1 /*- 2 * Copyright (c) 2003 Silicon Graphics International Corp. 3 * Copyright (c) 2009-2011 Spectra Logic Corporation 4 * Copyright (c) 2012 The FreeBSD Foundation 5 * All rights reserved. 6 * 7 * Portions of this software were developed by Edward Tomasz Napierala 8 * under sponsorship from the FreeBSD Foundation. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions, and the following disclaimer, 15 * without modification. 16 * 2. Redistributions in binary form must reproduce at minimum a disclaimer 17 * substantially similar to the "NO WARRANTY" disclaimer below 18 * ("Disclaimer") and any redistribution must be conditioned upon 19 * including a substantially similar Disclaimer requirement for further 20 * binary redistribution. 21 * 22 * NO WARRANTY 23 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 24 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 25 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR 26 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 27 * HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 28 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 29 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 30 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 31 * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 32 * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 33 * POSSIBILITY OF SUCH DAMAGES. 34 * 35 * $Id: //depot/users/kenm/FreeBSD-test2/sys/cam/ctl/ctl_backend_block.c#5 $ 36 */ 37 /* 38 * CAM Target Layer driver backend for block devices. 39 * 40 * Author: Ken Merry <ken@FreeBSD.org> 41 */ 42 #include <sys/cdefs.h> 43 __FBSDID("$FreeBSD$"); 44 45 #include <sys/param.h> 46 #include <sys/systm.h> 47 #include <sys/kernel.h> 48 #include <sys/types.h> 49 #include <sys/kthread.h> 50 #include <sys/bio.h> 51 #include <sys/fcntl.h> 52 #include <sys/limits.h> 53 #include <sys/lock.h> 54 #include <sys/mutex.h> 55 #include <sys/condvar.h> 56 #include <sys/malloc.h> 57 #include <sys/conf.h> 58 #include <sys/ioccom.h> 59 #include <sys/queue.h> 60 #include <sys/sbuf.h> 61 #include <sys/endian.h> 62 #include <sys/uio.h> 63 #include <sys/buf.h> 64 #include <sys/taskqueue.h> 65 #include <sys/vnode.h> 66 #include <sys/namei.h> 67 #include <sys/mount.h> 68 #include <sys/disk.h> 69 #include <sys/fcntl.h> 70 #include <sys/filedesc.h> 71 #include <sys/proc.h> 72 #include <sys/pcpu.h> 73 #include <sys/module.h> 74 #include <sys/sdt.h> 75 #include <sys/devicestat.h> 76 #include <sys/sysctl.h> 77 78 #include <geom/geom.h> 79 80 #include <cam/cam.h> 81 #include <cam/scsi/scsi_all.h> 82 #include <cam/scsi/scsi_da.h> 83 #include <cam/ctl/ctl_io.h> 84 #include <cam/ctl/ctl.h> 85 #include <cam/ctl/ctl_backend.h> 86 #include <cam/ctl/ctl_frontend_internal.h> 87 #include <cam/ctl/ctl_ioctl.h> 88 #include <cam/ctl/ctl_scsi_all.h> 89 #include <cam/ctl/ctl_error.h> 90 91 /* 92 * The idea here is that we'll allocate enough S/G space to hold a 1MB 93 * I/O. If we get an I/O larger than that, we'll split it. 94 */ 95 #define CTLBLK_HALF_IO_SIZE (512 * 1024) 96 #define CTLBLK_MAX_IO_SIZE (CTLBLK_HALF_IO_SIZE * 2) 97 #define CTLBLK_MAX_SEG MAXPHYS 98 #define CTLBLK_HALF_SEGS MAX(CTLBLK_HALF_IO_SIZE / CTLBLK_MAX_SEG, 1) 99 #define CTLBLK_MAX_SEGS (CTLBLK_HALF_SEGS * 2) 100 101 #ifdef CTLBLK_DEBUG 102 #define DPRINTF(fmt, args...) \ 103 printf("cbb(%s:%d): " fmt, __FUNCTION__, __LINE__, ##args) 104 #else 105 #define DPRINTF(fmt, args...) do {} while(0) 106 #endif 107 108 #define PRIV(io) \ 109 ((struct ctl_ptr_len_flags *)&(io)->io_hdr.ctl_private[CTL_PRIV_BACKEND]) 110 #define ARGS(io) \ 111 ((struct ctl_lba_len_flags *)&(io)->io_hdr.ctl_private[CTL_PRIV_LBA_LEN]) 112 113 SDT_PROVIDER_DEFINE(cbb); 114 115 typedef enum { 116 CTL_BE_BLOCK_LUN_UNCONFIGURED = 0x01, 117 CTL_BE_BLOCK_LUN_CONFIG_ERR = 0x02, 118 CTL_BE_BLOCK_LUN_WAITING = 0x04, 119 CTL_BE_BLOCK_LUN_MULTI_THREAD = 0x08 120 } ctl_be_block_lun_flags; 121 122 typedef enum { 123 CTL_BE_BLOCK_NONE, 124 CTL_BE_BLOCK_DEV, 125 CTL_BE_BLOCK_FILE 126 } ctl_be_block_type; 127 128 struct ctl_be_block_devdata { 129 struct cdev *cdev; 130 struct cdevsw *csw; 131 int dev_ref; 132 }; 133 134 struct ctl_be_block_filedata { 135 struct ucred *cred; 136 }; 137 138 union ctl_be_block_bedata { 139 struct ctl_be_block_devdata dev; 140 struct ctl_be_block_filedata file; 141 }; 142 143 struct ctl_be_block_io; 144 struct ctl_be_block_lun; 145 146 typedef void (*cbb_dispatch_t)(struct ctl_be_block_lun *be_lun, 147 struct ctl_be_block_io *beio); 148 typedef uint64_t (*cbb_getattr_t)(struct ctl_be_block_lun *be_lun, 149 const char *attrname); 150 151 /* 152 * Backend LUN structure. There is a 1:1 mapping between a block device 153 * and a backend block LUN, and between a backend block LUN and a CTL LUN. 154 */ 155 struct ctl_be_block_lun { 156 struct ctl_lun_create_params params; 157 struct ctl_block_disk *disk; 158 char lunname[32]; 159 char *dev_path; 160 ctl_be_block_type dev_type; 161 struct vnode *vn; 162 union ctl_be_block_bedata backend; 163 cbb_dispatch_t dispatch; 164 cbb_dispatch_t lun_flush; 165 cbb_dispatch_t unmap; 166 cbb_getattr_t getattr; 167 uma_zone_t lun_zone; 168 uint64_t size_blocks; 169 uint64_t size_bytes; 170 uint32_t blocksize; 171 int blocksize_shift; 172 uint16_t pblockexp; 173 uint16_t pblockoff; 174 struct ctl_be_block_softc *softc; 175 struct devstat *disk_stats; 176 ctl_be_block_lun_flags flags; 177 STAILQ_ENTRY(ctl_be_block_lun) links; 178 struct ctl_be_lun ctl_be_lun; 179 struct taskqueue *io_taskqueue; 180 struct task io_task; 181 int num_threads; 182 STAILQ_HEAD(, ctl_io_hdr) input_queue; 183 STAILQ_HEAD(, ctl_io_hdr) config_write_queue; 184 STAILQ_HEAD(, ctl_io_hdr) datamove_queue; 185 struct mtx_padalign io_lock; 186 struct mtx_padalign queue_lock; 187 }; 188 189 /* 190 * Overall softc structure for the block backend module. 191 */ 192 struct ctl_be_block_softc { 193 struct mtx lock; 194 int num_disks; 195 STAILQ_HEAD(, ctl_block_disk) disk_list; 196 int num_luns; 197 STAILQ_HEAD(, ctl_be_block_lun) lun_list; 198 }; 199 200 static struct ctl_be_block_softc backend_block_softc; 201 202 /* 203 * Per-I/O information. 204 */ 205 struct ctl_be_block_io { 206 union ctl_io *io; 207 struct ctl_sg_entry sg_segs[CTLBLK_MAX_SEGS]; 208 struct iovec xiovecs[CTLBLK_MAX_SEGS]; 209 int bio_cmd; 210 int num_segs; 211 int num_bios_sent; 212 int num_bios_done; 213 int send_complete; 214 int num_errors; 215 struct bintime ds_t0; 216 devstat_tag_type ds_tag_type; 217 devstat_trans_flags ds_trans_type; 218 uint64_t io_len; 219 uint64_t io_offset; 220 struct ctl_be_block_softc *softc; 221 struct ctl_be_block_lun *lun; 222 void (*beio_cont)(struct ctl_be_block_io *beio); /* to continue processing */ 223 }; 224 225 static int cbb_num_threads = 14; 226 SYSCTL_NODE(_kern_cam_ctl, OID_AUTO, block, CTLFLAG_RD, 0, 227 "CAM Target Layer Block Backend"); 228 SYSCTL_INT(_kern_cam_ctl_block, OID_AUTO, num_threads, CTLFLAG_RWTUN, 229 &cbb_num_threads, 0, "Number of threads per backing file"); 230 231 static struct ctl_be_block_io *ctl_alloc_beio(struct ctl_be_block_softc *softc); 232 static void ctl_free_beio(struct ctl_be_block_io *beio); 233 static void ctl_complete_beio(struct ctl_be_block_io *beio); 234 static int ctl_be_block_move_done(union ctl_io *io); 235 static void ctl_be_block_biodone(struct bio *bio); 236 static void ctl_be_block_flush_file(struct ctl_be_block_lun *be_lun, 237 struct ctl_be_block_io *beio); 238 static void ctl_be_block_dispatch_file(struct ctl_be_block_lun *be_lun, 239 struct ctl_be_block_io *beio); 240 static void ctl_be_block_flush_dev(struct ctl_be_block_lun *be_lun, 241 struct ctl_be_block_io *beio); 242 static void ctl_be_block_unmap_dev(struct ctl_be_block_lun *be_lun, 243 struct ctl_be_block_io *beio); 244 static void ctl_be_block_dispatch_dev(struct ctl_be_block_lun *be_lun, 245 struct ctl_be_block_io *beio); 246 static uint64_t ctl_be_block_getattr_dev(struct ctl_be_block_lun *be_lun, 247 const char *attrname); 248 static void ctl_be_block_cw_dispatch(struct ctl_be_block_lun *be_lun, 249 union ctl_io *io); 250 static void ctl_be_block_dispatch(struct ctl_be_block_lun *be_lun, 251 union ctl_io *io); 252 static void ctl_be_block_worker(void *context, int pending); 253 static int ctl_be_block_submit(union ctl_io *io); 254 static int ctl_be_block_ioctl(struct cdev *dev, u_long cmd, caddr_t addr, 255 int flag, struct thread *td); 256 static int ctl_be_block_open_file(struct ctl_be_block_lun *be_lun, 257 struct ctl_lun_req *req); 258 static int ctl_be_block_open_dev(struct ctl_be_block_lun *be_lun, 259 struct ctl_lun_req *req); 260 static int ctl_be_block_close(struct ctl_be_block_lun *be_lun); 261 static int ctl_be_block_open(struct ctl_be_block_softc *softc, 262 struct ctl_be_block_lun *be_lun, 263 struct ctl_lun_req *req); 264 static int ctl_be_block_create(struct ctl_be_block_softc *softc, 265 struct ctl_lun_req *req); 266 static int ctl_be_block_rm(struct ctl_be_block_softc *softc, 267 struct ctl_lun_req *req); 268 static int ctl_be_block_modify_file(struct ctl_be_block_lun *be_lun, 269 struct ctl_lun_req *req); 270 static int ctl_be_block_modify_dev(struct ctl_be_block_lun *be_lun, 271 struct ctl_lun_req *req); 272 static int ctl_be_block_modify(struct ctl_be_block_softc *softc, 273 struct ctl_lun_req *req); 274 static void ctl_be_block_lun_shutdown(void *be_lun); 275 static void ctl_be_block_lun_config_status(void *be_lun, 276 ctl_lun_config_status status); 277 static int ctl_be_block_config_write(union ctl_io *io); 278 static int ctl_be_block_config_read(union ctl_io *io); 279 static int ctl_be_block_lun_info(void *be_lun, struct sbuf *sb); 280 static uint64_t ctl_be_block_lun_attr(void *be_lun, const char *attrname); 281 int ctl_be_block_init(void); 282 283 static struct ctl_backend_driver ctl_be_block_driver = 284 { 285 .name = "block", 286 .flags = CTL_BE_FLAG_HAS_CONFIG, 287 .init = ctl_be_block_init, 288 .data_submit = ctl_be_block_submit, 289 .data_move_done = ctl_be_block_move_done, 290 .config_read = ctl_be_block_config_read, 291 .config_write = ctl_be_block_config_write, 292 .ioctl = ctl_be_block_ioctl, 293 .lun_info = ctl_be_block_lun_info, 294 .lun_attr = ctl_be_block_lun_attr 295 }; 296 297 MALLOC_DEFINE(M_CTLBLK, "ctlblk", "Memory used for CTL block backend"); 298 CTL_BACKEND_DECLARE(cbb, ctl_be_block_driver); 299 300 static uma_zone_t beio_zone; 301 302 static struct ctl_be_block_io * 303 ctl_alloc_beio(struct ctl_be_block_softc *softc) 304 { 305 struct ctl_be_block_io *beio; 306 307 beio = uma_zalloc(beio_zone, M_WAITOK | M_ZERO); 308 beio->softc = softc; 309 return (beio); 310 } 311 312 static void 313 ctl_free_beio(struct ctl_be_block_io *beio) 314 { 315 int duplicate_free; 316 int i; 317 318 duplicate_free = 0; 319 320 for (i = 0; i < beio->num_segs; i++) { 321 if (beio->sg_segs[i].addr == NULL) 322 duplicate_free++; 323 324 uma_zfree(beio->lun->lun_zone, beio->sg_segs[i].addr); 325 beio->sg_segs[i].addr = NULL; 326 327 /* For compare we had two equal S/G lists. */ 328 if (ARGS(beio->io)->flags & CTL_LLF_COMPARE) { 329 uma_zfree(beio->lun->lun_zone, 330 beio->sg_segs[i + CTLBLK_HALF_SEGS].addr); 331 beio->sg_segs[i + CTLBLK_HALF_SEGS].addr = NULL; 332 } 333 } 334 335 if (duplicate_free > 0) { 336 printf("%s: %d duplicate frees out of %d segments\n", __func__, 337 duplicate_free, beio->num_segs); 338 } 339 340 uma_zfree(beio_zone, beio); 341 } 342 343 static void 344 ctl_complete_beio(struct ctl_be_block_io *beio) 345 { 346 union ctl_io *io = beio->io; 347 348 if (beio->beio_cont != NULL) { 349 beio->beio_cont(beio); 350 } else { 351 ctl_free_beio(beio); 352 ctl_data_submit_done(io); 353 } 354 } 355 356 static int 357 ctl_be_block_move_done(union ctl_io *io) 358 { 359 struct ctl_be_block_io *beio; 360 struct ctl_be_block_lun *be_lun; 361 struct ctl_lba_len_flags *lbalen; 362 #ifdef CTL_TIME_IO 363 struct bintime cur_bt; 364 #endif 365 int i; 366 367 beio = (struct ctl_be_block_io *)PRIV(io)->ptr; 368 be_lun = beio->lun; 369 370 DPRINTF("entered\n"); 371 372 #ifdef CTL_TIME_IO 373 getbintime(&cur_bt); 374 bintime_sub(&cur_bt, &io->io_hdr.dma_start_bt); 375 bintime_add(&io->io_hdr.dma_bt, &cur_bt); 376 io->io_hdr.num_dmas++; 377 #endif 378 io->scsiio.kern_rel_offset += io->scsiio.kern_data_len; 379 380 /* 381 * We set status at this point for read commands, and write 382 * commands with errors. 383 */ 384 if ((io->io_hdr.port_status == 0) && 385 ((io->io_hdr.flags & CTL_FLAG_ABORT) == 0) && 386 ((io->io_hdr.status & CTL_STATUS_MASK) == CTL_STATUS_NONE)) { 387 lbalen = ARGS(beio->io); 388 if (lbalen->flags & CTL_LLF_READ) { 389 ctl_set_success(&io->scsiio); 390 } else if (lbalen->flags & CTL_LLF_COMPARE) { 391 /* We have two data blocks ready for comparison. */ 392 for (i = 0; i < beio->num_segs; i++) { 393 if (memcmp(beio->sg_segs[i].addr, 394 beio->sg_segs[i + CTLBLK_HALF_SEGS].addr, 395 beio->sg_segs[i].len) != 0) 396 break; 397 } 398 if (i < beio->num_segs) 399 ctl_set_sense(&io->scsiio, 400 /*current_error*/ 1, 401 /*sense_key*/ SSD_KEY_MISCOMPARE, 402 /*asc*/ 0x1D, 403 /*ascq*/ 0x00, 404 SSD_ELEM_NONE); 405 else 406 ctl_set_success(&io->scsiio); 407 } 408 } 409 else if ((io->io_hdr.port_status != 0) 410 && ((io->io_hdr.flags & CTL_FLAG_ABORT) == 0) 411 && ((io->io_hdr.status & CTL_STATUS_MASK) == CTL_STATUS_NONE)) { 412 /* 413 * For hardware error sense keys, the sense key 414 * specific value is defined to be a retry count, 415 * but we use it to pass back an internal FETD 416 * error code. XXX KDM Hopefully the FETD is only 417 * using 16 bits for an error code, since that's 418 * all the space we have in the sks field. 419 */ 420 ctl_set_internal_failure(&io->scsiio, 421 /*sks_valid*/ 1, 422 /*retry_count*/ 423 io->io_hdr.port_status); 424 } 425 426 /* 427 * If this is a read, or a write with errors, it is done. 428 */ 429 if ((beio->bio_cmd == BIO_READ) 430 || ((io->io_hdr.flags & CTL_FLAG_ABORT) != 0) 431 || ((io->io_hdr.status & CTL_STATUS_MASK) != CTL_STATUS_NONE)) { 432 ctl_complete_beio(beio); 433 return (0); 434 } 435 436 /* 437 * At this point, we have a write and the DMA completed 438 * successfully. We now have to queue it to the task queue to 439 * execute the backend I/O. That is because we do blocking 440 * memory allocations, and in the file backing case, blocking I/O. 441 * This move done routine is generally called in the SIM's 442 * interrupt context, and therefore we cannot block. 443 */ 444 mtx_lock(&be_lun->queue_lock); 445 /* 446 * XXX KDM make sure that links is okay to use at this point. 447 * Otherwise, we either need to add another field to ctl_io_hdr, 448 * or deal with resource allocation here. 449 */ 450 STAILQ_INSERT_TAIL(&be_lun->datamove_queue, &io->io_hdr, links); 451 mtx_unlock(&be_lun->queue_lock); 452 453 taskqueue_enqueue(be_lun->io_taskqueue, &be_lun->io_task); 454 455 return (0); 456 } 457 458 static void 459 ctl_be_block_biodone(struct bio *bio) 460 { 461 struct ctl_be_block_io *beio; 462 struct ctl_be_block_lun *be_lun; 463 union ctl_io *io; 464 int error; 465 466 beio = bio->bio_caller1; 467 be_lun = beio->lun; 468 io = beio->io; 469 470 DPRINTF("entered\n"); 471 472 error = bio->bio_error; 473 mtx_lock(&be_lun->io_lock); 474 if (error != 0) 475 beio->num_errors++; 476 477 beio->num_bios_done++; 478 479 /* 480 * XXX KDM will this cause WITNESS to complain? Holding a lock 481 * during the free might cause it to complain. 482 */ 483 g_destroy_bio(bio); 484 485 /* 486 * If the send complete bit isn't set, or we aren't the last I/O to 487 * complete, then we're done. 488 */ 489 if ((beio->send_complete == 0) 490 || (beio->num_bios_done < beio->num_bios_sent)) { 491 mtx_unlock(&be_lun->io_lock); 492 return; 493 } 494 495 /* 496 * At this point, we've verified that we are the last I/O to 497 * complete, so it's safe to drop the lock. 498 */ 499 devstat_end_transaction(beio->lun->disk_stats, beio->io_len, 500 beio->ds_tag_type, beio->ds_trans_type, 501 /*now*/ NULL, /*then*/&beio->ds_t0); 502 mtx_unlock(&be_lun->io_lock); 503 504 /* 505 * If there are any errors from the backing device, we fail the 506 * entire I/O with a medium error. 507 */ 508 if (beio->num_errors > 0) { 509 if (error == EOPNOTSUPP) { 510 ctl_set_invalid_opcode(&io->scsiio); 511 } else if (error == ENOSPC) { 512 ctl_set_space_alloc_fail(&io->scsiio); 513 } else if (beio->bio_cmd == BIO_FLUSH) { 514 /* XXX KDM is there is a better error here? */ 515 ctl_set_internal_failure(&io->scsiio, 516 /*sks_valid*/ 1, 517 /*retry_count*/ 0xbad2); 518 } else 519 ctl_set_medium_error(&io->scsiio); 520 ctl_complete_beio(beio); 521 return; 522 } 523 524 /* 525 * If this is a write, a flush, a delete or verify, we're all done. 526 * If this is a read, we can now send the data to the user. 527 */ 528 if ((beio->bio_cmd == BIO_WRITE) 529 || (beio->bio_cmd == BIO_FLUSH) 530 || (beio->bio_cmd == BIO_DELETE) 531 || (ARGS(io)->flags & CTL_LLF_VERIFY)) { 532 ctl_set_success(&io->scsiio); 533 ctl_complete_beio(beio); 534 } else { 535 #ifdef CTL_TIME_IO 536 getbintime(&io->io_hdr.dma_start_bt); 537 #endif 538 ctl_datamove(io); 539 } 540 } 541 542 static void 543 ctl_be_block_flush_file(struct ctl_be_block_lun *be_lun, 544 struct ctl_be_block_io *beio) 545 { 546 union ctl_io *io = beio->io; 547 struct mount *mountpoint; 548 int error, lock_flags; 549 550 DPRINTF("entered\n"); 551 552 binuptime(&beio->ds_t0); 553 mtx_lock(&be_lun->io_lock); 554 devstat_start_transaction(beio->lun->disk_stats, &beio->ds_t0); 555 mtx_unlock(&be_lun->io_lock); 556 557 (void) vn_start_write(be_lun->vn, &mountpoint, V_WAIT); 558 559 if (MNT_SHARED_WRITES(mountpoint) 560 || ((mountpoint == NULL) 561 && MNT_SHARED_WRITES(be_lun->vn->v_mount))) 562 lock_flags = LK_SHARED; 563 else 564 lock_flags = LK_EXCLUSIVE; 565 566 vn_lock(be_lun->vn, lock_flags | LK_RETRY); 567 568 error = VOP_FSYNC(be_lun->vn, MNT_WAIT, curthread); 569 VOP_UNLOCK(be_lun->vn, 0); 570 571 vn_finished_write(mountpoint); 572 573 mtx_lock(&be_lun->io_lock); 574 devstat_end_transaction(beio->lun->disk_stats, beio->io_len, 575 beio->ds_tag_type, beio->ds_trans_type, 576 /*now*/ NULL, /*then*/&beio->ds_t0); 577 mtx_unlock(&be_lun->io_lock); 578 579 if (error == 0) 580 ctl_set_success(&io->scsiio); 581 else { 582 /* XXX KDM is there is a better error here? */ 583 ctl_set_internal_failure(&io->scsiio, 584 /*sks_valid*/ 1, 585 /*retry_count*/ 0xbad1); 586 } 587 588 ctl_complete_beio(beio); 589 } 590 591 SDT_PROBE_DEFINE1(cbb, kernel, read, file_start, "uint64_t"); 592 SDT_PROBE_DEFINE1(cbb, kernel, write, file_start, "uint64_t"); 593 SDT_PROBE_DEFINE1(cbb, kernel, read, file_done,"uint64_t"); 594 SDT_PROBE_DEFINE1(cbb, kernel, write, file_done, "uint64_t"); 595 596 static void 597 ctl_be_block_dispatch_file(struct ctl_be_block_lun *be_lun, 598 struct ctl_be_block_io *beio) 599 { 600 struct ctl_be_block_filedata *file_data; 601 union ctl_io *io; 602 struct uio xuio; 603 struct iovec *xiovec; 604 int flags; 605 int error, i; 606 607 DPRINTF("entered\n"); 608 609 file_data = &be_lun->backend.file; 610 io = beio->io; 611 flags = 0; 612 if (ARGS(io)->flags & CTL_LLF_DPO) 613 flags |= IO_DIRECT; 614 if (beio->bio_cmd == BIO_WRITE && ARGS(io)->flags & CTL_LLF_FUA) 615 flags |= IO_SYNC; 616 617 bzero(&xuio, sizeof(xuio)); 618 if (beio->bio_cmd == BIO_READ) { 619 SDT_PROBE(cbb, kernel, read, file_start, 0, 0, 0, 0, 0); 620 xuio.uio_rw = UIO_READ; 621 } else { 622 SDT_PROBE(cbb, kernel, write, file_start, 0, 0, 0, 0, 0); 623 xuio.uio_rw = UIO_WRITE; 624 } 625 xuio.uio_offset = beio->io_offset; 626 xuio.uio_resid = beio->io_len; 627 xuio.uio_segflg = UIO_SYSSPACE; 628 xuio.uio_iov = beio->xiovecs; 629 xuio.uio_iovcnt = beio->num_segs; 630 xuio.uio_td = curthread; 631 632 for (i = 0, xiovec = xuio.uio_iov; i < xuio.uio_iovcnt; i++, xiovec++) { 633 xiovec->iov_base = beio->sg_segs[i].addr; 634 xiovec->iov_len = beio->sg_segs[i].len; 635 } 636 637 binuptime(&beio->ds_t0); 638 mtx_lock(&be_lun->io_lock); 639 devstat_start_transaction(beio->lun->disk_stats, &beio->ds_t0); 640 mtx_unlock(&be_lun->io_lock); 641 642 if (beio->bio_cmd == BIO_READ) { 643 vn_lock(be_lun->vn, LK_SHARED | LK_RETRY); 644 645 /* 646 * UFS pays attention to IO_DIRECT for reads. If the 647 * DIRECTIO option is configured into the kernel, it calls 648 * ffs_rawread(). But that only works for single-segment 649 * uios with user space addresses. In our case, with a 650 * kernel uio, it still reads into the buffer cache, but it 651 * will just try to release the buffer from the cache later 652 * on in ffs_read(). 653 * 654 * ZFS does not pay attention to IO_DIRECT for reads. 655 * 656 * UFS does not pay attention to IO_SYNC for reads. 657 * 658 * ZFS pays attention to IO_SYNC (which translates into the 659 * Solaris define FRSYNC for zfs_read()) for reads. It 660 * attempts to sync the file before reading. 661 * 662 * So, to attempt to provide some barrier semantics in the 663 * BIO_ORDERED case, set both IO_DIRECT and IO_SYNC. 664 */ 665 error = VOP_READ(be_lun->vn, &xuio, flags, file_data->cred); 666 667 VOP_UNLOCK(be_lun->vn, 0); 668 SDT_PROBE(cbb, kernel, read, file_done, 0, 0, 0, 0, 0); 669 } else { 670 struct mount *mountpoint; 671 int lock_flags; 672 673 (void)vn_start_write(be_lun->vn, &mountpoint, V_WAIT); 674 675 if (MNT_SHARED_WRITES(mountpoint) 676 || ((mountpoint == NULL) 677 && MNT_SHARED_WRITES(be_lun->vn->v_mount))) 678 lock_flags = LK_SHARED; 679 else 680 lock_flags = LK_EXCLUSIVE; 681 682 vn_lock(be_lun->vn, lock_flags | LK_RETRY); 683 684 /* 685 * UFS pays attention to IO_DIRECT for writes. The write 686 * is done asynchronously. (Normally the write would just 687 * get put into cache. 688 * 689 * UFS pays attention to IO_SYNC for writes. It will 690 * attempt to write the buffer out synchronously if that 691 * flag is set. 692 * 693 * ZFS does not pay attention to IO_DIRECT for writes. 694 * 695 * ZFS pays attention to IO_SYNC (a.k.a. FSYNC or FRSYNC) 696 * for writes. It will flush the transaction from the 697 * cache before returning. 698 * 699 * So if we've got the BIO_ORDERED flag set, we want 700 * IO_SYNC in either the UFS or ZFS case. 701 */ 702 error = VOP_WRITE(be_lun->vn, &xuio, flags, file_data->cred); 703 VOP_UNLOCK(be_lun->vn, 0); 704 705 vn_finished_write(mountpoint); 706 SDT_PROBE(cbb, kernel, write, file_done, 0, 0, 0, 0, 0); 707 } 708 709 mtx_lock(&be_lun->io_lock); 710 devstat_end_transaction(beio->lun->disk_stats, beio->io_len, 711 beio->ds_tag_type, beio->ds_trans_type, 712 /*now*/ NULL, /*then*/&beio->ds_t0); 713 mtx_unlock(&be_lun->io_lock); 714 715 /* 716 * If we got an error, set the sense data to "MEDIUM ERROR" and 717 * return the I/O to the user. 718 */ 719 if (error != 0) { 720 char path_str[32]; 721 722 ctl_scsi_path_string(io, path_str, sizeof(path_str)); 723 printf("%s%s command returned errno %d\n", path_str, 724 (beio->bio_cmd == BIO_READ) ? "READ" : "WRITE", error); 725 if (error == ENOSPC) { 726 ctl_set_space_alloc_fail(&io->scsiio); 727 } else 728 ctl_set_medium_error(&io->scsiio); 729 ctl_complete_beio(beio); 730 return; 731 } 732 733 /* 734 * If this is a write or a verify, we're all done. 735 * If this is a read, we can now send the data to the user. 736 */ 737 if ((beio->bio_cmd == BIO_WRITE) || 738 (ARGS(io)->flags & CTL_LLF_VERIFY)) { 739 ctl_set_success(&io->scsiio); 740 ctl_complete_beio(beio); 741 } else { 742 #ifdef CTL_TIME_IO 743 getbintime(&io->io_hdr.dma_start_bt); 744 #endif 745 ctl_datamove(io); 746 } 747 } 748 749 static void 750 ctl_be_block_dispatch_zvol(struct ctl_be_block_lun *be_lun, 751 struct ctl_be_block_io *beio) 752 { 753 struct ctl_be_block_devdata *dev_data; 754 union ctl_io *io; 755 struct uio xuio; 756 struct iovec *xiovec; 757 int flags; 758 int error, i; 759 760 DPRINTF("entered\n"); 761 762 dev_data = &be_lun->backend.dev; 763 io = beio->io; 764 flags = 0; 765 if (ARGS(io)->flags & CTL_LLF_DPO) 766 flags |= IO_DIRECT; 767 if (beio->bio_cmd == BIO_WRITE && ARGS(io)->flags & CTL_LLF_FUA) 768 flags |= IO_SYNC; 769 770 bzero(&xuio, sizeof(xuio)); 771 if (beio->bio_cmd == BIO_READ) { 772 SDT_PROBE(cbb, kernel, read, file_start, 0, 0, 0, 0, 0); 773 xuio.uio_rw = UIO_READ; 774 } else { 775 SDT_PROBE(cbb, kernel, write, file_start, 0, 0, 0, 0, 0); 776 xuio.uio_rw = UIO_WRITE; 777 } 778 xuio.uio_offset = beio->io_offset; 779 xuio.uio_resid = beio->io_len; 780 xuio.uio_segflg = UIO_SYSSPACE; 781 xuio.uio_iov = beio->xiovecs; 782 xuio.uio_iovcnt = beio->num_segs; 783 xuio.uio_td = curthread; 784 785 for (i = 0, xiovec = xuio.uio_iov; i < xuio.uio_iovcnt; i++, xiovec++) { 786 xiovec->iov_base = beio->sg_segs[i].addr; 787 xiovec->iov_len = beio->sg_segs[i].len; 788 } 789 790 binuptime(&beio->ds_t0); 791 mtx_lock(&be_lun->io_lock); 792 devstat_start_transaction(beio->lun->disk_stats, &beio->ds_t0); 793 mtx_unlock(&be_lun->io_lock); 794 795 if (beio->bio_cmd == BIO_READ) { 796 error = (*dev_data->csw->d_read)(dev_data->cdev, &xuio, flags); 797 SDT_PROBE(cbb, kernel, read, file_done, 0, 0, 0, 0, 0); 798 } else { 799 error = (*dev_data->csw->d_write)(dev_data->cdev, &xuio, flags); 800 SDT_PROBE(cbb, kernel, write, file_done, 0, 0, 0, 0, 0); 801 } 802 803 mtx_lock(&be_lun->io_lock); 804 devstat_end_transaction(beio->lun->disk_stats, beio->io_len, 805 beio->ds_tag_type, beio->ds_trans_type, 806 /*now*/ NULL, /*then*/&beio->ds_t0); 807 mtx_unlock(&be_lun->io_lock); 808 809 /* 810 * If we got an error, set the sense data to "MEDIUM ERROR" and 811 * return the I/O to the user. 812 */ 813 if (error != 0) { 814 if (error == ENOSPC) { 815 ctl_set_space_alloc_fail(&io->scsiio); 816 } else 817 ctl_set_medium_error(&io->scsiio); 818 ctl_complete_beio(beio); 819 return; 820 } 821 822 /* 823 * If this is a write or a verify, we're all done. 824 * If this is a read, we can now send the data to the user. 825 */ 826 if ((beio->bio_cmd == BIO_WRITE) || 827 (ARGS(io)->flags & CTL_LLF_VERIFY)) { 828 ctl_set_success(&io->scsiio); 829 ctl_complete_beio(beio); 830 } else { 831 #ifdef CTL_TIME_IO 832 getbintime(&io->io_hdr.dma_start_bt); 833 #endif 834 ctl_datamove(io); 835 } 836 } 837 838 static void 839 ctl_be_block_flush_dev(struct ctl_be_block_lun *be_lun, 840 struct ctl_be_block_io *beio) 841 { 842 struct bio *bio; 843 union ctl_io *io; 844 struct ctl_be_block_devdata *dev_data; 845 846 dev_data = &be_lun->backend.dev; 847 io = beio->io; 848 849 DPRINTF("entered\n"); 850 851 /* This can't fail, it's a blocking allocation. */ 852 bio = g_alloc_bio(); 853 854 bio->bio_cmd = BIO_FLUSH; 855 bio->bio_flags |= BIO_ORDERED; 856 bio->bio_dev = dev_data->cdev; 857 bio->bio_offset = 0; 858 bio->bio_data = 0; 859 bio->bio_done = ctl_be_block_biodone; 860 bio->bio_caller1 = beio; 861 bio->bio_pblkno = 0; 862 863 /* 864 * We don't need to acquire the LUN lock here, because we are only 865 * sending one bio, and so there is no other context to synchronize 866 * with. 867 */ 868 beio->num_bios_sent = 1; 869 beio->send_complete = 1; 870 871 binuptime(&beio->ds_t0); 872 mtx_lock(&be_lun->io_lock); 873 devstat_start_transaction(be_lun->disk_stats, &beio->ds_t0); 874 mtx_unlock(&be_lun->io_lock); 875 876 (*dev_data->csw->d_strategy)(bio); 877 } 878 879 static void 880 ctl_be_block_unmap_dev_range(struct ctl_be_block_lun *be_lun, 881 struct ctl_be_block_io *beio, 882 uint64_t off, uint64_t len, int last) 883 { 884 struct bio *bio; 885 struct ctl_be_block_devdata *dev_data; 886 uint64_t maxlen; 887 888 dev_data = &be_lun->backend.dev; 889 maxlen = LONG_MAX - (LONG_MAX % be_lun->blocksize); 890 while (len > 0) { 891 bio = g_alloc_bio(); 892 bio->bio_cmd = BIO_DELETE; 893 bio->bio_dev = dev_data->cdev; 894 bio->bio_offset = off; 895 bio->bio_length = MIN(len, maxlen); 896 bio->bio_data = 0; 897 bio->bio_done = ctl_be_block_biodone; 898 bio->bio_caller1 = beio; 899 bio->bio_pblkno = off / be_lun->blocksize; 900 901 off += bio->bio_length; 902 len -= bio->bio_length; 903 904 mtx_lock(&be_lun->io_lock); 905 beio->num_bios_sent++; 906 if (last && len == 0) 907 beio->send_complete = 1; 908 mtx_unlock(&be_lun->io_lock); 909 910 (*dev_data->csw->d_strategy)(bio); 911 } 912 } 913 914 static void 915 ctl_be_block_unmap_dev(struct ctl_be_block_lun *be_lun, 916 struct ctl_be_block_io *beio) 917 { 918 union ctl_io *io; 919 struct ctl_be_block_devdata *dev_data; 920 struct ctl_ptr_len_flags *ptrlen; 921 struct scsi_unmap_desc *buf, *end; 922 uint64_t len; 923 924 dev_data = &be_lun->backend.dev; 925 io = beio->io; 926 927 DPRINTF("entered\n"); 928 929 binuptime(&beio->ds_t0); 930 mtx_lock(&be_lun->io_lock); 931 devstat_start_transaction(be_lun->disk_stats, &beio->ds_t0); 932 mtx_unlock(&be_lun->io_lock); 933 934 if (beio->io_offset == -1) { 935 beio->io_len = 0; 936 ptrlen = (struct ctl_ptr_len_flags *)&io->io_hdr.ctl_private[CTL_PRIV_LBA_LEN]; 937 buf = (struct scsi_unmap_desc *)ptrlen->ptr; 938 end = buf + ptrlen->len / sizeof(*buf); 939 for (; buf < end; buf++) { 940 len = (uint64_t)scsi_4btoul(buf->length) * 941 be_lun->blocksize; 942 beio->io_len += len; 943 ctl_be_block_unmap_dev_range(be_lun, beio, 944 scsi_8btou64(buf->lba) * be_lun->blocksize, len, 945 (end - buf < 2) ? TRUE : FALSE); 946 } 947 } else 948 ctl_be_block_unmap_dev_range(be_lun, beio, 949 beio->io_offset, beio->io_len, TRUE); 950 } 951 952 static void 953 ctl_be_block_dispatch_dev(struct ctl_be_block_lun *be_lun, 954 struct ctl_be_block_io *beio) 955 { 956 TAILQ_HEAD(, bio) queue = TAILQ_HEAD_INITIALIZER(queue); 957 int i; 958 struct bio *bio; 959 struct ctl_be_block_devdata *dev_data; 960 off_t cur_offset; 961 int max_iosize; 962 963 DPRINTF("entered\n"); 964 965 dev_data = &be_lun->backend.dev; 966 967 /* 968 * We have to limit our I/O size to the maximum supported by the 969 * backend device. Hopefully it is MAXPHYS. If the driver doesn't 970 * set it properly, use DFLTPHYS. 971 */ 972 max_iosize = dev_data->cdev->si_iosize_max; 973 if (max_iosize < PAGE_SIZE) 974 max_iosize = DFLTPHYS; 975 976 cur_offset = beio->io_offset; 977 for (i = 0; i < beio->num_segs; i++) { 978 size_t cur_size; 979 uint8_t *cur_ptr; 980 981 cur_size = beio->sg_segs[i].len; 982 cur_ptr = beio->sg_segs[i].addr; 983 984 while (cur_size > 0) { 985 /* This can't fail, it's a blocking allocation. */ 986 bio = g_alloc_bio(); 987 988 KASSERT(bio != NULL, ("g_alloc_bio() failed!\n")); 989 990 bio->bio_cmd = beio->bio_cmd; 991 bio->bio_dev = dev_data->cdev; 992 bio->bio_caller1 = beio; 993 bio->bio_length = min(cur_size, max_iosize); 994 bio->bio_offset = cur_offset; 995 bio->bio_data = cur_ptr; 996 bio->bio_done = ctl_be_block_biodone; 997 bio->bio_pblkno = cur_offset / be_lun->blocksize; 998 999 cur_offset += bio->bio_length; 1000 cur_ptr += bio->bio_length; 1001 cur_size -= bio->bio_length; 1002 1003 TAILQ_INSERT_TAIL(&queue, bio, bio_queue); 1004 beio->num_bios_sent++; 1005 } 1006 } 1007 binuptime(&beio->ds_t0); 1008 mtx_lock(&be_lun->io_lock); 1009 devstat_start_transaction(be_lun->disk_stats, &beio->ds_t0); 1010 beio->send_complete = 1; 1011 mtx_unlock(&be_lun->io_lock); 1012 1013 /* 1014 * Fire off all allocated requests! 1015 */ 1016 while ((bio = TAILQ_FIRST(&queue)) != NULL) { 1017 TAILQ_REMOVE(&queue, bio, bio_queue); 1018 (*dev_data->csw->d_strategy)(bio); 1019 } 1020 } 1021 1022 static uint64_t 1023 ctl_be_block_getattr_dev(struct ctl_be_block_lun *be_lun, const char *attrname) 1024 { 1025 struct ctl_be_block_devdata *dev_data = &be_lun->backend.dev; 1026 struct diocgattr_arg arg; 1027 int error; 1028 1029 if (dev_data->csw == NULL || dev_data->csw->d_ioctl == NULL) 1030 return (UINT64_MAX); 1031 strlcpy(arg.name, attrname, sizeof(arg.name)); 1032 arg.len = sizeof(arg.value.off); 1033 error = dev_data->csw->d_ioctl(dev_data->cdev, 1034 DIOCGATTR, (caddr_t)&arg, FREAD, curthread); 1035 if (error != 0) 1036 return (UINT64_MAX); 1037 return (arg.value.off); 1038 } 1039 1040 static void 1041 ctl_be_block_cw_done_ws(struct ctl_be_block_io *beio) 1042 { 1043 union ctl_io *io; 1044 1045 io = beio->io; 1046 ctl_free_beio(beio); 1047 if ((io->io_hdr.flags & CTL_FLAG_ABORT) || 1048 ((io->io_hdr.status & CTL_STATUS_MASK) != CTL_STATUS_NONE && 1049 (io->io_hdr.status & CTL_STATUS_MASK) != CTL_SUCCESS)) { 1050 ctl_config_write_done(io); 1051 return; 1052 } 1053 1054 ctl_be_block_config_write(io); 1055 } 1056 1057 static void 1058 ctl_be_block_cw_dispatch_ws(struct ctl_be_block_lun *be_lun, 1059 union ctl_io *io) 1060 { 1061 struct ctl_be_block_io *beio; 1062 struct ctl_be_block_softc *softc; 1063 struct ctl_lba_len_flags *lbalen; 1064 uint64_t len_left, lba; 1065 int i, seglen; 1066 uint8_t *buf, *end; 1067 1068 DPRINTF("entered\n"); 1069 1070 beio = (struct ctl_be_block_io *)PRIV(io)->ptr; 1071 softc = be_lun->softc; 1072 lbalen = ARGS(beio->io); 1073 1074 if (lbalen->flags & ~(SWS_LBDATA | SWS_UNMAP | SWS_ANCHOR | SWS_NDOB) || 1075 (lbalen->flags & (SWS_UNMAP | SWS_ANCHOR) && be_lun->unmap == NULL)) { 1076 ctl_free_beio(beio); 1077 ctl_set_invalid_field(&io->scsiio, 1078 /*sks_valid*/ 1, 1079 /*command*/ 1, 1080 /*field*/ 1, 1081 /*bit_valid*/ 0, 1082 /*bit*/ 0); 1083 ctl_config_write_done(io); 1084 return; 1085 } 1086 1087 switch (io->scsiio.tag_type) { 1088 case CTL_TAG_ORDERED: 1089 beio->ds_tag_type = DEVSTAT_TAG_ORDERED; 1090 break; 1091 case CTL_TAG_HEAD_OF_QUEUE: 1092 beio->ds_tag_type = DEVSTAT_TAG_HEAD; 1093 break; 1094 case CTL_TAG_UNTAGGED: 1095 case CTL_TAG_SIMPLE: 1096 case CTL_TAG_ACA: 1097 default: 1098 beio->ds_tag_type = DEVSTAT_TAG_SIMPLE; 1099 break; 1100 } 1101 1102 if (lbalen->flags & (SWS_UNMAP | SWS_ANCHOR)) { 1103 beio->io_offset = lbalen->lba * be_lun->blocksize; 1104 beio->io_len = (uint64_t)lbalen->len * be_lun->blocksize; 1105 beio->bio_cmd = BIO_DELETE; 1106 beio->ds_trans_type = DEVSTAT_FREE; 1107 1108 be_lun->unmap(be_lun, beio); 1109 return; 1110 } 1111 1112 beio->bio_cmd = BIO_WRITE; 1113 beio->ds_trans_type = DEVSTAT_WRITE; 1114 1115 DPRINTF("WRITE SAME at LBA %jx len %u\n", 1116 (uintmax_t)lbalen->lba, lbalen->len); 1117 1118 len_left = (uint64_t)lbalen->len * be_lun->blocksize; 1119 for (i = 0, lba = 0; i < CTLBLK_MAX_SEGS && len_left > 0; i++) { 1120 1121 /* 1122 * Setup the S/G entry for this chunk. 1123 */ 1124 seglen = MIN(CTLBLK_MAX_SEG, len_left); 1125 seglen -= seglen % be_lun->blocksize; 1126 beio->sg_segs[i].len = seglen; 1127 beio->sg_segs[i].addr = uma_zalloc(be_lun->lun_zone, M_WAITOK); 1128 1129 DPRINTF("segment %d addr %p len %zd\n", i, 1130 beio->sg_segs[i].addr, beio->sg_segs[i].len); 1131 1132 beio->num_segs++; 1133 len_left -= seglen; 1134 1135 buf = beio->sg_segs[i].addr; 1136 end = buf + seglen; 1137 for (; buf < end; buf += be_lun->blocksize) { 1138 memcpy(buf, io->scsiio.kern_data_ptr, be_lun->blocksize); 1139 if (lbalen->flags & SWS_LBDATA) 1140 scsi_ulto4b(lbalen->lba + lba, buf); 1141 lba++; 1142 } 1143 } 1144 1145 beio->io_offset = lbalen->lba * be_lun->blocksize; 1146 beio->io_len = lba * be_lun->blocksize; 1147 1148 /* We can not do all in one run. Correct and schedule rerun. */ 1149 if (len_left > 0) { 1150 lbalen->lba += lba; 1151 lbalen->len -= lba; 1152 beio->beio_cont = ctl_be_block_cw_done_ws; 1153 } 1154 1155 be_lun->dispatch(be_lun, beio); 1156 } 1157 1158 static void 1159 ctl_be_block_cw_dispatch_unmap(struct ctl_be_block_lun *be_lun, 1160 union ctl_io *io) 1161 { 1162 struct ctl_be_block_io *beio; 1163 struct ctl_be_block_softc *softc; 1164 struct ctl_ptr_len_flags *ptrlen; 1165 1166 DPRINTF("entered\n"); 1167 1168 beio = (struct ctl_be_block_io *)PRIV(io)->ptr; 1169 softc = be_lun->softc; 1170 ptrlen = (struct ctl_ptr_len_flags *)&io->io_hdr.ctl_private[CTL_PRIV_LBA_LEN]; 1171 1172 if ((ptrlen->flags & ~SU_ANCHOR) != 0 || be_lun->unmap == NULL) { 1173 ctl_free_beio(beio); 1174 ctl_set_invalid_field(&io->scsiio, 1175 /*sks_valid*/ 0, 1176 /*command*/ 1, 1177 /*field*/ 0, 1178 /*bit_valid*/ 0, 1179 /*bit*/ 0); 1180 ctl_config_write_done(io); 1181 return; 1182 } 1183 1184 switch (io->scsiio.tag_type) { 1185 case CTL_TAG_ORDERED: 1186 beio->ds_tag_type = DEVSTAT_TAG_ORDERED; 1187 break; 1188 case CTL_TAG_HEAD_OF_QUEUE: 1189 beio->ds_tag_type = DEVSTAT_TAG_HEAD; 1190 break; 1191 case CTL_TAG_UNTAGGED: 1192 case CTL_TAG_SIMPLE: 1193 case CTL_TAG_ACA: 1194 default: 1195 beio->ds_tag_type = DEVSTAT_TAG_SIMPLE; 1196 break; 1197 } 1198 1199 beio->io_len = 0; 1200 beio->io_offset = -1; 1201 1202 beio->bio_cmd = BIO_DELETE; 1203 beio->ds_trans_type = DEVSTAT_FREE; 1204 1205 DPRINTF("UNMAP\n"); 1206 1207 be_lun->unmap(be_lun, beio); 1208 } 1209 1210 static void 1211 ctl_be_block_cw_done(struct ctl_be_block_io *beio) 1212 { 1213 union ctl_io *io; 1214 1215 io = beio->io; 1216 ctl_free_beio(beio); 1217 ctl_config_write_done(io); 1218 } 1219 1220 static void 1221 ctl_be_block_cw_dispatch(struct ctl_be_block_lun *be_lun, 1222 union ctl_io *io) 1223 { 1224 struct ctl_be_block_io *beio; 1225 struct ctl_be_block_softc *softc; 1226 1227 DPRINTF("entered\n"); 1228 1229 softc = be_lun->softc; 1230 beio = ctl_alloc_beio(softc); 1231 beio->io = io; 1232 beio->lun = be_lun; 1233 beio->beio_cont = ctl_be_block_cw_done; 1234 PRIV(io)->ptr = (void *)beio; 1235 1236 switch (io->scsiio.cdb[0]) { 1237 case SYNCHRONIZE_CACHE: 1238 case SYNCHRONIZE_CACHE_16: 1239 beio->bio_cmd = BIO_FLUSH; 1240 beio->ds_trans_type = DEVSTAT_NO_DATA; 1241 beio->ds_tag_type = DEVSTAT_TAG_ORDERED; 1242 beio->io_len = 0; 1243 be_lun->lun_flush(be_lun, beio); 1244 break; 1245 case WRITE_SAME_10: 1246 case WRITE_SAME_16: 1247 ctl_be_block_cw_dispatch_ws(be_lun, io); 1248 break; 1249 case UNMAP: 1250 ctl_be_block_cw_dispatch_unmap(be_lun, io); 1251 break; 1252 default: 1253 panic("Unhandled CDB type %#x", io->scsiio.cdb[0]); 1254 break; 1255 } 1256 } 1257 1258 SDT_PROBE_DEFINE1(cbb, kernel, read, start, "uint64_t"); 1259 SDT_PROBE_DEFINE1(cbb, kernel, write, start, "uint64_t"); 1260 SDT_PROBE_DEFINE1(cbb, kernel, read, alloc_done, "uint64_t"); 1261 SDT_PROBE_DEFINE1(cbb, kernel, write, alloc_done, "uint64_t"); 1262 1263 static void 1264 ctl_be_block_next(struct ctl_be_block_io *beio) 1265 { 1266 struct ctl_be_block_lun *be_lun; 1267 union ctl_io *io; 1268 1269 io = beio->io; 1270 be_lun = beio->lun; 1271 ctl_free_beio(beio); 1272 if ((io->io_hdr.flags & CTL_FLAG_ABORT) || 1273 ((io->io_hdr.status & CTL_STATUS_MASK) != CTL_STATUS_NONE && 1274 (io->io_hdr.status & CTL_STATUS_MASK) != CTL_SUCCESS)) { 1275 ctl_data_submit_done(io); 1276 return; 1277 } 1278 1279 io->io_hdr.status &= ~CTL_STATUS_MASK; 1280 io->io_hdr.status |= CTL_STATUS_NONE; 1281 1282 mtx_lock(&be_lun->queue_lock); 1283 /* 1284 * XXX KDM make sure that links is okay to use at this point. 1285 * Otherwise, we either need to add another field to ctl_io_hdr, 1286 * or deal with resource allocation here. 1287 */ 1288 STAILQ_INSERT_TAIL(&be_lun->input_queue, &io->io_hdr, links); 1289 mtx_unlock(&be_lun->queue_lock); 1290 1291 taskqueue_enqueue(be_lun->io_taskqueue, &be_lun->io_task); 1292 } 1293 1294 static void 1295 ctl_be_block_dispatch(struct ctl_be_block_lun *be_lun, 1296 union ctl_io *io) 1297 { 1298 struct ctl_be_block_io *beio; 1299 struct ctl_be_block_softc *softc; 1300 struct ctl_lba_len_flags *lbalen; 1301 struct ctl_ptr_len_flags *bptrlen; 1302 uint64_t len_left, lbas; 1303 int i; 1304 1305 softc = be_lun->softc; 1306 1307 DPRINTF("entered\n"); 1308 1309 lbalen = ARGS(io); 1310 if (lbalen->flags & CTL_LLF_WRITE) { 1311 SDT_PROBE(cbb, kernel, write, start, 0, 0, 0, 0, 0); 1312 } else { 1313 SDT_PROBE(cbb, kernel, read, start, 0, 0, 0, 0, 0); 1314 } 1315 1316 beio = ctl_alloc_beio(softc); 1317 beio->io = io; 1318 beio->lun = be_lun; 1319 bptrlen = PRIV(io); 1320 bptrlen->ptr = (void *)beio; 1321 1322 switch (io->scsiio.tag_type) { 1323 case CTL_TAG_ORDERED: 1324 beio->ds_tag_type = DEVSTAT_TAG_ORDERED; 1325 break; 1326 case CTL_TAG_HEAD_OF_QUEUE: 1327 beio->ds_tag_type = DEVSTAT_TAG_HEAD; 1328 break; 1329 case CTL_TAG_UNTAGGED: 1330 case CTL_TAG_SIMPLE: 1331 case CTL_TAG_ACA: 1332 default: 1333 beio->ds_tag_type = DEVSTAT_TAG_SIMPLE; 1334 break; 1335 } 1336 1337 if (lbalen->flags & CTL_LLF_WRITE) { 1338 beio->bio_cmd = BIO_WRITE; 1339 beio->ds_trans_type = DEVSTAT_WRITE; 1340 } else { 1341 beio->bio_cmd = BIO_READ; 1342 beio->ds_trans_type = DEVSTAT_READ; 1343 } 1344 1345 DPRINTF("%s at LBA %jx len %u @%ju\n", 1346 (beio->bio_cmd == BIO_READ) ? "READ" : "WRITE", 1347 (uintmax_t)lbalen->lba, lbalen->len, bptrlen->len); 1348 if (lbalen->flags & CTL_LLF_COMPARE) 1349 lbas = CTLBLK_HALF_IO_SIZE; 1350 else 1351 lbas = CTLBLK_MAX_IO_SIZE; 1352 lbas = MIN(lbalen->len - bptrlen->len, lbas / be_lun->blocksize); 1353 beio->io_offset = (lbalen->lba + bptrlen->len) * be_lun->blocksize; 1354 beio->io_len = lbas * be_lun->blocksize; 1355 bptrlen->len += lbas; 1356 1357 for (i = 0, len_left = beio->io_len; len_left > 0; i++) { 1358 KASSERT(i < CTLBLK_MAX_SEGS, ("Too many segs (%d >= %d)", 1359 i, CTLBLK_MAX_SEGS)); 1360 1361 /* 1362 * Setup the S/G entry for this chunk. 1363 */ 1364 beio->sg_segs[i].len = min(CTLBLK_MAX_SEG, len_left); 1365 beio->sg_segs[i].addr = uma_zalloc(be_lun->lun_zone, M_WAITOK); 1366 1367 DPRINTF("segment %d addr %p len %zd\n", i, 1368 beio->sg_segs[i].addr, beio->sg_segs[i].len); 1369 1370 /* Set up second segment for compare operation. */ 1371 if (lbalen->flags & CTL_LLF_COMPARE) { 1372 beio->sg_segs[i + CTLBLK_HALF_SEGS].len = 1373 beio->sg_segs[i].len; 1374 beio->sg_segs[i + CTLBLK_HALF_SEGS].addr = 1375 uma_zalloc(be_lun->lun_zone, M_WAITOK); 1376 } 1377 1378 beio->num_segs++; 1379 len_left -= beio->sg_segs[i].len; 1380 } 1381 if (bptrlen->len < lbalen->len) 1382 beio->beio_cont = ctl_be_block_next; 1383 io->scsiio.be_move_done = ctl_be_block_move_done; 1384 /* For compare we have separate S/G lists for read and datamove. */ 1385 if (lbalen->flags & CTL_LLF_COMPARE) 1386 io->scsiio.kern_data_ptr = (uint8_t *)&beio->sg_segs[CTLBLK_HALF_SEGS]; 1387 else 1388 io->scsiio.kern_data_ptr = (uint8_t *)beio->sg_segs; 1389 io->scsiio.kern_data_len = beio->io_len; 1390 io->scsiio.kern_data_resid = 0; 1391 io->scsiio.kern_sg_entries = beio->num_segs; 1392 io->io_hdr.flags |= CTL_FLAG_ALLOCATED | CTL_FLAG_KDPTR_SGLIST; 1393 1394 /* 1395 * For the read case, we need to read the data into our buffers and 1396 * then we can send it back to the user. For the write case, we 1397 * need to get the data from the user first. 1398 */ 1399 if (beio->bio_cmd == BIO_READ) { 1400 SDT_PROBE(cbb, kernel, read, alloc_done, 0, 0, 0, 0, 0); 1401 be_lun->dispatch(be_lun, beio); 1402 } else { 1403 SDT_PROBE(cbb, kernel, write, alloc_done, 0, 0, 0, 0, 0); 1404 #ifdef CTL_TIME_IO 1405 getbintime(&io->io_hdr.dma_start_bt); 1406 #endif 1407 ctl_datamove(io); 1408 } 1409 } 1410 1411 static void 1412 ctl_be_block_worker(void *context, int pending) 1413 { 1414 struct ctl_be_block_lun *be_lun; 1415 struct ctl_be_block_softc *softc; 1416 union ctl_io *io; 1417 1418 be_lun = (struct ctl_be_block_lun *)context; 1419 softc = be_lun->softc; 1420 1421 DPRINTF("entered\n"); 1422 1423 mtx_lock(&be_lun->queue_lock); 1424 for (;;) { 1425 io = (union ctl_io *)STAILQ_FIRST(&be_lun->datamove_queue); 1426 if (io != NULL) { 1427 struct ctl_be_block_io *beio; 1428 1429 DPRINTF("datamove queue\n"); 1430 1431 STAILQ_REMOVE(&be_lun->datamove_queue, &io->io_hdr, 1432 ctl_io_hdr, links); 1433 1434 mtx_unlock(&be_lun->queue_lock); 1435 1436 beio = (struct ctl_be_block_io *)PRIV(io)->ptr; 1437 1438 be_lun->dispatch(be_lun, beio); 1439 1440 mtx_lock(&be_lun->queue_lock); 1441 continue; 1442 } 1443 io = (union ctl_io *)STAILQ_FIRST(&be_lun->config_write_queue); 1444 if (io != NULL) { 1445 1446 DPRINTF("config write queue\n"); 1447 1448 STAILQ_REMOVE(&be_lun->config_write_queue, &io->io_hdr, 1449 ctl_io_hdr, links); 1450 1451 mtx_unlock(&be_lun->queue_lock); 1452 1453 ctl_be_block_cw_dispatch(be_lun, io); 1454 1455 mtx_lock(&be_lun->queue_lock); 1456 continue; 1457 } 1458 io = (union ctl_io *)STAILQ_FIRST(&be_lun->input_queue); 1459 if (io != NULL) { 1460 DPRINTF("input queue\n"); 1461 1462 STAILQ_REMOVE(&be_lun->input_queue, &io->io_hdr, 1463 ctl_io_hdr, links); 1464 mtx_unlock(&be_lun->queue_lock); 1465 1466 /* 1467 * We must drop the lock, since this routine and 1468 * its children may sleep. 1469 */ 1470 ctl_be_block_dispatch(be_lun, io); 1471 1472 mtx_lock(&be_lun->queue_lock); 1473 continue; 1474 } 1475 1476 /* 1477 * If we get here, there is no work left in the queues, so 1478 * just break out and let the task queue go to sleep. 1479 */ 1480 break; 1481 } 1482 mtx_unlock(&be_lun->queue_lock); 1483 } 1484 1485 /* 1486 * Entry point from CTL to the backend for I/O. We queue everything to a 1487 * work thread, so this just puts the I/O on a queue and wakes up the 1488 * thread. 1489 */ 1490 static int 1491 ctl_be_block_submit(union ctl_io *io) 1492 { 1493 struct ctl_be_block_lun *be_lun; 1494 struct ctl_be_lun *ctl_be_lun; 1495 1496 DPRINTF("entered\n"); 1497 1498 ctl_be_lun = (struct ctl_be_lun *)io->io_hdr.ctl_private[ 1499 CTL_PRIV_BACKEND_LUN].ptr; 1500 be_lun = (struct ctl_be_block_lun *)ctl_be_lun->be_lun; 1501 1502 /* 1503 * Make sure we only get SCSI I/O. 1504 */ 1505 KASSERT(io->io_hdr.io_type == CTL_IO_SCSI, ("Non-SCSI I/O (type " 1506 "%#x) encountered", io->io_hdr.io_type)); 1507 1508 PRIV(io)->len = 0; 1509 1510 mtx_lock(&be_lun->queue_lock); 1511 /* 1512 * XXX KDM make sure that links is okay to use at this point. 1513 * Otherwise, we either need to add another field to ctl_io_hdr, 1514 * or deal with resource allocation here. 1515 */ 1516 STAILQ_INSERT_TAIL(&be_lun->input_queue, &io->io_hdr, links); 1517 mtx_unlock(&be_lun->queue_lock); 1518 taskqueue_enqueue(be_lun->io_taskqueue, &be_lun->io_task); 1519 1520 return (CTL_RETVAL_COMPLETE); 1521 } 1522 1523 static int 1524 ctl_be_block_ioctl(struct cdev *dev, u_long cmd, caddr_t addr, 1525 int flag, struct thread *td) 1526 { 1527 struct ctl_be_block_softc *softc; 1528 int error; 1529 1530 softc = &backend_block_softc; 1531 1532 error = 0; 1533 1534 switch (cmd) { 1535 case CTL_LUN_REQ: { 1536 struct ctl_lun_req *lun_req; 1537 1538 lun_req = (struct ctl_lun_req *)addr; 1539 1540 switch (lun_req->reqtype) { 1541 case CTL_LUNREQ_CREATE: 1542 error = ctl_be_block_create(softc, lun_req); 1543 break; 1544 case CTL_LUNREQ_RM: 1545 error = ctl_be_block_rm(softc, lun_req); 1546 break; 1547 case CTL_LUNREQ_MODIFY: 1548 error = ctl_be_block_modify(softc, lun_req); 1549 break; 1550 default: 1551 lun_req->status = CTL_LUN_ERROR; 1552 snprintf(lun_req->error_str, sizeof(lun_req->error_str), 1553 "invalid LUN request type %d", 1554 lun_req->reqtype); 1555 break; 1556 } 1557 break; 1558 } 1559 default: 1560 error = ENOTTY; 1561 break; 1562 } 1563 1564 return (error); 1565 } 1566 1567 static int 1568 ctl_be_block_open_file(struct ctl_be_block_lun *be_lun, struct ctl_lun_req *req) 1569 { 1570 struct ctl_be_block_filedata *file_data; 1571 struct ctl_lun_create_params *params; 1572 struct vattr vattr; 1573 off_t pss; 1574 int error; 1575 1576 error = 0; 1577 file_data = &be_lun->backend.file; 1578 params = &be_lun->params; 1579 1580 be_lun->dev_type = CTL_BE_BLOCK_FILE; 1581 be_lun->dispatch = ctl_be_block_dispatch_file; 1582 be_lun->lun_flush = ctl_be_block_flush_file; 1583 1584 error = VOP_GETATTR(be_lun->vn, &vattr, curthread->td_ucred); 1585 if (error != 0) { 1586 snprintf(req->error_str, sizeof(req->error_str), 1587 "error calling VOP_GETATTR() for file %s", 1588 be_lun->dev_path); 1589 return (error); 1590 } 1591 1592 /* 1593 * Verify that we have the ability to upgrade to exclusive 1594 * access on this file so we can trap errors at open instead 1595 * of reporting them during first access. 1596 */ 1597 if (VOP_ISLOCKED(be_lun->vn) != LK_EXCLUSIVE) { 1598 vn_lock(be_lun->vn, LK_UPGRADE | LK_RETRY); 1599 if (be_lun->vn->v_iflag & VI_DOOMED) { 1600 error = EBADF; 1601 snprintf(req->error_str, sizeof(req->error_str), 1602 "error locking file %s", be_lun->dev_path); 1603 return (error); 1604 } 1605 } 1606 1607 1608 file_data->cred = crhold(curthread->td_ucred); 1609 if (params->lun_size_bytes != 0) 1610 be_lun->size_bytes = params->lun_size_bytes; 1611 else 1612 be_lun->size_bytes = vattr.va_size; 1613 /* 1614 * We set the multi thread flag for file operations because all 1615 * filesystems (in theory) are capable of allowing multiple readers 1616 * of a file at once. So we want to get the maximum possible 1617 * concurrency. 1618 */ 1619 be_lun->flags |= CTL_BE_BLOCK_LUN_MULTI_THREAD; 1620 1621 /* 1622 * For files we can use any logical block size. Prefer 512 bytes 1623 * for compatibility reasons. If file's vattr.va_blocksize 1624 * (preferred I/O block size) is bigger and multiple to chosen 1625 * logical block size -- report it as physical block size. 1626 */ 1627 if (params->blocksize_bytes != 0) 1628 be_lun->blocksize = params->blocksize_bytes; 1629 else 1630 be_lun->blocksize = 512; 1631 pss = vattr.va_blocksize / be_lun->blocksize; 1632 if ((pss > 0) && (pss * be_lun->blocksize == vattr.va_blocksize) && 1633 ((pss & (pss - 1)) == 0)) { 1634 be_lun->pblockexp = fls(pss) - 1; 1635 be_lun->pblockoff = 0; 1636 } 1637 1638 /* 1639 * Sanity check. The media size has to be at least one 1640 * sector long. 1641 */ 1642 if (be_lun->size_bytes < be_lun->blocksize) { 1643 error = EINVAL; 1644 snprintf(req->error_str, sizeof(req->error_str), 1645 "file %s size %ju < block size %u", be_lun->dev_path, 1646 (uintmax_t)be_lun->size_bytes, be_lun->blocksize); 1647 } 1648 return (error); 1649 } 1650 1651 static int 1652 ctl_be_block_open_dev(struct ctl_be_block_lun *be_lun, struct ctl_lun_req *req) 1653 { 1654 struct ctl_lun_create_params *params; 1655 struct vattr vattr; 1656 struct cdev *dev; 1657 struct cdevsw *devsw; 1658 int error; 1659 off_t ps, pss, po, pos; 1660 1661 params = &be_lun->params; 1662 1663 be_lun->dev_type = CTL_BE_BLOCK_DEV; 1664 be_lun->backend.dev.cdev = be_lun->vn->v_rdev; 1665 be_lun->backend.dev.csw = dev_refthread(be_lun->backend.dev.cdev, 1666 &be_lun->backend.dev.dev_ref); 1667 if (be_lun->backend.dev.csw == NULL) 1668 panic("Unable to retrieve device switch"); 1669 if (strcmp(be_lun->backend.dev.csw->d_name, "zvol") == 0) 1670 be_lun->dispatch = ctl_be_block_dispatch_zvol; 1671 else 1672 be_lun->dispatch = ctl_be_block_dispatch_dev; 1673 be_lun->lun_flush = ctl_be_block_flush_dev; 1674 be_lun->unmap = ctl_be_block_unmap_dev; 1675 be_lun->getattr = ctl_be_block_getattr_dev; 1676 1677 error = VOP_GETATTR(be_lun->vn, &vattr, NOCRED); 1678 if (error) { 1679 snprintf(req->error_str, sizeof(req->error_str), 1680 "error getting vnode attributes for device %s", 1681 be_lun->dev_path); 1682 return (error); 1683 } 1684 1685 dev = be_lun->vn->v_rdev; 1686 devsw = dev->si_devsw; 1687 if (!devsw->d_ioctl) { 1688 snprintf(req->error_str, sizeof(req->error_str), 1689 "no d_ioctl for device %s!", 1690 be_lun->dev_path); 1691 return (ENODEV); 1692 } 1693 1694 error = devsw->d_ioctl(dev, DIOCGSECTORSIZE, 1695 (caddr_t)&be_lun->blocksize, FREAD, 1696 curthread); 1697 if (error) { 1698 snprintf(req->error_str, sizeof(req->error_str), 1699 "error %d returned for DIOCGSECTORSIZE ioctl " 1700 "on %s!", error, be_lun->dev_path); 1701 return (error); 1702 } 1703 1704 /* 1705 * If the user has asked for a blocksize that is greater than the 1706 * backing device's blocksize, we can do it only if the blocksize 1707 * the user is asking for is an even multiple of the underlying 1708 * device's blocksize. 1709 */ 1710 if ((params->blocksize_bytes != 0) 1711 && (params->blocksize_bytes > be_lun->blocksize)) { 1712 uint32_t bs_multiple, tmp_blocksize; 1713 1714 bs_multiple = params->blocksize_bytes / be_lun->blocksize; 1715 1716 tmp_blocksize = bs_multiple * be_lun->blocksize; 1717 1718 if (tmp_blocksize == params->blocksize_bytes) { 1719 be_lun->blocksize = params->blocksize_bytes; 1720 } else { 1721 snprintf(req->error_str, sizeof(req->error_str), 1722 "requested blocksize %u is not an even " 1723 "multiple of backing device blocksize %u", 1724 params->blocksize_bytes, 1725 be_lun->blocksize); 1726 return (EINVAL); 1727 1728 } 1729 } else if ((params->blocksize_bytes != 0) 1730 && (params->blocksize_bytes != be_lun->blocksize)) { 1731 snprintf(req->error_str, sizeof(req->error_str), 1732 "requested blocksize %u < backing device " 1733 "blocksize %u", params->blocksize_bytes, 1734 be_lun->blocksize); 1735 return (EINVAL); 1736 } 1737 1738 error = devsw->d_ioctl(dev, DIOCGMEDIASIZE, 1739 (caddr_t)&be_lun->size_bytes, FREAD, 1740 curthread); 1741 if (error) { 1742 snprintf(req->error_str, sizeof(req->error_str), 1743 "error %d returned for DIOCGMEDIASIZE " 1744 " ioctl on %s!", error, 1745 be_lun->dev_path); 1746 return (error); 1747 } 1748 1749 if (params->lun_size_bytes != 0) { 1750 if (params->lun_size_bytes > be_lun->size_bytes) { 1751 snprintf(req->error_str, sizeof(req->error_str), 1752 "requested LUN size %ju > backing device " 1753 "size %ju", 1754 (uintmax_t)params->lun_size_bytes, 1755 (uintmax_t)be_lun->size_bytes); 1756 return (EINVAL); 1757 } 1758 1759 be_lun->size_bytes = params->lun_size_bytes; 1760 } 1761 1762 error = devsw->d_ioctl(dev, DIOCGSTRIPESIZE, 1763 (caddr_t)&ps, FREAD, curthread); 1764 if (error) 1765 ps = po = 0; 1766 else { 1767 error = devsw->d_ioctl(dev, DIOCGSTRIPEOFFSET, 1768 (caddr_t)&po, FREAD, curthread); 1769 if (error) 1770 po = 0; 1771 } 1772 pss = ps / be_lun->blocksize; 1773 pos = po / be_lun->blocksize; 1774 if ((pss > 0) && (pss * be_lun->blocksize == ps) && (pss >= pos) && 1775 ((pss & (pss - 1)) == 0) && (pos * be_lun->blocksize == po)) { 1776 be_lun->pblockexp = fls(pss) - 1; 1777 be_lun->pblockoff = (pss - pos) % pss; 1778 } 1779 1780 return (0); 1781 } 1782 1783 static int 1784 ctl_be_block_close(struct ctl_be_block_lun *be_lun) 1785 { 1786 DROP_GIANT(); 1787 if (be_lun->vn) { 1788 int flags = FREAD | FWRITE; 1789 1790 switch (be_lun->dev_type) { 1791 case CTL_BE_BLOCK_DEV: 1792 if (be_lun->backend.dev.csw) { 1793 dev_relthread(be_lun->backend.dev.cdev, 1794 be_lun->backend.dev.dev_ref); 1795 be_lun->backend.dev.csw = NULL; 1796 be_lun->backend.dev.cdev = NULL; 1797 } 1798 break; 1799 case CTL_BE_BLOCK_FILE: 1800 break; 1801 case CTL_BE_BLOCK_NONE: 1802 break; 1803 default: 1804 panic("Unexpected backend type."); 1805 break; 1806 } 1807 1808 (void)vn_close(be_lun->vn, flags, NOCRED, curthread); 1809 be_lun->vn = NULL; 1810 1811 switch (be_lun->dev_type) { 1812 case CTL_BE_BLOCK_DEV: 1813 break; 1814 case CTL_BE_BLOCK_FILE: 1815 if (be_lun->backend.file.cred != NULL) { 1816 crfree(be_lun->backend.file.cred); 1817 be_lun->backend.file.cred = NULL; 1818 } 1819 break; 1820 case CTL_BE_BLOCK_NONE: 1821 break; 1822 default: 1823 panic("Unexpected backend type."); 1824 break; 1825 } 1826 be_lun->dev_type = CTL_BE_BLOCK_NONE; 1827 } 1828 PICKUP_GIANT(); 1829 1830 return (0); 1831 } 1832 1833 static int 1834 ctl_be_block_open(struct ctl_be_block_softc *softc, 1835 struct ctl_be_block_lun *be_lun, struct ctl_lun_req *req) 1836 { 1837 struct nameidata nd; 1838 int flags; 1839 int error; 1840 1841 /* 1842 * XXX KDM allow a read-only option? 1843 */ 1844 flags = FREAD | FWRITE; 1845 error = 0; 1846 1847 if (rootvnode == NULL) { 1848 snprintf(req->error_str, sizeof(req->error_str), 1849 "Root filesystem is not mounted"); 1850 return (1); 1851 } 1852 1853 if (!curthread->td_proc->p_fd->fd_cdir) { 1854 curthread->td_proc->p_fd->fd_cdir = rootvnode; 1855 VREF(rootvnode); 1856 } 1857 if (!curthread->td_proc->p_fd->fd_rdir) { 1858 curthread->td_proc->p_fd->fd_rdir = rootvnode; 1859 VREF(rootvnode); 1860 } 1861 if (!curthread->td_proc->p_fd->fd_jdir) { 1862 curthread->td_proc->p_fd->fd_jdir = rootvnode; 1863 VREF(rootvnode); 1864 } 1865 1866 again: 1867 NDINIT(&nd, LOOKUP, FOLLOW, UIO_SYSSPACE, be_lun->dev_path, curthread); 1868 error = vn_open(&nd, &flags, 0, NULL); 1869 if (error) { 1870 /* 1871 * This is the only reasonable guess we can make as far as 1872 * path if the user doesn't give us a fully qualified path. 1873 * If they want to specify a file, they need to specify the 1874 * full path. 1875 */ 1876 if (be_lun->dev_path[0] != '/') { 1877 char *dev_path = "/dev/"; 1878 char *dev_name; 1879 1880 /* Try adding device path at beginning of name */ 1881 dev_name = malloc(strlen(be_lun->dev_path) 1882 + strlen(dev_path) + 1, 1883 M_CTLBLK, M_WAITOK); 1884 if (dev_name) { 1885 sprintf(dev_name, "%s%s", dev_path, 1886 be_lun->dev_path); 1887 free(be_lun->dev_path, M_CTLBLK); 1888 be_lun->dev_path = dev_name; 1889 goto again; 1890 } 1891 } 1892 snprintf(req->error_str, sizeof(req->error_str), 1893 "error opening %s: %d", be_lun->dev_path, error); 1894 return (error); 1895 } 1896 1897 NDFREE(&nd, NDF_ONLY_PNBUF); 1898 1899 be_lun->vn = nd.ni_vp; 1900 1901 /* We only support disks and files. */ 1902 if (vn_isdisk(be_lun->vn, &error)) { 1903 error = ctl_be_block_open_dev(be_lun, req); 1904 } else if (be_lun->vn->v_type == VREG) { 1905 error = ctl_be_block_open_file(be_lun, req); 1906 } else { 1907 error = EINVAL; 1908 snprintf(req->error_str, sizeof(req->error_str), 1909 "%s is not a disk or plain file", be_lun->dev_path); 1910 } 1911 VOP_UNLOCK(be_lun->vn, 0); 1912 1913 if (error != 0) { 1914 ctl_be_block_close(be_lun); 1915 return (error); 1916 } 1917 1918 be_lun->blocksize_shift = fls(be_lun->blocksize) - 1; 1919 be_lun->size_blocks = be_lun->size_bytes >> be_lun->blocksize_shift; 1920 1921 return (0); 1922 } 1923 1924 static int 1925 ctl_be_block_create(struct ctl_be_block_softc *softc, struct ctl_lun_req *req) 1926 { 1927 struct ctl_be_block_lun *be_lun; 1928 struct ctl_lun_create_params *params; 1929 char num_thread_str[16]; 1930 char tmpstr[32]; 1931 char *value; 1932 int retval, num_threads, unmap; 1933 int tmp_num_threads; 1934 1935 params = &req->reqdata.create; 1936 retval = 0; 1937 req->status = CTL_LUN_OK; 1938 1939 num_threads = cbb_num_threads; 1940 1941 be_lun = malloc(sizeof(*be_lun), M_CTLBLK, M_ZERO | M_WAITOK); 1942 1943 be_lun->params = req->reqdata.create; 1944 be_lun->softc = softc; 1945 STAILQ_INIT(&be_lun->input_queue); 1946 STAILQ_INIT(&be_lun->config_write_queue); 1947 STAILQ_INIT(&be_lun->datamove_queue); 1948 sprintf(be_lun->lunname, "cblk%d", softc->num_luns); 1949 mtx_init(&be_lun->io_lock, "cblk io lock", NULL, MTX_DEF); 1950 mtx_init(&be_lun->queue_lock, "cblk queue lock", NULL, MTX_DEF); 1951 ctl_init_opts(&be_lun->ctl_be_lun.options, 1952 req->num_be_args, req->kern_be_args); 1953 1954 be_lun->lun_zone = uma_zcreate(be_lun->lunname, CTLBLK_MAX_SEG, 1955 NULL, NULL, NULL, NULL, /*align*/ 0, /*flags*/0); 1956 1957 if (be_lun->lun_zone == NULL) { 1958 snprintf(req->error_str, sizeof(req->error_str), 1959 "error allocating UMA zone"); 1960 goto bailout_error; 1961 } 1962 1963 if (params->flags & CTL_LUN_FLAG_DEV_TYPE) 1964 be_lun->ctl_be_lun.lun_type = params->device_type; 1965 else 1966 be_lun->ctl_be_lun.lun_type = T_DIRECT; 1967 1968 if (be_lun->ctl_be_lun.lun_type == T_DIRECT) { 1969 value = ctl_get_opt(&be_lun->ctl_be_lun.options, "file"); 1970 if (value == NULL) { 1971 snprintf(req->error_str, sizeof(req->error_str), 1972 "no file argument specified"); 1973 goto bailout_error; 1974 } 1975 be_lun->dev_path = strdup(value, M_CTLBLK); 1976 be_lun->blocksize = 512; 1977 be_lun->blocksize_shift = fls(be_lun->blocksize) - 1; 1978 1979 retval = ctl_be_block_open(softc, be_lun, req); 1980 if (retval != 0) { 1981 retval = 0; 1982 req->status = CTL_LUN_WARNING; 1983 } 1984 } else { 1985 /* 1986 * For processor devices, we don't have any size. 1987 */ 1988 be_lun->blocksize = 0; 1989 be_lun->pblockexp = 0; 1990 be_lun->pblockoff = 0; 1991 be_lun->size_blocks = 0; 1992 be_lun->size_bytes = 0; 1993 be_lun->ctl_be_lun.maxlba = 0; 1994 1995 /* 1996 * Default to just 1 thread for processor devices. 1997 */ 1998 num_threads = 1; 1999 } 2000 2001 /* 2002 * XXX This searching loop might be refactored to be combined with 2003 * the loop above, 2004 */ 2005 value = ctl_get_opt(&be_lun->ctl_be_lun.options, "num_threads"); 2006 if (value != NULL) { 2007 tmp_num_threads = strtol(value, NULL, 0); 2008 2009 /* 2010 * We don't let the user specify less than one 2011 * thread, but hope he's clueful enough not to 2012 * specify 1000 threads. 2013 */ 2014 if (tmp_num_threads < 1) { 2015 snprintf(req->error_str, sizeof(req->error_str), 2016 "invalid number of threads %s", 2017 num_thread_str); 2018 goto bailout_error; 2019 } 2020 num_threads = tmp_num_threads; 2021 } 2022 unmap = (be_lun->dispatch == ctl_be_block_dispatch_zvol); 2023 value = ctl_get_opt(&be_lun->ctl_be_lun.options, "unmap"); 2024 if (value != NULL) 2025 unmap = (strcmp(value, "on") == 0); 2026 2027 be_lun->flags = CTL_BE_BLOCK_LUN_UNCONFIGURED; 2028 be_lun->ctl_be_lun.flags = CTL_LUN_FLAG_PRIMARY; 2029 if (be_lun->vn == NULL) 2030 be_lun->ctl_be_lun.flags |= CTL_LUN_FLAG_OFFLINE; 2031 if (unmap) 2032 be_lun->ctl_be_lun.flags |= CTL_LUN_FLAG_UNMAP; 2033 be_lun->ctl_be_lun.be_lun = be_lun; 2034 be_lun->ctl_be_lun.maxlba = (be_lun->size_blocks == 0) ? 2035 0 : (be_lun->size_blocks - 1); 2036 be_lun->ctl_be_lun.blocksize = be_lun->blocksize; 2037 be_lun->ctl_be_lun.pblockexp = be_lun->pblockexp; 2038 be_lun->ctl_be_lun.pblockoff = be_lun->pblockoff; 2039 if (be_lun->dispatch == ctl_be_block_dispatch_zvol && 2040 be_lun->blocksize != 0) 2041 be_lun->ctl_be_lun.atomicblock = CTLBLK_MAX_IO_SIZE / 2042 be_lun->blocksize; 2043 /* Tell the user the blocksize we ended up using */ 2044 params->lun_size_bytes = be_lun->size_bytes; 2045 params->blocksize_bytes = be_lun->blocksize; 2046 if (params->flags & CTL_LUN_FLAG_ID_REQ) { 2047 be_lun->ctl_be_lun.req_lun_id = params->req_lun_id; 2048 be_lun->ctl_be_lun.flags |= CTL_LUN_FLAG_ID_REQ; 2049 } else 2050 be_lun->ctl_be_lun.req_lun_id = 0; 2051 2052 be_lun->ctl_be_lun.lun_shutdown = ctl_be_block_lun_shutdown; 2053 be_lun->ctl_be_lun.lun_config_status = 2054 ctl_be_block_lun_config_status; 2055 be_lun->ctl_be_lun.be = &ctl_be_block_driver; 2056 2057 if ((params->flags & CTL_LUN_FLAG_SERIAL_NUM) == 0) { 2058 snprintf(tmpstr, sizeof(tmpstr), "MYSERIAL%4d", 2059 softc->num_luns); 2060 strncpy((char *)be_lun->ctl_be_lun.serial_num, tmpstr, 2061 ctl_min(sizeof(be_lun->ctl_be_lun.serial_num), 2062 sizeof(tmpstr))); 2063 2064 /* Tell the user what we used for a serial number */ 2065 strncpy((char *)params->serial_num, tmpstr, 2066 ctl_min(sizeof(params->serial_num), sizeof(tmpstr))); 2067 } else { 2068 strncpy((char *)be_lun->ctl_be_lun.serial_num, 2069 params->serial_num, 2070 ctl_min(sizeof(be_lun->ctl_be_lun.serial_num), 2071 sizeof(params->serial_num))); 2072 } 2073 if ((params->flags & CTL_LUN_FLAG_DEVID) == 0) { 2074 snprintf(tmpstr, sizeof(tmpstr), "MYDEVID%4d", softc->num_luns); 2075 strncpy((char *)be_lun->ctl_be_lun.device_id, tmpstr, 2076 ctl_min(sizeof(be_lun->ctl_be_lun.device_id), 2077 sizeof(tmpstr))); 2078 2079 /* Tell the user what we used for a device ID */ 2080 strncpy((char *)params->device_id, tmpstr, 2081 ctl_min(sizeof(params->device_id), sizeof(tmpstr))); 2082 } else { 2083 strncpy((char *)be_lun->ctl_be_lun.device_id, 2084 params->device_id, 2085 ctl_min(sizeof(be_lun->ctl_be_lun.device_id), 2086 sizeof(params->device_id))); 2087 } 2088 2089 TASK_INIT(&be_lun->io_task, /*priority*/0, ctl_be_block_worker, be_lun); 2090 2091 be_lun->io_taskqueue = taskqueue_create(be_lun->lunname, M_WAITOK, 2092 taskqueue_thread_enqueue, /*context*/&be_lun->io_taskqueue); 2093 2094 if (be_lun->io_taskqueue == NULL) { 2095 snprintf(req->error_str, sizeof(req->error_str), 2096 "unable to create taskqueue"); 2097 goto bailout_error; 2098 } 2099 2100 /* 2101 * Note that we start the same number of threads by default for 2102 * both the file case and the block device case. For the file 2103 * case, we need multiple threads to allow concurrency, because the 2104 * vnode interface is designed to be a blocking interface. For the 2105 * block device case, ZFS zvols at least will block the caller's 2106 * context in many instances, and so we need multiple threads to 2107 * overcome that problem. Other block devices don't need as many 2108 * threads, but they shouldn't cause too many problems. 2109 * 2110 * If the user wants to just have a single thread for a block 2111 * device, he can specify that when the LUN is created, or change 2112 * the tunable/sysctl to alter the default number of threads. 2113 */ 2114 retval = taskqueue_start_threads(&be_lun->io_taskqueue, 2115 /*num threads*/num_threads, 2116 /*priority*/PWAIT, 2117 /*thread name*/ 2118 "%s taskq", be_lun->lunname); 2119 2120 if (retval != 0) 2121 goto bailout_error; 2122 2123 be_lun->num_threads = num_threads; 2124 2125 mtx_lock(&softc->lock); 2126 softc->num_luns++; 2127 STAILQ_INSERT_TAIL(&softc->lun_list, be_lun, links); 2128 2129 mtx_unlock(&softc->lock); 2130 2131 retval = ctl_add_lun(&be_lun->ctl_be_lun); 2132 if (retval != 0) { 2133 mtx_lock(&softc->lock); 2134 STAILQ_REMOVE(&softc->lun_list, be_lun, ctl_be_block_lun, 2135 links); 2136 softc->num_luns--; 2137 mtx_unlock(&softc->lock); 2138 snprintf(req->error_str, sizeof(req->error_str), 2139 "ctl_add_lun() returned error %d, see dmesg for " 2140 "details", retval); 2141 retval = 0; 2142 goto bailout_error; 2143 } 2144 2145 mtx_lock(&softc->lock); 2146 2147 /* 2148 * Tell the config_status routine that we're waiting so it won't 2149 * clean up the LUN in the event of an error. 2150 */ 2151 be_lun->flags |= CTL_BE_BLOCK_LUN_WAITING; 2152 2153 while (be_lun->flags & CTL_BE_BLOCK_LUN_UNCONFIGURED) { 2154 retval = msleep(be_lun, &softc->lock, PCATCH, "ctlblk", 0); 2155 if (retval == EINTR) 2156 break; 2157 } 2158 be_lun->flags &= ~CTL_BE_BLOCK_LUN_WAITING; 2159 2160 if (be_lun->flags & CTL_BE_BLOCK_LUN_CONFIG_ERR) { 2161 snprintf(req->error_str, sizeof(req->error_str), 2162 "LUN configuration error, see dmesg for details"); 2163 STAILQ_REMOVE(&softc->lun_list, be_lun, ctl_be_block_lun, 2164 links); 2165 softc->num_luns--; 2166 mtx_unlock(&softc->lock); 2167 goto bailout_error; 2168 } else { 2169 params->req_lun_id = be_lun->ctl_be_lun.lun_id; 2170 } 2171 2172 mtx_unlock(&softc->lock); 2173 2174 be_lun->disk_stats = devstat_new_entry("cbb", params->req_lun_id, 2175 be_lun->blocksize, 2176 DEVSTAT_ALL_SUPPORTED, 2177 be_lun->ctl_be_lun.lun_type 2178 | DEVSTAT_TYPE_IF_OTHER, 2179 DEVSTAT_PRIORITY_OTHER); 2180 2181 return (retval); 2182 2183 bailout_error: 2184 req->status = CTL_LUN_ERROR; 2185 2186 if (be_lun->io_taskqueue != NULL) 2187 taskqueue_free(be_lun->io_taskqueue); 2188 ctl_be_block_close(be_lun); 2189 if (be_lun->dev_path != NULL) 2190 free(be_lun->dev_path, M_CTLBLK); 2191 if (be_lun->lun_zone != NULL) 2192 uma_zdestroy(be_lun->lun_zone); 2193 ctl_free_opts(&be_lun->ctl_be_lun.options); 2194 mtx_destroy(&be_lun->queue_lock); 2195 mtx_destroy(&be_lun->io_lock); 2196 free(be_lun, M_CTLBLK); 2197 2198 return (retval); 2199 } 2200 2201 static int 2202 ctl_be_block_rm(struct ctl_be_block_softc *softc, struct ctl_lun_req *req) 2203 { 2204 struct ctl_lun_rm_params *params; 2205 struct ctl_be_block_lun *be_lun; 2206 int retval; 2207 2208 params = &req->reqdata.rm; 2209 2210 mtx_lock(&softc->lock); 2211 2212 be_lun = NULL; 2213 2214 STAILQ_FOREACH(be_lun, &softc->lun_list, links) { 2215 if (be_lun->ctl_be_lun.lun_id == params->lun_id) 2216 break; 2217 } 2218 mtx_unlock(&softc->lock); 2219 2220 if (be_lun == NULL) { 2221 snprintf(req->error_str, sizeof(req->error_str), 2222 "LUN %u is not managed by the block backend", 2223 params->lun_id); 2224 goto bailout_error; 2225 } 2226 2227 retval = ctl_disable_lun(&be_lun->ctl_be_lun); 2228 2229 if (retval != 0) { 2230 snprintf(req->error_str, sizeof(req->error_str), 2231 "error %d returned from ctl_disable_lun() for " 2232 "LUN %d", retval, params->lun_id); 2233 goto bailout_error; 2234 2235 } 2236 2237 retval = ctl_invalidate_lun(&be_lun->ctl_be_lun); 2238 if (retval != 0) { 2239 snprintf(req->error_str, sizeof(req->error_str), 2240 "error %d returned from ctl_invalidate_lun() for " 2241 "LUN %d", retval, params->lun_id); 2242 goto bailout_error; 2243 } 2244 2245 mtx_lock(&softc->lock); 2246 2247 be_lun->flags |= CTL_BE_BLOCK_LUN_WAITING; 2248 2249 while ((be_lun->flags & CTL_BE_BLOCK_LUN_UNCONFIGURED) == 0) { 2250 retval = msleep(be_lun, &softc->lock, PCATCH, "ctlblk", 0); 2251 if (retval == EINTR) 2252 break; 2253 } 2254 2255 be_lun->flags &= ~CTL_BE_BLOCK_LUN_WAITING; 2256 2257 if ((be_lun->flags & CTL_BE_BLOCK_LUN_UNCONFIGURED) == 0) { 2258 snprintf(req->error_str, sizeof(req->error_str), 2259 "interrupted waiting for LUN to be freed"); 2260 mtx_unlock(&softc->lock); 2261 goto bailout_error; 2262 } 2263 2264 STAILQ_REMOVE(&softc->lun_list, be_lun, ctl_be_block_lun, links); 2265 2266 softc->num_luns--; 2267 mtx_unlock(&softc->lock); 2268 2269 taskqueue_drain(be_lun->io_taskqueue, &be_lun->io_task); 2270 2271 taskqueue_free(be_lun->io_taskqueue); 2272 2273 ctl_be_block_close(be_lun); 2274 2275 if (be_lun->disk_stats != NULL) 2276 devstat_remove_entry(be_lun->disk_stats); 2277 2278 uma_zdestroy(be_lun->lun_zone); 2279 2280 ctl_free_opts(&be_lun->ctl_be_lun.options); 2281 free(be_lun->dev_path, M_CTLBLK); 2282 mtx_destroy(&be_lun->queue_lock); 2283 mtx_destroy(&be_lun->io_lock); 2284 free(be_lun, M_CTLBLK); 2285 2286 req->status = CTL_LUN_OK; 2287 2288 return (0); 2289 2290 bailout_error: 2291 2292 req->status = CTL_LUN_ERROR; 2293 2294 return (0); 2295 } 2296 2297 static int 2298 ctl_be_block_modify_file(struct ctl_be_block_lun *be_lun, 2299 struct ctl_lun_req *req) 2300 { 2301 struct vattr vattr; 2302 int error; 2303 struct ctl_lun_create_params *params = &be_lun->params; 2304 2305 if (params->lun_size_bytes != 0) { 2306 be_lun->size_bytes = params->lun_size_bytes; 2307 } else { 2308 vn_lock(be_lun->vn, LK_SHARED | LK_RETRY); 2309 error = VOP_GETATTR(be_lun->vn, &vattr, curthread->td_ucred); 2310 VOP_UNLOCK(be_lun->vn, 0); 2311 if (error != 0) { 2312 snprintf(req->error_str, sizeof(req->error_str), 2313 "error calling VOP_GETATTR() for file %s", 2314 be_lun->dev_path); 2315 return (error); 2316 } 2317 2318 be_lun->size_bytes = vattr.va_size; 2319 } 2320 2321 return (0); 2322 } 2323 2324 static int 2325 ctl_be_block_modify_dev(struct ctl_be_block_lun *be_lun, 2326 struct ctl_lun_req *req) 2327 { 2328 struct ctl_be_block_devdata *dev_data; 2329 int error; 2330 struct ctl_lun_create_params *params = &be_lun->params; 2331 uint64_t size_bytes; 2332 2333 dev_data = &be_lun->backend.dev; 2334 if (!dev_data->csw->d_ioctl) { 2335 snprintf(req->error_str, sizeof(req->error_str), 2336 "no d_ioctl for device %s!", be_lun->dev_path); 2337 return (ENODEV); 2338 } 2339 2340 error = dev_data->csw->d_ioctl(dev_data->cdev, DIOCGMEDIASIZE, 2341 (caddr_t)&size_bytes, FREAD, 2342 curthread); 2343 if (error) { 2344 snprintf(req->error_str, sizeof(req->error_str), 2345 "error %d returned for DIOCGMEDIASIZE ioctl " 2346 "on %s!", error, be_lun->dev_path); 2347 return (error); 2348 } 2349 2350 if (params->lun_size_bytes != 0) { 2351 if (params->lun_size_bytes > size_bytes) { 2352 snprintf(req->error_str, sizeof(req->error_str), 2353 "requested LUN size %ju > backing device " 2354 "size %ju", 2355 (uintmax_t)params->lun_size_bytes, 2356 (uintmax_t)size_bytes); 2357 return (EINVAL); 2358 } 2359 2360 be_lun->size_bytes = params->lun_size_bytes; 2361 } else { 2362 be_lun->size_bytes = size_bytes; 2363 } 2364 2365 return (0); 2366 } 2367 2368 static int 2369 ctl_be_block_modify(struct ctl_be_block_softc *softc, struct ctl_lun_req *req) 2370 { 2371 struct ctl_lun_modify_params *params; 2372 struct ctl_be_block_lun *be_lun; 2373 uint64_t oldsize; 2374 int error; 2375 2376 params = &req->reqdata.modify; 2377 2378 mtx_lock(&softc->lock); 2379 be_lun = NULL; 2380 STAILQ_FOREACH(be_lun, &softc->lun_list, links) { 2381 if (be_lun->ctl_be_lun.lun_id == params->lun_id) 2382 break; 2383 } 2384 mtx_unlock(&softc->lock); 2385 2386 if (be_lun == NULL) { 2387 snprintf(req->error_str, sizeof(req->error_str), 2388 "LUN %u is not managed by the block backend", 2389 params->lun_id); 2390 goto bailout_error; 2391 } 2392 2393 be_lun->params.lun_size_bytes = params->lun_size_bytes; 2394 2395 oldsize = be_lun->size_bytes; 2396 if (be_lun->vn == NULL) 2397 error = ctl_be_block_open(softc, be_lun, req); 2398 else if (be_lun->vn->v_type == VREG) 2399 error = ctl_be_block_modify_file(be_lun, req); 2400 else 2401 error = ctl_be_block_modify_dev(be_lun, req); 2402 2403 if (error == 0 && be_lun->size_bytes != oldsize) { 2404 be_lun->size_blocks = be_lun->size_bytes >> 2405 be_lun->blocksize_shift; 2406 2407 /* 2408 * The maximum LBA is the size - 1. 2409 * 2410 * XXX: Note that this field is being updated without locking, 2411 * which might cause problems on 32-bit architectures. 2412 */ 2413 be_lun->ctl_be_lun.maxlba = (be_lun->size_blocks == 0) ? 2414 0 : (be_lun->size_blocks - 1); 2415 be_lun->ctl_be_lun.blocksize = be_lun->blocksize; 2416 be_lun->ctl_be_lun.pblockexp = be_lun->pblockexp; 2417 be_lun->ctl_be_lun.pblockoff = be_lun->pblockoff; 2418 if (be_lun->dispatch == ctl_be_block_dispatch_zvol && 2419 be_lun->blocksize != 0) 2420 be_lun->ctl_be_lun.atomicblock = CTLBLK_MAX_IO_SIZE / 2421 be_lun->blocksize; 2422 ctl_lun_capacity_changed(&be_lun->ctl_be_lun); 2423 if (oldsize == 0 && be_lun->size_blocks != 0) 2424 ctl_lun_online(&be_lun->ctl_be_lun); 2425 } 2426 2427 /* Tell the user the exact size we ended up using */ 2428 params->lun_size_bytes = be_lun->size_bytes; 2429 2430 req->status = error ? CTL_LUN_WARNING : CTL_LUN_OK; 2431 2432 return (0); 2433 2434 bailout_error: 2435 req->status = CTL_LUN_ERROR; 2436 2437 return (0); 2438 } 2439 2440 static void 2441 ctl_be_block_lun_shutdown(void *be_lun) 2442 { 2443 struct ctl_be_block_lun *lun; 2444 struct ctl_be_block_softc *softc; 2445 2446 lun = (struct ctl_be_block_lun *)be_lun; 2447 2448 softc = lun->softc; 2449 2450 mtx_lock(&softc->lock); 2451 lun->flags |= CTL_BE_BLOCK_LUN_UNCONFIGURED; 2452 if (lun->flags & CTL_BE_BLOCK_LUN_WAITING) 2453 wakeup(lun); 2454 mtx_unlock(&softc->lock); 2455 2456 } 2457 2458 static void 2459 ctl_be_block_lun_config_status(void *be_lun, ctl_lun_config_status status) 2460 { 2461 struct ctl_be_block_lun *lun; 2462 struct ctl_be_block_softc *softc; 2463 2464 lun = (struct ctl_be_block_lun *)be_lun; 2465 softc = lun->softc; 2466 2467 if (status == CTL_LUN_CONFIG_OK) { 2468 mtx_lock(&softc->lock); 2469 lun->flags &= ~CTL_BE_BLOCK_LUN_UNCONFIGURED; 2470 if (lun->flags & CTL_BE_BLOCK_LUN_WAITING) 2471 wakeup(lun); 2472 mtx_unlock(&softc->lock); 2473 2474 /* 2475 * We successfully added the LUN, attempt to enable it. 2476 */ 2477 if (ctl_enable_lun(&lun->ctl_be_lun) != 0) { 2478 printf("%s: ctl_enable_lun() failed!\n", __func__); 2479 if (ctl_invalidate_lun(&lun->ctl_be_lun) != 0) { 2480 printf("%s: ctl_invalidate_lun() failed!\n", 2481 __func__); 2482 } 2483 } 2484 2485 return; 2486 } 2487 2488 2489 mtx_lock(&softc->lock); 2490 lun->flags &= ~CTL_BE_BLOCK_LUN_UNCONFIGURED; 2491 lun->flags |= CTL_BE_BLOCK_LUN_CONFIG_ERR; 2492 wakeup(lun); 2493 mtx_unlock(&softc->lock); 2494 } 2495 2496 2497 static int 2498 ctl_be_block_config_write(union ctl_io *io) 2499 { 2500 struct ctl_be_block_lun *be_lun; 2501 struct ctl_be_lun *ctl_be_lun; 2502 int retval; 2503 2504 retval = 0; 2505 2506 DPRINTF("entered\n"); 2507 2508 ctl_be_lun = (struct ctl_be_lun *)io->io_hdr.ctl_private[ 2509 CTL_PRIV_BACKEND_LUN].ptr; 2510 be_lun = (struct ctl_be_block_lun *)ctl_be_lun->be_lun; 2511 2512 switch (io->scsiio.cdb[0]) { 2513 case SYNCHRONIZE_CACHE: 2514 case SYNCHRONIZE_CACHE_16: 2515 case WRITE_SAME_10: 2516 case WRITE_SAME_16: 2517 case UNMAP: 2518 /* 2519 * The upper level CTL code will filter out any CDBs with 2520 * the immediate bit set and return the proper error. 2521 * 2522 * We don't really need to worry about what LBA range the 2523 * user asked to be synced out. When they issue a sync 2524 * cache command, we'll sync out the whole thing. 2525 */ 2526 mtx_lock(&be_lun->queue_lock); 2527 STAILQ_INSERT_TAIL(&be_lun->config_write_queue, &io->io_hdr, 2528 links); 2529 mtx_unlock(&be_lun->queue_lock); 2530 taskqueue_enqueue(be_lun->io_taskqueue, &be_lun->io_task); 2531 break; 2532 case START_STOP_UNIT: { 2533 struct scsi_start_stop_unit *cdb; 2534 2535 cdb = (struct scsi_start_stop_unit *)io->scsiio.cdb; 2536 2537 if (cdb->how & SSS_START) 2538 retval = ctl_start_lun(ctl_be_lun); 2539 else { 2540 retval = ctl_stop_lun(ctl_be_lun); 2541 /* 2542 * XXX KDM Copan-specific offline behavior. 2543 * Figure out a reasonable way to port this? 2544 */ 2545 #ifdef NEEDTOPORT 2546 if ((retval == 0) 2547 && (cdb->byte2 & SSS_ONOFFLINE)) 2548 retval = ctl_lun_offline(ctl_be_lun); 2549 #endif 2550 } 2551 2552 /* 2553 * In general, the above routines should not fail. They 2554 * just set state for the LUN. So we've got something 2555 * pretty wrong here if we can't start or stop the LUN. 2556 */ 2557 if (retval != 0) { 2558 ctl_set_internal_failure(&io->scsiio, 2559 /*sks_valid*/ 1, 2560 /*retry_count*/ 0xf051); 2561 retval = CTL_RETVAL_COMPLETE; 2562 } else { 2563 ctl_set_success(&io->scsiio); 2564 } 2565 ctl_config_write_done(io); 2566 break; 2567 } 2568 default: 2569 ctl_set_invalid_opcode(&io->scsiio); 2570 ctl_config_write_done(io); 2571 retval = CTL_RETVAL_COMPLETE; 2572 break; 2573 } 2574 2575 return (retval); 2576 2577 } 2578 2579 static int 2580 ctl_be_block_config_read(union ctl_io *io) 2581 { 2582 return (0); 2583 } 2584 2585 static int 2586 ctl_be_block_lun_info(void *be_lun, struct sbuf *sb) 2587 { 2588 struct ctl_be_block_lun *lun; 2589 int retval; 2590 2591 lun = (struct ctl_be_block_lun *)be_lun; 2592 retval = 0; 2593 2594 retval = sbuf_printf(sb, "\t<num_threads>"); 2595 2596 if (retval != 0) 2597 goto bailout; 2598 2599 retval = sbuf_printf(sb, "%d", lun->num_threads); 2600 2601 if (retval != 0) 2602 goto bailout; 2603 2604 retval = sbuf_printf(sb, "</num_threads>\n"); 2605 2606 bailout: 2607 2608 return (retval); 2609 } 2610 2611 static uint64_t 2612 ctl_be_block_lun_attr(void *be_lun, const char *attrname) 2613 { 2614 struct ctl_be_block_lun *lun = (struct ctl_be_block_lun *)be_lun; 2615 2616 if (lun->getattr == NULL) 2617 return (UINT64_MAX); 2618 return (lun->getattr(lun, attrname)); 2619 } 2620 2621 int 2622 ctl_be_block_init(void) 2623 { 2624 struct ctl_be_block_softc *softc; 2625 int retval; 2626 2627 softc = &backend_block_softc; 2628 retval = 0; 2629 2630 mtx_init(&softc->lock, "ctlblock", NULL, MTX_DEF); 2631 beio_zone = uma_zcreate("beio", sizeof(struct ctl_be_block_io), 2632 NULL, NULL, NULL, NULL, UMA_ALIGN_PTR, 0); 2633 STAILQ_INIT(&softc->disk_list); 2634 STAILQ_INIT(&softc->lun_list); 2635 2636 return (retval); 2637 } 2638