1 /*- 2 * SPDX-License-Identifier: (BSD-2-Clause-NetBSD AND BSD-3-Clause) 3 * 4 * Copyright (c) 2003 Poul-Henning Kamp. 5 * Copyright (c) 1996, 1997 The NetBSD Foundation, Inc. 6 * All rights reserved. 7 * 8 * This code is derived from software contributed to The NetBSD Foundation 9 * by Jason R. Thorpe. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 21 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 22 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 23 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 24 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 25 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 26 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 27 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 28 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 29 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 30 * POSSIBILITY OF SUCH DAMAGE. 31 * 32 * $NetBSD: ccd.c,v 1.22 1995/12/08 19:13:26 thorpej Exp $ 33 */ 34 35 /*- 36 * Copyright (c) 1988 University of Utah. 37 * Copyright (c) 1990, 1993 38 * The Regents of the University of California. All rights reserved. 39 * 40 * This code is derived from software contributed to Berkeley by 41 * the Systems Programming Group of the University of Utah Computer 42 * Science Department. 43 * 44 * Redistribution and use in source and binary forms, with or without 45 * modification, are permitted provided that the following conditions 46 * are met: 47 * 1. Redistributions of source code must retain the above copyright 48 * notice, this list of conditions and the following disclaimer. 49 * 2. Redistributions in binary form must reproduce the above copyright 50 * notice, this list of conditions and the following disclaimer in the 51 * documentation and/or other materials provided with the distribution. 52 * 3. Neither the name of the University nor the names of its contributors 53 * may be used to endorse or promote products derived from this software 54 * without specific prior written permission. 55 * 56 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 57 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 58 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 59 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 60 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 61 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 62 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 63 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 64 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 65 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 66 * SUCH DAMAGE. 67 * 68 * from: Utah $Hdr: cd.c 1.6 90/11/28$ 69 * 70 * @(#)cd.c 8.2 (Berkeley) 11/16/93 71 */ 72 73 /* 74 * Dynamic configuration and disklabel support by: 75 * Jason R. Thorpe <thorpej@nas.nasa.gov> 76 * Numerical Aerodynamic Simulation Facility 77 * Mail Stop 258-6 78 * NASA Ames Research Center 79 * Moffett Field, CA 94035 80 */ 81 82 #include <sys/cdefs.h> 83 __FBSDID("$FreeBSD$"); 84 85 #include <sys/param.h> 86 #include <sys/systm.h> 87 #include <sys/kernel.h> 88 #include <sys/module.h> 89 #include <sys/bio.h> 90 #include <sys/malloc.h> 91 #include <sys/sbuf.h> 92 #include <geom/geom.h> 93 94 /* 95 * Number of blocks to untouched in front of a component partition. 96 * This is to avoid violating its disklabel area when it starts at the 97 * beginning of the slice. 98 */ 99 #if !defined(CCD_OFFSET) 100 #define CCD_OFFSET 16 101 #endif 102 103 /* sc_flags */ 104 #define CCDF_UNIFORM 0x02 /* use LCCD of sizes for uniform interleave */ 105 #define CCDF_MIRROR 0x04 /* use mirroring */ 106 #define CCDF_NO_OFFSET 0x08 /* do not leave space in front */ 107 #define CCDF_LINUX 0x10 /* use Linux compatibility mode */ 108 109 /* Mask of user-settable ccd flags. */ 110 #define CCDF_USERMASK (CCDF_UNIFORM|CCDF_MIRROR) 111 112 /* 113 * Interleave description table. 114 * Computed at boot time to speed irregular-interleave lookups. 115 * The idea is that we interleave in "groups". First we interleave 116 * evenly over all component disks up to the size of the smallest 117 * component (the first group), then we interleave evenly over all 118 * remaining disks up to the size of the next-smallest (second group), 119 * and so on. 120 * 121 * Each table entry describes the interleave characteristics of one 122 * of these groups. For example if a concatenated disk consisted of 123 * three components of 5, 3, and 7 DEV_BSIZE blocks interleaved at 124 * DEV_BSIZE (1), the table would have three entries: 125 * 126 * ndisk startblk startoff dev 127 * 3 0 0 0, 1, 2 128 * 2 9 3 0, 2 129 * 1 13 5 2 130 * 0 - - - 131 * 132 * which says that the first nine blocks (0-8) are interleaved over 133 * 3 disks (0, 1, 2) starting at block offset 0 on any component disk, 134 * the next 4 blocks (9-12) are interleaved over 2 disks (0, 2) starting 135 * at component block 3, and the remaining blocks (13-14) are on disk 136 * 2 starting at offset 5. 137 */ 138 struct ccdiinfo { 139 int ii_ndisk; /* # of disks range is interleaved over */ 140 daddr_t ii_startblk; /* starting scaled block # for range */ 141 daddr_t ii_startoff; /* starting component offset (block #) */ 142 int *ii_index; /* ordered list of components in range */ 143 }; 144 145 /* 146 * Component info table. 147 * Describes a single component of a concatenated disk. 148 */ 149 struct ccdcinfo { 150 daddr_t ci_size; /* size */ 151 struct g_provider *ci_provider; /* provider */ 152 struct g_consumer *ci_consumer; /* consumer */ 153 }; 154 155 /* 156 * A concatenated disk is described by this structure. 157 */ 158 159 struct ccd_s { 160 LIST_ENTRY(ccd_s) list; 161 162 int sc_unit; /* logical unit number */ 163 int sc_flags; /* flags */ 164 daddr_t sc_size; /* size of ccd */ 165 int sc_ileave; /* interleave */ 166 u_int sc_ndisks; /* number of components */ 167 struct ccdcinfo *sc_cinfo; /* component info */ 168 struct ccdiinfo *sc_itable; /* interleave table */ 169 u_int32_t sc_secsize; /* # bytes per sector */ 170 int sc_pick; /* side of mirror picked */ 171 daddr_t sc_blk[2]; /* mirror localization */ 172 u_int32_t sc_offset; /* actual offset used */ 173 }; 174 175 static g_start_t g_ccd_start; 176 static void ccdiodone(struct bio *bp); 177 static void ccdinterleave(struct ccd_s *); 178 static int ccdinit(struct gctl_req *req, struct ccd_s *); 179 static int ccdbuffer(struct bio **ret, struct ccd_s *, 180 struct bio *, daddr_t, caddr_t, long); 181 182 static void 183 g_ccd_orphan(struct g_consumer *cp) 184 { 185 /* 186 * XXX: We don't do anything here. It is not obvious 187 * XXX: what DTRT would be, so we do what the previous 188 * XXX: code did: ignore it and let the user cope. 189 */ 190 } 191 192 static int 193 g_ccd_access(struct g_provider *pp, int dr, int dw, int de) 194 { 195 struct g_geom *gp; 196 struct g_consumer *cp1, *cp2; 197 int error; 198 199 de += dr; 200 de += dw; 201 202 gp = pp->geom; 203 error = ENXIO; 204 LIST_FOREACH(cp1, &gp->consumer, consumer) { 205 error = g_access(cp1, dr, dw, de); 206 if (error) { 207 LIST_FOREACH(cp2, &gp->consumer, consumer) { 208 if (cp1 == cp2) 209 break; 210 g_access(cp2, -dr, -dw, -de); 211 } 212 break; 213 } 214 } 215 return (error); 216 } 217 218 /* 219 * Free the softc and its substructures. 220 */ 221 static void 222 g_ccd_freesc(struct ccd_s *sc) 223 { 224 struct ccdiinfo *ii; 225 226 g_free(sc->sc_cinfo); 227 if (sc->sc_itable != NULL) { 228 for (ii = sc->sc_itable; ii->ii_ndisk > 0; ii++) 229 if (ii->ii_index != NULL) 230 g_free(ii->ii_index); 231 g_free(sc->sc_itable); 232 } 233 g_free(sc); 234 } 235 236 static int 237 ccdinit(struct gctl_req *req, struct ccd_s *cs) 238 { 239 struct ccdcinfo *ci; 240 daddr_t size; 241 int ix; 242 daddr_t minsize; 243 int maxsecsize; 244 off_t mediasize; 245 u_int sectorsize; 246 247 cs->sc_size = 0; 248 249 maxsecsize = 0; 250 minsize = 0; 251 252 if (cs->sc_flags & CCDF_LINUX) { 253 cs->sc_offset = 0; 254 cs->sc_ileave *= 2; 255 if (cs->sc_flags & CCDF_MIRROR && cs->sc_ndisks != 2) 256 gctl_error(req, "Mirror mode for Linux raids is " 257 "only supported with 2 devices"); 258 } else { 259 if (cs->sc_flags & CCDF_NO_OFFSET) 260 cs->sc_offset = 0; 261 else 262 cs->sc_offset = CCD_OFFSET; 263 } 264 for (ix = 0; ix < cs->sc_ndisks; ix++) { 265 ci = &cs->sc_cinfo[ix]; 266 267 mediasize = ci->ci_provider->mediasize; 268 sectorsize = ci->ci_provider->sectorsize; 269 if (sectorsize > maxsecsize) 270 maxsecsize = sectorsize; 271 size = mediasize / DEV_BSIZE - cs->sc_offset; 272 273 /* Truncate to interleave boundary */ 274 275 if (cs->sc_ileave > 1) 276 size -= size % cs->sc_ileave; 277 278 if (size == 0) { 279 gctl_error(req, "Component %s has effective size zero", 280 ci->ci_provider->name); 281 return(ENODEV); 282 } 283 284 if (minsize == 0 || size < minsize) 285 minsize = size; 286 ci->ci_size = size; 287 cs->sc_size += size; 288 } 289 290 /* 291 * Don't allow the interleave to be smaller than 292 * the biggest component sector. 293 */ 294 if ((cs->sc_ileave > 0) && 295 (cs->sc_ileave < (maxsecsize / DEV_BSIZE))) { 296 gctl_error(req, "Interleave to small for sector size"); 297 return(EINVAL); 298 } 299 300 /* 301 * If uniform interleave is desired set all sizes to that of 302 * the smallest component. This will guarantee that a single 303 * interleave table is generated. 304 * 305 * Lost space must be taken into account when calculating the 306 * overall size. Half the space is lost when CCDF_MIRROR is 307 * specified. 308 */ 309 if (cs->sc_flags & CCDF_UNIFORM) { 310 for (ix = 0; ix < cs->sc_ndisks; ix++) { 311 ci = &cs->sc_cinfo[ix]; 312 ci->ci_size = minsize; 313 } 314 cs->sc_size = cs->sc_ndisks * minsize; 315 } 316 317 if (cs->sc_flags & CCDF_MIRROR) { 318 /* 319 * Check to see if an even number of components 320 * have been specified. The interleave must also 321 * be non-zero in order for us to be able to 322 * guarantee the topology. 323 */ 324 if (cs->sc_ndisks % 2) { 325 gctl_error(req, 326 "Mirroring requires an even number of disks"); 327 return(EINVAL); 328 } 329 if (cs->sc_ileave == 0) { 330 gctl_error(req, 331 "An interleave must be specified when mirroring"); 332 return(EINVAL); 333 } 334 cs->sc_size = (cs->sc_ndisks/2) * minsize; 335 } 336 337 /* 338 * Construct the interleave table. 339 */ 340 ccdinterleave(cs); 341 342 /* 343 * Create pseudo-geometry based on 1MB cylinders. It's 344 * pretty close. 345 */ 346 cs->sc_secsize = maxsecsize; 347 348 return (0); 349 } 350 351 static void 352 ccdinterleave(struct ccd_s *cs) 353 { 354 struct ccdcinfo *ci, *smallci; 355 struct ccdiinfo *ii; 356 daddr_t bn, lbn; 357 int ix; 358 daddr_t size; 359 360 /* 361 * Allocate an interleave table. The worst case occurs when each 362 * of N disks is of a different size, resulting in N interleave 363 * tables. 364 * 365 * Chances are this is too big, but we don't care. 366 */ 367 size = (cs->sc_ndisks + 1) * sizeof(struct ccdiinfo); 368 cs->sc_itable = g_malloc(size, M_WAITOK | M_ZERO); 369 370 /* 371 * Trivial case: no interleave (actually interleave of disk size). 372 * Each table entry represents a single component in its entirety. 373 * 374 * An interleave of 0 may not be used with a mirror setup. 375 */ 376 if (cs->sc_ileave == 0) { 377 bn = 0; 378 ii = cs->sc_itable; 379 380 for (ix = 0; ix < cs->sc_ndisks; ix++) { 381 /* Allocate space for ii_index. */ 382 ii->ii_index = g_malloc(sizeof(int), M_WAITOK); 383 ii->ii_ndisk = 1; 384 ii->ii_startblk = bn; 385 ii->ii_startoff = 0; 386 ii->ii_index[0] = ix; 387 bn += cs->sc_cinfo[ix].ci_size; 388 ii++; 389 } 390 ii->ii_ndisk = 0; 391 return; 392 } 393 394 /* 395 * The following isn't fast or pretty; it doesn't have to be. 396 */ 397 size = 0; 398 bn = lbn = 0; 399 for (ii = cs->sc_itable; ; ii++) { 400 /* 401 * Allocate space for ii_index. We might allocate more then 402 * we use. 403 */ 404 ii->ii_index = g_malloc((sizeof(int) * cs->sc_ndisks), 405 M_WAITOK); 406 407 /* 408 * Locate the smallest of the remaining components 409 */ 410 smallci = NULL; 411 for (ci = cs->sc_cinfo; ci < &cs->sc_cinfo[cs->sc_ndisks]; 412 ci++) { 413 if (ci->ci_size > size && 414 (smallci == NULL || 415 ci->ci_size < smallci->ci_size)) { 416 smallci = ci; 417 } 418 } 419 420 /* 421 * Nobody left, all done 422 */ 423 if (smallci == NULL) { 424 ii->ii_ndisk = 0; 425 g_free(ii->ii_index); 426 ii->ii_index = NULL; 427 break; 428 } 429 430 /* 431 * Record starting logical block using an sc_ileave blocksize. 432 */ 433 ii->ii_startblk = bn / cs->sc_ileave; 434 435 /* 436 * Record starting component block using an sc_ileave 437 * blocksize. This value is relative to the beginning of 438 * a component disk. 439 */ 440 ii->ii_startoff = lbn; 441 442 /* 443 * Determine how many disks take part in this interleave 444 * and record their indices. 445 */ 446 ix = 0; 447 for (ci = cs->sc_cinfo; 448 ci < &cs->sc_cinfo[cs->sc_ndisks]; ci++) { 449 if (ci->ci_size >= smallci->ci_size) { 450 ii->ii_index[ix++] = ci - cs->sc_cinfo; 451 } 452 } 453 ii->ii_ndisk = ix; 454 bn += ix * (smallci->ci_size - size); 455 lbn = smallci->ci_size / cs->sc_ileave; 456 size = smallci->ci_size; 457 } 458 } 459 460 static void 461 g_ccd_start(struct bio *bp) 462 { 463 long bcount, rcount; 464 struct bio *cbp[2]; 465 caddr_t addr; 466 daddr_t bn; 467 int err; 468 struct ccd_s *cs; 469 470 cs = bp->bio_to->geom->softc; 471 472 /* 473 * Block all GETATTR requests, we wouldn't know which of our 474 * subdevices we should ship it off to. 475 * XXX: this may not be the right policy. 476 */ 477 if(bp->bio_cmd == BIO_GETATTR) { 478 g_io_deliver(bp, EINVAL); 479 return; 480 } 481 482 /* 483 * Translate the partition-relative block number to an absolute. 484 */ 485 bn = bp->bio_offset / cs->sc_secsize; 486 487 /* 488 * Allocate component buffers and fire off the requests 489 */ 490 addr = bp->bio_data; 491 for (bcount = bp->bio_length; bcount > 0; bcount -= rcount) { 492 err = ccdbuffer(cbp, cs, bp, bn, addr, bcount); 493 if (err) { 494 bp->bio_completed += bcount; 495 if (bp->bio_error == 0) 496 bp->bio_error = err; 497 if (bp->bio_completed == bp->bio_length) 498 g_io_deliver(bp, bp->bio_error); 499 return; 500 } 501 rcount = cbp[0]->bio_length; 502 503 if (cs->sc_flags & CCDF_MIRROR) { 504 /* 505 * Mirroring. Writes go to both disks, reads are 506 * taken from whichever disk seems most appropriate. 507 * 508 * We attempt to localize reads to the disk whos arm 509 * is nearest the read request. We ignore seeks due 510 * to writes when making this determination and we 511 * also try to avoid hogging. 512 */ 513 if (cbp[0]->bio_cmd != BIO_READ) { 514 g_io_request(cbp[0], cbp[0]->bio_from); 515 g_io_request(cbp[1], cbp[1]->bio_from); 516 } else { 517 int pick = cs->sc_pick; 518 daddr_t range = cs->sc_size / 16; 519 520 if (bn < cs->sc_blk[pick] - range || 521 bn > cs->sc_blk[pick] + range 522 ) { 523 cs->sc_pick = pick = 1 - pick; 524 } 525 cs->sc_blk[pick] = bn + btodb(rcount); 526 g_io_request(cbp[pick], cbp[pick]->bio_from); 527 } 528 } else { 529 /* 530 * Not mirroring 531 */ 532 g_io_request(cbp[0], cbp[0]->bio_from); 533 } 534 bn += btodb(rcount); 535 addr += rcount; 536 } 537 } 538 539 /* 540 * Build a component buffer header. 541 */ 542 static int 543 ccdbuffer(struct bio **cb, struct ccd_s *cs, struct bio *bp, daddr_t bn, caddr_t addr, long bcount) 544 { 545 struct ccdcinfo *ci, *ci2 = NULL; 546 struct bio *cbp; 547 daddr_t cbn, cboff; 548 off_t cbc; 549 550 /* 551 * Determine which component bn falls in. 552 */ 553 cbn = bn; 554 cboff = 0; 555 556 if (cs->sc_ileave == 0) { 557 /* 558 * Serially concatenated and neither a mirror nor a parity 559 * config. This is a special case. 560 */ 561 daddr_t sblk; 562 563 sblk = 0; 564 for (ci = cs->sc_cinfo; cbn >= sblk + ci->ci_size; ci++) 565 sblk += ci->ci_size; 566 cbn -= sblk; 567 } else { 568 struct ccdiinfo *ii; 569 int ccdisk, off; 570 571 /* 572 * Calculate cbn, the logical superblock (sc_ileave chunks), 573 * and cboff, a normal block offset (DEV_BSIZE chunks) relative 574 * to cbn. 575 */ 576 cboff = cbn % cs->sc_ileave; /* DEV_BSIZE gran */ 577 cbn = cbn / cs->sc_ileave; /* DEV_BSIZE * ileave gran */ 578 579 /* 580 * Figure out which interleave table to use. 581 */ 582 for (ii = cs->sc_itable; ii->ii_ndisk; ii++) { 583 if (ii->ii_startblk > cbn) 584 break; 585 } 586 ii--; 587 588 /* 589 * off is the logical superblock relative to the beginning 590 * of this interleave block. 591 */ 592 off = cbn - ii->ii_startblk; 593 594 /* 595 * We must calculate which disk component to use (ccdisk), 596 * and recalculate cbn to be the superblock relative to 597 * the beginning of the component. This is typically done by 598 * adding 'off' and ii->ii_startoff together. However, 'off' 599 * must typically be divided by the number of components in 600 * this interleave array to be properly convert it from a 601 * CCD-relative logical superblock number to a 602 * component-relative superblock number. 603 */ 604 if (ii->ii_ndisk == 1) { 605 /* 606 * When we have just one disk, it can't be a mirror 607 * or a parity config. 608 */ 609 ccdisk = ii->ii_index[0]; 610 cbn = ii->ii_startoff + off; 611 } else { 612 if (cs->sc_flags & CCDF_MIRROR) { 613 /* 614 * We have forced a uniform mapping, resulting 615 * in a single interleave array. We double 616 * up on the first half of the available 617 * components and our mirror is in the second 618 * half. This only works with a single 619 * interleave array because doubling up 620 * doubles the number of sectors, so there 621 * cannot be another interleave array because 622 * the next interleave array's calculations 623 * would be off. 624 */ 625 int ndisk2 = ii->ii_ndisk / 2; 626 ccdisk = ii->ii_index[off % ndisk2]; 627 cbn = ii->ii_startoff + off / ndisk2; 628 ci2 = &cs->sc_cinfo[ccdisk + ndisk2]; 629 } else { 630 ccdisk = ii->ii_index[off % ii->ii_ndisk]; 631 cbn = ii->ii_startoff + off / ii->ii_ndisk; 632 } 633 } 634 635 ci = &cs->sc_cinfo[ccdisk]; 636 637 /* 638 * Convert cbn from a superblock to a normal block so it 639 * can be used to calculate (along with cboff) the normal 640 * block index into this particular disk. 641 */ 642 cbn *= cs->sc_ileave; 643 } 644 645 /* 646 * Fill in the component buf structure. 647 */ 648 cbp = g_clone_bio(bp); 649 if (cbp == NULL) 650 return (ENOMEM); 651 cbp->bio_done = g_std_done; 652 cbp->bio_offset = dbtob(cbn + cboff + cs->sc_offset); 653 cbp->bio_data = addr; 654 if (cs->sc_ileave == 0) 655 cbc = dbtob((off_t)(ci->ci_size - cbn)); 656 else 657 cbc = dbtob((off_t)(cs->sc_ileave - cboff)); 658 cbp->bio_length = (cbc < bcount) ? cbc : bcount; 659 660 cbp->bio_from = ci->ci_consumer; 661 cb[0] = cbp; 662 663 if (cs->sc_flags & CCDF_MIRROR) { 664 cbp = g_clone_bio(bp); 665 if (cbp == NULL) 666 return (ENOMEM); 667 cbp->bio_done = cb[0]->bio_done = ccdiodone; 668 cbp->bio_offset = cb[0]->bio_offset; 669 cbp->bio_data = cb[0]->bio_data; 670 cbp->bio_length = cb[0]->bio_length; 671 cbp->bio_from = ci2->ci_consumer; 672 cbp->bio_caller1 = cb[0]; 673 cb[0]->bio_caller1 = cbp; 674 cb[1] = cbp; 675 } 676 return (0); 677 } 678 679 /* 680 * Called only for mirrored operations. 681 */ 682 static void 683 ccdiodone(struct bio *cbp) 684 { 685 struct bio *mbp, *pbp; 686 687 mbp = cbp->bio_caller1; 688 pbp = cbp->bio_parent; 689 690 if (pbp->bio_cmd == BIO_READ) { 691 if (cbp->bio_error == 0) { 692 /* We will not be needing the partner bio */ 693 if (mbp != NULL) { 694 pbp->bio_inbed++; 695 g_destroy_bio(mbp); 696 } 697 g_std_done(cbp); 698 return; 699 } 700 if (mbp != NULL) { 701 /* Try partner the bio instead */ 702 mbp->bio_caller1 = NULL; 703 pbp->bio_inbed++; 704 g_destroy_bio(cbp); 705 g_io_request(mbp, mbp->bio_from); 706 /* 707 * XXX: If this comes back OK, we should actually 708 * try to write the good data on the failed mirror 709 */ 710 return; 711 } 712 g_std_done(cbp); 713 return; 714 } 715 if (mbp != NULL) { 716 mbp->bio_caller1 = NULL; 717 pbp->bio_inbed++; 718 if (cbp->bio_error != 0 && pbp->bio_error == 0) 719 pbp->bio_error = cbp->bio_error; 720 g_destroy_bio(cbp); 721 return; 722 } 723 g_std_done(cbp); 724 } 725 726 static void 727 g_ccd_create(struct gctl_req *req, struct g_class *mp) 728 { 729 int *unit, *ileave, *nprovider; 730 struct g_geom *gp; 731 struct g_consumer *cp; 732 struct g_provider *pp; 733 struct ccd_s *sc; 734 struct sbuf *sb; 735 char buf[20]; 736 int i, error; 737 738 g_topology_assert(); 739 unit = gctl_get_paraml(req, "unit", sizeof (*unit)); 740 if (unit == NULL) { 741 gctl_error(req, "unit parameter not given"); 742 return; 743 } 744 ileave = gctl_get_paraml(req, "ileave", sizeof (*ileave)); 745 if (ileave == NULL) { 746 gctl_error(req, "ileave parameter not given"); 747 return; 748 } 749 nprovider = gctl_get_paraml(req, "nprovider", sizeof (*nprovider)); 750 if (nprovider == NULL) { 751 gctl_error(req, "nprovider parameter not given"); 752 return; 753 } 754 755 /* Check for duplicate unit */ 756 LIST_FOREACH(gp, &mp->geom, geom) { 757 sc = gp->softc; 758 if (sc != NULL && sc->sc_unit == *unit) { 759 gctl_error(req, "Unit %d already configured", *unit); 760 return; 761 } 762 } 763 764 if (*nprovider <= 0) { 765 gctl_error(req, "Bogus nprovider argument (= %d)", *nprovider); 766 return; 767 } 768 769 /* Check all providers are valid */ 770 for (i = 0; i < *nprovider; i++) { 771 snprintf(buf, sizeof(buf), "provider%d", i); 772 pp = gctl_get_provider(req, buf); 773 if (pp == NULL) 774 return; 775 } 776 777 gp = g_new_geomf(mp, "ccd%d", *unit); 778 sc = g_malloc(sizeof *sc, M_WAITOK | M_ZERO); 779 gp->softc = sc; 780 sc->sc_ndisks = *nprovider; 781 782 /* Allocate space for the component info. */ 783 sc->sc_cinfo = g_malloc(sc->sc_ndisks * sizeof(struct ccdcinfo), 784 M_WAITOK | M_ZERO); 785 786 /* Create consumers and attach to all providers */ 787 for (i = 0; i < *nprovider; i++) { 788 snprintf(buf, sizeof(buf), "provider%d", i); 789 pp = gctl_get_provider(req, buf); 790 cp = g_new_consumer(gp); 791 error = g_attach(cp, pp); 792 KASSERT(error == 0, ("attach to %s failed", pp->name)); 793 sc->sc_cinfo[i].ci_consumer = cp; 794 sc->sc_cinfo[i].ci_provider = pp; 795 } 796 797 sc->sc_unit = *unit; 798 sc->sc_ileave = *ileave; 799 800 if (gctl_get_param(req, "no_offset", NULL)) 801 sc->sc_flags |= CCDF_NO_OFFSET; 802 if (gctl_get_param(req, "linux", NULL)) 803 sc->sc_flags |= CCDF_LINUX; 804 805 if (gctl_get_param(req, "uniform", NULL)) 806 sc->sc_flags |= CCDF_UNIFORM; 807 if (gctl_get_param(req, "mirror", NULL)) 808 sc->sc_flags |= CCDF_MIRROR; 809 810 if (sc->sc_ileave == 0 && (sc->sc_flags & CCDF_MIRROR)) { 811 printf("%s: disabling mirror, interleave is 0\n", gp->name); 812 sc->sc_flags &= ~(CCDF_MIRROR); 813 } 814 815 if ((sc->sc_flags & CCDF_MIRROR) && !(sc->sc_flags & CCDF_UNIFORM)) { 816 printf("%s: mirror/parity forces uniform flag\n", gp->name); 817 sc->sc_flags |= CCDF_UNIFORM; 818 } 819 820 error = ccdinit(req, sc); 821 if (error != 0) { 822 g_ccd_freesc(sc); 823 gp->softc = NULL; 824 g_wither_geom(gp, ENXIO); 825 return; 826 } 827 828 pp = g_new_providerf(gp, "%s", gp->name); 829 pp->mediasize = sc->sc_size * (off_t)sc->sc_secsize; 830 pp->sectorsize = sc->sc_secsize; 831 g_error_provider(pp, 0); 832 833 sb = sbuf_new_auto(); 834 sbuf_printf(sb, "ccd%d: %d components ", sc->sc_unit, *nprovider); 835 for (i = 0; i < *nprovider; i++) { 836 sbuf_printf(sb, "%s%s", 837 i == 0 ? "(" : ", ", 838 sc->sc_cinfo[i].ci_provider->name); 839 } 840 sbuf_printf(sb, "), %jd blocks ", (off_t)pp->mediasize / DEV_BSIZE); 841 if (sc->sc_ileave != 0) 842 sbuf_printf(sb, "interleaved at %d blocks\n", 843 sc->sc_ileave); 844 else 845 sbuf_printf(sb, "concatenated\n"); 846 sbuf_finish(sb); 847 gctl_set_param_err(req, "output", sbuf_data(sb), sbuf_len(sb) + 1); 848 sbuf_delete(sb); 849 } 850 851 static int 852 g_ccd_destroy_geom(struct gctl_req *req, struct g_class *mp, struct g_geom *gp) 853 { 854 struct g_provider *pp; 855 struct ccd_s *sc; 856 857 g_topology_assert(); 858 sc = gp->softc; 859 pp = LIST_FIRST(&gp->provider); 860 if (sc == NULL || pp == NULL) 861 return (EBUSY); 862 if (pp->acr != 0 || pp->acw != 0 || pp->ace != 0) { 863 gctl_error(req, "%s is open(r%dw%de%d)", gp->name, 864 pp->acr, pp->acw, pp->ace); 865 return (EBUSY); 866 } 867 g_ccd_freesc(sc); 868 gp->softc = NULL; 869 g_wither_geom(gp, ENXIO); 870 return (0); 871 } 872 873 static void 874 g_ccd_list(struct gctl_req *req, struct g_class *mp) 875 { 876 struct sbuf *sb; 877 struct ccd_s *cs; 878 struct g_geom *gp; 879 int i, unit, *up; 880 881 up = gctl_get_paraml(req, "unit", sizeof (*up)); 882 if (up == NULL) { 883 gctl_error(req, "unit parameter not given"); 884 return; 885 } 886 unit = *up; 887 sb = sbuf_new_auto(); 888 LIST_FOREACH(gp, &mp->geom, geom) { 889 cs = gp->softc; 890 if (cs == NULL || (unit >= 0 && unit != cs->sc_unit)) 891 continue; 892 sbuf_printf(sb, "ccd%d\t\t%d\t%d\t", 893 cs->sc_unit, cs->sc_ileave, cs->sc_flags & CCDF_USERMASK); 894 895 for (i = 0; i < cs->sc_ndisks; ++i) { 896 sbuf_printf(sb, "%s/dev/%s", i == 0 ? "" : " ", 897 cs->sc_cinfo[i].ci_provider->name); 898 } 899 sbuf_printf(sb, "\n"); 900 } 901 sbuf_finish(sb); 902 gctl_set_param_err(req, "output", sbuf_data(sb), sbuf_len(sb) + 1); 903 sbuf_delete(sb); 904 } 905 906 static void 907 g_ccd_config(struct gctl_req *req, struct g_class *mp, char const *verb) 908 { 909 struct g_geom *gp; 910 911 g_topology_assert(); 912 if (!strcmp(verb, "create geom")) { 913 g_ccd_create(req, mp); 914 } else if (!strcmp(verb, "destroy geom")) { 915 gp = gctl_get_geom(req, mp, "geom"); 916 if (gp != NULL) 917 g_ccd_destroy_geom(req, mp, gp); 918 } else if (!strcmp(verb, "list")) { 919 g_ccd_list(req, mp); 920 } else { 921 gctl_error(req, "unknown verb"); 922 } 923 } 924 925 static struct g_class g_ccd_class = { 926 .name = "CCD", 927 .version = G_VERSION, 928 .ctlreq = g_ccd_config, 929 .destroy_geom = g_ccd_destroy_geom, 930 .start = g_ccd_start, 931 .orphan = g_ccd_orphan, 932 .access = g_ccd_access, 933 }; 934 935 DECLARE_GEOM_CLASS(g_ccd_class, g_ccd); 936 MODULE_VERSION(geom_ccd, 0); 937