lustre/lod/lod_qos.c

   1 /*
   2  * GPL HEADER START
   3  *
   4  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License version 2 only,
   8  * as published by the Free Software Foundation.
   9
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13  * GNU General Public License version 2 for more details.  A copy is
  14  * included in the COPYING file that accompanied this code.
  15
  16  * You should have received a copy of the GNU General Public License
  17  * along with this program; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  19  *
  20  * GPL HEADER END
  21  */
  22 /*
  23  * Copyright  2009 Sun Microsystems, Inc. All rights reserved
  24  * Use is subject to license terms.
  25  *
  26  * Copyright (c) 2012, Intel Corporation.
  27  */
  28 /*
  29  * This file is part of Lustre, http://www.lustre.org/
  30  * Lustre is a trademark of Sun Microsystems, Inc.
  31  *
  32  * lustre/lod/lod_qos.c
  33  *
  34  */
  35
  36 #define DEBUG_SUBSYSTEM S_LOV
  37
  38 #include <libcfs/libcfs.h>
  39 #include <obd_class.h>
  40 #include <obd_lov.h>
  41 #include <lustre/lustre_idl.h>
  42 #include "lod_internal.h"
  43
  44 /*
  45  * force QoS policy (not RR) to be used for testing purposes
  46  */
  47 #define FORCE_QOS_
  48
  49 #define D_QOS   D_OTHER
  50
  51 #if 0
  52 #define QOS_DEBUG(fmt, ...)     CDEBUG(D_OTHER, fmt, ## __VA_ARGS__)
  53 #define QOS_CONSOLE(fmt, ...)   LCONSOLE(D_OTHER, fmt, ## __VA_ARGS__)
  54 #else
  55 #define QOS_DEBUG(fmt, ...)
  56 #define QOS_CONSOLE(fmt, ...)
  57 #endif
  58
  59 #define TGT_BAVAIL(i) (OST_TGT(lod,i)->ltd_statfs.os_bavail * \
  60                        OST_TGT(lod,i)->ltd_statfs.os_bsize)
  61
  62 int qos_add_tgt(struct lod_device *lod, struct lod_tgt_desc *ost_desc)
  63 {
  64         struct lov_qos_oss *oss = NULL, *temposs;
  65         struct obd_export  *exp = ost_desc->ltd_exp;
  66         int                 rc = 0, found = 0;
  67         cfs_list_t         *list;
  68         ENTRY;
  69
  70         down_write(&lod->lod_qos.lq_rw_sem);
  71         /*
  72          * a bit hacky approach to learn NID of corresponding connection
  73          * but there is no official API to access information like this
  74          * with OSD API.
  75          */
  76         cfs_list_for_each_entry(oss, &lod->lod_qos.lq_oss_list, lqo_oss_list) {
  77                 if (obd_uuid_equals(&oss->lqo_uuid,
  78                                     &exp->exp_connection->c_remote_uuid)) {
  79                         found++;
  80                         break;
  81                 }
  82         }
  83
  84         if (!found) {
  85                 OBD_ALLOC_PTR(oss);
  86                 if (!oss)
  87                         GOTO(out, rc = -ENOMEM);
  88                 memcpy(&oss->lqo_uuid, &exp->exp_connection->c_remote_uuid,
  89                        sizeof(oss->lqo_uuid));
  90         } else {
  91                 /* Assume we have to move this one */
  92                 cfs_list_del(&oss->lqo_oss_list);
  93         }
  94
  95         oss->lqo_ost_count++;
  96         ost_desc->ltd_qos.ltq_oss = oss;
  97
  98         CDEBUG(D_QOS, "add tgt %s to OSS %s (%d OSTs)\n",
  99                obd_uuid2str(&ost_desc->ltd_uuid), obd_uuid2str(&oss->lqo_uuid),
 100                oss->lqo_ost_count);
 101
 102         /* Add sorted by # of OSTs.  Find the first entry that we're
 103            bigger than... */
 104         list = &lod->lod_qos.lq_oss_list;
 105         cfs_list_for_each_entry(temposs, list, lqo_oss_list) {
 106                 if (oss->lqo_ost_count > temposs->lqo_ost_count)
 107                         break;
 108         }
 109         /* ...and add before it.  If we're the first or smallest, temposs
 110            points to the list head, and we add to the end. */
 111         cfs_list_add_tail(&oss->lqo_oss_list, &temposs->lqo_oss_list);
 112
 113         lod->lod_qos.lq_dirty = 1;
 114         lod->lod_qos.lq_rr.lqr_dirty = 1;
 115
 116 out:
 117         up_write(&lod->lod_qos.lq_rw_sem);
 118         RETURN(rc);
 119 }
 120
 121 int qos_del_tgt(struct lod_device *lod, struct lod_tgt_desc *ost_desc)
 122 {
 123         struct lov_qos_oss *oss;
 124         int                 rc = 0;
 125         ENTRY;
 126
 127         down_write(&lod->lod_qos.lq_rw_sem);
 128         oss = ost_desc->ltd_qos.ltq_oss;
 129         if (!oss)
 130                 GOTO(out, rc = -ENOENT);
 131
 132         oss->lqo_ost_count--;
 133         if (oss->lqo_ost_count == 0) {
 134                 CDEBUG(D_QOS, "removing OSS %s\n",
 135                        obd_uuid2str(&oss->lqo_uuid));
 136                 cfs_list_del(&oss->lqo_oss_list);
 137                 ost_desc->ltd_qos.ltq_oss = NULL;
 138                 OBD_FREE_PTR(oss);
 139         }
 140
 141         lod->lod_qos.lq_dirty = 1;
 142         lod->lod_qos.lq_rr.lqr_dirty = 1;
 143 out:
 144         up_write(&lod->lod_qos.lq_rw_sem);
 145         RETURN(rc);
 146 }
 147
 148 static int lod_statfs_and_check(const struct lu_env *env, struct lod_device *d,
 149                                 int index, struct obd_statfs *sfs)
 150 {
 151         struct lod_tgt_desc *ost;
 152         int                  rc;
 153
 154         LASSERT(d);
 155         ost = OST_TGT(d,index);
 156         LASSERT(ost);
 157
 158         rc = dt_statfs(env, ost->ltd_ost, sfs);
 159         if (rc && rc != -ENOTCONN)
 160                 CERROR("%s: statfs: rc = %d\n", lod2obd(d)->obd_name, rc);
 161
 162         /* check whether device has changed state (active, inactive) */
 163         if (rc != 0 && ost->ltd_active) {
 164                 /* turned inactive? */
 165                 spin_lock(&d->lod_desc_lock);
 166                 if (ost->ltd_active) {
 167                         ost->ltd_active = 0;
 168                         LASSERT(d->lod_desc.ld_active_tgt_count > 0);
 169                         d->lod_desc.ld_active_tgt_count--;
 170                         d->lod_qos.lq_dirty = 1;
 171                         d->lod_qos.lq_rr.lqr_dirty = 1;
 172                         CDEBUG(D_CONFIG, "%s: turns inactive\n",
 173                                ost->ltd_exp->exp_obd->obd_name);
 174                 }
 175                 spin_unlock(&d->lod_desc_lock);
 176         } else if (rc == 0 && ost->ltd_active == 0) {
 177                 /* turned active? */
 178                 LASSERT(d->lod_desc.ld_active_tgt_count < d->lod_ostnr);
 179                 spin_lock(&d->lod_desc_lock);
 180                 if (ost->ltd_active == 0) {
 181                         ost->ltd_active = 1;
 182                         d->lod_desc.ld_active_tgt_count++;
 183                         d->lod_qos.lq_dirty = 1;
 184                         d->lod_qos.lq_rr.lqr_dirty = 1;
 185                         CDEBUG(D_CONFIG, "%s: turns active\n",
 186                                ost->ltd_exp->exp_obd->obd_name);
 187                 }
 188                 spin_unlock(&d->lod_desc_lock);
 189         }
 190
 191         return rc;
 192 }
 193
 194 static void lod_qos_statfs_update(const struct lu_env *env,
 195                                   struct lod_device *lod)
 196 {
 197         struct obd_device *obd = lod2obd(lod);
 198         struct ost_pool   *osts = &(lod->lod_pool_info);
 199         int                i, idx, rc = 0;
 200         __u64              max_age, avail;
 201         ENTRY;
 202
 203         max_age = cfs_time_shift_64(-2 * lod->lod_desc.ld_qos_maxage);
 204
 205         if (cfs_time_beforeq_64(max_age, obd->obd_osfs_age))
 206                 /* statfs data are quite recent, don't need to refresh it */
 207                 RETURN_EXIT;
 208
 209         down_write(&lod->lod_qos.lq_rw_sem);
 210         if (cfs_time_beforeq_64(max_age, obd->obd_osfs_age))
 211                 GOTO(out, rc = 0);
 212
 213         for (i = 0; i < osts->op_count; i++) {
 214                 idx = osts->op_array[i];
 215                 avail = OST_TGT(lod,idx)->ltd_statfs.os_bavail;
 216                 rc = lod_statfs_and_check(env, lod, idx,
 217                                           &OST_TGT(lod,idx)->ltd_statfs);
 218                 if (rc)
 219                         break;
 220                 if (OST_TGT(lod,idx)->ltd_statfs.os_bavail != avail)
 221                         /* recalculate weigths */
 222                         lod->lod_qos.lq_dirty = 1;
 223         }
 224         obd->obd_osfs_age = cfs_time_current_64();
 225
 226 out:
 227         up_write(&lod->lod_qos.lq_rw_sem);
 228 }
 229
 230 /* Recalculate per-object penalties for OSSs and OSTs,
 231    depends on size of each ost in an oss */
 232 static int lod_qos_calc_ppo(struct lod_device *lod)
 233 {
 234         struct lov_qos_oss *oss;
 235         __u64               ba_max, ba_min, temp;
 236         __u32               num_active;
 237         int                 rc, i, prio_wide;
 238         time_t              now, age;
 239         ENTRY;
 240
 241         if (!lod->lod_qos.lq_dirty)
 242                 GOTO(out, rc = 0);
 243
 244         num_active = lod->lod_desc.ld_active_tgt_count - 1;
 245         if (num_active < 1)
 246                 GOTO(out, rc = -EAGAIN);
 247
 248         /* find bavail on each OSS */
 249         cfs_list_for_each_entry(oss, &lod->lod_qos.lq_oss_list, lqo_oss_list)
 250                                 oss->lqo_bavail = 0;
 251         lod->lod_qos.lq_active_oss_count = 0;
 252
 253         /*
 254          * How badly user wants to select OSTs "widely" (not recently chosen
 255          * and not on recent OSS's).  As opposed to "freely" (free space
 256          * avail.) 0-256
 257          */
 258         prio_wide = 256 - lod->lod_qos.lq_prio_free;
 259
 260         ba_min = (__u64)(-1);
 261         ba_max = 0;
 262         now = cfs_time_current_sec();
 263         /* Calculate OST penalty per object
 264          * (lod ref taken in lod_qos_prep_create()) */
 265         cfs_foreach_bit(lod->lod_ost_bitmap, i) {
 266                 LASSERT(OST_TGT(lod,i));
 267                 temp = TGT_BAVAIL(i);
 268                 if (!temp)
 269                         continue;
 270                 ba_min = min(temp, ba_min);
 271                 ba_max = max(temp, ba_max);
 272
 273                 /* Count the number of usable OSS's */
 274                 if (OST_TGT(lod,i)->ltd_qos.ltq_oss->lqo_bavail == 0)
 275                         lod->lod_qos.lq_active_oss_count++;
 276                 OST_TGT(lod,i)->ltd_qos.ltq_oss->lqo_bavail += temp;
 277
 278                 /* per-OST penalty is prio * TGT_bavail / (num_ost - 1) / 2 */
 279                 temp >>= 1;
 280                 lov_do_div64(temp, num_active);
 281                 OST_TGT(lod,i)->ltd_qos.ltq_penalty_per_obj =
 282                         (temp * prio_wide) >> 8;
 283
 284                 age = (now - OST_TGT(lod,i)->ltd_qos.ltq_used) >> 3;
 285                 if (lod->lod_qos.lq_reset ||
 286                     age > 32 * lod->lod_desc.ld_qos_maxage)
 287                         OST_TGT(lod,i)->ltd_qos.ltq_penalty = 0;
 288                 else if (age > lod->lod_desc.ld_qos_maxage)
 289                         /* Decay the penalty by half for every 8x the update
 290                          * interval that the device has been idle.  That gives
 291                          * lots of time for the statfs information to be
 292                          * updated (which the penalty is only a proxy for),
 293                          * and avoids penalizing OSS/OSTs under light load. */
 294                         OST_TGT(lod,i)->ltd_qos.ltq_penalty >>=
 295                                 (age / lod->lod_desc.ld_qos_maxage);
 296         }
 297
 298         num_active = lod->lod_qos.lq_active_oss_count - 1;
 299         if (num_active < 1) {
 300                 /* If there's only 1 OSS, we can't penalize it, so instead
 301                    we have to double the OST penalty */
 302                 num_active = 1;
 303                 cfs_foreach_bit(lod->lod_ost_bitmap, i)
 304                         OST_TGT(lod,i)->ltd_qos.ltq_penalty_per_obj <<= 1;
 305         }
 306
 307         /* Per-OSS penalty is prio * oss_avail / oss_osts / (num_oss - 1) / 2 */
 308         cfs_list_for_each_entry(oss, &lod->lod_qos.lq_oss_list, lqo_oss_list) {
 309                 temp = oss->lqo_bavail >> 1;
 310                 lov_do_div64(temp, oss->lqo_ost_count * num_active);
 311                 oss->lqo_penalty_per_obj = (temp * prio_wide) >> 8;
 312
 313                 age = (now - oss->lqo_used) >> 3;
 314                 if (lod->lod_qos.lq_reset ||
 315                     age > 32 * lod->lod_desc.ld_qos_maxage)
 316                         oss->lqo_penalty = 0;
 317                 else if (age > lod->lod_desc.ld_qos_maxage)
 318                         /* Decay the penalty by half for every 8x the update
 319                          * interval that the device has been idle.  That gives
 320                          * lots of time for the statfs information to be
 321                          * updated (which the penalty is only a proxy for),
 322                          * and avoids penalizing OSS/OSTs under light load. */
 323                         oss->lqo_penalty >>= age / lod->lod_desc.ld_qos_maxage;
 324         }
 325
 326         lod->lod_qos.lq_dirty = 0;
 327         lod->lod_qos.lq_reset = 0;
 328
 329         /* If each ost has almost same free space,
 330          * do rr allocation for better creation performance */
 331         lod->lod_qos.lq_same_space = 0;
 332         if ((ba_max * (256 - lod->lod_qos.lq_threshold_rr)) >> 8 < ba_min) {
 333                 lod->lod_qos.lq_same_space = 1;
 334                 /* Reset weights for the next time we enter qos mode */
 335                 lod->lod_qos.lq_reset = 1;
 336         }
 337         rc = 0;
 338
 339 out:
 340 #ifndef FORCE_QOS
 341         if (!rc && lod->lod_qos.lq_same_space)
 342                 RETURN(-EAGAIN);
 343 #endif
 344         RETURN(rc);
 345 }
 346
 347 static int lod_qos_calc_weight(struct lod_device *lod, int i)
 348 {
 349         __u64 temp, temp2;
 350
 351         /* Final ost weight = TGT_BAVAIL - ost_penalty - oss_penalty */
 352         temp = TGT_BAVAIL(i);
 353         temp2 = OST_TGT(lod,i)->ltd_qos.ltq_penalty +
 354                 OST_TGT(lod,i)->ltd_qos.ltq_oss->lqo_penalty;
 355         if (temp < temp2)
 356                 OST_TGT(lod,i)->ltd_qos.ltq_weight = 0;
 357         else
 358                 OST_TGT(lod,i)->ltd_qos.ltq_weight = temp - temp2;
 359         return 0;
 360 }
 361
 362 /* We just used this index for a stripe; adjust everyone's weights */
 363 static int lod_qos_used(struct lod_device *lod, struct ost_pool *osts,
 364                         __u32 index, __u64 *total_wt)
 365 {
 366         struct lod_tgt_desc *ost;
 367         struct lov_qos_oss  *oss;
 368         int j;
 369         ENTRY;
 370
 371         ost = OST_TGT(lod,index);
 372         LASSERT(ost);
 373
 374         /* Don't allocate on this devuce anymore, until the next alloc_qos */
 375         ost->ltd_qos.ltq_usable = 0;
 376
 377         oss = ost->ltd_qos.ltq_oss;
 378
 379         /* Decay old penalty by half (we're adding max penalty, and don't
 380            want it to run away.) */
 381         ost->ltd_qos.ltq_penalty >>= 1;
 382         oss->lqo_penalty >>= 1;
 383
 384         /* mark the OSS and OST as recently used */
 385         ost->ltd_qos.ltq_used = oss->lqo_used = cfs_time_current_sec();
 386
 387         /* Set max penalties for this OST and OSS */
 388         ost->ltd_qos.ltq_penalty +=
 389                 ost->ltd_qos.ltq_penalty_per_obj * lod->lod_ostnr;
 390         oss->lqo_penalty += oss->lqo_penalty_per_obj *
 391                 lod->lod_qos.lq_active_oss_count;
 392
 393         /* Decrease all OSS penalties */
 394         cfs_list_for_each_entry(oss, &lod->lod_qos.lq_oss_list, lqo_oss_list) {
 395                 if (oss->lqo_penalty < oss->lqo_penalty_per_obj)
 396                         oss->lqo_penalty = 0;
 397                 else
 398                         oss->lqo_penalty -= oss->lqo_penalty_per_obj;
 399         }
 400
 401         *total_wt = 0;
 402         /* Decrease all OST penalties */
 403         for (j = 0; j < osts->op_count; j++) {
 404                 int i;
 405
 406                 i = osts->op_array[j];
 407                 if (!cfs_bitmap_check(lod->lod_ost_bitmap, i))
 408                         continue;
 409
 410                 ost = OST_TGT(lod,i);
 411                 LASSERT(ost);
 412
 413                 if (ost->ltd_qos.ltq_penalty <
 414                                 ost->ltd_qos.ltq_penalty_per_obj)
 415                         ost->ltd_qos.ltq_penalty = 0;
 416                 else
 417                         ost->ltd_qos.ltq_penalty -=
 418                                 ost->ltd_qos.ltq_penalty_per_obj;
 419
 420                 lod_qos_calc_weight(lod, i);
 421
 422                 /* Recalc the total weight of usable osts */
 423                 if (ost->ltd_qos.ltq_usable)
 424                         *total_wt += ost->ltd_qos.ltq_weight;
 425
 426                 QOS_DEBUG("recalc tgt %d usable=%d avail="LPU64
 427                           " ostppo="LPU64" ostp="LPU64" ossppo="LPU64
 428                           " ossp="LPU64" wt="LPU64"\n",
 429                           i, ost->ltd_qos.ltq_usable, TGT_BAVAIL(i) >> 10,
 430                           ost->ltd_qos.ltq_penalty_per_obj >> 10,
 431                           ost->ltd_qos.ltq_penalty >> 10,
 432                           ost->ltd_qos.ltq_oss->lqo_penalty_per_obj >> 10,
 433                           ost->ltd_qos.ltq_oss->lqo_penalty >> 10,
 434                           ost->ltd_qos.ltq_weight >> 10);
 435         }
 436
 437         RETURN(0);
 438 }
 439
 440 #define LOV_QOS_EMPTY ((__u32)-1)
 441 /* compute optimal round-robin order, based on OSTs per OSS */
 442 static int lod_qos_calc_rr(struct lod_device *lod, struct ost_pool *src_pool,
 443                            struct lov_qos_rr *lqr)
 444 {
 445         struct lov_qos_oss  *oss;
 446         struct lod_tgt_desc *ost;
 447         unsigned placed, real_count;
 448         int i, rc;
 449         ENTRY;
 450
 451         if (!lqr->lqr_dirty) {
 452                 LASSERT(lqr->lqr_pool.op_size);
 453                 RETURN(0);
 454         }
 455
 456         /* Do actual allocation. */
 457         down_write(&lod->lod_qos.lq_rw_sem);
 458
 459         /*
 460          * Check again. While we were sleeping on @lq_rw_sem something could
 461          * change.
 462          */
 463         if (!lqr->lqr_dirty) {
 464                 LASSERT(lqr->lqr_pool.op_size);
 465                 up_write(&lod->lod_qos.lq_rw_sem);
 466                 RETURN(0);
 467         }
 468
 469         real_count = src_pool->op_count;
 470
 471         /* Zero the pool array */
 472         /* alloc_rr is holding a read lock on the pool, so nobody is adding/
 473            deleting from the pool. The lq_rw_sem insures that nobody else
 474            is reading. */
 475         lqr->lqr_pool.op_count = real_count;
 476         rc = lod_ost_pool_extend(&lqr->lqr_pool, real_count);
 477         if (rc) {
 478                 up_write(&lod->lod_qos.lq_rw_sem);
 479                 RETURN(rc);
 480         }
 481         for (i = 0; i < lqr->lqr_pool.op_count; i++)
 482                 lqr->lqr_pool.op_array[i] = LOV_QOS_EMPTY;
 483
 484         /* Place all the OSTs from 1 OSS at the same time. */
 485         placed = 0;
 486         cfs_list_for_each_entry(oss, &lod->lod_qos.lq_oss_list, lqo_oss_list) {
 487                 int j = 0;
 488
 489                 for (i = 0; i < lqr->lqr_pool.op_count; i++) {
 490                         int next;
 491
 492                         if (!cfs_bitmap_check(lod->lod_ost_bitmap,
 493                                                 src_pool->op_array[i]))
 494                                 continue;
 495
 496                         ost = OST_TGT(lod,src_pool->op_array[i]);
 497                         LASSERT(ost && ost->ltd_ost);
 498                         if (ost->ltd_qos.ltq_oss != oss)
 499                                 continue;
 500
 501                         /* Evenly space these OSTs across arrayspace */
 502                         next = j * lqr->lqr_pool.op_count / oss->lqo_ost_count;
 503                         while (lqr->lqr_pool.op_array[next] != LOV_QOS_EMPTY)
 504                                 next = (next + 1) % lqr->lqr_pool.op_count;
 505
 506                         lqr->lqr_pool.op_array[next] = src_pool->op_array[i];
 507                         j++;
 508                         placed++;
 509                 }
 510         }
 511
 512         lqr->lqr_dirty = 0;
 513         up_write(&lod->lod_qos.lq_rw_sem);
 514
 515         if (placed != real_count) {
 516                 /* This should never happen */
 517                 LCONSOLE_ERROR_MSG(0x14e, "Failed to place all OSTs in the "
 518                                    "round-robin list (%d of %d).\n",
 519                                    placed, real_count);
 520                 for (i = 0; i < lqr->lqr_pool.op_count; i++) {
 521                         LCONSOLE(D_WARNING, "rr #%d ost idx=%d\n", i,
 522                                  lqr->lqr_pool.op_array[i]);
 523                 }
 524                 lqr->lqr_dirty = 1;
 525                 RETURN(-EAGAIN);
 526         }
 527
 528 #if 0
 529         for (i = 0; i < lqr->lqr_pool.op_count; i++)
 530                 QOS_CONSOLE("rr #%d ost idx=%d\n", i, lqr->lqr_pool.op_array[i]);
 531 #endif
 532
 533         RETURN(0);
 534 }
 535
 536 /**
 537  * A helper function to:
 538  *   create in-core lu object on the specified OSP
 539  *   declare creation of the object
 540  * IMPORTANT: at this stage object is anonymouos - it has no fid assigned
 541  *            this is a workaround till we have natural FIDs on OST
 542  *
 543  *            at this point we want to declare (reserve) object for us as
 544  *            we can't block at execution (when create method is called).
 545  *            otherwise we'd block whole transaction batch
 546  */
 547 static struct dt_object *lod_qos_declare_object_on(const struct lu_env *env,
 548                                                    struct lod_device *d,
 549                                                    int ost_idx,
 550                                                    struct thandle *th)
 551 {
 552         struct lod_tgt_desc *ost;
 553         struct lu_object *o, *n;
 554         struct lu_device *nd;
 555         struct dt_object *dt;
 556         int               rc;
 557         ENTRY;
 558
 559         LASSERT(d);
 560         LASSERT(ost_idx >= 0);
 561         LASSERT(ost_idx < d->lod_osts_size);
 562         ost = OST_TGT(d,ost_idx);
 563         LASSERT(ost);
 564         LASSERT(ost->ltd_ost);
 565
 566         nd = &ost->ltd_ost->dd_lu_dev;
 567
 568         /*
 569          * allocate anonymous object with zero fid, real fid
 570          * will be assigned by OSP within transaction
 571          * XXX: to be fixed with fully-functional OST fids
 572          */
 573         o = lu_object_anon(env, nd, NULL);
 574         if (IS_ERR(o))
 575                 GOTO(out, dt = ERR_PTR(PTR_ERR(o)));
 576
 577         n = lu_object_locate(o->lo_header, nd->ld_type);
 578         if (unlikely(n == NULL)) {
 579                 CERROR("can't find slice\n");
 580                 lu_object_put(env, o);
 581                 GOTO(out, dt = ERR_PTR(-EINVAL));
 582         }
 583
 584         dt = container_of(n, struct dt_object, do_lu);
 585
 586         rc = dt_declare_create(env, dt, NULL, NULL, NULL, th);
 587         if (rc) {
 588                 CDEBUG(D_OTHER, "can't declare creation on #%u: %d\n",
 589                        ost_idx, rc);
 590                 lu_object_put(env, o);
 591                 dt = ERR_PTR(rc);
 592         }
 593
 594 out:
 595         RETURN(dt);
 596 }
 597
 598 static int min_stripe_count(int stripe_cnt, int flags)
 599 {
 600         return (flags & LOV_USES_DEFAULT_STRIPE ?
 601                         stripe_cnt - (stripe_cnt / 4) : stripe_cnt);
 602 }
 603
 604 #define LOV_CREATE_RESEED_MULT 30
 605 #define LOV_CREATE_RESEED_MIN  2000
 606
 607 static int inline lod_qos_dev_is_full(struct obd_statfs *msfs)
 608 {
 609         __u64 used;
 610         int   bs = msfs->os_bsize;
 611
 612         LASSERT(((bs - 1) & bs) == 0);
 613
 614         /* the minimum of 0.1% used blocks and 1GB bytes. */
 615         used = min_t(__u64, (msfs->os_blocks - msfs->os_bfree) >> 10,
 616                         1 << (31 - ffs(bs)));
 617         return (msfs->os_bavail < used);
 618 }
 619
 620 int lod_ea_store_resize(struct lod_thread_info *info, int size);
 621
 622 static inline int lod_qos_ost_in_use_clear(const struct lu_env *env, int stripes)
 623 {
 624         struct lod_thread_info *info = lod_env_info(env);
 625
 626         if (info->lti_ea_store_size < sizeof(int) * stripes)
 627                 lod_ea_store_resize(info, stripes * sizeof(int));
 628         if (info->lti_ea_store_size < sizeof(int) * stripes) {
 629                 CERROR("can't allocate memory for ost-in-use array\n");
 630                 return -ENOMEM;
 631         }
 632         memset(info->lti_ea_store, 0, sizeof(int) * stripes);
 633         return 0;
 634 }
 635
 636 static inline void lod_qos_ost_in_use(const struct lu_env *env, int idx, int ost)
 637 {
 638         struct lod_thread_info *info = lod_env_info(env);
 639         int *osts = info->lti_ea_store;
 640
 641         LASSERT(info->lti_ea_store_size >= idx * sizeof(int));
 642         osts[idx] = ost;
 643 }
 644
 645 static int lod_qos_is_ost_used(const struct lu_env *env, int ost, int stripes)
 646 {
 647         struct lod_thread_info *info = lod_env_info(env);
 648         int *osts = info->lti_ea_store;
 649         int j;
 650
 651         for (j = 0; j < stripes; j++) {
 652                 if (osts[j] == ost)
 653                         return 1;
 654         }
 655         return 0;
 656 }
 657
 658 /* Allocate objects on OSTs with round-robin algorithm */
 659 static int lod_alloc_rr(const struct lu_env *env, struct lod_object *lo,
 660                         struct dt_object **stripe, int flags,
 661                         struct thandle *th)
 662 {
 663         struct lod_device *m = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
 664         struct obd_statfs *sfs = &lod_env_info(env)->lti_osfs;
 665         struct pool_desc  *pool = NULL;
 666         struct ost_pool   *osts;
 667         struct lov_qos_rr *lqr;
 668         struct dt_object  *o;
 669         unsigned           array_idx;
 670         int                i, rc;
 671         int                ost_start_idx_temp;
 672         int                speed = 0;
 673         int                stripe_idx = 0;
 674         int                stripe_cnt = lo->ldo_stripenr;
 675         int                stripe_cnt_min = min_stripe_count(stripe_cnt, flags);
 676         __u32              ost_idx;
 677         ENTRY;
 678
 679         if (lo->ldo_pool)
 680                 pool = lod_find_pool(m, lo->ldo_pool);
 681
 682         if (pool != NULL) {
 683                 down_read(&pool_tgt_rw_sem(pool));
 684                 osts = &(pool->pool_obds);
 685                 lqr = &(pool->pool_rr);
 686         } else {
 687                 osts = &(m->lod_pool_info);
 688                 lqr = &(m->lod_qos.lq_rr);
 689         }
 690
 691         rc = lod_qos_calc_rr(m, osts, lqr);
 692         if (rc)
 693                 GOTO(out, rc);
 694
 695         rc = lod_qos_ost_in_use_clear(env, lo->ldo_stripenr);
 696         if (rc)
 697                 GOTO(out, rc);
 698
 699         if (--lqr->lqr_start_count <= 0) {
 700                 lqr->lqr_start_idx = cfs_rand() % osts->op_count;
 701                 lqr->lqr_start_count =
 702                         (LOV_CREATE_RESEED_MIN / max(osts->op_count, 1U) +
 703                          LOV_CREATE_RESEED_MULT) * max(osts->op_count, 1U);
 704         } else if (stripe_cnt_min >= osts->op_count ||
 705                         lqr->lqr_start_idx > osts->op_count) {
 706                 /* If we have allocated from all of the OSTs, slowly
 707                  * precess the next start if the OST/stripe count isn't
 708                  * already doing this for us. */
 709                 lqr->lqr_start_idx %= osts->op_count;
 710                 if (stripe_cnt > 1 && (osts->op_count % stripe_cnt) != 1)
 711                         ++lqr->lqr_offset_idx;
 712         }
 713         down_read(&m->lod_qos.lq_rw_sem);
 714         ost_start_idx_temp = lqr->lqr_start_idx;
 715
 716 repeat_find:
 717         array_idx = (lqr->lqr_start_idx + lqr->lqr_offset_idx) %
 718                         osts->op_count;
 719
 720         QOS_DEBUG("pool '%s' want %d startidx %d startcnt %d offset %d "
 721                   "active %d count %d arrayidx %d\n",
 722                   lo->ldo_pool ? lo->ldo_pool : "",
 723                   stripe_cnt, lqr->lqr_start_idx, lqr->lqr_start_count,
 724                   lqr->lqr_offset_idx, osts->op_count, osts->op_count,
 725                   array_idx);
 726
 727         for (i = 0; i < osts->op_count && stripe_idx < lo->ldo_stripenr;
 728              i++, array_idx = (array_idx + 1) % osts->op_count) {
 729                 ++lqr->lqr_start_idx;
 730                 ost_idx = lqr->lqr_pool.op_array[array_idx];
 731
 732                 QOS_DEBUG("#%d strt %d act %d strp %d ary %d idx %d\n",
 733                           i, lqr->lqr_start_idx, /* XXX: active*/ 0,
 734                           stripe_idx, array_idx, ost_idx);
 735
 736                 if ((ost_idx == LOV_QOS_EMPTY) ||
 737                     !cfs_bitmap_check(m->lod_ost_bitmap, ost_idx))
 738                         continue;
 739
 740                 /* Fail Check before osc_precreate() is called
 741                    so we can only 'fail' single OSC. */
 742                 if (OBD_FAIL_CHECK(OBD_FAIL_MDS_OSC_PRECREATE) && ost_idx == 0)
 743                         continue;
 744
 745                 rc = lod_statfs_and_check(env, m, ost_idx, sfs);
 746                 if (rc) {
 747                         /* this OSP doesn't feel well */
 748                         continue;
 749                 }
 750
 751                 /*
 752                  * skip full devices
 753                  */
 754                 if (lod_qos_dev_is_full(sfs)) {
 755                         QOS_DEBUG("#%d is full\n", ost_idx);
 756                         continue;
 757                 }
 758
 759                 /*
 760                  * We expect number of precreated objects in f_ffree at
 761                  * the first iteration, skip OSPs with no objects ready
 762                  */
 763                 if (sfs->os_fprecreated == 0 && speed == 0) {
 764                         QOS_DEBUG("#%d: precreation is empty\n", ost_idx);
 765                         continue;
 766                 }
 767
 768                 /*
 769                  * try to use another OSP if this one is degraded
 770                  */
 771                 if (sfs->os_state == OS_STATE_DEGRADED && speed < 2) {
 772                         QOS_DEBUG("#%d: degraded\n", ost_idx);
 773                         continue;
 774                 }
 775
 776                 /*
 777                  * do not put >1 objects on a single OST
 778                  */
 779                 if (speed && lod_qos_is_ost_used(env, ost_idx, stripe_idx))
 780                         continue;
 781
 782                 o = lod_qos_declare_object_on(env, m, ost_idx, th);
 783                 if (IS_ERR(o)) {
 784                         CDEBUG(D_OTHER, "can't declare new object on #%u: %d\n",
 785                                ost_idx, (int) PTR_ERR(o));
 786                         rc = PTR_ERR(o);
 787                         continue;
 788                 }
 789
 790                 /*
 791                  * We've successfuly declared (reserved) an object
 792                  */
 793                 lod_qos_ost_in_use(env, stripe_idx, ost_idx);
 794                 stripe[stripe_idx] = o;
 795                 stripe_idx++;
 796
 797         }
 798         if ((speed < 2) && (stripe_idx < stripe_cnt_min)) {
 799                 /* Try again, allowing slower OSCs */
 800                 speed++;
 801                 lqr->lqr_start_idx = ost_start_idx_temp;
 802                 goto repeat_find;
 803         }
 804
 805         up_read(&m->lod_qos.lq_rw_sem);
 806
 807         if (stripe_idx) {
 808                 lo->ldo_stripenr = stripe_idx;
 809                 /* at least one stripe is allocated */
 810                 rc = 0;
 811         } else {
 812                 /* nobody provided us with a single object */
 813                 rc = -ENOSPC;
 814         }
 815
 816 out:
 817         if (pool != NULL) {
 818                 up_read(&pool_tgt_rw_sem(pool));
 819                 /* put back ref got by lod_find_pool() */
 820                 lod_pool_putref(pool);
 821         }
 822
 823         RETURN(rc);
 824 }
 825
 826 /* alloc objects on osts with specific stripe offset */
 827 static int lod_alloc_specific(const struct lu_env *env, struct lod_object *lo,
 828                               struct dt_object **stripe, int flags,
 829                               struct thandle *th)
 830 {
 831         struct lod_device *m = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
 832         struct obd_statfs *sfs = &lod_env_info(env)->lti_osfs;
 833         struct dt_object  *o;
 834         unsigned           ost_idx, array_idx, ost_count;
 835         int                i, rc, stripe_num = 0;
 836         int                speed = 0;
 837         struct pool_desc  *pool = NULL;
 838         struct ost_pool   *osts;
 839         ENTRY;
 840
 841         rc = lod_qos_ost_in_use_clear(env, lo->ldo_stripenr);
 842         if (rc)
 843                 GOTO(out, rc);
 844
 845         if (lo->ldo_pool)
 846                 pool = lod_find_pool(m, lo->ldo_pool);
 847
 848         if (pool != NULL) {
 849                 down_read(&pool_tgt_rw_sem(pool));
 850                 osts = &(pool->pool_obds);
 851         } else {
 852                 osts = &(m->lod_pool_info);
 853         }
 854
 855         ost_count = osts->op_count;
 856
 857 repeat_find:
 858         /* search loi_ost_idx in ost array */
 859         array_idx = 0;
 860         for (i = 0; i < ost_count; i++) {
 861                 if (osts->op_array[i] == lo->ldo_def_stripe_offset) {
 862                         array_idx = i;
 863                         break;
 864                 }
 865         }
 866         if (i == ost_count) {
 867                 CERROR("Start index %d not found in pool '%s'\n",
 868                        lo->ldo_def_stripe_offset,
 869                        lo->ldo_pool ? lo->ldo_pool : "");
 870                 GOTO(out, rc = -EINVAL);
 871         }
 872
 873         for (i = 0; i < ost_count;
 874                         i++, array_idx = (array_idx + 1) % ost_count) {
 875                 ost_idx = osts->op_array[array_idx];
 876
 877                 if (!cfs_bitmap_check(m->lod_ost_bitmap, ost_idx))
 878                         continue;
 879
 880                 /* Fail Check before osc_precreate() is called
 881                    so we can only 'fail' single OSC. */
 882                 if (OBD_FAIL_CHECK(OBD_FAIL_MDS_OSC_PRECREATE) && ost_idx == 0)
 883                         continue;
 884
 885                 /*
 886                  * do not put >1 objects on a single OST
 887                  */
 888                 if (lod_qos_is_ost_used(env, ost_idx, stripe_num))
 889                         continue;
 890
 891                 /* Drop slow OSCs if we can, but not for requested start idx.
 892                  *
 893                  * This means "if OSC is slow and it is not the requested
 894                  * start OST, then it can be skipped, otherwise skip it only
 895                  * if it is inactive/recovering/out-of-space." */
 896
 897                 rc = lod_statfs_and_check(env, m, ost_idx, sfs);
 898                 if (rc) {
 899                         /* this OSP doesn't feel well */
 900                         continue;
 901                 }
 902
 903                 /*
 904                  * We expect number of precreated objects in f_ffree at
 905                  * the first iteration, skip OSPs with no objects ready
 906                  * don't apply this logic to OST specified with stripe_offset
 907                  */
 908                 if (i != 0 && sfs->os_fprecreated == 0 && speed == 0)
 909                         continue;
 910
 911                 o = lod_qos_declare_object_on(env, m, ost_idx, th);
 912                 if (IS_ERR(o)) {
 913                         CDEBUG(D_OTHER, "can't declare new object on #%u: %d\n",
 914                                ost_idx, (int) PTR_ERR(o));
 915                         continue;
 916                 }
 917
 918                 /*
 919                  * We've successfuly declared (reserved) an object
 920                  */
 921                 stripe[stripe_num] = o;
 922                 stripe_num++;
 923
 924                 /* We have enough stripes */
 925                 if (stripe_num == lo->ldo_stripenr)
 926                         GOTO(out, rc = 0);
 927         }
 928         if (speed < 2) {
 929                 /* Try again, allowing slower OSCs */
 930                 speed++;
 931                 goto repeat_find;
 932         }
 933
 934         /* If we were passed specific striping params, then a failure to
 935          * meet those requirements is an error, since we can't reallocate
 936          * that memory (it might be part of a larger array or something).
 937          *
 938          * We can only get here if lsm_stripe_count was originally > 1.
 939          */
 940         CERROR("can't lstripe objid "DFID": have %d want %u\n",
 941                PFID(lu_object_fid(lod2lu_obj(lo))), stripe_num,
 942                lo->ldo_stripenr);
 943         rc = -EFBIG;
 944 out:
 945         if (pool != NULL) {
 946                 up_read(&pool_tgt_rw_sem(pool));
 947                 /* put back ref got by lod_find_pool() */
 948                 lod_pool_putref(pool);
 949         }
 950
 951         RETURN(rc);
 952 }
 953
 954 static inline int lod_qos_is_usable(struct lod_device *lod)
 955 {
 956 #ifdef FORCE_QOS
 957         /* to be able to debug QoS code */
 958         return 1;
 959 #endif
 960
 961         /* Detect -EAGAIN early, before expensive lock is taken. */
 962         if (!lod->lod_qos.lq_dirty && lod->lod_qos.lq_same_space)
 963                 return 0;
 964
 965         if (lod->lod_desc.ld_active_tgt_count < 2)
 966                 return 0;
 967
 968         return 1;
 969 }
 970
 971 /* Alloc objects on OSTs with optimization based on:
 972    - free space
 973    - network resources (shared OSS's)
 974  */
 975 static int lod_alloc_qos(const struct lu_env *env, struct lod_object *lo,
 976                          struct dt_object **stripe, int flags,
 977                          struct thandle *th)
 978 {
 979         struct lod_device   *m = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
 980         struct obd_statfs   *sfs = &lod_env_info(env)->lti_osfs;
 981         struct lod_tgt_desc *ost;
 982         struct dt_object    *o;
 983         __u64                total_weight = 0;
 984         int                  nfound, good_osts, i, rc = 0;
 985         int                  stripe_cnt = lo->ldo_stripenr;
 986         int                  stripe_cnt_min;
 987         struct pool_desc    *pool = NULL;
 988         struct ost_pool    *osts;
 989         ENTRY;
 990
 991         stripe_cnt_min = min_stripe_count(stripe_cnt, flags);
 992         if (stripe_cnt_min < 1)
 993                 RETURN(-EINVAL);
 994
 995         if (lo->ldo_pool)
 996                 pool = lod_find_pool(m, lo->ldo_pool);
 997
 998         if (pool != NULL) {
 999                 down_read(&pool_tgt_rw_sem(pool));
1000                 osts = &(pool->pool_obds);
1001         } else {
1002                 osts = &(m->lod_pool_info);
1003         }
1004
1005         /* Detect -EAGAIN early, before expensive lock is taken. */
1006         if (!lod_qos_is_usable(m))
1007                 GOTO(out_nolock, rc = -EAGAIN);
1008
1009         /* Do actual allocation, use write lock here. */
1010         down_write(&m->lod_qos.lq_rw_sem);
1011
1012         /*
1013          * Check again, while we were sleeping on @lq_rw_sem things could
1014          * change.
1015          */
1016         if (!lod_qos_is_usable(m))
1017                 GOTO(out, rc = -EAGAIN);
1018
1019         rc = lod_qos_calc_ppo(m);
1020         if (rc)
1021                 GOTO(out, rc);
1022
1023         rc = lod_qos_ost_in_use_clear(env, lo->ldo_stripenr);
1024         if (rc)
1025                 GOTO(out, rc);
1026
1027         good_osts = 0;
1028         /* Find all the OSTs that are valid stripe candidates */
1029         for (i = 0; i < osts->op_count; i++) {
1030                 if (!cfs_bitmap_check(m->lod_ost_bitmap, osts->op_array[i]))
1031                         continue;
1032
1033                 rc = lod_statfs_and_check(env, m, osts->op_array[i], sfs);
1034                 if (rc) {
1035                         /* this OSP doesn't feel well */
1036                         continue;
1037                 }
1038
1039                 /*
1040                  * skip full devices
1041                  */
1042                 if (lod_qos_dev_is_full(sfs))
1043                         continue;
1044
1045                 /* Fail Check before osc_precreate() is called
1046                    so we can only 'fail' single OSC. */
1047                 if (OBD_FAIL_CHECK(OBD_FAIL_MDS_OSC_PRECREATE) &&
1048                                    osts->op_array[i] == 0)
1049                         continue;
1050
1051                 ost = OST_TGT(m,osts->op_array[i]);
1052                 ost->ltd_qos.ltq_usable = 1;
1053                 lod_qos_calc_weight(m, osts->op_array[i]);
1054                 total_weight += ost->ltd_qos.ltq_weight;
1055
1056                 good_osts++;
1057         }
1058
1059         QOS_DEBUG("found %d good osts\n", good_osts);
1060
1061         if (good_osts < stripe_cnt_min)
1062                 GOTO(out, rc = -EAGAIN);
1063
1064         /* We have enough osts */
1065         if (good_osts < stripe_cnt)
1066                 stripe_cnt = good_osts;
1067
1068         /* Find enough OSTs with weighted random allocation. */
1069         nfound = 0;
1070         while (nfound < stripe_cnt) {
1071                 __u64 rand, cur_weight;
1072
1073                 cur_weight = 0;
1074                 rc = -ENOSPC;
1075
1076                 if (total_weight) {
1077 #if BITS_PER_LONG == 32
1078                         rand = cfs_rand() % (unsigned)total_weight;
1079                         /* If total_weight > 32-bit, first generate the high
1080                          * 32 bits of the random number, then add in the low
1081                          * 32 bits (truncated to the upper limit, if needed) */
1082                         if (total_weight > 0xffffffffULL)
1083                                 rand = (__u64)(cfs_rand() %
1084                                         (unsigned)(total_weight >> 32)) << 32;
1085                         else
1086                                 rand = 0;
1087
1088                         if (rand == (total_weight & 0xffffffff00000000ULL))
1089                                 rand |= cfs_rand() % (unsigned)total_weight;
1090                         else
1091                                 rand |= cfs_rand();
1092
1093 #else
1094                         rand = ((__u64)cfs_rand() << 32 | cfs_rand()) %
1095                                 total_weight;
1096 #endif
1097                 } else {
1098                         rand = 0;
1099                 }
1100
1101                 /* On average, this will hit larger-weighted osts more often.
1102                    0-weight osts will always get used last (only when rand=0) */
1103                 for (i = 0; i < osts->op_count; i++) {
1104                         int idx = osts->op_array[i];
1105
1106                         if (!cfs_bitmap_check(m->lod_ost_bitmap, idx))
1107                                 continue;
1108
1109                         ost = OST_TGT(m,idx);
1110
1111                         if (!ost->ltd_qos.ltq_usable)
1112                                 continue;
1113
1114                         cur_weight += ost->ltd_qos.ltq_weight;
1115                         QOS_DEBUG("stripe_cnt=%d nfound=%d cur_weight="LPU64
1116                                   " rand="LPU64" total_weight="LPU64"\n",
1117                                   stripe_cnt, nfound, cur_weight, rand,
1118                                   total_weight);
1119
1120                         if (cur_weight < rand)
1121                                 continue;
1122
1123                         QOS_DEBUG("stripe=%d to idx=%d\n", nfound, idx);
1124
1125                         /*
1126                          * do not put >1 objects on a single OST
1127                          */
1128                         if (lod_qos_is_ost_used(env, idx, nfound))
1129                                 continue;
1130                         lod_qos_ost_in_use(env, nfound, idx);
1131
1132                         o = lod_qos_declare_object_on(env, m, idx, th);
1133                         if (IS_ERR(o)) {
1134                                 QOS_DEBUG("can't declare object on #%u: %d\n",
1135                                           idx, (int) PTR_ERR(o));
1136                                 continue;
1137                         }
1138                         stripe[nfound++] = o;
1139                         lod_qos_used(m, osts, idx, &total_weight);
1140                         rc = 0;
1141                         break;
1142                 }
1143
1144                 if (rc) {
1145                         /* no OST found on this iteration, give up */
1146                         break;
1147                 }
1148         }
1149
1150         if (unlikely(nfound != stripe_cnt)) {
1151                 /*
1152                  * when the decision to use weighted algorithm was made
1153                  * we had enough appropriate OSPs, but this state can
1154                  * change anytime (no space on OST, broken connection, etc)
1155                  * so it's possible OSP won't be able to provide us with
1156                  * an object due to just changed state
1157                  */
1158                 LCONSOLE_INFO("wanted %d, found %d\n", stripe_cnt, nfound);
1159                 for (i = 0; i < nfound; i++) {
1160                         LASSERT(stripe[i] != NULL);
1161                         lu_object_put(env, &stripe[i]->do_lu);
1162                         stripe[i] = NULL;
1163                 }
1164
1165                 /* makes sense to rebalance next time */
1166                 m->lod_qos.lq_dirty = 1;
1167                 m->lod_qos.lq_same_space = 0;
1168
1169                 rc = -EAGAIN;
1170         }
1171
1172 out:
1173         up_write(&m->lod_qos.lq_rw_sem);
1174
1175 out_nolock:
1176         if (pool != NULL) {
1177                 up_read(&pool_tgt_rw_sem(pool));
1178                 /* put back ref got by lod_find_pool() */
1179                 lod_pool_putref(pool);
1180         }
1181
1182         RETURN(rc);
1183 }
1184
1185 /* Find the max stripecount we should use */
1186 static __u16 lod_get_stripecnt(struct lod_device *lod, __u32 magic,
1187                                __u16 stripe_count)
1188 {
1189         __u32 max_stripes = LOV_MAX_STRIPE_COUNT_OLD;
1190
1191         if (!stripe_count)
1192                 stripe_count = lod->lod_desc.ld_default_stripe_count;
1193         if (stripe_count > lod->lod_desc.ld_active_tgt_count)
1194                 stripe_count = lod->lod_desc.ld_active_tgt_count;
1195         if (!stripe_count)
1196                 stripe_count = 1;
1197
1198         /* stripe count is based on whether OSD can handle larger EA sizes */
1199         if (lod->lod_osd_max_easize > 0)
1200                 max_stripes = lov_mds_md_stripecnt(lod->lod_osd_max_easize,
1201                                                    magic);
1202
1203         return (stripe_count < max_stripes) ? stripe_count : max_stripes;
1204 }
1205
1206 static int lod_use_defined_striping(const struct lu_env *env,
1207                                     struct lod_object *mo,
1208                                     const struct lu_buf *buf)
1209 {
1210         struct lod_device      *d = lu2lod_dev(lod2lu_obj(mo)->lo_dev);
1211         struct lov_mds_md_v1   *v1 = buf->lb_buf;
1212         struct lov_mds_md_v3   *v3 = buf->lb_buf;
1213         struct lov_ost_data_v1 *objs;
1214         __u32                   magic;
1215         int                     rc;
1216         ENTRY;
1217
1218         rc = lod_verify_striping(d, buf, 1);
1219         if (rc)
1220                 RETURN(rc);
1221
1222         magic = le32_to_cpu(v1->lmm_magic);
1223         if (magic == LOV_MAGIC_V1_DEF) {
1224                 objs = &v1->lmm_objects[0];
1225         } else if (magic == LOV_MAGIC_V3_DEF) {
1226                 objs = &v3->lmm_objects[0];
1227                 lod_object_set_pool(mo, v3->lmm_pool_name);
1228         } else {
1229                 GOTO(out, rc = -EINVAL);
1230         }
1231
1232         /*
1233          * LOD shouldn't be aware of recovery at all,
1234          * but this track recovery status (to some extent)
1235          * to be do additional checks like this one
1236          */
1237         LASSERT(d->lod_recovery_completed == 0);
1238
1239         mo->ldo_stripe_size = le32_to_cpu(v1->lmm_stripe_size);
1240         mo->ldo_stripenr = le16_to_cpu(v1->lmm_stripe_count);
1241         mo->ldo_layout_gen = le16_to_cpu(v1->lmm_layout_gen);
1242         LASSERT(buf->lb_len >= lov_mds_md_size(mo->ldo_stripenr, magic));
1243
1244         rc = lod_initialize_objects(env, mo, objs);
1245
1246 out:
1247         RETURN(rc);
1248 }
1249
1250 static int lod_qos_parse_config(const struct lu_env *env,
1251                                 struct lod_object *lo,
1252                                 const struct lu_buf *buf)
1253 {
1254         struct lod_device     *d = lu2lod_dev(lod2lu_obj(lo)->lo_dev);
1255         struct lov_user_md_v1 *v1 = NULL;
1256         struct lov_user_md_v3 *v3 = NULL;
1257         struct pool_desc      *pool;
1258         __u32                  magic;
1259         int                    rc;
1260         ENTRY;
1261
1262         if (buf == NULL || buf->lb_buf == NULL || buf->lb_len == 0)
1263                 RETURN(0);
1264
1265         v1 = buf->lb_buf;
1266         magic = v1->lmm_magic;
1267
1268         if (magic == __swab32(LOV_USER_MAGIC_V1)) {
1269                 lustre_swab_lov_user_md_v1(v1);
1270                 magic = v1->lmm_magic;
1271         } else if (magic == __swab32(LOV_USER_MAGIC_V3)) {
1272                 v3 = buf->lb_buf;
1273                 lustre_swab_lov_user_md_v3(v3);
1274                 magic = v3->lmm_magic;
1275         }
1276
1277         if (unlikely(magic != LOV_MAGIC_V1 && magic != LOV_MAGIC_V3)) {
1278                 /* try to use as fully defined striping */
1279                 rc = lod_use_defined_striping(env, lo, buf);
1280                 RETURN(rc);
1281         }
1282
1283         if (unlikely(buf->lb_len < sizeof(*v1))) {
1284                 CERROR("wrong size: %u\n", (unsigned) buf->lb_len);
1285                 RETURN(-EINVAL);
1286         }
1287
1288         if (v1->lmm_pattern != 0 && v1->lmm_pattern != LOV_PATTERN_RAID0) {
1289                 CERROR("invalid pattern: %x\n", v1->lmm_pattern);
1290                 RETURN(-EINVAL);
1291         }
1292
1293         if (v1->lmm_stripe_size)
1294                 lo->ldo_stripe_size = v1->lmm_stripe_size;
1295         if (lo->ldo_stripe_size & (LOV_MIN_STRIPE_SIZE - 1))
1296                 lo->ldo_stripe_size = LOV_MIN_STRIPE_SIZE;
1297
1298         if (v1->lmm_stripe_count)
1299                 lo->ldo_stripenr = v1->lmm_stripe_count;
1300
1301         if ((v1->lmm_stripe_offset >= d->lod_desc.ld_tgt_count) &&
1302             (v1->lmm_stripe_offset != (typeof(v1->lmm_stripe_offset))(-1))) {
1303                 CERROR("invalid offset: %x\n", v1->lmm_stripe_offset);
1304                 RETURN(-EINVAL);
1305         }
1306         lo->ldo_def_stripe_offset = v1->lmm_stripe_offset;
1307
1308         CDEBUG(D_OTHER, "lsm: %u size, %u stripes, %u offset\n",
1309                v1->lmm_stripe_size, v1->lmm_stripe_count,
1310                v1->lmm_stripe_offset);
1311
1312         if (v1->lmm_magic == LOV_MAGIC_V3) {
1313                 if (buf->lb_len < sizeof(*v3)) {
1314                         CERROR("wrong size: %u\n", (unsigned) buf->lb_len);
1315                         RETURN(-EINVAL);
1316                 }
1317
1318                 v3 = buf->lb_buf;
1319                 lod_object_set_pool(lo, v3->lmm_pool_name);
1320
1321                 /* In the function below, .hs_keycmp resolves to
1322                  * pool_hashkey_keycmp() */
1323                 /* coverity[overrun-buffer-val] */
1324                 pool = lod_find_pool(d, v3->lmm_pool_name);
1325                 if (pool != NULL) {
1326                         if (lo->ldo_def_stripe_offset !=
1327                             (typeof(v1->lmm_stripe_offset))(-1)) {
1328                                 rc = lo->ldo_def_stripe_offset;
1329                                 rc = lod_check_index_in_pool(rc, pool);
1330                                 if (rc < 0) {
1331                                         lod_pool_putref(pool);
1332                                         CERROR("invalid offset\n");
1333                                         RETURN(-EINVAL);
1334                                 }
1335                         }
1336
1337                         if (lo->ldo_stripenr > pool_tgt_count(pool))
1338                                 lo->ldo_stripenr= pool_tgt_count(pool);
1339
1340                         lod_pool_putref(pool);
1341                 }
1342         } else
1343                 lod_object_set_pool(lo, NULL);
1344
1345         RETURN(0);
1346 }
1347
1348 /*
1349  * buf should be NULL or contain striping settings
1350  */
1351 int lod_qos_prep_create(const struct lu_env *env, struct lod_object *lo,
1352                         struct lu_attr *attr, const struct lu_buf *buf,
1353                         struct thandle *th)
1354 {
1355         struct lod_device      *d = lu2lod_dev(lod2lu_obj(lo)->lo_dev);
1356         struct dt_object      **stripe;
1357         int                     stripe_len;
1358         int                     flag = LOV_USES_ASSIGNED_STRIPE;
1359         int                     i, rc;
1360         ENTRY;
1361
1362         LASSERT(lo);
1363
1364         /* no OST available */
1365         /* XXX: should we be waiting a bit to prevent failures during
1366          * cluster initialization? */
1367         if (d->lod_ostnr == 0)
1368                 GOTO(out, rc = -EIO);
1369
1370         /*
1371          * by this time, the object's ldo_stripenr and ldo_stripe_size
1372          * contain default value for striping: taken from the parent
1373          * or from filesystem defaults
1374          *
1375          * in case the caller is passing lovea with new striping config,
1376          * we may need to parse lovea and apply new configuration
1377          */
1378         rc = lod_qos_parse_config(env, lo, buf);
1379         if (rc)
1380                 GOTO(out, rc);
1381
1382         if (likely(lo->ldo_stripe == NULL)) {
1383                 /*
1384                  * no striping has been created so far
1385                  */
1386                 LASSERT(lo->ldo_stripenr > 0);
1387                 lo->ldo_stripenr = lod_get_stripecnt(d, LOV_MAGIC,
1388                                 lo->ldo_stripenr);
1389
1390                 stripe_len = lo->ldo_stripenr;
1391                 OBD_ALLOC(stripe, sizeof(stripe[0]) * stripe_len);
1392                 if (stripe == NULL)
1393                         GOTO(out, rc = -ENOMEM);
1394
1395                 lod_getref(&d->lod_ost_descs);
1396                 /* XXX: support for non-0 files w/o objects */
1397                 if (lo->ldo_def_stripe_offset >= d->lod_desc.ld_tgt_count) {
1398                         lod_qos_statfs_update(env, d);
1399                         rc = lod_alloc_qos(env, lo, stripe, flag, th);
1400                         if (rc == -EAGAIN)
1401                                 rc = lod_alloc_rr(env, lo, stripe, flag, th);
1402                 } else {
1403                         rc = lod_alloc_specific(env, lo, stripe, flag, th);
1404                 }
1405                 lod_putref(d, &d->lod_ost_descs);
1406
1407                 if (rc < 0) {
1408                         for (i = 0; i < stripe_len; i++)
1409                                 if (stripe[i] != NULL)
1410                                         lu_object_put(env, &stripe[i]->do_lu);
1411
1412                         OBD_FREE(stripe, sizeof(stripe[0]) * stripe_len);
1413                 } else {
1414                         lo->ldo_stripe = stripe;
1415                         lo->ldo_stripes_allocated = stripe_len;
1416                 }
1417         } else {
1418                 /*
1419                  * lod_qos_parse_config() found supplied buf as a predefined
1420                  * striping (not a hint), so it allocated all the object
1421                  * now we need to create them
1422                  */
1423                 for (i = 0; i < lo->ldo_stripenr; i++) {
1424                         struct dt_object  *o;
1425
1426                         o = lo->ldo_stripe[i];
1427                         LASSERT(o);
1428
1429                         rc = dt_declare_create(env, o, attr, NULL, NULL, th);
1430                         if (rc) {
1431                                 CERROR("can't declare create: %d\n", rc);
1432                                 break;
1433                         }
1434                 }
1435         }
1436
1437 out:
1438         RETURN(rc);
1439 }
1440