808 md++; 8809 8810 if (t4_read_reg(sc, A_LE_DB_CONFIG) & F_HASHEN) { 8811 if (chip_id(sc) <= CHELSIO_T5) 8812 md->base = t4_read_reg(sc, A_LE_DB_HASH_TID_BASE); 8813 else 8814 md->base = t4_read_reg(sc, A_LE_DB_HASH_TBL_BASE_ADDR); 8815 md->limit = 0; 8816 } else { 8817 md->base = 0; 8818 md->idx = nitems(region); /* hide it */ 8819 } 8820 md++; 8821 8822 #define ulp_region(reg) \ 8823 md->base = t4_read_reg(sc, A_ULP_ ## reg ## _LLIMIT);\ 8824 (md++)->limit = t4_read_reg(sc, A_ULP_ ## reg ## _ULIMIT) 8825 8826 ulp_region(RX_ISCSI); 8827 ulp_region(RX_TDDP); 8828 ulp_region(TX_TPT); 8829 ulp_region(RX_STAG); 8830 ulp_region(RX_RQ); 8831 ulp_region(RX_RQUDP); 8832 ulp_region(RX_PBL); 8833 ulp_region(TX_PBL); 8834 #undef ulp_region 8835 8836 md->base = 0; 8837 md->idx = nitems(region); 8838 if (!is_t4(sc)) { 8839 uint32_t size = 0; 8840 uint32_t sge_ctrl = t4_read_reg(sc, A_SGE_CONTROL2); 8841 uint32_t fifo_size = t4_read_reg(sc, A_SGE_DBVFIFO_SIZE); 8842 8843 if (is_t5(sc)) { 8844 if (sge_ctrl & F_VFIFO_ENABLE) 8845 size = G_DBVFIFO_SIZE(fifo_size); 8846 } else 8847 size = G_T6_DBVFIFO_SIZE(fifo_size); 8848 8849 if (size) { 8850 md->base = G_BASEADDR(t4_read_reg(sc, 8851 A_SGE_DBVFIFO_BADDR)); 8852 md->limit = md->base + (size << 2) - 1; 8853 } 8854 } 8855 md++; 8856 8857 md->base = t4_read_reg(sc, A_ULP_RX_CTX_BASE); 8858 md->limit = 0; 8859 md++; 8860 md->base = t4_read_reg(sc, A_ULP_TX_ERR_TABLE_BASE); 8861 md->limit = 0; 8862 md++; 8863 8864 md->base = sc->vres.ocq.start; 8865 if (sc->vres.ocq.size) 8866 md->limit = md->base + sc->vres.ocq.size - 1; 8867 else 8868 md->idx = nitems(region); /* hide it */ 8869 md++; 8870 8871 md->base = sc->vres.key.start; 8872 if (sc->vres.key.size) 8873 md->limit = md->base + sc->vres.key.size - 1; 8874 else 8875 md->idx = nitems(region); /* hide it */ 8876 md++; 8877 8878 /* add any address-space holes, there can be up to 3 */ 8879 for (n = 0; n < i - 1; n++) 8880 if (avail[n].limit < avail[n + 1].base) 8881 (md++)->base = avail[n].limit; 8882 if (avail[n].limit) 8883 (md++)->base = avail[n].limit; 8884 8885 n = md - mem; 8886 qsort(mem, n, sizeof(struct mem_desc), mem_desc_cmp); 8887 8888 for (lo = 0; lo < i; lo++) 8889 mem_region_show(sb, memory[avail[lo].idx], avail[lo].base, 8890 avail[lo].limit - 1); 8891 8892 sbuf_printf(sb, "\n"); 8893 for (i = 0; i < n; i++) { 8894 if (mem[i].idx >= nitems(region)) 8895 continue; /* skip holes */ 8896 if (!mem[i].limit) 8897 mem[i].limit = i < n - 1 ? mem[i + 1].base - 1 : ~0; 8898 mem_region_show(sb, region[mem[i].idx], mem[i].base, 8899 mem[i].limit); 8900 } 8901 8902 sbuf_printf(sb, "\n"); 8903 lo = t4_read_reg(sc, A_CIM_SDRAM_BASE_ADDR); 8904 hi = t4_read_reg(sc, A_CIM_SDRAM_ADDR_SIZE) + lo - 1; 8905 mem_region_show(sb, "uP RAM:", lo, hi); 8906 8907 lo = t4_read_reg(sc, A_CIM_EXTMEM2_BASE_ADDR); 8908 hi = t4_read_reg(sc, A_CIM_EXTMEM2_ADDR_SIZE) + lo - 1; 8909 mem_region_show(sb, "uP Extmem2:", lo, hi); 8910 8911 lo = t4_read_reg(sc, A_TP_PMM_RX_MAX_PAGE); 8912 sbuf_printf(sb, "\n%u Rx pages of size %uKiB for %u channels\n", 8913 G_PMRXMAXPAGE(lo), 8914 t4_read_reg(sc, A_TP_PMM_RX_PAGE_SIZE) >> 10, 8915 (lo & F_PMRXNUMCHN) ? 2 : 1); 8916 8917 lo = t4_read_reg(sc, A_TP_PMM_TX_MAX_PAGE); 8918 hi = t4_read_reg(sc, A_TP_PMM_TX_PAGE_SIZE); 8919 sbuf_printf(sb, "%u Tx pages of size %u%ciB for %u channels\n", 8920 G_PMTXMAXPAGE(lo), 8921 hi >= (1 << 20) ? (hi >> 20) : (hi >> 10), 8922 hi >= (1 << 20) ? 'M' : 'K', 1 << G_PMTXNUMCHN(lo)); 8923 sbuf_printf(sb, "%u p-structs\n", 8924 t4_read_reg(sc, A_TP_CMM_MM_MAX_PSTRUCT)); 8925 8926 for (i = 0; i < 4; i++) { 8927 if (chip_id(sc) > CHELSIO_T5) 8928 lo = t4_read_reg(sc, A_MPS_RX_MAC_BG_PG_CNT0 + i * 4); 8929 else 8930 lo = t4_read_reg(sc, A_MPS_RX_PG_RSV0 + i * 4); 8931 if (is_t5(sc)) { 8932 used = G_T5_USED(lo); 8933 alloc = G_T5_ALLOC(lo); 8934 } else { 8935 used = G_USED(lo); 8936 alloc = G_ALLOC(lo); 8937 } 8938 /* For T6 these are MAC buffer groups */ 8939 sbuf_printf(sb, "\nPort %d using %u pages out of %u allocated", 8940 i, used, alloc); 8941 } 8942 for (i = 0; i < sc->chip_params->nchan; i++) { 8943 if (chip_id(sc) > CHELSIO_T5) 8944 lo = t4_read_reg(sc, A_MPS_RX_LPBK_BG_PG_CNT0 + i * 4); 8945 else 8946 lo = t4_read_reg(sc, A_MPS_RX_PG_RSV4 + i * 4); 8947 if (is_t5(sc)) { 8948 used = G_T5_USED(lo); 8949 alloc = G_T5_ALLOC(lo); 8950 } else { 8951 used = G_USED(lo); 8952 alloc = G_ALLOC(lo); 8953 } 8954 /* For T6 these are MAC buffer groups */ 8955 sbuf_printf(sb, 8956 "\nLoopback %d using %u pages out of %u allocated", 8957 i, used, alloc); 8958 } 8959 8960 rc = sbuf_finish(sb); 8961 sbuf_delete(sb); 8962 8963 return (rc); 8964 } 8965 8966 static inline void 8967 tcamxy2valmask(uint64_t x, uint64_t y, uint8_t *addr, uint64_t *mask) 8968 { 8969 *mask = x | y; 8970 y = htobe64(y); 8971 memcpy(addr, (char *)&y + 2, ETHER_ADDR_LEN); 8972 } 8973 8974 static int 8975 sysctl_mps_tcam(SYSCTL_HANDLER_ARGS) 8976 { 8977 struct adapter *sc = arg1; 8978 struct sbuf *sb; 8979 int rc, i; 8980 8981 MPASS(chip_id(sc) <= CHELSIO_T5); 8982 8983 rc = sysctl_wire_old_buffer(req, 0); 8984 if (rc != 0) 8985 return (rc); 8986 8987 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 8988 if (sb == NULL) 8989 return (ENOMEM); 8990 8991 sbuf_printf(sb, 8992 "Idx Ethernet address Mask Vld Ports PF" 8993 " VF Replication P0 P1 P2 P3 ML"); 8994 for (i = 0; i < sc->chip_params->mps_tcam_size; i++) { 8995 uint64_t tcamx, tcamy, mask; 8996 uint32_t cls_lo, cls_hi; 8997 uint8_t addr[ETHER_ADDR_LEN]; 8998 8999 tcamy = t4_read_reg64(sc, MPS_CLS_TCAM_Y_L(i)); 9000 tcamx = t4_read_reg64(sc, MPS_CLS_TCAM_X_L(i)); 9001 if (tcamx & tcamy) 9002 continue; 9003 tcamxy2valmask(tcamx, tcamy, addr, &mask); 9004 cls_lo = t4_read_reg(sc, MPS_CLS_SRAM_L(i)); 9005 cls_hi = t4_read_reg(sc, MPS_CLS_SRAM_H(i)); 9006 sbuf_printf(sb, "\n%3u %02x:%02x:%02x:%02x:%02x:%02x %012jx" 9007 " %c %#x%4u%4d", i, addr[0], addr[1], addr[2], 9008 addr[3], addr[4], addr[5], (uintmax_t)mask, 9009 (cls_lo & F_SRAM_VLD) ? 'Y' : 'N', 9010 G_PORTMAP(cls_hi), G_PF(cls_lo), 9011 (cls_lo & F_VF_VALID) ? G_VF(cls_lo) : -1); 9012 9013 if (cls_lo & F_REPLICATE) { 9014 struct fw_ldst_cmd ldst_cmd; 9015 9016 memset(&ldst_cmd, 0, sizeof(ldst_cmd)); 9017 ldst_cmd.op_to_addrspace = 9018 htobe32(V_FW_CMD_OP(FW_LDST_CMD) | 9019 F_FW_CMD_REQUEST | F_FW_CMD_READ | 9020 V_FW_LDST_CMD_ADDRSPACE(FW_LDST_ADDRSPC_MPS)); 9021 ldst_cmd.cycles_to_len16 = htobe32(FW_LEN16(ldst_cmd)); 9022 ldst_cmd.u.mps.rplc.fid_idx = 9023 htobe16(V_FW_LDST_CMD_FID(FW_LDST_MPS_RPLC) | 9024 V_FW_LDST_CMD_IDX(i)); 9025 9026 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, 9027 "t4mps"); 9028 if (rc) 9029 break; 9030 rc = -t4_wr_mbox(sc, sc->mbox, &ldst_cmd, 9031 sizeof(ldst_cmd), &ldst_cmd); 9032 end_synchronized_op(sc, 0); 9033 9034 if (rc != 0) { 9035 sbuf_printf(sb, "%36d", rc); 9036 rc = 0; 9037 } else { 9038 sbuf_printf(sb, " %08x %08x %08x %08x", 9039 be32toh(ldst_cmd.u.mps.rplc.rplc127_96), 9040 be32toh(ldst_cmd.u.mps.rplc.rplc95_64), 9041 be32toh(ldst_cmd.u.mps.rplc.rplc63_32), 9042 be32toh(ldst_cmd.u.mps.rplc.rplc31_0)); 9043 } 9044 } else 9045 sbuf_printf(sb, "%36s", ""); 9046 9047 sbuf_printf(sb, "%4u%3u%3u%3u %#3x", G_SRAM_PRIO0(cls_lo), 9048 G_SRAM_PRIO1(cls_lo), G_SRAM_PRIO2(cls_lo), 9049 G_SRAM_PRIO3(cls_lo), (cls_lo >> S_MULTILISTEN0) & 0xf); 9050 } 9051 9052 if (rc) 9053 (void) sbuf_finish(sb); 9054 else 9055 rc = sbuf_finish(sb); 9056 sbuf_delete(sb); 9057 9058 return (rc); 9059 } 9060 9061 static int 9062 sysctl_mps_tcam_t6(SYSCTL_HANDLER_ARGS) 9063 { 9064 struct adapter *sc = arg1; 9065 struct sbuf *sb; 9066 int rc, i; 9067 9068 MPASS(chip_id(sc) > CHELSIO_T5); 9069 9070 rc = sysctl_wire_old_buffer(req, 0); 9071 if (rc != 0) 9072 return (rc); 9073 9074 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 9075 if (sb == NULL) 9076 return (ENOMEM); 9077 9078 sbuf_printf(sb, "Idx Ethernet address Mask VNI Mask" 9079 " IVLAN Vld DIP_Hit Lookup Port Vld Ports PF VF" 9080 " Replication" 9081 " P0 P1 P2 P3 ML\n"); 9082 9083 for (i = 0; i < sc->chip_params->mps_tcam_size; i++) { 9084 uint8_t dip_hit, vlan_vld, lookup_type, port_num; 9085 uint16_t ivlan; 9086 uint64_t tcamx, tcamy, val, mask; 9087 uint32_t cls_lo, cls_hi, ctl, data2, vnix, vniy; 9088 uint8_t addr[ETHER_ADDR_LEN]; 9089 9090 ctl = V_CTLREQID(1) | V_CTLCMDTYPE(0) | V_CTLXYBITSEL(0); 9091 if (i < 256) 9092 ctl |= V_CTLTCAMINDEX(i) | V_CTLTCAMSEL(0); 9093 else 9094 ctl |= V_CTLTCAMINDEX(i - 256) | V_CTLTCAMSEL(1); 9095 t4_write_reg(sc, A_MPS_CLS_TCAM_DATA2_CTL, ctl); 9096 val = t4_read_reg(sc, A_MPS_CLS_TCAM_RDATA1_REQ_ID1); 9097 tcamy = G_DMACH(val) << 32; 9098 tcamy |= t4_read_reg(sc, A_MPS_CLS_TCAM_RDATA0_REQ_ID1); 9099 data2 = t4_read_reg(sc, A_MPS_CLS_TCAM_RDATA2_REQ_ID1); 9100 lookup_type = G_DATALKPTYPE(data2); 9101 port_num = G_DATAPORTNUM(data2); 9102 if (lookup_type && lookup_type != M_DATALKPTYPE) { 9103 /* Inner header VNI */ 9104 vniy = ((data2 & F_DATAVIDH2) << 23) | 9105 (G_DATAVIDH1(data2) << 16) | G_VIDL(val); 9106 dip_hit = data2 & F_DATADIPHIT; 9107 vlan_vld = 0; 9108 } else { 9109 vniy = 0; 9110 dip_hit = 0; 9111 vlan_vld = data2 & F_DATAVIDH2; 9112 ivlan = G_VIDL(val); 9113 } 9114 9115 ctl |= V_CTLXYBITSEL(1); 9116 t4_write_reg(sc, A_MPS_CLS_TCAM_DATA2_CTL, ctl); 9117 val = t4_read_reg(sc, A_MPS_CLS_TCAM_RDATA1_REQ_ID1); 9118 tcamx = G_DMACH(val) << 32; 9119 tcamx |= t4_read_reg(sc, A_MPS_CLS_TCAM_RDATA0_REQ_ID1); 9120 data2 = t4_read_reg(sc, A_MPS_CLS_TCAM_RDATA2_REQ_ID1); 9121 if (lookup_type && lookup_type != M_DATALKPTYPE) { 9122 /* Inner header VNI mask */ 9123 vnix = ((data2 & F_DATAVIDH2) << 23) | 9124 (G_DATAVIDH1(data2) << 16) | G_VIDL(val); 9125 } else 9126 vnix = 0; 9127 9128 if (tcamx & tcamy) 9129 continue; 9130 tcamxy2valmask(tcamx, tcamy, addr, &mask); 9131 9132 cls_lo = t4_read_reg(sc, MPS_CLS_SRAM_L(i)); 9133 cls_hi = t4_read_reg(sc, MPS_CLS_SRAM_H(i)); 9134 9135 if (lookup_type && lookup_type != M_DATALKPTYPE) { 9136 sbuf_printf(sb, "\n%3u %02x:%02x:%02x:%02x:%02x:%02x " 9137 "%012jx %06x %06x - - %3c" 9138 " 'I' %4x %3c %#x%4u%4d", i, addr[0], 9139 addr[1], addr[2], addr[3], addr[4], addr[5], 9140 (uintmax_t)mask, vniy, vnix, dip_hit ? 'Y' : 'N', 9141 port_num, cls_lo & F_T6_SRAM_VLD ? 'Y' : 'N', 9142 G_PORTMAP(cls_hi), G_T6_PF(cls_lo), 9143 cls_lo & F_T6_VF_VALID ? G_T6_VF(cls_lo) : -1); 9144 } else { 9145 sbuf_printf(sb, "\n%3u %02x:%02x:%02x:%02x:%02x:%02x " 9146 "%012jx - - ", i, addr[0], addr[1], 9147 addr[2], addr[3], addr[4], addr[5], 9148 (uintmax_t)mask); 9149 9150 if (vlan_vld) 9151 sbuf_printf(sb, "%4u Y ", ivlan); 9152 else 9153 sbuf_printf(sb, " - N "); 9154 9155 sbuf_printf(sb, "- %3c %4x %3c %#x%4u%4d", 9156 lookup_type ? 'I' : 'O', port_num, 9157 cls_lo & F_T6_SRAM_VLD ? 'Y' : 'N', 9158 G_PORTMAP(cls_hi), G_T6_PF(cls_lo), 9159 cls_lo & F_T6_VF_VALID ? G_T6_VF(cls_lo) : -1); 9160 } 9161 9162 9163 if (cls_lo & F_T6_REPLICATE) { 9164 struct fw_ldst_cmd ldst_cmd; 9165 9166 memset(&ldst_cmd, 0, sizeof(ldst_cmd)); 9167 ldst_cmd.op_to_addrspace = 9168 htobe32(V_FW_CMD_OP(FW_LDST_CMD) | 9169 F_FW_CMD_REQUEST | F_FW_CMD_READ | 9170 V_FW_LDST_CMD_ADDRSPACE(FW_LDST_ADDRSPC_MPS)); 9171 ldst_cmd.cycles_to_len16 = htobe32(FW_LEN16(ldst_cmd)); 9172 ldst_cmd.u.mps.rplc.fid_idx = 9173 htobe16(V_FW_LDST_CMD_FID(FW_LDST_MPS_RPLC) | 9174 V_FW_LDST_CMD_IDX(i)); 9175 9176 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, 9177 "t6mps"); 9178 if (rc) 9179 break; 9180 rc = -t4_wr_mbox(sc, sc->mbox, &ldst_cmd, 9181 sizeof(ldst_cmd), &ldst_cmd); 9182 end_synchronized_op(sc, 0); 9183 9184 if (rc != 0) { 9185 sbuf_printf(sb, "%72d", rc); 9186 rc = 0; 9187 } else { 9188 sbuf_printf(sb, " %08x %08x %08x %08x" 9189 " %08x %08x %08x %08x", 9190 be32toh(ldst_cmd.u.mps.rplc.rplc255_224), 9191 be32toh(ldst_cmd.u.mps.rplc.rplc223_192), 9192 be32toh(ldst_cmd.u.mps.rplc.rplc191_160), 9193 be32toh(ldst_cmd.u.mps.rplc.rplc159_128), 9194 be32toh(ldst_cmd.u.mps.rplc.rplc127_96), 9195 be32toh(ldst_cmd.u.mps.rplc.rplc95_64), 9196 be32toh(ldst_cmd.u.mps.rplc.rplc63_32), 9197 be32toh(ldst_cmd.u.mps.rplc.rplc31_0)); 9198 } 9199 } else 9200 sbuf_printf(sb, "%72s", ""); 9201 9202 sbuf_printf(sb, "%4u%3u%3u%3u %#x", 9203 G_T6_SRAM_PRIO0(cls_lo), G_T6_SRAM_PRIO1(cls_lo), 9204 G_T6_SRAM_PRIO2(cls_lo), G_T6_SRAM_PRIO3(cls_lo), 9205 (cls_lo >> S_T6_MULTILISTEN0) & 0xf); 9206 } 9207 9208 if (rc) 9209 (void) sbuf_finish(sb); 9210 else 9211 rc = sbuf_finish(sb); 9212 sbuf_delete(sb); 9213 9214 return (rc); 9215 } 9216 9217 static int 9218 sysctl_path_mtus(SYSCTL_HANDLER_ARGS) 9219 { 9220 struct adapter *sc = arg1; 9221 struct sbuf *sb; 9222 int rc; 9223 uint16_t mtus[NMTUS]; 9224 9225 rc = sysctl_wire_old_buffer(req, 0); 9226 if (rc != 0) 9227 return (rc); 9228 9229 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9230 if (sb == NULL) 9231 return (ENOMEM); 9232 9233 t4_read_mtu_tbl(sc, mtus, NULL); 9234 9235 sbuf_printf(sb, "%u %u %u %u %u %u %u %u %u %u %u %u %u %u %u %u", 9236 mtus[0], mtus[1], mtus[2], mtus[3], mtus[4], mtus[5], mtus[6], 9237 mtus[7], mtus[8], mtus[9], mtus[10], mtus[11], mtus[12], mtus[13], 9238 mtus[14], mtus[15]); 9239 9240 rc = sbuf_finish(sb); 9241 sbuf_delete(sb); 9242 9243 return (rc); 9244 } 9245 9246 static int 9247 sysctl_pm_stats(SYSCTL_HANDLER_ARGS) 9248 { 9249 struct adapter *sc = arg1; 9250 struct sbuf *sb; 9251 int rc, i; 9252 uint32_t tx_cnt[MAX_PM_NSTATS], rx_cnt[MAX_PM_NSTATS]; 9253 uint64_t tx_cyc[MAX_PM_NSTATS], rx_cyc[MAX_PM_NSTATS]; 9254 static const char *tx_stats[MAX_PM_NSTATS] = { 9255 "Read:", "Write bypass:", "Write mem:", "Bypass + mem:", 9256 "Tx FIFO wait", NULL, "Tx latency" 9257 }; 9258 static const char *rx_stats[MAX_PM_NSTATS] = { 9259 "Read:", "Write bypass:", "Write mem:", "Flush:", 9260 "Rx FIFO wait", NULL, "Rx latency" 9261 }; 9262 9263 rc = sysctl_wire_old_buffer(req, 0); 9264 if (rc != 0) 9265 return (rc); 9266 9267 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9268 if (sb == NULL) 9269 return (ENOMEM); 9270 9271 t4_pmtx_get_stats(sc, tx_cnt, tx_cyc); 9272 t4_pmrx_get_stats(sc, rx_cnt, rx_cyc); 9273 9274 sbuf_printf(sb, " Tx pcmds Tx bytes"); 9275 for (i = 0; i < 4; i++) { 9276 sbuf_printf(sb, "\n%-13s %10u %20ju", tx_stats[i], tx_cnt[i], 9277 tx_cyc[i]); 9278 } 9279 9280 sbuf_printf(sb, "\n Rx pcmds Rx bytes"); 9281 for (i = 0; i < 4; i++) { 9282 sbuf_printf(sb, "\n%-13s %10u %20ju", rx_stats[i], rx_cnt[i], 9283 rx_cyc[i]); 9284 } 9285 9286 if (chip_id(sc) > CHELSIO_T5) { 9287 sbuf_printf(sb, 9288 "\n Total wait Total occupancy"); 9289 sbuf_printf(sb, "\n%-13s %10u %20ju", tx_stats[i], tx_cnt[i], 9290 tx_cyc[i]); 9291 sbuf_printf(sb, "\n%-13s %10u %20ju", rx_stats[i], rx_cnt[i], 9292 rx_cyc[i]); 9293 9294 i += 2; 9295 MPASS(i < nitems(tx_stats)); 9296 9297 sbuf_printf(sb, 9298 "\n Reads Total wait"); 9299 sbuf_printf(sb, "\n%-13s %10u %20ju", tx_stats[i], tx_cnt[i], 9300 tx_cyc[i]); 9301 sbuf_printf(sb, "\n%-13s %10u %20ju", rx_stats[i], rx_cnt[i], 9302 rx_cyc[i]); 9303 } 9304 9305 rc = sbuf_finish(sb); 9306 sbuf_delete(sb); 9307 9308 return (rc); 9309 } 9310 9311 static int 9312 sysctl_rdma_stats(SYSCTL_HANDLER_ARGS) 9313 { 9314 struct adapter *sc = arg1; 9315 struct sbuf *sb; 9316 int rc; 9317 struct tp_rdma_stats stats; 9318 9319 rc = sysctl_wire_old_buffer(req, 0); 9320 if (rc != 0) 9321 return (rc); 9322 9323 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9324 if (sb == NULL) 9325 return (ENOMEM); 9326 9327 mtx_lock(&sc->reg_lock); 9328 t4_tp_get_rdma_stats(sc, &stats, 0); 9329 mtx_unlock(&sc->reg_lock); 9330 9331 sbuf_printf(sb, "NoRQEModDefferals: %u\n", stats.rqe_dfr_mod); 9332 sbuf_printf(sb, "NoRQEPktDefferals: %u", stats.rqe_dfr_pkt); 9333 9334 rc = sbuf_finish(sb); 9335 sbuf_delete(sb); 9336 9337 return (rc); 9338 } 9339 9340 static int 9341 sysctl_tcp_stats(SYSCTL_HANDLER_ARGS) 9342 { 9343 struct adapter *sc = arg1; 9344 struct sbuf *sb; 9345 int rc; 9346 struct tp_tcp_stats v4, v6; 9347 9348 rc = sysctl_wire_old_buffer(req, 0); 9349 if (rc != 0) 9350 return (rc); 9351 9352 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9353 if (sb == NULL) 9354 return (ENOMEM); 9355 9356 mtx_lock(&sc->reg_lock); 9357 t4_tp_get_tcp_stats(sc, &v4, &v6, 0); 9358 mtx_unlock(&sc->reg_lock); 9359 9360 sbuf_printf(sb, 9361 " IP IPv6\n"); 9362 sbuf_printf(sb, "OutRsts: %20u %20u\n", 9363 v4.tcp_out_rsts, v6.tcp_out_rsts); 9364 sbuf_printf(sb, "InSegs: %20ju %20ju\n", 9365 v4.tcp_in_segs, v6.tcp_in_segs); 9366 sbuf_printf(sb, "OutSegs: %20ju %20ju\n", 9367 v4.tcp_out_segs, v6.tcp_out_segs); 9368 sbuf_printf(sb, "RetransSegs: %20ju %20ju", 9369 v4.tcp_retrans_segs, v6.tcp_retrans_segs); 9370 9371 rc = sbuf_finish(sb); 9372 sbuf_delete(sb); 9373 9374 return (rc); 9375 } 9376 9377 static int 9378 sysctl_tids(SYSCTL_HANDLER_ARGS) 9379 { 9380 struct adapter *sc = arg1; 9381 struct sbuf *sb; 9382 int rc; 9383 struct tid_info *t = &sc->tids; 9384 9385 rc = sysctl_wire_old_buffer(req, 0); 9386 if (rc != 0) 9387 return (rc); 9388 9389 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9390 if (sb == NULL) 9391 return (ENOMEM); 9392 9393 if (t->natids) { 9394 sbuf_printf(sb, "ATID range: 0-%u, in use: %u\n", t->natids - 1, 9395 t->atids_in_use); 9396 } 9397 9398 if (t->nhpftids) { 9399 sbuf_printf(sb, "HPFTID range: %u-%u, in use: %u\n", 9400 t->hpftid_base, t->hpftid_end, t->hpftids_in_use); 9401 } 9402 9403 if (t->ntids) { 9404 sbuf_printf(sb, "TID range: "); 9405 if (t4_read_reg(sc, A_LE_DB_CONFIG) & F_HASHEN) { 9406 uint32_t b, hb; 9407 9408 if (chip_id(sc) <= CHELSIO_T5) { 9409 b = t4_read_reg(sc, A_LE_DB_SERVER_INDEX) / 4; 9410 hb = t4_read_reg(sc, A_LE_DB_TID_HASHBASE) / 4; 9411 } else { 9412 b = t4_read_reg(sc, A_LE_DB_SRVR_START_INDEX); 9413 hb = t4_read_reg(sc, A_T6_LE_DB_HASH_TID_BASE); 9414 } 9415 9416 if (b) 9417 sbuf_printf(sb, "%u-%u, ", t->tid_base, b - 1); 9418 sbuf_printf(sb, "%u-%u", hb, t->ntids - 1); 9419 } else { 9420 sbuf_printf(sb, "%u-%u", t->tid_base, t->tid_base + 9421 t->ntids - 1); 9422 } 9423 sbuf_printf(sb, ", in use: %u\n", 9424 atomic_load_acq_int(&t->tids_in_use)); 9425 } 9426 9427 if (t->nstids) { 9428 sbuf_printf(sb, "STID range: %u-%u, in use: %u\n", t->stid_base, 9429 t->stid_base + t->nstids - 1, t->stids_in_use); 9430 } 9431 9432 if (t->nftids) { 9433 sbuf_printf(sb, "FTID range: %u-%u, in use: %u\n", t->ftid_base, 9434 t->ftid_end, t->ftids_in_use); 9435 } 9436 9437 if (t->netids) { 9438 sbuf_printf(sb, "ETID range: %u-%u, in use: %u\n", t->etid_base, 9439 t->etid_base + t->netids - 1, t->etids_in_use); 9440 } 9441 9442 sbuf_printf(sb, "HW TID usage: %u IP users, %u IPv6 users", 9443 t4_read_reg(sc, A_LE_DB_ACT_CNT_IPV4), 9444 t4_read_reg(sc, A_LE_DB_ACT_CNT_IPV6)); 9445 9446 rc = sbuf_finish(sb); 9447 sbuf_delete(sb); 9448 9449 return (rc); 9450 } 9451 9452 static int 9453 sysctl_tp_err_stats(SYSCTL_HANDLER_ARGS) 9454 { 9455 struct adapter *sc = arg1; 9456 struct sbuf *sb; 9457 int rc; 9458 struct tp_err_stats stats; 9459 9460 rc = sysctl_wire_old_buffer(req, 0); 9461 if (rc != 0) 9462 return (rc); 9463 9464 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9465 if (sb == NULL) 9466 return (ENOMEM); 9467 9468 mtx_lock(&sc->reg_lock); 9469 t4_tp_get_err_stats(sc, &stats, 0); 9470 mtx_unlock(&sc->reg_lock); 9471 9472 if (sc->chip_params->nchan > 2) { 9473 sbuf_printf(sb, " channel 0 channel 1" 9474 " channel 2 channel 3\n"); 9475 sbuf_printf(sb, "macInErrs: %10u %10u %10u %10u\n", 9476 stats.mac_in_errs[0], stats.mac_in_errs[1], 9477 stats.mac_in_errs[2], stats.mac_in_errs[3]); 9478 sbuf_printf(sb, "hdrInErrs: %10u %10u %10u %10u\n", 9479 stats.hdr_in_errs[0], stats.hdr_in_errs[1], 9480 stats.hdr_in_errs[2], stats.hdr_in_errs[3]); 9481 sbuf_printf(sb, "tcpInErrs: %10u %10u %10u %10u\n", 9482 stats.tcp_in_errs[0], stats.tcp_in_errs[1], 9483 stats.tcp_in_errs[2], stats.tcp_in_errs[3]); 9484 sbuf_printf(sb, "tcp6InErrs: %10u %10u %10u %10u\n", 9485 stats.tcp6_in_errs[0], stats.tcp6_in_errs[1], 9486 stats.tcp6_in_errs[2], stats.tcp6_in_errs[3]); 9487 sbuf_printf(sb, "tnlCongDrops: %10u %10u %10u %10u\n", 9488 stats.tnl_cong_drops[0], stats.tnl_cong_drops[1], 9489 stats.tnl_cong_drops[2], stats.tnl_cong_drops[3]); 9490 sbuf_printf(sb, "tnlTxDrops: %10u %10u %10u %10u\n", 9491 stats.tnl_tx_drops[0], stats.tnl_tx_drops[1], 9492 stats.tnl_tx_drops[2], stats.tnl_tx_drops[3]); 9493 sbuf_printf(sb, "ofldVlanDrops: %10u %10u %10u %10u\n", 9494 stats.ofld_vlan_drops[0], stats.ofld_vlan_drops[1], 9495 stats.ofld_vlan_drops[2], stats.ofld_vlan_drops[3]); 9496 sbuf_printf(sb, "ofldChanDrops: %10u %10u %10u %10u\n\n", 9497 stats.ofld_chan_drops[0], stats.ofld_chan_drops[1], 9498 stats.ofld_chan_drops[2], stats.ofld_chan_drops[3]); 9499 } else { 9500 sbuf_printf(sb, " channel 0 channel 1\n"); 9501 sbuf_printf(sb, "macInErrs: %10u %10u\n", 9502 stats.mac_in_errs[0], stats.mac_in_errs[1]); 9503 sbuf_printf(sb, "hdrInErrs: %10u %10u\n", 9504 stats.hdr_in_errs[0], stats.hdr_in_errs[1]); 9505 sbuf_printf(sb, "tcpInErrs: %10u %10u\n", 9506 stats.tcp_in_errs[0], stats.tcp_in_errs[1]); 9507 sbuf_printf(sb, "tcp6InErrs: %10u %10u\n", 9508 stats.tcp6_in_errs[0], stats.tcp6_in_errs[1]); 9509 sbuf_printf(sb, "tnlCongDrops: %10u %10u\n", 9510 stats.tnl_cong_drops[0], stats.tnl_cong_drops[1]); 9511 sbuf_printf(sb, "tnlTxDrops: %10u %10u\n", 9512 stats.tnl_tx_drops[0], stats.tnl_tx_drops[1]); 9513 sbuf_printf(sb, "ofldVlanDrops: %10u %10u\n", 9514 stats.ofld_vlan_drops[0], stats.ofld_vlan_drops[1]); 9515 sbuf_printf(sb, "ofldChanDrops: %10u %10u\n\n", 9516 stats.ofld_chan_drops[0], stats.ofld_chan_drops[1]); 9517 } 9518 9519 sbuf_printf(sb, "ofldNoNeigh: %u\nofldCongDefer: %u", 9520 stats.ofld_no_neigh, stats.ofld_cong_defer); 9521 9522 rc = sbuf_finish(sb); 9523 sbuf_delete(sb); 9524 9525 return (rc); 9526 } 9527 9528 static int 9529 sysctl_tnl_stats(SYSCTL_HANDLER_ARGS) 9530 { 9531 struct adapter *sc = arg1; 9532 struct sbuf *sb; 9533 int rc; 9534 struct tp_tnl_stats stats; 9535 9536 rc = sysctl_wire_old_buffer(req, 0); 9537 if (rc != 0) 9538 return(rc); 9539 9540 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9541 if (sb == NULL) 9542 return (ENOMEM); 9543 9544 mtx_lock(&sc->reg_lock); 9545 t4_tp_get_tnl_stats(sc, &stats, 1); 9546 mtx_unlock(&sc->reg_lock); 9547 9548 if (sc->chip_params->nchan > 2) { 9549 sbuf_printf(sb, " channel 0 channel 1" 9550 " channel 2 channel 3\n"); 9551 sbuf_printf(sb, "OutPkts: %10u %10u %10u %10u\n", 9552 stats.out_pkt[0], stats.out_pkt[1], 9553 stats.out_pkt[2], stats.out_pkt[3]); 9554 sbuf_printf(sb, "InPkts: %10u %10u %10u %10u", 9555 stats.in_pkt[0], stats.in_pkt[1], 9556 stats.in_pkt[2], stats.in_pkt[3]); 9557 } else { 9558 sbuf_printf(sb, " channel 0 channel 1\n"); 9559 sbuf_printf(sb, "OutPkts: %10u %10u\n", 9560 stats.out_pkt[0], stats.out_pkt[1]); 9561 sbuf_printf(sb, "InPkts: %10u %10u", 9562 stats.in_pkt[0], stats.in_pkt[1]); 9563 } 9564 9565 rc = sbuf_finish(sb); 9566 sbuf_delete(sb); 9567 9568 return (rc); 9569 } 9570 9571 static int 9572 sysctl_tp_la_mask(SYSCTL_HANDLER_ARGS) 9573 { 9574 struct adapter *sc = arg1; 9575 struct tp_params *tpp = &sc->params.tp; 9576 u_int mask; 9577 int rc; 9578 9579 mask = tpp->la_mask >> 16; 9580 rc = sysctl_handle_int(oidp, &mask, 0, req); 9581 if (rc != 0 || req->newptr == NULL) 9582 return (rc); 9583 if (mask > 0xffff) 9584 return (EINVAL); 9585 tpp->la_mask = mask << 16; 9586 t4_set_reg_field(sc, A_TP_DBG_LA_CONFIG, 0xffff0000U, tpp->la_mask); 9587 9588 return (0); 9589 } 9590 9591 struct field_desc { 9592 const char *name; 9593 u_int start; 9594 u_int width; 9595 }; 9596 9597 static void 9598 field_desc_show(struct sbuf *sb, uint64_t v, const struct field_desc *f) 9599 { 9600 char buf[32]; 9601 int line_size = 0; 9602 9603 while (f->name) { 9604 uint64_t mask = (1ULL << f->width) - 1; 9605 int len = snprintf(buf, sizeof(buf), "%s: %ju", f->name, 9606 ((uintmax_t)v >> f->start) & mask); 9607 9608 if (line_size + len >= 79) { 9609 line_size = 8; 9610 sbuf_printf(sb, "\n "); 9611 } 9612 sbuf_printf(sb, "%s ", buf); 9613 line_size += len + 1; 9614 f++; 9615 } 9616 sbuf_printf(sb, "\n"); 9617 } 9618 9619 static const struct field_desc tp_la0[] = { 9620 { "RcfOpCodeOut", 60, 4 }, 9621 { "State", 56, 4 }, 9622 { "WcfState", 52, 4 }, 9623 { "RcfOpcSrcOut", 50, 2 }, 9624 { "CRxError", 49, 1 }, 9625 { "ERxError", 48, 1 }, 9626 { "SanityFailed", 47, 1 }, 9627 { "SpuriousMsg", 46, 1 }, 9628 { "FlushInputMsg", 45, 1 }, 9629 { "FlushInputCpl", 44, 1 }, 9630 { "RssUpBit", 43, 1 }, 9631 { "RssFilterHit", 42, 1 }, 9632 { "Tid", 32, 10 }, 9633 { "InitTcb", 31, 1 }, 9634 { "LineNumber", 24, 7 }, 9635 { "Emsg", 23, 1 }, 9636 { "EdataOut", 22, 1 }, 9637 { "Cmsg", 21, 1 }, 9638 { "CdataOut", 20, 1 }, 9639 { "EreadPdu", 19, 1 }, 9640 { "CreadPdu", 18, 1 }, 9641 { "TunnelPkt", 17, 1 }, 9642 { "RcfPeerFin", 16, 1 }, 9643 { "RcfReasonOut", 12, 4 }, 9644 { "TxCchannel", 10, 2 }, 9645 { "RcfTxChannel", 8, 2 }, 9646 { "RxEchannel", 6, 2 }, 9647 { "RcfRxChannel", 5, 1 }, 9648 { "RcfDataOutSrdy", 4, 1 }, 9649 { "RxDvld", 3, 1 }, 9650 { "RxOoDvld", 2, 1 }, 9651 { "RxCongestion", 1, 1 }, 9652 { "TxCongestion", 0, 1 }, 9653 { NULL } 9654 }; 9655 9656 static const struct field_desc tp_la1[] = { 9657 { "CplCmdIn", 56, 8 }, 9658 { "CplCmdOut", 48, 8 }, 9659 { "ESynOut", 47, 1 }, 9660 { "EAckOut", 46, 1 }, 9661 { "EFinOut", 45, 1 }, 9662 { "ERstOut", 44, 1 }, 9663 { "SynIn", 43, 1 }, 9664 { "AckIn", 42, 1 }, 9665 { "FinIn", 41, 1 }, 9666 { "RstIn", 40, 1 }, 9667 { "DataIn", 39, 1 }, 9668 { "DataInVld", 38, 1 }, 9669 { "PadIn", 37, 1 }, 9670 { "RxBufEmpty", 36, 1 }, 9671 { "RxDdp", 35, 1 }, 9672 { "RxFbCongestion", 34, 1 }, 9673 { "TxFbCongestion", 33, 1 }, 9674 { "TxPktSumSrdy", 32, 1 }, 9675 { "RcfUlpType", 28, 4 }, 9676 { "Eread", 27, 1 }, 9677 { "Ebypass", 26, 1 }, 9678 { "Esave", 25, 1 }, 9679 { "Static0", 24, 1 }, 9680 { "Cread", 23, 1 }, 9681 { "Cbypass", 22, 1 }, 9682 { "Csave", 21, 1 }, 9683 { "CPktOut", 20, 1 }, 9684 { "RxPagePoolFull", 18, 2 }, 9685 { "RxLpbkPkt", 17, 1 }, 9686 { "TxLpbkPkt", 16, 1 }, 9687 { "RxVfValid", 15, 1 }, 9688 { "SynLearned", 14, 1 }, 9689 { "SetDelEntry", 13, 1 }, 9690 { "SetInvEntry", 12, 1 }, 9691 { "CpcmdDvld", 11, 1 }, 9692 { "CpcmdSave", 10, 1 }, 9693 { "RxPstructsFull", 8, 2 }, 9694 { "EpcmdDvld", 7, 1 }, 9695 { "EpcmdFlush", 6, 1 }, 9696 { "EpcmdTrimPrefix", 5, 1 }, 9697 { "EpcmdTrimPostfix", 4, 1 }, 9698 { "ERssIp4Pkt", 3, 1 }, 9699 { "ERssIp6Pkt", 2, 1 }, 9700 { "ERssTcpUdpPkt", 1, 1 }, 9701 { "ERssFceFipPkt", 0, 1 }, 9702 { NULL } 9703 }; 9704 9705 static const struct field_desc tp_la2[] = { 9706 { "CplCmdIn", 56, 8 }, 9707 { "MpsVfVld", 55, 1 }, 9708 { "MpsPf", 52, 3 }, 9709 { "MpsVf", 44, 8 }, 9710 { "SynIn", 43, 1 }, 9711 { "AckIn", 42, 1 }, 9712 { "FinIn", 41, 1 }, 9713 { "RstIn", 40, 1 }, 9714 { "DataIn", 39, 1 }, 9715 { "DataInVld", 38, 1 }, 9716 { "PadIn", 37, 1 }, 9717 { "RxBufEmpty", 36, 1 }, 9718 { "RxDdp", 35, 1 }, 9719 { "RxFbCongestion", 34, 1 }, 9720 { "TxFbCongestion", 33, 1 }, 9721 { "TxPktSumSrdy", 32, 1 }, 9722 { "RcfUlpType", 28, 4 }, 9723 { "Eread", 27, 1 }, 9724 { "Ebypass", 26, 1 }, 9725 { "Esave", 25, 1 }, 9726 { "Static0", 24, 1 }, 9727 { "Cread", 23, 1 }, 9728 { "Cbypass", 22, 1 }, 9729 { "Csave", 21, 1 }, 9730 { "CPktOut", 20, 1 }, 9731 { "RxPagePoolFull", 18, 2 }, 9732 { "RxLpbkPkt", 17, 1 }, 9733 { "TxLpbkPkt", 16, 1 }, 9734 { "RxVfValid", 15, 1 }, 9735 { "SynLearned", 14, 1 }, 9736 { "SetDelEntry", 13, 1 }, 9737 { "SetInvEntry", 12, 1 }, 9738 { "CpcmdDvld", 11, 1 }, 9739 { "CpcmdSave", 10, 1 }, 9740 { "RxPstructsFull", 8, 2 }, 9741 { "EpcmdDvld", 7, 1 }, 9742 { "EpcmdFlush", 6, 1 }, 9743 { "EpcmdTrimPrefix", 5, 1 }, 9744 { "EpcmdTrimPostfix", 4, 1 }, 9745 { "ERssIp4Pkt", 3, 1 }, 9746 { "ERssIp6Pkt", 2, 1 }, 9747 { "ERssTcpUdpPkt", 1, 1 }, 9748 { "ERssFceFipPkt", 0, 1 }, 9749 { NULL } 9750 }; 9751 9752 static void 9753 tp_la_show(struct sbuf *sb, uint64_t *p, int idx) 9754 { 9755 9756 field_desc_show(sb, *p, tp_la0); 9757 } 9758 9759 static void 9760 tp_la_show2(struct sbuf *sb, uint64_t *p, int idx) 9761 { 9762 9763 if (idx) 9764 sbuf_printf(sb, "\n"); 9765 field_desc_show(sb, p[0], tp_la0); 9766 if (idx < (TPLA_SIZE / 2 - 1) || p[1] != ~0ULL) 9767 field_desc_show(sb, p[1], tp_la0); 9768 } 9769 9770 static void 9771 tp_la_show3(struct sbuf *sb, uint64_t *p, int idx) 9772 { 9773 9774 if (idx) 9775 sbuf_printf(sb, "\n"); 9776 field_desc_show(sb, p[0], tp_la0); 9777 if (idx < (TPLA_SIZE / 2 - 1) || p[1] != ~0ULL) 9778 field_desc_show(sb, p[1], (p[0] & (1 << 17)) ? tp_la2 : tp_la1); 9779 } 9780 9781 static int 9782 sysctl_tp_la(SYSCTL_HANDLER_ARGS) 9783 { 9784 struct adapter *sc = arg1; 9785 struct sbuf *sb; 9786 uint64_t *buf, *p; 9787 int rc; 9788 u_int i, inc; 9789 void (*show_func)(struct sbuf *, uint64_t *, int); 9790 9791 rc = sysctl_wire_old_buffer(req, 0); 9792 if (rc != 0) 9793 return (rc); 9794 9795 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 9796 if (sb == NULL) 9797 return (ENOMEM); 9798 9799 buf = malloc(TPLA_SIZE * sizeof(uint64_t), M_CXGBE, M_ZERO | M_WAITOK); 9800 9801 t4_tp_read_la(sc, buf, NULL); 9802 p = buf; 9803 9804 switch (G_DBGLAMODE(t4_read_reg(sc, A_TP_DBG_LA_CONFIG))) { 9805 case 2: 9806 inc = 2; 9807 show_func = tp_la_show2; 9808 break; 9809 case 3: 9810 inc = 2; 9811 show_func = tp_la_show3; 9812 break; 9813 default: 9814 inc = 1; 9815 show_func = tp_la_show; 9816 } 9817 9818 for (i = 0; i < TPLA_SIZE / inc; i++, p += inc) 9819 (*show_func)(sb, p, i); 9820 9821 rc = sbuf_finish(sb); 9822 sbuf_delete(sb); 9823 free(buf, M_CXGBE); 9824 return (rc); 9825 } 9826 9827 static int 9828 sysctl_tx_rate(SYSCTL_HANDLER_ARGS) 9829 { 9830 struct adapter *sc = arg1; 9831 struct sbuf *sb; 9832 int rc; 9833 u64 nrate[MAX_NCHAN], orate[MAX_NCHAN]; 9834 9835 rc = sysctl_wire_old_buffer(req, 0); 9836 if (rc != 0) 9837 return (rc); 9838 9839 sb = sbuf_new_for_sysctl(NULL, NULL, 256, req); 9840 if (sb == NULL) 9841 return (ENOMEM); 9842 9843 t4_get_chan_txrate(sc, nrate, orate); 9844 9845 if (sc->chip_params->nchan > 2) { 9846 sbuf_printf(sb, " channel 0 channel 1" 9847 " channel 2 channel 3\n"); 9848 sbuf_printf(sb, "NIC B/s: %10ju %10ju %10ju %10ju\n", 9849 nrate[0], nrate[1], nrate[2], nrate[3]); 9850 sbuf_printf(sb, "Offload B/s: %10ju %10ju %10ju %10ju", 9851 orate[0], orate[1], orate[2], orate[3]); 9852 } else { 9853 sbuf_printf(sb, " channel 0 channel 1\n"); 9854 sbuf_printf(sb, "NIC B/s: %10ju %10ju\n", 9855 nrate[0], nrate[1]); 9856 sbuf_printf(sb, "Offload B/s: %10ju %10ju", 9857 orate[0], orate[1]); 9858 } 9859 9860 rc = sbuf_finish(sb); 9861 sbuf_delete(sb); 9862 9863 return (rc); 9864 } 9865 9866 static int 9867 sysctl_ulprx_la(SYSCTL_HANDLER_ARGS) 9868 { 9869 struct adapter *sc = arg1; 9870 struct sbuf *sb; 9871 uint32_t *buf, *p; 9872 int rc, i; 9873 9874 rc = sysctl_wire_old_buffer(req, 0); 9875 if (rc != 0) 9876 return (rc); 9877 9878 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 9879 if (sb == NULL) 9880 return (ENOMEM); 9881 9882 buf = malloc(ULPRX_LA_SIZE * 8 * sizeof(uint32_t), M_CXGBE, 9883 M_ZERO | M_WAITOK); 9884 9885 t4_ulprx_read_la(sc, buf); 9886 p = buf; 9887 9888 sbuf_printf(sb, " Pcmd Type Message" 9889 " Data"); 9890 for (i = 0; i < ULPRX_LA_SIZE; i++, p += 8) { 9891 sbuf_printf(sb, "\n%08x%08x %4x %08x %08x%08x%08x%08x", 9892 p[1], p[0], p[2], p[3], p[7], p[6], p[5], p[4]); 9893 } 9894 9895 rc = sbuf_finish(sb); 9896 sbuf_delete(sb); 9897 free(buf, M_CXGBE); 9898 return (rc); 9899 } 9900 9901 static int 9902 sysctl_wcwr_stats(SYSCTL_HANDLER_ARGS) 9903 { 9904 struct adapter *sc = arg1; 9905 struct sbuf *sb; 9906 int rc, v; 9907 9908 MPASS(chip_id(sc) >= CHELSIO_T5); 9909 9910 rc = sysctl_wire_old_buffer(req, 0); 9911 if (rc != 0) 9912 return (rc); 9913 9914 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 9915 if (sb == NULL) 9916 return (ENOMEM); 9917 9918 v = t4_read_reg(sc, A_SGE_STAT_CFG); 9919 if (G_STATSOURCE_T5(v) == 7) { 9920 int mode; 9921 9922 mode = is_t5(sc) ? G_STATMODE(v) : G_T6_STATMODE(v); 9923 if (mode == 0) { 9924 sbuf_printf(sb, "total %d, incomplete %d", 9925 t4_read_reg(sc, A_SGE_STAT_TOTAL), 9926 t4_read_reg(sc, A_SGE_STAT_MATCH)); 9927 } else if (mode == 1) { 9928 sbuf_printf(sb, "total %d, data overflow %d", 9929 t4_read_reg(sc, A_SGE_STAT_TOTAL), 9930 t4_read_reg(sc, A_SGE_STAT_MATCH)); 9931 } else { 9932 sbuf_printf(sb, "unknown mode %d", mode); 9933 } 9934 } 9935 rc = sbuf_finish(sb); 9936 sbuf_delete(sb); 9937 9938 return (rc); 9939 } 9940 9941 static int 9942 sysctl_cpus(SYSCTL_HANDLER_ARGS) 9943 { 9944 struct adapter *sc = arg1; 9945 enum cpu_sets op = arg2; 9946 cpuset_t cpuset; 9947 struct sbuf *sb; 9948 int i, rc; 9949 9950 MPASS(op == LOCAL_CPUS || op == INTR_CPUS); 9951 9952 CPU_ZERO(&cpuset); 9953 rc = bus_get_cpus(sc->dev, op, sizeof(cpuset), &cpuset); 9954 if (rc != 0) 9955 return (rc); 9956 9957 rc = sysctl_wire_old_buffer(req, 0); 9958 if (rc != 0) 9959 return (rc); 9960 9961 sb = sbuf_new_for_sysctl(NULL, NULL, 4096, req); 9962 if (sb == NULL) 9963 return (ENOMEM); 9964 9965 CPU_FOREACH(i) 9966 sbuf_printf(sb, "%d ", i); 9967 rc = sbuf_finish(sb); 9968 sbuf_delete(sb); 9969 9970 return (rc); 9971 } 9972 9973 #ifdef TCP_OFFLOAD 9974 static int 9975 sysctl_tls(SYSCTL_HANDLER_ARGS) 9976 { 9977 struct adapter *sc = arg1; 9978 int i, j, v, rc; 9979 struct vi_info *vi; 9980 9981 v = sc->tt.tls; 9982 rc = sysctl_handle_int(oidp, &v, 0, req); 9983 if (rc != 0 || req->newptr == NULL) 9984 return (rc); 9985 9986 if (v != 0 && !(sc->cryptocaps & FW_CAPS_CONFIG_TLSKEYS)) 9987 return (ENOTSUP); 9988 9989 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4stls"); 9990 if (rc) 9991 return (rc); 9992 sc->tt.tls = !!v; 9993 for_each_port(sc, i) { 9994 for_each_vi(sc->port[i], j, vi) { 9995 if (vi->flags & VI_INIT_DONE) 9996 t4_update_fl_bufsize(vi->ifp); 9997 } 9998 } 9999 end_synchronized_op(sc, 0); 10000 10001 return (0); 10002 10003 } 10004 10005 static int 10006 sysctl_tls_rx_ports(SYSCTL_HANDLER_ARGS) 10007 { 10008 struct adapter *sc = arg1; 10009 int *old_ports, *new_ports; 10010 int i, new_count, rc; 10011 10012 if (req->newptr == NULL && req->oldptr == NULL) 10013 return (SYSCTL_OUT(req, NULL, imax(sc->tt.num_tls_rx_ports, 1) * 10014 sizeof(sc->tt.tls_rx_ports[0]))); 10015 10016 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4tlsrx"); 10017 if (rc) 10018 return (rc); 10019 10020 if (sc->tt.num_tls_rx_ports == 0) { 10021 i = -1; 10022 rc = SYSCTL_OUT(req, &i, sizeof(i)); 10023 } else 10024 rc = SYSCTL_OUT(req, sc->tt.tls_rx_ports, 10025 sc->tt.num_tls_rx_ports * sizeof(sc->tt.tls_rx_ports[0])); 10026 if (rc == 0 && req->newptr != NULL) { 10027 new_count = req->newlen / sizeof(new_ports[0]); 10028 new_ports = malloc(new_count * sizeof(new_ports[0]), M_CXGBE, 10029 M_WAITOK); 10030 rc = SYSCTL_IN(req, new_ports, new_count * 10031 sizeof(new_ports[0])); 10032 if (rc) 10033 goto err; 10034 10035 /* Allow setting to a single '-1' to clear the list. */ 10036 if (new_count == 1 && new_ports[0] == -1) { 10037 ADAPTER_LOCK(sc); 10038 old_ports = sc->tt.tls_rx_ports; 10039 sc->tt.tls_rx_ports = NULL; 10040 sc->tt.num_tls_rx_ports = 0; 10041 ADAPTER_UNLOCK(sc); 10042 free(old_ports, M_CXGBE); 10043 } else { 10044 for (i = 0; i < new_count; i++) { 10045 if (new_ports[i] < 1 || 10046 new_ports[i] > IPPORT_MAX) { 10047 rc = EINVAL; 10048 goto err; 10049 } 10050 } 10051 10052 ADAPTER_LOCK(sc); 10053 old_ports = sc->tt.tls_rx_ports; 10054 sc->tt.tls_rx_ports = new_ports; 10055 sc->tt.num_tls_rx_ports = new_count; 10056 ADAPTER_UNLOCK(sc); 10057 free(old_ports, M_CXGBE); 10058 new_ports = NULL; 10059 } 10060 err: 10061 free(new_ports, M_CXGBE); 10062 } 10063 end_synchronized_op(sc, 0); 10064 return (rc); 10065 } 10066 10067 static int 10068 sysctl_tls_rx_timeout(SYSCTL_HANDLER_ARGS) 10069 { 10070 struct adapter *sc = arg1; 10071 int v, rc; 10072 10073 v = sc->tt.tls_rx_timeout; 10074 rc = sysctl_handle_int(oidp, &v, 0, req); 10075 if (rc != 0 || req->newptr == NULL) 10076 return (rc); 10077 10078 if (v < 0) 10079 return (EINVAL); 10080 10081 if (v != 0 && !(sc->cryptocaps & FW_CAPS_CONFIG_TLSKEYS)) 10082 return (ENOTSUP); 10083 10084 sc->tt.tls_rx_timeout = v; 10085 10086 return (0); 10087 10088 } 10089 10090 static void 10091 unit_conv(char *buf, size_t len, u_int val, u_int factor) 10092 { 10093 u_int rem = val % factor; 10094 10095 if (rem == 0) 10096 snprintf(buf, len, "%u", val / factor); 10097 else { 10098 while (rem % 10 == 0) 10099 rem /= 10; 10100 snprintf(buf, len, "%u.%u", val / factor, rem); 10101 } 10102 } 10103 10104 static int 10105 sysctl_tp_tick(SYSCTL_HANDLER_ARGS) 10106 { 10107 struct adapter *sc = arg1; 10108 char buf[16]; 10109 u_int res, re; 10110 u_int cclk_ps = 1000000000 / sc->params.vpd.cclk; 10111 10112 res = t4_read_reg(sc, A_TP_TIMER_RESOLUTION); 10113 switch (arg2) { 10114 case 0: 10115 /* timer_tick */ 10116 re = G_TIMERRESOLUTION(res); 10117 break; 10118 case 1: 10119 /* TCP timestamp tick */ 10120 re = G_TIMESTAMPRESOLUTION(res); 10121 break; 10122 case 2: 10123 /* DACK tick */ 10124 re = G_DELAYEDACKRESOLUTION(res); 10125 break; 10126 default: 10127 return (EDOOFUS); 10128 } 10129 10130 unit_conv(buf, sizeof(buf), (cclk_ps << re), 1000000); 10131 10132 return (sysctl_handle_string(oidp, buf, sizeof(buf), req)); 10133 } 10134 10135 static int 10136 sysctl_tp_dack_timer(SYSCTL_HANDLER_ARGS) 10137 { 10138 struct adapter *sc = arg1; 10139 u_int res, dack_re, v; 10140 u_int cclk_ps = 1000000000 / sc->params.vpd.cclk; 10141 10142 res = t4_read_reg(sc, A_TP_TIMER_RESOLUTION); 10143 dack_re = G_DELAYEDACKRESOLUTION(res); 10144 v = ((cclk_ps << dack_re) / 1000000) * t4_read_reg(sc, A_TP_DACK_TIMER); 10145 10146 return (sysctl_handle_int(oidp, &v, 0, req)); 10147 } 10148 10149 static int 10150 sysctl_tp_timer(SYSCTL_HANDLER_ARGS) 10151 { 10152 struct adapter *sc = arg1; 10153 int reg = arg2; 10154 u_int tre; 10155 u_long tp_tick_us, v; 10156 u_int cclk_ps = 1000000000 / sc->params.vpd.cclk; 10157 10158 MPASS(reg == A_TP_RXT_MIN || reg == A_TP_RXT_MAX || 10159 reg == A_TP_PERS_MIN || reg == A_TP_PERS_MAX || 10160 reg == A_TP_KEEP_IDLE || reg == A_TP_KEEP_INTVL || 10161 reg == A_TP_INIT_SRTT || reg == A_TP_FINWAIT2_TIMER); 10162 10163 tre = G_TIMERRESOLUTION(t4_read_reg(sc, A_TP_TIMER_RESOLUTION)); 10164 tp_tick_us = (cclk_ps << tre) / 1000000; 10165 10166 if (reg == A_TP_INIT_SRTT) 10167 v = tp_tick_us * G_INITSRTT(t4_read_reg(sc, reg)); 10168 else 10169 v = tp_tick_us * t4_read_reg(sc, reg); 10170 10171 return (sysctl_handle_long(oidp, &v, 0, req)); 10172 } 10173 10174 /* 10175 * All fields in TP_SHIFT_CNT are 4b and the starting location of the field is 10176 * passed to this function. 10177 */ 10178 static int 10179 sysctl_tp_shift_cnt(SYSCTL_HANDLER_ARGS) 10180 { 10181 struct adapter *sc = arg1; 10182 int idx = arg2; 10183 u_int v; 10184 10185 MPASS(idx >= 0 && idx <= 24); 10186 10187 v = (t4_read_reg(sc, A_TP_SHIFT_CNT) >> idx) & 0xf; 10188 10189 return (sysctl_handle_int(oidp, &v, 0, req)); 10190 } 10191 10192 static int 10193 sysctl_tp_backoff(SYSCTL_HANDLER_ARGS) 10194 { 10195 struct adapter *sc = arg1; 10196 int idx = arg2; 10197 u_int shift, v, r; 10198 10199 MPASS(idx >= 0 && idx < 16); 10200 10201 r = A_TP_TCP_BACKOFF_REG0 + (idx & ~3); 10202 shift = (idx & 3) << 3; 10203 v = (t4_read_reg(sc, r) >> shift) & M_TIMERBACKOFFINDEX0; 10204 10205 return (sysctl_handle_int(oidp, &v, 0, req)); 10206 } 10207 10208 static int 10209 sysctl_holdoff_tmr_idx_ofld(SYSCTL_HANDLER_ARGS) 10210 { 10211 struct vi_info *vi = arg1; 10212 struct adapter *sc = vi->adapter; 10213 int idx, rc, i; 10214 struct sge_ofld_rxq *ofld_rxq; 10215 uint8_t v; 10216 10217 idx = vi->ofld_tmr_idx; 10218 10219 rc = sysctl_handle_int(oidp, &idx, 0, req); 10220 if (rc != 0 || req->newptr == NULL) 10221 return (rc); 10222 10223 if (idx < 0 || idx >= SGE_NTIMERS) 10224 return (EINVAL); 10225 10226 rc = begin_synchronized_op(sc, vi, HOLD_LOCK | SLEEP_OK | INTR_OK, 10227 "t4otmr"); 10228 if (rc) 10229 return (rc); 10230 10231 v = V_QINTR_TIMER_IDX(idx) | V_QINTR_CNT_EN(vi->ofld_pktc_idx != -1); 10232 for_each_ofld_rxq(vi, i, ofld_rxq) { 10233 #ifdef atomic_store_rel_8 10234 atomic_store_rel_8(&ofld_rxq->iq.intr_params, v); 10235 #else 10236 ofld_rxq->iq.intr_params = v; 10237 #endif 10238 } 10239 vi->ofld_tmr_idx = idx; 10240 10241 end_synchronized_op(sc, LOCK_HELD); 10242 return (0); 10243 } 10244 10245 static int 10246 sysctl_holdoff_pktc_idx_ofld(SYSCTL_HANDLER_ARGS) 10247 { 10248 struct vi_info *vi = arg1; 10249 struct adapter *sc = vi->adapter; 10250 int idx, rc; 10251 10252 idx = vi->ofld_pktc_idx; 10253 10254 rc = sysctl_handle_int(oidp, &idx, 0, req); 10255 if (rc != 0 || req->newptr == NULL) 10256 return (rc); 10257 10258 if (idx < -1 || idx >= SGE_NCOUNTERS) 10259 return (EINVAL); 10260 10261 rc = begin_synchronized_op(sc, vi, HOLD_LOCK | SLEEP_OK | INTR_OK, 10262 "t4opktc"); 10263 if (rc) 10264 return (rc); 10265 10266 if (vi->flags & VI_INIT_DONE) 10267 rc = EBUSY; /* cannot be changed once the queues are created */ 10268 else 10269 vi->ofld_pktc_idx = idx; 10270 10271 end_synchronized_op(sc, LOCK_HELD); 10272 return (rc); 10273 } 10274 #endif 10275 10276 static int 10277 get_sge_context(struct adapter *sc, struct t4_sge_context *cntxt) 10278 { 10279 int rc; 10280 10281 if (cntxt->cid > M_CTXTQID) 10282 return (EINVAL); 10283 10284 if (cntxt->mem_id != CTXT_EGRESS && cntxt->mem_id != CTXT_INGRESS && 10285 cntxt->mem_id != CTXT_FLM && cntxt->mem_id != CTXT_CNM) 10286 return (EINVAL); 10287 10288 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4ctxt"); 10289 if (rc) 10290 return (rc); 10291 10292 if (sc->flags & FW_OK) { 10293 rc = -t4_sge_ctxt_rd(sc, sc->mbox, cntxt->cid, cntxt->mem_id, 10294 &cntxt->data[0]); 10295 if (rc == 0) 10296 goto done; 10297 } 10298 10299 /* 10300 * Read via firmware failed or wasn't even attempted. Read directly via 10301 * the backdoor. 10302 */ 10303 rc = -t4_sge_ctxt_rd_bd(sc, cntxt->cid, cntxt->mem_id, &cntxt->data[0]); 10304 done: 10305 end_synchronized_op(sc, 0); 10306 return (rc); 10307 } 10308 10309 static int 10310 load_fw(struct adapter *sc, struct t4_data *fw) 10311 { 10312 int rc; 10313 uint8_t *fw_data; 10314 10315 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4ldfw"); 10316 if (rc) 10317 return (rc); 10318 10319 /* 10320 * The firmware, with the sole exception of the memory parity error 10321 * handler, runs from memory and not flash. It is almost always safe to 10322 * install a new firmware on a running system. Just set bit 1 in 10323 * hw.cxgbe.dflags or dev.<nexus>.<n>.dflags first. 10324 */ 10325 if (sc->flags & FULL_INIT_DONE && 10326 (sc->debug_flags & DF_LOAD_FW_ANYTIME) == 0) { 10327 rc = EBUSY; 10328 goto done; 10329 } 10330 10331 fw_data = malloc(fw->len, M_CXGBE, M_WAITOK); 10332 10333 rc = copyin(fw->data, fw_data, fw->len); 10334 if (rc == 0) 10335 rc = -t4_load_fw(sc, fw_data, fw->len); 10336 10337 free(fw_data, M_CXGBE); 10338 done: 10339 end_synchronized_op(sc, 0); 10340 return (rc); 10341 } 10342 10343 static int 10344 load_cfg(struct adapter *sc, struct t4_data *cfg) 10345 { 10346 int rc; 10347 uint8_t *cfg_data = NULL; 10348 10349 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4ldcf"); 10350 if (rc) 10351 return (rc); 10352 10353 if (cfg->len == 0) { 10354 /* clear */ 10355 rc = -t4_load_cfg(sc, NULL, 0); 10356 goto done; 10357 } 10358 10359 cfg_data = malloc(cfg->len, M_CXGBE, M_WAITOK); 10360 10361 rc = copyin(cfg->data, cfg_data, cfg->len); 10362 if (rc == 0) 10363 rc = -t4_load_cfg(sc, cfg_data, cfg->len); 10364 10365 free(cfg_data, M_CXGBE); 10366 done: 10367 end_synchronized_op(sc, 0); 10368 return (rc); 10369 } 10370 10371 static int 10372 load_boot(struct adapter *sc, struct t4_bootrom *br) 10373 { 10374 int rc; 10375 uint8_t *br_data = NULL; 10376 u_int offset; 10377 10378 if (br->len > 1024 * 1024) 10379 return (EFBIG); 10380 10381 if (br->pf_offset == 0) { 10382 /* pfidx */ 10383 if (br->pfidx_addr > 7) 10384 return (EINVAL); 10385 offset = G_OFFSET(t4_read_reg(sc, PF_REG(br->pfidx_addr, 10386 A_PCIE_PF_EXPROM_OFST))); 10387 } else if (br->pf_offset == 1) { 10388 /* offset */ 10389 offset = G_OFFSET(br->pfidx_addr); 10390 } else { 10391 return (EINVAL); 10392 } 10393 10394 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4ldbr"); 10395 if (rc) 10396 return (rc); 10397 10398 if (br->len == 0) { 10399 /* clear */ 10400 rc = -t4_load_boot(sc, NULL, offset, 0); 10401 goto done; 10402 } 10403 10404 br_data = malloc(br->len, M_CXGBE, M_WAITOK); 10405 10406 rc = copyin(br->data, br_data, br->len); 10407 if (rc == 0) 10408 rc = -t4_load_boot(sc, br_data, offset, br->len); 10409 10410 free(br_data, M_CXGBE); 10411 done: 10412 end_synchronized_op(sc, 0); 10413 return (rc); 10414 } 10415 10416 static int 10417 load_bootcfg(struct adapter *sc, struct t4_data *bc) 10418 { 10419 int rc; 10420 uint8_t *bc_data = NULL; 10421 10422 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4ldcf"); 10423 if (rc) 10424 return (rc); 10425 10426 if (bc->len == 0) { 10427 /* clear */ 10428 rc = -t4_load_bootcfg(sc, NULL, 0); 10429 goto done; 10430 } 10431 10432 bc_data = malloc(bc->len, M_CXGBE, M_WAITOK); 10433 10434 rc = copyin(bc->data, bc_data, bc->len); 10435 if (rc == 0) 10436 rc = -t4_load_bootcfg(sc, bc_data, bc->len); 10437 10438 free(bc_data, M_CXGBE); 10439 done: 10440 end_synchronized_op(sc, 0); 10441 return (rc); 10442 } 10443 10444 static int 10445 cudbg_dump(struct adapter *sc, struct t4_cudbg_dump *dump) 10446 { 10447 int rc; 10448 struct cudbg_init *cudbg; 10449 void *handle, *buf; 10450 10451 /* buf is large, don't block if no memory is available */ 10452 buf = malloc(dump->len, M_CXGBE, M_NOWAIT | M_ZERO); 10453 if (buf == NULL) 10454 return (ENOMEM); 10455 10456 handle = cudbg_alloc_handle(); 10457 if (handle == NULL) { 10458 rc = ENOMEM; 10459 goto done; 10460 } 10461 10462 cudbg = cudbg_get_init(handle); 10463 cudbg->adap = sc; 10464 cudbg->print = (cudbg_print_cb)printf; 10465 10466 #ifndef notyet 10467 device_printf(sc->dev, "%s: wr_flash %u, len %u, data %p.\n", 10468 __func__, dump->wr_flash, dump->len, dump->data); 10469 #endif 10470 10471 if (dump->wr_flash) 10472 cudbg->use_flash = 1; 10473 MPASS(sizeof(cudbg->dbg_bitmap) == sizeof(dump->bitmap)); 10474 memcpy(cudbg->dbg_bitmap, dump->bitmap, sizeof(cudbg->dbg_bitmap)); 10475 10476 rc = cudbg_collect(handle, buf, &dump->len); 10477 if (rc != 0) 10478 goto done; 10479 10480 rc = copyout(buf, dump->data, dump->len); 10481 done: 10482 cudbg_free_handle(handle); 10483 free(buf, M_CXGBE); 10484 return (rc); 10485 } 10486 10487 static void 10488 free_offload_policy(struct t4_offload_policy *op) 10489 { 10490 struct offload_rule *r; 10491 int i; 10492 10493 if (op == NULL) 10494 return; 10495 10496 r = &op->rule[0]; 10497 for (i = 0; i < op->nrules; i++, r++) { 10498 free(r->bpf_prog.bf_insns, M_CXGBE); 10499 } 10500 free(op->rule, M_CXGBE); 10501 free(op, M_CXGBE); 10502 } 10503 10504 static int 10505 set_offload_policy(struct adapter *sc, struct t4_offload_policy *uop) 10506 { 10507 int i, rc, len; 10508 struct t4_offload_policy *op, *old; 10509 struct bpf_program *bf; 10510 const struct offload_settings *s; 10511 struct offload_rule *r; 10512 void *u; 10513 10514 if (!is_offload(sc)) 10515 return (ENODEV); 10516 10517 if (uop->nrules == 0) { 10518 /* Delete installed policies. */ 10519 op = NULL; 10520 goto set_policy; 10521 } else if (uop->nrules > 256) { /* arbitrary */ 10522 return (E2BIG); 10523 } 10524 10525 /* Copy userspace offload policy to kernel */ 10526 op = malloc(sizeof(*op), M_CXGBE, M_ZERO | M_WAITOK); 10527 op->nrules = uop->nrules; 10528 len = op->nrules * sizeof(struct offload_rule); 10529 op->rule = malloc(len, M_CXGBE, M_ZERO | M_WAITOK); 10530 rc = copyin(uop->rule, op->rule, len); 10531 if (rc) { 10532 free(op->rule, M_CXGBE); 10533 free(op, M_CXGBE); 10534 return (rc); 10535 } 10536 10537 r = &op->rule[0]; 10538 for (i = 0; i < op->nrules; i++, r++) { 10539 10540 /* Validate open_type */ 10541 if (r->open_type != OPEN_TYPE_LISTEN && 10542 r->open_type != OPEN_TYPE_ACTIVE && 10543 r->open_type != OPEN_TYPE_PASSIVE && 10544 r->open_type != OPEN_TYPE_DONTCARE) { 10545 error: 10546 /* 10547 * Rules 0 to i have malloc'd filters that need to be 10548 * freed. Rules i+1 to nrules have userspace pointers 10549 * and should be left alone. 10550 */ 10551 op->nrules = i; 10552 free_offload_policy(op); 10553 return (rc); 10554 } 10555 10556 /* Validate settings */ 10557 s = &r->settings; 10558 if ((s->offload != 0 && s->offload != 1) || 10559 s->cong_algo < -1 || s->cong_algo > CONG_ALG_HIGHSPEED || 10560 s->sched_class < -1 || 10561 s->sched_class >= sc->chip_params->nsched_cls) { 10562 rc = EINVAL; 10563 goto error; 10564 } 10565 10566 bf = &r->bpf_prog; 10567 u = bf->bf_insns; /* userspace ptr */ 10568 bf->bf_insns = NULL; 10569 if (bf->bf_len == 0) { 10570 /* legal, matches everything */ 10571 continue; 10572 } 10573 len = bf->bf_len * sizeof(*bf->bf_insns); 10574 bf->bf_insns = malloc(len, M_CXGBE, M_ZERO | M_WAITOK); 10575 rc = copyin(u, bf->bf_insns, len); 10576 if (rc != 0) 10577 goto error; 10578 10579 if (!bpf_validate(bf->bf_insns, bf->bf_len)) { 10580 rc = EINVAL; 10581 goto error; 10582 } 10583 } 10584 set_policy: 10585 rw_wlock(&sc->policy_lock); 10586 old = sc->policy; 10587 sc->policy = op; 10588 rw_wunlock(&sc->policy_lock); 10589 free_offload_policy(old); 10590 10591 return (0); 10592 } 10593 10594 #define MAX_READ_BUF_SIZE (128 * 1024) 10595 static int 10596 read_card_mem(struct adapter *sc, int win, struct t4_mem_range *mr) 10597 { 10598 uint32_t addr, remaining, n; 10599 uint32_t *buf; 10600 int rc; 10601 uint8_t *dst; 10602 10603 rc = validate_mem_range(sc, mr->addr, mr->len); 10604 if (rc != 0) 10605 return (rc); 10606 10607 buf = malloc(min(mr->len, MAX_READ_BUF_SIZE), M_CXGBE, M_WAITOK); 10608 addr = mr->addr; 10609 remaining = mr->len; 10610 dst = (void *)mr->data; 10611 10612 while (remaining) { 10613 n = min(remaining, MAX_READ_BUF_SIZE); 10614 read_via_memwin(sc, 2, addr, buf, n); 10615 10616 rc = copyout(buf, dst, n); 10617 if (rc != 0) 10618 break; 10619 10620 dst += n; 10621 remaining -= n; 10622 addr += n; 10623 } 10624 10625 free(buf, M_CXGBE); 10626 return (rc); 10627 } 10628 #undef MAX_READ_BUF_SIZE 10629 10630 static int 10631 read_i2c(struct adapter *sc, struct t4_i2c_data *i2cd) 10632 { 10633 int rc; 10634 10635 if (i2cd->len == 0 || i2cd->port_id >= sc->params.nports) 10636 return (EINVAL); 10637 10638 if (i2cd->len > sizeof(i2cd->data)) 10639 return (EFBIG); 10640 10641 rc = begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4i2crd"); 10642 if (rc) 10643 return (rc); 10644 rc = -t4_i2c_rd(sc, sc->mbox, i2cd->port_id, i2cd->dev_addr, 10645 i2cd->offset, i2cd->len, &i2cd->data[0]); 10646 end_synchronized_op(sc, 0); 10647 10648 return (rc); 10649 } 10650 10651 static int 10652 clear_stats(struct adapter *sc, u_int port_id) 10653 { 10654 int i, v, chan_map; 10655 struct port_info *pi; 10656 struct vi_info *vi; 10657 struct sge_rxq *rxq; 10658 struct sge_txq *txq; 10659 struct sge_wrq *wrq; 10660 #ifdef TCP_OFFLOAD 10661 struct sge_ofld_rxq *ofld_rxq; 10662 #endif 10663 10664 if (port_id >= sc->params.nports) 10665 return (EINVAL); 10666 pi = sc->port[port_id]; 10667 if (pi == NULL) 10668 return (EIO); 10669 10670 /* MAC stats */ 10671 t4_clr_port_stats(sc, pi->tx_chan); 10672 if (is_t6(sc)) { 10673 if (pi->fcs_reg != -1) 10674 pi->fcs_base = t4_read_reg64(sc, pi->fcs_reg); 10675 else 10676 pi->stats.rx_fcs_err = 0; 10677 } 10678 pi->tx_parse_error = 0; 10679 pi->tnl_cong_drops = 0; 10680 mtx_lock(&sc->reg_lock); 10681 for_each_vi(pi, v, vi) { 10682 if (vi->flags & VI_INIT_DONE) 10683 t4_clr_vi_stats(sc, vi->vin); 10684 } 10685 chan_map = pi->rx_e_chan_map; 10686 v = 0; /* reuse */ 10687 while (chan_map) { 10688 i = ffs(chan_map) - 1; 10689 t4_write_indirect(sc, A_TP_MIB_INDEX, A_TP_MIB_DATA, &v, 10690 1, A_TP_MIB_TNL_CNG_DROP_0 + i); 10691 chan_map &= ~(1 << i); 10692 } 10693 mtx_unlock(&sc->reg_lock); 10694 10695 /* 10696 * Since this command accepts a port, clear stats for 10697 * all VIs on this port. 10698 */ 10699 for_each_vi(pi, v, vi) { 10700 if (vi->flags & VI_INIT_DONE) { 10701 10702 for_each_rxq(vi, i, rxq) { 10703 #if defined(INET) || defined(INET6) 10704 rxq->lro.lro_queued = 0; 10705 rxq->lro.lro_flushed = 0; 10706 #endif 10707 rxq->rxcsum = 0; 10708 rxq->vlan_extraction = 0; 10709 rxq->vxlan_rxcsum = 0; 10710 10711 rxq->fl.cl_allocated = 0; 10712 rxq->fl.cl_recycled = 0; 10713 rxq->fl.cl_fast_recycled = 0; 10714 } 10715 10716 for_each_txq(vi, i, txq) { 10717 txq->txcsum = 0; 10718 txq->tso_wrs = 0; 10719 txq->vlan_insertion = 0; 10720 txq->imm_wrs = 0; 10721 txq->sgl_wrs = 0; 10722 txq->txpkt_wrs = 0; 10723 txq->txpkts0_wrs = 0; 10724 txq->txpkts1_wrs = 0; 10725 txq->txpkts0_pkts = 0; 10726 txq->txpkts1_pkts = 0; 10727 txq->txpkts_flush = 0; 10728 txq->raw_wrs = 0; 10729 txq->vxlan_tso_wrs = 0; 10730 txq->vxlan_txcsum = 0; 10731 txq->kern_tls_records = 0; 10732 txq->kern_tls_short = 0; 10733 txq->kern_tls_partial = 0; 10734 txq->kern_tls_full = 0; 10735 txq->kern_tls_octets = 0; 10736 txq->kern_tls_waste = 0; 10737 txq->kern_tls_options = 0; 10738 txq->kern_tls_header = 0; 10739 txq->kern_tls_fin = 0; 10740 txq->kern_tls_fin_short = 0; 10741 txq->kern_tls_cbc = 0; 10742 txq->kern_tls_gcm = 0; 10743 mp_ring_reset_stats(txq->r); 10744 } 10745 10746 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 10747 for_each_ofld_txq(vi, i, wrq) { 10748 wrq->tx_wrs_direct = 0; 10749 wrq->tx_wrs_copied = 0; 10750 } 10751 #endif 10752 #ifdef TCP_OFFLOAD 10753 for_each_ofld_rxq(vi, i, ofld_rxq) { 10754 ofld_rxq->fl.cl_allocated = 0; 10755 ofld_rxq->fl.cl_recycled = 0; 10756 ofld_rxq->fl.cl_fast_recycled = 0; 10757 } 10758 #endif 10759 10760 if (IS_MAIN_VI(vi)) { 10761 wrq = &sc->sge.ctrlq[pi->port_id]; 10762 wrq->tx_wrs_direct = 0; 10763 wrq->tx_wrs_copied = 0; 10764 } 10765 } 10766 } 10767 10768 return (0); 10769 } 10770 10771 int 10772 t4_os_find_pci_capability(struct adapter *sc, int cap) 10773 { 10774 int i; 10775 10776 return (pci_find_cap(sc->dev, cap, &i) == 0 ? i : 0); 10777 } 10778 10779 int 10780 t4_os_pci_save_state(struct adapter *sc) 10781 { 10782 device_t dev; 10783 struct pci_devinfo *dinfo; 10784 10785 dev = sc->dev; 10786 dinfo = device_get_ivars(dev); 10787 10788 pci_cfg_save(dev, dinfo, 0); 10789 return (0); 10790 } 10791 10792 int 10793 t4_os_pci_restore_state(struct adapter *sc) 10794 { 10795 device_t dev; 10796 struct pci_devinfo *dinfo; 10797 10798 dev = sc->dev; 10799 dinfo = device_get_ivars(dev); 10800 10801 pci_cfg_restore(dev, dinfo); 10802 return (0); 10803 } 10804 10805 void 10806 t4_os_portmod_changed(struct port_info *pi) 10807 { 10808 struct adapter *sc = pi->adapter; 10809 struct vi_info *vi; 10810 struct ifnet *ifp; 10811 static const char *mod_str[] = { 10812 NULL, "LR", "SR", "ER", "TWINAX", "active TWINAX", "LRM" 10813 }; 10814 10815 KASSERT((pi->flags & FIXED_IFMEDIA) == 0, 10816 ("%s: port_type %u", __func__, pi->port_type)); 10817 10818 vi = &pi->vi[0]; 10819 if (begin_synchronized_op(sc, vi, HOLD_LOCK, "t4mod") == 0) { 10820 PORT_LOCK(pi); 10821 build_medialist(pi); 10822 if (pi->mod_type != FW_PORT_MOD_TYPE_NONE) { 10823 fixup_link_config(pi); 10824 apply_link_config(pi); 10825 } 10826 PORT_UNLOCK(pi); 10827 end_synchronized_op(sc, LOCK_HELD); 10828 } 10829 10830 ifp = vi->ifp; 10831 if (pi->mod_type == FW_PORT_MOD_TYPE_NONE) 10832 if_printf(ifp, "transceiver unplugged.\n"); 10833 else if (pi->mod_type == FW_PORT_MOD_TYPE_UNKNOWN) 10834 if_printf(ifp, "unknown transceiver inserted.\n"); 10835 else if (pi->mod_type == FW_PORT_MOD_TYPE_NOTSUPPORTED) 10836 if_printf(ifp, "unsupported transceiver inserted.\n"); 10837 else if (pi->mod_type > 0 && pi->mod_type < nitems(mod_str)) { 10838 if_printf(ifp, "%dGbps %s transceiver inserted.\n", 10839 port_top_speed(pi), mod_str[pi->mod_type]); 10840 } else { 10841 if_printf(ifp, "transceiver (type %d) inserted.\n", 10842 pi->mod_type); 10843 } 10844 } 10845 10846 void 10847 t4_os_link_changed(struct port_info *pi) 10848 { 10849 struct vi_info *vi; 10850 struct ifnet *ifp; 10851 struct link_config *lc = &pi->link_cfg; 10852 struct adapter *sc = pi->adapter; 10853 int v; 10854 10855 PORT_LOCK_ASSERT_OWNED(pi); 10856 10857 if (is_t6(sc)) { 10858 if (lc->link_ok) { 10859 if (lc->speed > 25000 || 10860 (lc->speed == 25000 && lc->fec == FEC_RS)) { 10861 pi->fcs_reg = T5_PORT_REG(pi->tx_chan, 10862 A_MAC_PORT_AFRAMECHECKSEQUENCEERRORS); 10863 } else { 10864 pi->fcs_reg = T5_PORT_REG(pi->tx_chan, 10865 A_MAC_PORT_MTIP_1G10G_RX_CRCERRORS); 10866 } 10867 pi->fcs_base = t4_read_reg64(sc, pi->fcs_reg); 10868 pi->stats.rx_fcs_err = 0; 10869 } else { 10870 pi->fcs_reg = -1; 10871 } 10872 } else { 10873 MPASS(pi->fcs_reg != -1); 10874 MPASS(pi->fcs_base == 0); 10875 } 10876 10877 for_each_vi(pi, v, vi) { 10878 ifp = vi->ifp; 10879 if (ifp == NULL) 10880 continue; 10881 10882 if (lc->link_ok) { 10883 ifp->if_baudrate = IF_Mbps(lc->speed); 10884 if_link_state_change(ifp, LINK_STATE_UP); 10885 } else { 10886 if_link_state_change(ifp, LINK_STATE_DOWN); 10887 } 10888 } 10889 } 10890 10891 void 10892 t4_iterate(void (*func)(struct adapter *, void *), void *arg) 10893 { 10894 struct adapter *sc; 10895 10896 sx_slock(&t4_list_lock); 10897 SLIST_FOREACH(sc, &t4_list, link) { 10898 /* 10899 * func should not make any assumptions about what state sc is 10900 * in - the only guarantee is that sc->sc_lock is a valid lock. 10901 */ 10902 func(sc, arg); 10903 } 10904 sx_sunlock(&t4_list_lock); 10905 } 10906 10907 static int 10908 t4_ioctl(struct cdev *dev, unsigned long cmd, caddr_t data, int fflag, 10909 struct thread *td) 10910 { 10911 int rc; 10912 struct adapter *sc = dev->si_drv1; 10913 10914 rc = priv_check(td, PRIV_DRIVER); 10915 if (rc != 0) 10916 return (rc); 10917 10918 switch (cmd) { 10919 case CHELSIO_T4_GETREG: { 10920 struct t4_reg *edata = (struct t4_reg *)data; 10921 10922 if ((edata->addr & 0x3) != 0 || edata->addr >= sc->mmio_len) 10923 return (EFAULT); 10924 10925 if (edata->size == 4) 10926 edata->val = t4_read_reg(sc, edata->addr); 10927 else if (edata->size == 8) 10928 edata->val = t4_read_reg64(sc, edata->addr); 10929 else 10930 return (EINVAL); 10931 10932 break; 10933 } 10934 case CHELSIO_T4_SETREG: { 10935 struct t4_reg *edata = (struct t4_reg *)data; 10936 10937 if ((edata->addr & 0x3) != 0 || edata->addr >= sc->mmio_len) 10938 return (EFAULT); 10939 10940 if (edata->size == 4) { 10941 if (edata->val & 0xffffffff00000000) 10942 return (EINVAL); 10943 t4_write_reg(sc, edata->addr, (uint32_t) edata->val); 10944 } else if (edata->size == 8) 10945 t4_write_reg64(sc, edata->addr, edata->val); 10946 else 10947 return (EINVAL); 10948 break; 10949 } 10950 case CHELSIO_T4_REGDUMP: { 10951 struct t4_regdump *regs = (struct t4_regdump *)data; 10952 int reglen = t4_get_regs_len(sc); 10953 uint8_t *buf; 10954 10955 if (regs->len < reglen) { 10956 regs->len = reglen; /* hint to the caller */ 10957 return (ENOBUFS); 10958 } 10959 10960 regs->len = reglen; 10961 buf = malloc(reglen, M_CXGBE, M_WAITOK | M_ZERO); 10962 get_regs(sc, regs, buf); 10963 rc = copyout(buf, regs->data, reglen); 10964 free(buf, M_CXGBE); 10965 break; 10966 } 10967 case CHELSIO_T4_GET_FILTER_MODE: 10968 rc = get_filter_mode(sc, (uint32_t *)data); 10969 break; 10970 case CHELSIO_T4_SET_FILTER_MODE: 10971 rc = set_filter_mode(sc, *(uint32_t *)data); 10972 break; 10973 case CHELSIO_T4_GET_FILTER: 10974 rc = get_filter(sc, (struct t4_filter *)data); 10975 break; 10976 case CHELSIO_T4_SET_FILTER: 10977 rc = set_filter(sc, (struct t4_filter *)data); 10978 break; 10979 case CHELSIO_T4_DEL_FILTER: 10980 rc = del_filter(sc, (struct t4_filter *)data); 10981 break; 10982 case CHELSIO_T4_GET_SGE_CONTEXT: 10983 rc = get_sge_context(sc, (struct t4_sge_context *)data); 10984 break; 10985 case CHELSIO_T4_LOAD_FW: 10986 rc = load_fw(sc, (struct t4_data *)data); 10987 break; 10988 case CHELSIO_T4_GET_MEM: 10989 rc = read_card_mem(sc, 2, (struct t4_mem_range *)data); 10990 break; 10991 case CHELSIO_T4_GET_I2C: 10992 rc = read_i2c(sc, (struct t4_i2c_data *)data); 10993 break; 10994 case CHELSIO_T4_CLEAR_STATS: 10995 rc = clear_stats(sc, *(uint32_t *)data); 10996 break; 10997 case CHELSIO_T4_SCHED_CLASS: 10998 rc = t4_set_sched_class(sc, (struct t4_sched_params *)data); 10999 break; 11000 case CHELSIO_T4_SCHED_QUEUE: 11001 rc = t4_set_sched_queue(sc, (struct t4_sched_queue *)data); 11002 break; 11003 case CHELSIO_T4_GET_TRACER: 11004 rc = t4_get_tracer(sc, (struct t4_tracer *)data); 11005 break; 11006 case CHELSIO_T4_SET_TRACER: 11007 rc = t4_set_tracer(sc, (struct t4_tracer *)data); 11008 break; 11009 case CHELSIO_T4_LOAD_CFG: 11010 rc = load_cfg(sc, (struct t4_data *)data); 11011 break; 11012 case CHELSIO_T4_LOAD_BOOT: 11013 rc = load_boot(sc, (struct t4_bootrom *)data); 11014 break; 11015 case CHELSIO_T4_LOAD_BOOTCFG: 11016 rc = load_bootcfg(sc, (struct t4_data *)data); 11017 break; 11018 case CHELSIO_T4_CUDBG_DUMP: 11019 rc = cudbg_dump(sc, (struct t4_cudbg_dump *)data); 11020 break; 11021 case CHELSIO_T4_SET_OFLD_POLICY: 11022 rc = set_offload_policy(sc, (struct t4_offload_policy *)data); 11023 break; 11024 default: 11025 rc = ENOTTY; 11026 } 11027 11028 return (rc); 11029 } 11030 11031 #ifdef TCP_OFFLOAD 11032 static int 11033 toe_capability(struct vi_info *vi, int enable) 11034 { 11035 int rc; 11036 struct port_info *pi = vi->pi; 11037 struct adapter *sc = pi->adapter; 11038 11039 ASSERT_SYNCHRONIZED_OP(sc); 11040 11041 if (!is_offload(sc)) 11042 return (ENODEV); 11043 11044 if (enable) { 11045 if ((vi->ifp->if_capenable & IFCAP_TOE) != 0) { 11046 /* TOE is already enabled. */ 11047 return (0); 11048 } 11049 11050 /* 11051 * We need the port's queues around so that we're able to send 11052 * and receive CPLs to/from the TOE even if the ifnet for this 11053 * port has never been UP'd administratively. 11054 */ 11055 if (!(vi->flags & VI_INIT_DONE)) { 11056 rc = vi_full_init(vi); 11057 if (rc) 11058 return (rc); 11059 } 11060 if (!(pi->vi[0].flags & VI_INIT_DONE)) { 11061 rc = vi_full_init(&pi->vi[0]); 11062 if (rc) 11063 return (rc); 11064 } 11065 11066 if (isset(&sc->offload_map, pi->port_id)) { 11067 /* TOE is enabled on another VI of this port. */ 11068 pi->uld_vis++; 11069 return (0); 11070 } 11071 11072 if (!uld_active(sc, ULD_TOM)) { 11073 rc = t4_activate_uld(sc, ULD_TOM); 11074 if (rc == EAGAIN) { 11075 log(LOG_WARNING, 11076 "You must kldload t4_tom.ko before trying " 11077 "to enable TOE on a cxgbe interface.\n"); 11078 } 11079 if (rc != 0) 11080 return (rc); 11081 KASSERT(sc->tom_softc != NULL, 11082 ("%s: TOM activated but softc NULL", __func__)); 11083 KASSERT(uld_active(sc, ULD_TOM), 11084 ("%s: TOM activated but flag not set", __func__)); 11085 } 11086 11087 /* Activate iWARP and iSCSI too, if the modules are loaded. */ 11088 if (!uld_active(sc, ULD_IWARP)) 11089 (void) t4_activate_uld(sc, ULD_IWARP); 11090 if (!uld_active(sc, ULD_ISCSI)) 11091 (void) t4_activate_uld(sc, ULD_ISCSI); 11092 11093 pi->uld_vis++; 11094 setbit(&sc->offload_map, pi->port_id); 11095 } else { 11096 pi->uld_vis--; 11097 11098 if (!isset(&sc->offload_map, pi->port_id) || pi->uld_vis > 0) 11099 return (0); 11100 11101 KASSERT(uld_active(sc, ULD_TOM), 11102 ("%s: TOM never initialized?", __func__)); 11103 clrbit(&sc->offload_map, pi->port_id); 11104 } 11105 11106 return (0); 11107 } 11108 11109 /* 11110 * Add an upper layer driver to the global list. 11111 */ 11112 int 11113 t4_register_uld(struct uld_info *ui) 11114 { 11115 int rc = 0; 11116 struct uld_info *u; 11117 11118 sx_xlock(&t4_uld_list_lock); 11119 SLIST_FOREACH(u, &t4_uld_list, link) { 11120 if (u->uld_id == ui->uld_id) { 11121 rc = EEXIST; 11122 goto done; 11123 } 11124 } 11125 11126 SLIST_INSERT_HEAD(&t4_uld_list, ui, link); 11127 ui->refcount = 0; 11128 done: 11129 sx_xunlock(&t4_uld_list_lock); 11130 return (rc); 11131 } 11132 11133 int 11134 t4_unregister_uld(struct uld_info *ui) 11135 { 11136 int rc = EINVAL; 11137 struct uld_info *u; 11138 11139 sx_xlock(&t4_uld_list_lock); 11140 11141 SLIST_FOREACH(u, &t4_uld_list, link) { 11142 if (u == ui) { 11143 if (ui->refcount > 0) { 11144 rc = EBUSY; 11145 goto done; 11146 } 11147 11148 SLIST_REMOVE(&t4_uld_list, ui, uld_info, link); 11149 rc = 0; 11150 goto done; 11151 } 11152 } 11153 done: 11154 sx_xunlock(&t4_uld_list_lock); 11155 return (rc); 11156 } 11157 11158 int 11159 t4_activate_uld(struct adapter *sc, int id) 11160 { 11161 int rc; 11162 struct uld_info *ui; 11163 11164 ASSERT_SYNCHRONIZED_OP(sc); 11165 11166 if (id < 0 || id > ULD_MAX) 11167 return (EINVAL); 11168 rc = EAGAIN; /* kldoad the module with this ULD and try again. */ 11169 11170 sx_slock(&t4_uld_list_lock); 11171 11172 SLIST_FOREACH(ui, &t4_uld_list, link) { 11173 if (ui->uld_id == id) { 11174 if (!(sc->flags & FULL_INIT_DONE)) { 11175 rc = adapter_full_init(sc); 11176 if (rc != 0) 11177 break; 11178 } 11179 11180 rc = ui->activate(sc); 11181 if (rc == 0) { 11182 setbit(&sc->active_ulds, id); 11183 ui->refcount++; 11184 } 11185 break; 11186 } 11187 } 11188 11189 sx_sunlock(&t4_uld_list_lock); 11190 11191 return (rc); 11192 } 11193 11194 int 11195 t4_deactivate_uld(struct adapter *sc, int id) 11196 { 11197 int rc; 11198 struct uld_info *ui; 11199 11200 ASSERT_SYNCHRONIZED_OP(sc); 11201 11202 if (id < 0 || id > ULD_MAX) 11203 return (EINVAL); 11204 rc = ENXIO; 11205 11206 sx_slock(&t4_uld_list_lock); 11207 11208 SLIST_FOREACH(ui, &t4_uld_list, link) { 11209 if (ui->uld_id == id) { 11210 rc = ui->deactivate(sc); 11211 if (rc == 0) { 11212 clrbit(&sc->active_ulds, id); 11213 ui->refcount--; 11214 } 11215 break; 11216 } 11217 } 11218 11219 sx_sunlock(&t4_uld_list_lock); 11220 11221 return (rc); 11222 } 11223 11224 static void 11225 t4_async_event(void *arg, int n) 11226 { 11227 struct uld_info *ui; 11228 struct adapter *sc = (struct adapter *)arg; 11229 11230 if (begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4async") != 0) 11231 return; 11232 sx_slock(&t4_uld_list_lock); 11233 SLIST_FOREACH(ui, &t4_uld_list, link) { 11234 if (ui->uld_id == ULD_IWARP) { 11235 ui->async_event(sc); 11236 break; 11237 } 11238 } 11239 sx_sunlock(&t4_uld_list_lock); 11240 end_synchronized_op(sc, 0); 11241 } 11242 11243 int 11244 uld_active(struct adapter *sc, int uld_id) 11245 { 11246 11247 MPASS(uld_id >= 0 && uld_id <= ULD_MAX); 11248 11249 return (isset(&sc->active_ulds, uld_id)); 11250 } 11251 #endif 11252 11253 /* 11254 * t = ptr to tunable. 11255 * nc = number of CPUs. 11256 * c = compiled in default for that tunable. 11257 */ 11258 static void 11259 calculate_nqueues(int *t, int nc, const int c) 11260 { 11261 int nq; 11262 11263 if (*t > 0) 11264 return; 11265 nq = *t < 0 ? -*t : c; 11266 *t = min(nc, nq); 11267 } 11268 11269 /* 11270 * Come up with reasonable defaults for some of the tunables, provided they're 11271 * not set by the user (in which case we'll use the values as is). 11272 */ 11273 static void 11274 tweak_tunables(void) 11275 { 11276 int nc = mp_ncpus; /* our snapshot of the number of CPUs */ 11277 11278 if (t4_ntxq < 1) { 11279 #ifdef RSS 11280 t4_ntxq = rss_getnumbuckets(); 11281 #else 11282 calculate_nqueues(&t4_ntxq, nc, NTXQ); 11283 #endif 11284 } 11285 11286 calculate_nqueues(&t4_ntxq_vi, nc, NTXQ_VI); 11287 11288 if (t4_nrxq < 1) { 11289 #ifdef RSS 11290 t4_nrxq = rss_getnumbuckets(); 11291 #else 11292 calculate_nqueues(&t4_nrxq, nc, NRXQ); 11293 #endif 11294 } 11295 11296 calculate_nqueues(&t4_nrxq_vi, nc, NRXQ_VI); 11297 11298 #if defined(TCP_OFFLOAD) || defined(RATELIMIT) 11299 calculate_nqueues(&t4_nofldtxq, nc, NOFLDTXQ); 11300 calculate_nqueues(&t4_nofldtxq_vi, nc, NOFLDTXQ_VI); 11301 #endif 11302 #ifdef TCP_OFFLOAD 11303 calculate_nqueues(&t4_nofldrxq, nc, NOFLDRXQ); 11304 calculate_nqueues(&t4_nofldrxq_vi, nc, NOFLDRXQ_VI); 11305 #endif 11306 11307 #if defined(TCP_OFFLOAD) || defined(KERN_TLS) 11308 if (t4_toecaps_allowed == -1) 11309 t4_toecaps_allowed = FW_CAPS_CONFIG_TOE; 11310 #else 11311 if (t4_toecaps_allowed == -1) 11312 t4_toecaps_allowed = 0; 11313 #endif 11314 11315 #ifdef TCP_OFFLOAD 11316 if (t4_rdmacaps_allowed == -1) { 11317 t4_rdmacaps_allowed = FW_CAPS_CONFIG_RDMA_RDDP | 11318 FW_CAPS_CONFIG_RDMA_RDMAC; 11319 } 11320 11321 if (t4_iscsicaps_allowed == -1) { 11322 t4_iscsicaps_allowed = FW_CAPS_CONFIG_ISCSI_INITIATOR_PDU | 11323 FW_CAPS_CONFIG_ISCSI_TARGET_PDU | 11324 FW_CAPS_CONFIG_ISCSI_T10DIF; 11325 } 11326 11327 if (t4_tmr_idx_ofld < 0 || t4_tmr_idx_ofld >= SGE_NTIMERS) 11328 t4_tmr_idx_ofld = TMR_IDX_OFLD; 11329 11330 if (t4_pktc_idx_ofld < -1 || t4_pktc_idx_ofld >= SGE_NCOUNTERS) 11331 t4_pktc_idx_ofld = PKTC_IDX_OFLD; 11332 11333 if (t4_toe_tls_rx_timeout < 0) 11334 t4_toe_tls_rx_timeout = 0; 11335 #else 11336 if (t4_rdmacaps_allowed == -1) 11337 t4_rdmacaps_allowed = 0; 11338 11339 if (t4_iscsicaps_allowed == -1) 11340 t4_iscsicaps_allowed = 0; 11341 #endif 11342 11343 #ifdef DEV_NETMAP 11344 calculate_nqueues(&t4_nnmtxq, nc, NNMTXQ); 11345 calculate_nqueues(&t4_nnmrxq, nc, NNMRXQ); 11346 calculate_nqueues(&t4_nnmtxq_vi, nc, NNMTXQ_VI); 11347 calculate_nqueues(&t4_nnmrxq_vi, nc, NNMRXQ_VI); 11348 #endif 11349 11350 if (t4_tmr_idx < 0 || t4_tmr_idx >= SGE_NTIMERS) 11351 t4_tmr_idx = TMR_IDX; 11352 11353 if (t4_pktc_idx < -1 || t4_pktc_idx >= SGE_NCOUNTERS) 11354 t4_pktc_idx = PKTC_IDX; 11355 11356 if (t4_qsize_txq < 128) 11357 t4_qsize_txq = 128; 11358 11359 if (t4_qsize_rxq < 128) 11360 t4_qsize_rxq = 128; 11361 while (t4_qsize_rxq & 7) 11362 t4_qsize_rxq++; 11363 11364 t4_intr_types &= INTR_MSIX | INTR_MSI | INTR_INTX; 11365 11366 /* 11367 * Number of VIs to create per-port. The first VI is the "main" regular 11368 * VI for the port. The rest are additional virtual interfaces on the 11369 * same physical port. Note that the main VI does not have native 11370 * netmap support but the extra VIs do. 11371 * 11372 * Limit the number of VIs per port to the number of available 11373 * MAC addresses per port. 11374 */ 11375 if (t4_num_vis < 1) 11376 t4_num_vis = 1; 11377 if (t4_num_vis > nitems(vi_mac_funcs)) { 11378 t4_num_vis = nitems(vi_mac_funcs); 11379 printf("cxgbe: number of VIs limited to %d\n", t4_num_vis); 11380 } 11381 11382 if (pcie_relaxed_ordering < 0 || pcie_relaxed_ordering > 2) { 11383 pcie_relaxed_ordering = 1; 11384 #if defined(__i386__) || defined(__amd64__) 11385 if (cpu_vendor_id == CPU_VENDOR_INTEL) 11386 pcie_relaxed_ordering = 0; 11387 #endif 11388 } 11389 } 11390 11391 #ifdef DDB 11392 static void 11393 t4_dump_tcb(struct adapter *sc, int tid) 11394 { 11395 uint32_t base, i, j, off, pf, reg, save, tcb_addr, win_pos; 11396 11397 reg = PCIE_MEM_ACCESS_REG(A_PCIE_MEM_ACCESS_OFFSET, 2); 11398 save = t4_read_reg(sc, reg); 11399 base = sc->memwin[2].mw_base; 11400 11401 /* Dump TCB for the tid */ 11402 tcb_addr = t4_read_reg(sc, A_TP_CMM_TCB_BASE); 11403 tcb_addr += tid * TCB_SIZE; 11404 11405 if (is_t4(sc)) { 11406 pf = 0; 11407 win_pos = tcb_addr & ~0xf; /* start must be 16B aligned */ 11408 } else { 11409 pf = V_PFNUM(sc->pf); 11410 win_pos = tcb_addr & ~0x7f; /* start must be 128B aligned */ 11411 } 11412 t4_write_reg(sc, reg, win_pos | pf); 11413 t4_read_reg(sc, reg); 11414 11415 off = tcb_addr - win_pos; 11416 for (i = 0; i < 4; i++) { 11417 uint32_t buf[8]; 11418 for (j = 0; j < 8; j++, off += 4) 11419 buf[j] = htonl(t4_read_reg(sc, base + off)); 11420 11421 db_printf("%08x %08x %08x %08x %08x %08x %08x %08x\n", 11422 buf[0], buf[1], buf[2], buf[3], buf[4], buf[5], buf[6], 11423 buf[7]); 11424 } 11425 11426 t4_write_reg(sc, reg, save); 11427 t4_read_reg(sc, reg); 11428 } 11429 11430 static void 11431 t4_dump_devlog(struct adapter *sc) 11432 { 11433 struct devlog_params *dparams = &sc->params.devlog; 11434 struct fw_devlog_e e; 11435 int i, first, j, m, nentries, rc; 11436 uint64_t ftstamp = UINT64_MAX; 11437 11438 if (dparams->start == 0) { 11439 db_printf("devlog params not valid\n"); 11440 return; 11441 } 11442 11443 nentries = dparams->size / sizeof(struct fw_devlog_e); 11444 m = fwmtype_to_hwmtype(dparams->memtype); 11445 11446 /* Find the first entry. */ 11447 first = -1; 11448 for (i = 0; i < nentries && !db_pager_quit; i++) { 11449 rc = -t4_mem_read(sc, m, dparams->start + i * sizeof(e), 11450 sizeof(e), (void *)&e); 11451 if (rc != 0) 11452 break; 11453 11454 if (e.timestamp == 0) 11455 break; 11456 11457 e.timestamp = be64toh(e.timestamp); 11458 if (e.timestamp < ftstamp) { 11459 ftstamp = e.timestamp; 11460 first = i; 11461 } 11462 } 11463 11464 if (first == -1) 11465 return; 11466 11467 i = first; 11468 do { 11469 rc = -t4_mem_read(sc, m, dparams->start + i * sizeof(e), 11470 sizeof(e), (void *)&e); 11471 if (rc != 0) 11472 return; 11473 11474 if (e.timestamp == 0) 11475 return; 11476 11477 e.timestamp = be64toh(e.timestamp); 11478 e.seqno = be32toh(e.seqno); 11479 for (j = 0; j < 8; j++) 11480 e.params[j] = be32toh(e.params[j]); 11481 11482 db_printf("%10d %15ju %8s %8s ", 11483 e.seqno, e.timestamp, 11484 (e.level < nitems(devlog_level_strings) ? 11485 devlog_level_strings[e.level] : "UNKNOWN"), 11486 (e.facility < nitems(devlog_facility_strings) ? 11487 devlog_facility_strings[e.facility] : "UNKNOWN")); 11488 db_printf(e.fmt, e.params[0], e.params[1], e.params[2], 11489 e.params[3], e.params[4], e.params[5], e.params[6], 11490 e.params[7]); 11491 11492 if (++i == nentries) 11493 i = 0; 11494 } while (i != first && !db_pager_quit); 11495 } 11496 11497 static struct command_table db_t4_table = LIST_HEAD_INITIALIZER(db_t4_table); 11498 _DB_SET(_show, t4, NULL, db_show_table, 0, &db_t4_table); 11499 11500 DB_FUNC(devlog, db_show_devlog, db_t4_table, CS_OWN, NULL) 11501 { 11502 device_t dev; 11503 int t; 11504 bool valid; 11505 11506 valid = false; 11507 t = db_read_token(); 11508 if (t == tIDENT) { 11509 dev = device_lookup_by_name(db_tok_string); 11510 valid = true; 11511 } 11512 db_skip_to_eol(); 11513 if (!valid) { 11514 db_printf("usage: show t4 devlog <nexus>\n"); 11515 return; 11516 } 11517 11518 if (dev == NULL) { 11519 db_printf("device not found\n"); 11520 return; 11521 } 11522 11523 t4_dump_devlog(device_get_softc(dev)); 11524 } 11525 11526 DB_FUNC(tcb, db_show_t4tcb, db_t4_table, CS_OWN, NULL) 11527 { 11528 device_t dev; 11529 int radix, tid, t; 11530 bool valid; 11531 11532 valid = false; 11533 radix = db_radix; 11534 db_radix = 10; 11535 t = db_read_token(); 11536 if (t == tIDENT) { 11537 dev = device_lookup_by_name(db_tok_string); 11538 t = db_read_token(); 11539 if (t == tNUMBER) { 11540 tid = db_tok_number; 11541 valid = true; 11542 } 11543 } 11544 db_radix = radix; 11545 db_skip_to_eol(); 11546 if (!valid) { 11547 db_printf("usage: show t4 tcb <nexus> <tid>\n"); 11548 return; 11549 } 11550 11551 if (dev == NULL) { 11552 db_printf("device not found\n"); 11553 return; 11554 } 11555 if (tid < 0) { 11556 db_printf("invalid tid\n"); 11557 return; 11558 } 11559 11560 t4_dump_tcb(device_get_softc(dev), tid); 11561 } 11562 #endif 11563 11564 static eventhandler_tag vxlan_start_evtag; 11565 static eventhandler_tag vxlan_stop_evtag; 11566 11567 struct vxlan_evargs { 11568 struct ifnet *ifp; 11569 uint16_t port; 11570 }; 11571 11572 static void 11573 t4_vxlan_start(struct adapter *sc, void *arg) 11574 { 11575 struct vxlan_evargs *v = arg; 11576 struct port_info *pi; 11577 uint8_t match_all_mac[ETHER_ADDR_LEN] = {0}; 11578 int i, rc; 11579 11580 if (sc->nrawf == 0 || chip_id(sc) <= CHELSIO_T5) 11581 return; 11582 if (begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4vxst") != 0) 11583 return; 11584 11585 if (sc->vxlan_refcount == 0) { 11586 sc->vxlan_port = v->port; 11587 sc->vxlan_refcount = 1; 11588 t4_write_reg(sc, A_MPS_RX_VXLAN_TYPE, 11589 V_VXLAN(v->port) | F_VXLAN_EN); 11590 for_each_port(sc, i) { 11591 pi = sc->port[i]; 11592 if (pi->vxlan_tcam_entry == true) 11593 continue; 11594 rc = t4_alloc_raw_mac_filt(sc, pi->vi[0].viid, 11595 match_all_mac, match_all_mac, 11596 sc->rawf_base + pi->port_id, 1, pi->port_id, true); 11597 if (rc < 0) { 11598 rc = -rc; 11599 log(LOG_ERR, 11600 "%s: failed to add VXLAN TCAM entry: %d.\n", 11601 device_get_name(pi->vi[0].dev), rc); 11602 } else { 11603 MPASS(rc == sc->rawf_base + pi->port_id); 11604 rc = 0; 11605 pi->vxlan_tcam_entry = true; 11606 } 11607 } 11608 } else if (sc->vxlan_port == v->port) { 11609 sc->vxlan_refcount++; 11610 } else { 11611 log(LOG_ERR, "%s: VXLAN already configured on port %d; " 11612 "ignoring attempt to configure it on port %d\n", 11613 device_get_nameunit(sc->dev), sc->vxlan_port, v->port); 11614 } 11615 end_synchronized_op(sc, 0); 11616 } 11617 11618 static void 11619 t4_vxlan_stop(struct adapter *sc, void *arg) 11620 { 11621 struct vxlan_evargs *v = arg; 11622 11623 if (sc->nrawf == 0 || chip_id(sc) <= CHELSIO_T5) 11624 return; 11625 if (begin_synchronized_op(sc, NULL, SLEEP_OK | INTR_OK, "t4vxsp") != 0) 11626 return; 11627 11628 /* 11629 * VXLANs may have been configured before the driver was loaded so we 11630 * may see more stops than starts. This is not handled cleanly but at 11631 * least we keep the refcount sane. 11632 */ 11633 if (sc->vxlan_port != v->port) 11634 goto done; 11635 if (sc->vxlan_refcount == 0) { 11636 log(LOG_ERR, 11637 "%s: VXLAN operation on port %d was stopped earlier; " 11638 "ignoring attempt to stop it again.\n", 11639 device_get_nameunit(sc->dev), sc->vxlan_port); 11640 } else if (--sc->vxlan_refcount == 0) { 11641 t4_set_reg_field(sc, A_MPS_RX_VXLAN_TYPE, F_VXLAN_EN, 0); 11642 } 11643 done: 11644 end_synchronized_op(sc, 0); 11645 } 11646 11647 static void 11648 t4_vxlan_start_handler(void *arg __unused, struct ifnet *ifp, 11649 sa_family_t family, u_int port) 11650 { 11651 struct vxlan_evargs v; 11652 11653 MPASS(family == AF_INET || family == AF_INET6); 11654 v.ifp = ifp; 11655 v.port = port; 11656 11657 t4_iterate(t4_vxlan_start, &v); 11658 } 11659 11660 static void 11661 t4_vxlan_stop_handler(void *arg __unused, struct ifnet *ifp, sa_family_t family, 11662 u_int port) 11663 { 11664 struct vxlan_evargs v; 11665 11666 MPASS(family == AF_INET || family == AF_INET6); 11667 v.ifp = ifp; 11668 v.port = port; 11669 11670 t4_iterate(t4_vxlan_stop, &v); 11671 } 11672 11673 11674 static struct sx mlu; /* mod load unload */ 11675 SX_SYSINIT(cxgbe_mlu, &mlu, "cxgbe mod load/unload"); 11676 11677 static int 11678 mod_event(module_t mod, int cmd, void *arg) 11679 { 11680 int rc = 0; 11681 static int loaded = 0; 11682 11683 switch (cmd) { 11684 case MOD_LOAD: 11685 sx_xlock(&mlu); 11686 if (loaded++ == 0) { 11687 t4_sge_modload(); 11688 t4_register_shared_cpl_handler(CPL_SET_TCB_RPL, 11689 t4_filter_rpl, CPL_COOKIE_FILTER); 11690 t4_register_shared_cpl_handler(CPL_L2T_WRITE_RPL, 11691 do_l2t_write_rpl, CPL_COOKIE_FILTER); 11692 t4_register_shared_cpl_handler(CPL_ACT_OPEN_RPL, 11693 t4_hashfilter_ao_rpl, CPL_COOKIE_HASHFILTER); 11694 t4_register_shared_cpl_handler(CPL_SET_TCB_RPL, 11695 t4_hashfilter_tcb_rpl, CPL_COOKIE_HASHFILTER); 11696 t4_register_shared_cpl_handler(CPL_ABORT_RPL_RSS, 11697 t4_del_hashfilter_rpl, CPL_COOKIE_HASHFILTER); 11698 t4_register_cpl_handler(CPL_TRACE_PKT, t4_trace_pkt); 11699 t4_register_cpl_handler(CPL_T5_TRACE_PKT, t5_trace_pkt); 11700 t4_register_cpl_handler(CPL_SMT_WRITE_RPL, 11701 do_smt_write_rpl); 11702 sx_init(&t4_list_lock, "T4/T5 adapters"); 11703 SLIST_INIT(&t4_list); 11704 callout_init(&fatal_callout, 1); 11705 #ifdef TCP_OFFLOAD 11706 sx_init(&t4_uld_list_lock, "T4/T5 ULDs"); 11707 SLIST_INIT(&t4_uld_list); 11708 #endif 11709 #ifdef INET6 11710 t4_clip_modload(); 11711 #endif 11712 #ifdef KERN_TLS 11713 t6_ktls_modload(); 11714 #endif 11715 t4_tracer_modload(); 11716 tweak_tunables(); 11717 vxlan_start_evtag = 11718 EVENTHANDLER_REGISTER(vxlan_start, 11719 t4_vxlan_start_handler, NULL, 11720 EVENTHANDLER_PRI_ANY); 11721 vxlan_stop_evtag = 11722 EVENTHANDLER_REGISTER(vxlan_stop, 11723 t4_vxlan_stop_handler, NULL, 11724 EVENTHANDLER_PRI_ANY); 11725 } 11726 sx_xunlock(&mlu); 11727 break; 11728 11729 case MOD_UNLOAD: 11730 sx_xlock(&mlu); 11731 if (--loaded == 0) { 11732 int tries; 11733 11734 sx_slock(&t4_list_lock); 11735 if (!SLIST_EMPTY(&t4_list)) { 11736 rc = EBUSY; 11737 sx_sunlock(&t4_list_lock); 11738 goto done_unload; 11739 } 11740 #ifdef TCP_OFFLOAD 11741 sx_slock(&t4_uld_list_lock); 11742 if (!SLIST_EMPTY(&t4_uld_list)) { 11743 rc = EBUSY; 11744 sx_sunlock(&t4_uld_list_lock); 11745 sx_sunlock(&t4_list_lock); 11746 goto done_unload; 11747 } 11748 #endif 11749 tries = 0; 11750 while (tries++ < 5 && t4_sge_extfree_refs() != 0) { 11751 uprintf("%ju clusters with custom free routine " 11752 "still is use.\n", t4_sge_extfree_refs()); 11753 pause("t4unload", 2 * hz); 11754 } 11755 #ifdef TCP_OFFLOAD 11756 sx_sunlock(&t4_uld_list_lock); 11757 #endif 11758 sx_sunlock(&t4_list_lock); 11759 11760 if (t4_sge_extfree_refs() == 0) { 11761 EVENTHANDLER_DEREGISTER(vxlan_start, 11762 vxlan_start_evtag); 11763 EVENTHANDLER_DEREGISTER(vxlan_stop, 11764 vxlan_stop_evtag); 11765 t4_tracer_modunload(); 11766 #ifdef KERN_TLS 11767 t6_ktls_modunload(); 11768 #endif 11769 #ifdef INET6 11770 t4_clip_modunload(); 11771 #endif 11772 #ifdef TCP_OFFLOAD 11773 sx_destroy(&t4_uld_list_lock); 11774 #endif 11775 sx_destroy(&t4_list_lock); 11776 t4_sge_modunload(); 11777 loaded = 0; 11778 } else { 11779 rc = EBUSY; 11780 loaded++; /* undo earlier decrement */ 11781 } 11782 } 11783 done_unload: 11784 sx_xunlock(&mlu); 11785 break; 11786 } 11787 11788 return (rc); 11789 } 11790 11791 static devclass_t t4_devclass, t5_devclass, t6_devclass; 11792 static devclass_t cxgbe_devclass, cxl_devclass, cc_devclass; 11793 static devclass_t vcxgbe_devclass, vcxl_devclass, vcc_devclass; 11794 11795 DRIVER_MODULE(t4nex, pci, t4_driver, t4_devclass, mod_event, 0); 11796 MODULE_VERSION(t4nex, 1); 11797 MODULE_DEPEND(t4nex, firmware, 1, 1, 1); 11798 #ifdef DEV_NETMAP 11799 MODULE_DEPEND(t4nex, netmap, 1, 1, 1); 11800 #endif /* DEV_NETMAP */ 11801 11802 DRIVER_MODULE(t5nex, pci, t5_driver, t5_devclass, mod_event, 0); 11803 MODULE_VERSION(t5nex, 1); 11804 MODULE_DEPEND(t5nex, firmware, 1, 1, 1); 11805 #ifdef DEV_NETMAP 11806 MODULE_DEPEND(t5nex, netmap, 1, 1, 1); 11807 #endif /* DEV_NETMAP */ 11808 11809 DRIVER_MODULE(t6nex, pci, t6_driver, t6_devclass, mod_event, 0); 11810 MODULE_VERSION(t6nex, 1); 11811 MODULE_DEPEND(t6nex, firmware, 1, 1, 1); 11812 #ifdef DEV_NETMAP 11813 MODULE_DEPEND(t6nex, netmap, 1, 1, 1); 11814 #endif /* DEV_NETMAP */ 11815 11816 DRIVER_MODULE(cxgbe, t4nex, cxgbe_driver, cxgbe_devclass, 0, 0); 11817 MODULE_VERSION(cxgbe, 1); 11818 11819 DRIVER_MODULE(cxl, t5nex, cxl_driver, cxl_devclass, 0, 0); 11820 MODULE_VERSION(cxl, 1); 11821 11822 DRIVER_MODULE(cc, t6nex, cc_driver, cc_devclass, 0, 0); 11823 MODULE_VERSION(cc, 1); 11824 11825 DRIVER_MODULE(vcxgbe, cxgbe, vcxgbe_driver, vcxgbe_devclass, 0, 0); 11826 MODULE_VERSION(vcxgbe, 1); 11827 11828 DRIVER_MODULE(vcxl, cxl, vcxl_driver, vcxl_devclass, 0, 0); 11829 MODULE_VERSION(vcxl, 1); 11830 11831 DRIVER_MODULE(vcc, cc, vcc_driver, vcc_devclass, 0, 0); 11832 MODULE_VERSION(vcc, 1); 11833