lustre/osc/osc_request.c

   1 /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
   2  * vim:expandtab:shiftwidth=8:tabstop=8:
   3  *
   4  * GPL HEADER START
   5  *
   6  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License version 2 only,
  10  * as published by the Free Software Foundation.
  11  *
  12  * This program is distributed in the hope that it will be useful, but
  13  * WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15  * General Public License version 2 for more details (a copy is included
  16  * in the LICENSE file that accompanied this code).
  17  *
  18  * You should have received a copy of the GNU General Public License
  19  * version 2 along with this program; If not, see
  20  * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
  21  *
  22  * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
  23  * CA 95054 USA or visit www.sun.com if you need additional information or
  24  * have any questions.
  25  *
  26  * GPL HEADER END
  27  */
  28 /*
  29  * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
  30  * Use is subject to license terms.
  31  */
  32 /*
  33  * Copyright (c) 2011 Whamcloud, Inc.
  34  */
  35 /*
  36  * This file is part of Lustre, http://www.lustre.org/
  37  * Lustre is a trademark of Sun Microsystems, Inc.
  38  */
  39
  40 #ifndef EXPORT_SYMTAB
  41 # define EXPORT_SYMTAB
  42 #endif
  43 #define DEBUG_SUBSYSTEM S_OSC
  44
  45 #include <libcfs/libcfs.h>
  46
  47 #ifndef __KERNEL__
  48 # include <liblustre.h>
  49 #endif
  50
  51 #include <lustre_dlm.h>
  52 #include <lustre_net.h>
  53 #include <lustre/lustre_user.h>
  54 #include <obd_cksum.h>
  55 #include <obd_ost.h>
  56 #include <obd_lov.h>
  57
  58 #ifdef  __CYGWIN__
  59 # include <ctype.h>
  60 #endif
  61
  62 #include <lustre_ha.h>
  63 #include <lprocfs_status.h>
  64 #include <lustre_log.h>
  65 #include <lustre_debug.h>
  66 #include <lustre_param.h>
  67 #include "osc_internal.h"
  68
  69 static quota_interface_t *quota_interface = NULL;
  70 extern quota_interface_t osc_quota_interface;
  71
  72 static void osc_release_ppga(struct brw_page **ppga, obd_count count);
  73 static int brw_interpret(const struct lu_env *env,
  74                          struct ptlrpc_request *req, void *data, int rc);
  75 int osc_cleanup(struct obd_device *obd);
  76
  77 /* Pack OSC object metadata for disk storage (LE byte order). */
  78 static int osc_packmd(struct obd_export *exp, struct lov_mds_md **lmmp,
  79                       struct lov_stripe_md *lsm)
  80 {
  81         int lmm_size;
  82         ENTRY;
  83
  84         lmm_size = sizeof(**lmmp);
  85         if (!lmmp)
  86                 RETURN(lmm_size);
  87
  88         if (*lmmp && !lsm) {
  89                 OBD_FREE(*lmmp, lmm_size);
  90                 *lmmp = NULL;
  91                 RETURN(0);
  92         }
  93
  94         if (!*lmmp) {
  95                 OBD_ALLOC(*lmmp, lmm_size);
  96                 if (!*lmmp)
  97                         RETURN(-ENOMEM);
  98         }
  99
 100         if (lsm) {
 101                 LASSERT(lsm->lsm_object_id);
 102                 LASSERT_SEQ_IS_MDT(lsm->lsm_object_seq);
 103                 (*lmmp)->lmm_object_id = cpu_to_le64(lsm->lsm_object_id);
 104                 (*lmmp)->lmm_object_seq = cpu_to_le64(lsm->lsm_object_seq);
 105         }
 106
 107         RETURN(lmm_size);
 108 }
 109
 110 /* Unpack OSC object metadata from disk storage (LE byte order). */
 111 static int osc_unpackmd(struct obd_export *exp, struct lov_stripe_md **lsmp,
 112                         struct lov_mds_md *lmm, int lmm_bytes)
 113 {
 114         int lsm_size;
 115         struct obd_import *imp = class_exp2cliimp(exp);
 116         ENTRY;
 117
 118         if (lmm != NULL) {
 119                 if (lmm_bytes < sizeof (*lmm)) {
 120                         CERROR("lov_mds_md too small: %d, need %d\n",
 121                                lmm_bytes, (int)sizeof(*lmm));
 122                         RETURN(-EINVAL);
 123                 }
 124                 /* XXX LOV_MAGIC etc check? */
 125
 126                 if (lmm->lmm_object_id == 0) {
 127                         CERROR("lov_mds_md: zero lmm_object_id\n");
 128                         RETURN(-EINVAL);
 129                 }
 130         }
 131
 132         lsm_size = lov_stripe_md_size(1);
 133         if (lsmp == NULL)
 134                 RETURN(lsm_size);
 135
 136         if (*lsmp != NULL && lmm == NULL) {
 137                 OBD_FREE((*lsmp)->lsm_oinfo[0], sizeof(struct lov_oinfo));
 138                 OBD_FREE(*lsmp, lsm_size);
 139                 *lsmp = NULL;
 140                 RETURN(0);
 141         }
 142
 143         if (*lsmp == NULL) {
 144                 OBD_ALLOC(*lsmp, lsm_size);
 145                 if (*lsmp == NULL)
 146                         RETURN(-ENOMEM);
 147                 OBD_ALLOC((*lsmp)->lsm_oinfo[0], sizeof(struct lov_oinfo));
 148                 if ((*lsmp)->lsm_oinfo[0] == NULL) {
 149                         OBD_FREE(*lsmp, lsm_size);
 150                         RETURN(-ENOMEM);
 151                 }
 152                 loi_init((*lsmp)->lsm_oinfo[0]);
 153         }
 154
 155         if (lmm != NULL) {
 156                 /* XXX zero *lsmp? */
 157                 (*lsmp)->lsm_object_id = le64_to_cpu (lmm->lmm_object_id);
 158                 (*lsmp)->lsm_object_seq = le64_to_cpu (lmm->lmm_object_seq);
 159                 LASSERT((*lsmp)->lsm_object_id);
 160                 LASSERT_SEQ_IS_MDT((*lsmp)->lsm_object_seq);
 161         }
 162
 163         if (imp != NULL &&
 164             (imp->imp_connect_data.ocd_connect_flags & OBD_CONNECT_MAXBYTES))
 165                 (*lsmp)->lsm_maxbytes = imp->imp_connect_data.ocd_maxbytes;
 166         else
 167                 (*lsmp)->lsm_maxbytes = LUSTRE_STRIPE_MAXBYTES;
 168
 169         RETURN(lsm_size);
 170 }
 171
 172 static inline void osc_pack_capa(struct ptlrpc_request *req,
 173                                  struct ost_body *body, void *capa)
 174 {
 175         struct obd_capa *oc = (struct obd_capa *)capa;
 176         struct lustre_capa *c;
 177
 178         if (!capa)
 179                 return;
 180
 181         c = req_capsule_client_get(&req->rq_pill, &RMF_CAPA1);
 182         LASSERT(c);
 183         capa_cpy(c, oc);
 184         body->oa.o_valid |= OBD_MD_FLOSSCAPA;
 185         DEBUG_CAPA(D_SEC, c, "pack");
 186 }
 187
 188 static inline void osc_pack_req_body(struct ptlrpc_request *req,
 189                                      struct obd_info *oinfo)
 190 {
 191         struct ost_body *body;
 192
 193         body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
 194         LASSERT(body);
 195
 196         lustre_set_wire_obdo(&body->oa, oinfo->oi_oa);
 197         osc_pack_capa(req, body, oinfo->oi_capa);
 198 }
 199
 200 static inline void osc_set_capa_size(struct ptlrpc_request *req,
 201                                      const struct req_msg_field *field,
 202                                      struct obd_capa *oc)
 203 {
 204         if (oc == NULL)
 205                 req_capsule_set_size(&req->rq_pill, field, RCL_CLIENT, 0);
 206         else
 207                 /* it is already calculated as sizeof struct obd_capa */
 208                 ;
 209 }
 210
 211 static int osc_getattr_interpret(const struct lu_env *env,
 212                                  struct ptlrpc_request *req,
 213                                  struct osc_async_args *aa, int rc)
 214 {
 215         struct ost_body *body;
 216         ENTRY;
 217
 218         if (rc != 0)
 219                 GOTO(out, rc);
 220
 221         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
 222         if (body) {
 223                 CDEBUG(D_INODE, "mode: %o\n", body->oa.o_mode);
 224                 lustre_get_wire_obdo(aa->aa_oi->oi_oa, &body->oa);
 225
 226                 /* This should really be sent by the OST */
 227                 aa->aa_oi->oi_oa->o_blksize = PTLRPC_MAX_BRW_SIZE;
 228                 aa->aa_oi->oi_oa->o_valid |= OBD_MD_FLBLKSZ;
 229         } else {
 230                 CDEBUG(D_INFO, "can't unpack ost_body\n");
 231                 rc = -EPROTO;
 232                 aa->aa_oi->oi_oa->o_valid = 0;
 233         }
 234 out:
 235         rc = aa->aa_oi->oi_cb_up(aa->aa_oi, rc);
 236         RETURN(rc);
 237 }
 238
 239 static int osc_getattr_async(struct obd_export *exp, struct obd_info *oinfo,
 240                              struct ptlrpc_request_set *set)
 241 {
 242         struct ptlrpc_request *req;
 243         struct osc_async_args *aa;
 244         int                    rc;
 245         ENTRY;
 246
 247         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_GETATTR);
 248         if (req == NULL)
 249                 RETURN(-ENOMEM);
 250
 251         osc_set_capa_size(req, &RMF_CAPA1, oinfo->oi_capa);
 252         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GETATTR);
 253         if (rc) {
 254                 ptlrpc_request_free(req);
 255                 RETURN(rc);
 256         }
 257
 258         osc_pack_req_body(req, oinfo);
 259
 260         ptlrpc_request_set_replen(req);
 261         req->rq_interpret_reply = (ptlrpc_interpterer_t)osc_getattr_interpret;
 262
 263         CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
 264         aa = ptlrpc_req_async_args(req);
 265         aa->aa_oi = oinfo;
 266
 267         ptlrpc_set_add_req(set, req);
 268         RETURN(0);
 269 }
 270
 271 static int osc_getattr(struct obd_export *exp, struct obd_info *oinfo)
 272 {
 273         struct ptlrpc_request *req;
 274         struct ost_body       *body;
 275         int                    rc;
 276         ENTRY;
 277
 278         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_GETATTR);
 279         if (req == NULL)
 280                 RETURN(-ENOMEM);
 281
 282         osc_set_capa_size(req, &RMF_CAPA1, oinfo->oi_capa);
 283         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GETATTR);
 284         if (rc) {
 285                 ptlrpc_request_free(req);
 286                 RETURN(rc);
 287         }
 288
 289         osc_pack_req_body(req, oinfo);
 290
 291         ptlrpc_request_set_replen(req);
 292
 293         rc = ptlrpc_queue_wait(req);
 294         if (rc)
 295                 GOTO(out, rc);
 296
 297         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
 298         if (body == NULL)
 299                 GOTO(out, rc = -EPROTO);
 300
 301         CDEBUG(D_INODE, "mode: %o\n", body->oa.o_mode);
 302         lustre_get_wire_obdo(oinfo->oi_oa, &body->oa);
 303
 304         /* This should really be sent by the OST */
 305         oinfo->oi_oa->o_blksize = PTLRPC_MAX_BRW_SIZE;
 306         oinfo->oi_oa->o_valid |= OBD_MD_FLBLKSZ;
 307
 308         EXIT;
 309  out:
 310         ptlrpc_req_finished(req);
 311         return rc;
 312 }
 313
 314 static int osc_setattr(struct obd_export *exp, struct obd_info *oinfo,
 315                        struct obd_trans_info *oti)
 316 {
 317         struct ptlrpc_request *req;
 318         struct ost_body       *body;
 319         int                    rc;
 320         ENTRY;
 321
 322         LASSERT(oinfo->oi_oa->o_valid & OBD_MD_FLGROUP);
 323
 324         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SETATTR);
 325         if (req == NULL)
 326                 RETURN(-ENOMEM);
 327
 328         osc_set_capa_size(req, &RMF_CAPA1, oinfo->oi_capa);
 329         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SETATTR);
 330         if (rc) {
 331                 ptlrpc_request_free(req);
 332                 RETURN(rc);
 333         }
 334
 335         osc_pack_req_body(req, oinfo);
 336
 337         ptlrpc_request_set_replen(req);
 338
 339         rc = ptlrpc_queue_wait(req);
 340         if (rc)
 341                 GOTO(out, rc);
 342
 343         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
 344         if (body == NULL)
 345                 GOTO(out, rc = -EPROTO);
 346
 347         lustre_get_wire_obdo(oinfo->oi_oa, &body->oa);
 348
 349         EXIT;
 350 out:
 351         ptlrpc_req_finished(req);
 352         RETURN(rc);
 353 }
 354
 355 static int osc_setattr_interpret(const struct lu_env *env,
 356                                  struct ptlrpc_request *req,
 357                                  struct osc_setattr_args *sa, int rc)
 358 {
 359         struct ost_body *body;
 360         ENTRY;
 361
 362         if (rc != 0)
 363                 GOTO(out, rc);
 364
 365         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
 366         if (body == NULL)
 367                 GOTO(out, rc = -EPROTO);
 368
 369         lustre_get_wire_obdo(sa->sa_oa, &body->oa);
 370 out:
 371         rc = sa->sa_upcall(sa->sa_cookie, rc);
 372         RETURN(rc);
 373 }
 374
 375 int osc_setattr_async_base(struct obd_export *exp, struct obd_info *oinfo,
 376                            struct obd_trans_info *oti,
 377                            obd_enqueue_update_f upcall, void *cookie,
 378                            struct ptlrpc_request_set *rqset)
 379 {
 380         struct ptlrpc_request   *req;
 381         struct osc_setattr_args *sa;
 382         int                      rc;
 383         ENTRY;
 384
 385         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SETATTR);
 386         if (req == NULL)
 387                 RETURN(-ENOMEM);
 388
 389         osc_set_capa_size(req, &RMF_CAPA1, oinfo->oi_capa);
 390         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SETATTR);
 391         if (rc) {
 392                 ptlrpc_request_free(req);
 393                 RETURN(rc);
 394         }
 395
 396         if (oti && oinfo->oi_oa->o_valid & OBD_MD_FLCOOKIE)
 397                 oinfo->oi_oa->o_lcookie = *oti->oti_logcookies;
 398
 399         osc_pack_req_body(req, oinfo);
 400
 401         ptlrpc_request_set_replen(req);
 402
 403         /* do mds to ost setattr asynchronously */
 404         if (!rqset) {
 405                 /* Do not wait for response. */
 406                 ptlrpcd_add_req(req, PSCOPE_OTHER);
 407         } else {
 408                 req->rq_interpret_reply =
 409                         (ptlrpc_interpterer_t)osc_setattr_interpret;
 410
 411                 CLASSERT (sizeof(*sa) <= sizeof(req->rq_async_args));
 412                 sa = ptlrpc_req_async_args(req);
 413                 sa->sa_oa = oinfo->oi_oa;
 414                 sa->sa_upcall = upcall;
 415                 sa->sa_cookie = cookie;
 416
 417                 if (rqset == PTLRPCD_SET)
 418                         ptlrpcd_add_req(req, PSCOPE_OTHER);
 419                 else
 420                         ptlrpc_set_add_req(rqset, req);
 421         }
 422
 423         RETURN(0);
 424 }
 425
 426 static int osc_setattr_async(struct obd_export *exp, struct obd_info *oinfo,
 427                              struct obd_trans_info *oti,
 428                              struct ptlrpc_request_set *rqset)
 429 {
 430         return osc_setattr_async_base(exp, oinfo, oti,
 431                                       oinfo->oi_cb_up, oinfo, rqset);
 432 }
 433
 434 int osc_real_create(struct obd_export *exp, struct obdo *oa,
 435                     struct lov_stripe_md **ea, struct obd_trans_info *oti)
 436 {
 437         struct ptlrpc_request *req;
 438         struct ost_body       *body;
 439         struct lov_stripe_md  *lsm;
 440         int                    rc;
 441         ENTRY;
 442
 443         LASSERT(oa);
 444         LASSERT(ea);
 445
 446         lsm = *ea;
 447         if (!lsm) {
 448                 rc = obd_alloc_memmd(exp, &lsm);
 449                 if (rc < 0)
 450                         RETURN(rc);
 451         }
 452
 453         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_CREATE);
 454         if (req == NULL)
 455                 GOTO(out, rc = -ENOMEM);
 456
 457         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_CREATE);
 458         if (rc) {
 459                 ptlrpc_request_free(req);
 460                 GOTO(out, rc);
 461         }
 462
 463         body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
 464         LASSERT(body);
 465         lustre_set_wire_obdo(&body->oa, oa);
 466
 467         ptlrpc_request_set_replen(req);
 468
 469         if ((oa->o_valid & OBD_MD_FLFLAGS) &&
 470             oa->o_flags == OBD_FL_DELORPHAN) {
 471                 DEBUG_REQ(D_HA, req,
 472                           "delorphan from OST integration");
 473                 /* Don't resend the delorphan req */
 474                 req->rq_no_resend = req->rq_no_delay = 1;
 475         }
 476
 477         rc = ptlrpc_queue_wait(req);
 478         if (rc)
 479                 GOTO(out_req, rc);
 480
 481         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
 482         if (body == NULL)
 483                 GOTO(out_req, rc = -EPROTO);
 484
 485         lustre_get_wire_obdo(oa, &body->oa);
 486
 487         /* This should really be sent by the OST */
 488         oa->o_blksize = PTLRPC_MAX_BRW_SIZE;
 489         oa->o_valid |= OBD_MD_FLBLKSZ;
 490
 491         /* XXX LOV STACKING: the lsm that is passed to us from LOV does not
 492          * have valid lsm_oinfo data structs, so don't go touching that.
 493          * This needs to be fixed in a big way.
 494          */
 495         lsm->lsm_object_id = oa->o_id;
 496         lsm->lsm_object_seq = oa->o_seq;
 497         *ea = lsm;
 498
 499         if (oti != NULL) {
 500                 oti->oti_transno = lustre_msg_get_transno(req->rq_repmsg);
 501
 502                 if (oa->o_valid & OBD_MD_FLCOOKIE) {
 503                         if (!oti->oti_logcookies)
 504                                 oti_alloc_cookies(oti, 1);
 505                         *oti->oti_logcookies = oa->o_lcookie;
 506                 }
 507         }
 508
 509         CDEBUG(D_HA, "transno: "LPD64"\n",
 510                lustre_msg_get_transno(req->rq_repmsg));
 511 out_req:
 512         ptlrpc_req_finished(req);
 513 out:
 514         if (rc && !*ea)
 515                 obd_free_memmd(exp, &lsm);
 516         RETURN(rc);
 517 }
 518
 519 int osc_punch_base(struct obd_export *exp, struct obd_info *oinfo,
 520                    obd_enqueue_update_f upcall, void *cookie,
 521                    struct ptlrpc_request_set *rqset)
 522 {
 523         struct ptlrpc_request   *req;
 524         struct osc_setattr_args *sa;
 525         struct ost_body         *body;
 526         int                      rc;
 527         ENTRY;
 528
 529         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_PUNCH);
 530         if (req == NULL)
 531                 RETURN(-ENOMEM);
 532
 533         osc_set_capa_size(req, &RMF_CAPA1, oinfo->oi_capa);
 534         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_PUNCH);
 535         if (rc) {
 536                 ptlrpc_request_free(req);
 537                 RETURN(rc);
 538         }
 539         req->rq_request_portal = OST_IO_PORTAL; /* bug 7198 */
 540         ptlrpc_at_set_req_timeout(req);
 541
 542         body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
 543         LASSERT(body);
 544         lustre_set_wire_obdo(&body->oa, oinfo->oi_oa);
 545         osc_pack_capa(req, body, oinfo->oi_capa);
 546
 547         ptlrpc_request_set_replen(req);
 548
 549
 550         req->rq_interpret_reply = (ptlrpc_interpterer_t)osc_setattr_interpret;
 551         CLASSERT (sizeof(*sa) <= sizeof(req->rq_async_args));
 552         sa = ptlrpc_req_async_args(req);
 553         sa->sa_oa     = oinfo->oi_oa;
 554         sa->sa_upcall = upcall;
 555         sa->sa_cookie = cookie;
 556         if (rqset == PTLRPCD_SET)
 557                 ptlrpcd_add_req(req, PSCOPE_OTHER);
 558         else
 559                 ptlrpc_set_add_req(rqset, req);
 560
 561         RETURN(0);
 562 }
 563
 564 static int osc_punch(struct obd_export *exp, struct obd_info *oinfo,
 565                      struct obd_trans_info *oti,
 566                      struct ptlrpc_request_set *rqset)
 567 {
 568         oinfo->oi_oa->o_size   = oinfo->oi_policy.l_extent.start;
 569         oinfo->oi_oa->o_blocks = oinfo->oi_policy.l_extent.end;
 570         oinfo->oi_oa->o_valid |= OBD_MD_FLSIZE | OBD_MD_FLBLOCKS;
 571         return osc_punch_base(exp, oinfo,
 572                               oinfo->oi_cb_up, oinfo, rqset);
 573 }
 574
 575 static int osc_sync_interpret(const struct lu_env *env,
 576                               struct ptlrpc_request *req,
 577                               void *arg, int rc)
 578 {
 579         struct osc_async_args *aa = arg;
 580         struct ost_body *body;
 581         ENTRY;
 582
 583         if (rc)
 584                 GOTO(out, rc);
 585
 586         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
 587         if (body == NULL) {
 588                 CERROR ("can't unpack ost_body\n");
 589                 GOTO(out, rc = -EPROTO);
 590         }
 591
 592         *aa->aa_oi->oi_oa = body->oa;
 593 out:
 594         rc = aa->aa_oi->oi_cb_up(aa->aa_oi, rc);
 595         RETURN(rc);
 596 }
 597
 598 static int osc_sync(struct obd_export *exp, struct obd_info *oinfo,
 599                     obd_size start, obd_size end,
 600                     struct ptlrpc_request_set *set)
 601 {
 602         struct ptlrpc_request *req;
 603         struct ost_body       *body;
 604         struct osc_async_args *aa;
 605         int                    rc;
 606         ENTRY;
 607
 608         if (!oinfo->oi_oa) {
 609                 CDEBUG(D_INFO, "oa NULL\n");
 610                 RETURN(-EINVAL);
 611         }
 612
 613         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SYNC);
 614         if (req == NULL)
 615                 RETURN(-ENOMEM);
 616
 617         osc_set_capa_size(req, &RMF_CAPA1, oinfo->oi_capa);
 618         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SYNC);
 619         if (rc) {
 620                 ptlrpc_request_free(req);
 621                 RETURN(rc);
 622         }
 623
 624         /* overload the size and blocks fields in the oa with start/end */
 625         body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
 626         LASSERT(body);
 627         lustre_set_wire_obdo(&body->oa, oinfo->oi_oa);
 628         body->oa.o_size = start;
 629         body->oa.o_blocks = end;
 630         body->oa.o_valid |= (OBD_MD_FLSIZE | OBD_MD_FLBLOCKS);
 631         osc_pack_capa(req, body, oinfo->oi_capa);
 632
 633         ptlrpc_request_set_replen(req);
 634         req->rq_interpret_reply = osc_sync_interpret;
 635
 636         CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
 637         aa = ptlrpc_req_async_args(req);
 638         aa->aa_oi = oinfo;
 639
 640         ptlrpc_set_add_req(set, req);
 641         RETURN (0);
 642 }
 643
 644 /* Find and cancel locally locks matched by @mode in the resource found by
 645  * @objid. Found locks are added into @cancel list. Returns the amount of
 646  * locks added to @cancels list. */
 647 static int osc_resource_get_unused(struct obd_export *exp, struct obdo *oa,
 648                                    cfs_list_t *cancels,
 649                                    ldlm_mode_t mode, int lock_flags)
 650 {
 651         struct ldlm_namespace *ns = exp->exp_obd->obd_namespace;
 652         struct ldlm_res_id res_id;
 653         struct ldlm_resource *res;
 654         int count;
 655         ENTRY;
 656
 657         osc_build_res_name(oa->o_id, oa->o_seq, &res_id);
 658         res = ldlm_resource_get(ns, NULL, &res_id, 0, 0);
 659         if (res == NULL)
 660                 RETURN(0);
 661
 662         LDLM_RESOURCE_ADDREF(res);
 663         count = ldlm_cancel_resource_local(res, cancels, NULL, mode,
 664                                            lock_flags, 0, NULL);
 665         LDLM_RESOURCE_DELREF(res);
 666         ldlm_resource_putref(res);
 667         RETURN(count);
 668 }
 669
 670 static int osc_destroy_interpret(const struct lu_env *env,
 671                                  struct ptlrpc_request *req, void *data,
 672                                  int rc)
 673 {
 674         struct client_obd *cli = &req->rq_import->imp_obd->u.cli;
 675
 676         cfs_atomic_dec(&cli->cl_destroy_in_flight);
 677         cfs_waitq_signal(&cli->cl_destroy_waitq);
 678         return 0;
 679 }
 680
 681 static int osc_can_send_destroy(struct client_obd *cli)
 682 {
 683         if (cfs_atomic_inc_return(&cli->cl_destroy_in_flight) <=
 684             cli->cl_max_rpcs_in_flight) {
 685                 /* The destroy request can be sent */
 686                 return 1;
 687         }
 688         if (cfs_atomic_dec_return(&cli->cl_destroy_in_flight) <
 689             cli->cl_max_rpcs_in_flight) {
 690                 /*
 691                  * The counter has been modified between the two atomic
 692                  * operations.
 693                  */
 694                 cfs_waitq_signal(&cli->cl_destroy_waitq);
 695         }
 696         return 0;
 697 }
 698
 699 /* Destroy requests can be async always on the client, and we don't even really
 700  * care about the return code since the client cannot do anything at all about
 701  * a destroy failure.
 702  * When the MDS is unlinking a filename, it saves the file objects into a
 703  * recovery llog, and these object records are cancelled when the OST reports
 704  * they were destroyed and sync'd to disk (i.e. transaction committed).
 705  * If the client dies, or the OST is down when the object should be destroyed,
 706  * the records are not cancelled, and when the OST reconnects to the MDS next,
 707  * it will retrieve the llog unlink logs and then sends the log cancellation
 708  * cookies to the MDS after committing destroy transactions. */
 709 static int osc_destroy(struct obd_export *exp, struct obdo *oa,
 710                        struct lov_stripe_md *ea, struct obd_trans_info *oti,
 711                        struct obd_export *md_export, void *capa)
 712 {
 713         struct client_obd     *cli = &exp->exp_obd->u.cli;
 714         struct ptlrpc_request *req;
 715         struct ost_body       *body;
 716         CFS_LIST_HEAD(cancels);
 717         int rc, count;
 718         ENTRY;
 719
 720         if (!oa) {
 721                 CDEBUG(D_INFO, "oa NULL\n");
 722                 RETURN(-EINVAL);
 723         }
 724
 725         count = osc_resource_get_unused(exp, oa, &cancels, LCK_PW,
 726                                         LDLM_FL_DISCARD_DATA);
 727
 728         req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_DESTROY);
 729         if (req == NULL) {
 730                 ldlm_lock_list_put(&cancels, l_bl_ast, count);
 731                 RETURN(-ENOMEM);
 732         }
 733
 734         osc_set_capa_size(req, &RMF_CAPA1, (struct obd_capa *)capa);
 735         rc = ldlm_prep_elc_req(exp, req, LUSTRE_OST_VERSION, OST_DESTROY,
 736                                0, &cancels, count);
 737         if (rc) {
 738                 ptlrpc_request_free(req);
 739                 RETURN(rc);
 740         }
 741
 742         req->rq_request_portal = OST_IO_PORTAL; /* bug 7198 */
 743         ptlrpc_at_set_req_timeout(req);
 744
 745         if (oti != NULL && oa->o_valid & OBD_MD_FLCOOKIE)
 746                 oa->o_lcookie = *oti->oti_logcookies;
 747         body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
 748         LASSERT(body);
 749         lustre_set_wire_obdo(&body->oa, oa);
 750
 751         osc_pack_capa(req, body, (struct obd_capa *)capa);
 752         ptlrpc_request_set_replen(req);
 753
 754         /* don't throttle destroy RPCs for the MDT */
 755         if (!(cli->cl_import->imp_connect_flags_orig & OBD_CONNECT_MDS)) {
 756                 req->rq_interpret_reply = osc_destroy_interpret;
 757                 if (!osc_can_send_destroy(cli)) {
 758                         struct l_wait_info lwi = LWI_INTR(LWI_ON_SIGNAL_NOOP,
 759                                                           NULL);
 760
 761                         /*
 762                          * Wait until the number of on-going destroy RPCs drops
 763                          * under max_rpc_in_flight
 764                          */
 765                         l_wait_event_exclusive(cli->cl_destroy_waitq,
 766                                                osc_can_send_destroy(cli), &lwi);
 767                 }
 768         }
 769
 770         /* Do not wait for response */
 771         ptlrpcd_add_req(req, PSCOPE_OTHER);
 772         RETURN(0);
 773 }
 774
 775 static void osc_announce_cached(struct client_obd *cli, struct obdo *oa,
 776                                 long writing_bytes)
 777 {
 778         obd_flag bits = OBD_MD_FLBLOCKS|OBD_MD_FLGRANT;
 779
 780         LASSERT(!(oa->o_valid & bits));
 781
 782         oa->o_valid |= bits;
 783         client_obd_list_lock(&cli->cl_loi_list_lock);
 784         oa->o_dirty = cli->cl_dirty;
 785         if (cli->cl_dirty - cli->cl_dirty_transit > cli->cl_dirty_max) {
 786                 CERROR("dirty %lu - %lu > dirty_max %lu\n",
 787                        cli->cl_dirty, cli->cl_dirty_transit, cli->cl_dirty_max);
 788                 oa->o_undirty = 0;
 789         } else if (cfs_atomic_read(&obd_dirty_pages) -
 790                    cfs_atomic_read(&obd_dirty_transit_pages) >
 791                    obd_max_dirty_pages + 1){
 792                 /* The cfs_atomic_read() allowing the cfs_atomic_inc() are
 793                  * not covered by a lock thus they may safely race and trip
 794                  * this CERROR() unless we add in a small fudge factor (+1). */
 795                 CERROR("dirty %d - %d > system dirty_max %d\n",
 796                        cfs_atomic_read(&obd_dirty_pages),
 797                        cfs_atomic_read(&obd_dirty_transit_pages),
 798                        obd_max_dirty_pages);
 799                 oa->o_undirty = 0;
 800         } else if (cli->cl_dirty_max - cli->cl_dirty > 0x7fffffff) {
 801                 CERROR("dirty %lu - dirty_max %lu too big???\n",
 802                        cli->cl_dirty, cli->cl_dirty_max);
 803                 oa->o_undirty = 0;
 804         } else {
 805                 long max_in_flight = (cli->cl_max_pages_per_rpc << CFS_PAGE_SHIFT)*
 806                                 (cli->cl_max_rpcs_in_flight + 1);
 807                 oa->o_undirty = max(cli->cl_dirty_max, max_in_flight);
 808         }
 809         oa->o_grant = cli->cl_avail_grant;
 810         oa->o_dropped = cli->cl_lost_grant;
 811         cli->cl_lost_grant = 0;
 812         client_obd_list_unlock(&cli->cl_loi_list_lock);
 813         CDEBUG(D_CACHE,"dirty: "LPU64" undirty: %u dropped %u grant: "LPU64"\n",
 814                oa->o_dirty, oa->o_undirty, oa->o_dropped, oa->o_grant);
 815
 816 }
 817
 818 static void osc_update_next_shrink(struct client_obd *cli)
 819 {
 820         cli->cl_next_shrink_grant =
 821                 cfs_time_shift(cli->cl_grant_shrink_interval);
 822         CDEBUG(D_CACHE, "next time %ld to shrink grant \n",
 823                cli->cl_next_shrink_grant);
 824 }
 825
 826 /* caller must hold loi_list_lock */
 827 static void osc_consume_write_grant(struct client_obd *cli,
 828                                     struct brw_page *pga)
 829 {
 830         LASSERT_SPIN_LOCKED(&cli->cl_loi_list_lock.lock);
 831         LASSERT(!(pga->flag & OBD_BRW_FROM_GRANT));
 832         cfs_atomic_inc(&obd_dirty_pages);
 833         cli->cl_dirty += CFS_PAGE_SIZE;
 834         cli->cl_avail_grant -= CFS_PAGE_SIZE;
 835         pga->flag |= OBD_BRW_FROM_GRANT;
 836         CDEBUG(D_CACHE, "using %lu grant credits for brw %p page %p\n",
 837                CFS_PAGE_SIZE, pga, pga->pg);
 838         LASSERT(cli->cl_avail_grant >= 0);
 839         osc_update_next_shrink(cli);
 840 }
 841
 842 /* the companion to osc_consume_write_grant, called when a brw has completed.
 843  * must be called with the loi lock held. */
 844 static void osc_release_write_grant(struct client_obd *cli,
 845                                     struct brw_page *pga, int sent)
 846 {
 847         int blocksize = cli->cl_import->imp_obd->obd_osfs.os_bsize ? : 4096;
 848         ENTRY;
 849
 850         LASSERT_SPIN_LOCKED(&cli->cl_loi_list_lock.lock);
 851         if (!(pga->flag & OBD_BRW_FROM_GRANT)) {
 852                 EXIT;
 853                 return;
 854         }
 855
 856         pga->flag &= ~OBD_BRW_FROM_GRANT;
 857         cfs_atomic_dec(&obd_dirty_pages);
 858         cli->cl_dirty -= CFS_PAGE_SIZE;
 859         if (pga->flag & OBD_BRW_NOCACHE) {
 860                 pga->flag &= ~OBD_BRW_NOCACHE;
 861                 cfs_atomic_dec(&obd_dirty_transit_pages);
 862                 cli->cl_dirty_transit -= CFS_PAGE_SIZE;
 863         }
 864         if (!sent) {
 865                 cli->cl_lost_grant += CFS_PAGE_SIZE;
 866                 CDEBUG(D_CACHE, "lost grant: %lu avail grant: %lu dirty: %lu\n",
 867                        cli->cl_lost_grant, cli->cl_avail_grant, cli->cl_dirty);
 868         } else if (CFS_PAGE_SIZE != blocksize && pga->count != CFS_PAGE_SIZE) {
 869                 /* For short writes we shouldn't count parts of pages that
 870                  * span a whole block on the OST side, or our accounting goes
 871                  * wrong.  Should match the code in filter_grant_check. */
 872                 int offset = pga->off & ~CFS_PAGE_MASK;
 873                 int count = pga->count + (offset & (blocksize - 1));
 874                 int end = (offset + pga->count) & (blocksize - 1);
 875                 if (end)
 876                         count += blocksize - end;
 877
 878                 cli->cl_lost_grant += CFS_PAGE_SIZE - count;
 879                 CDEBUG(D_CACHE, "lost %lu grant: %lu avail: %lu dirty: %lu\n",
 880                        CFS_PAGE_SIZE - count, cli->cl_lost_grant,
 881                        cli->cl_avail_grant, cli->cl_dirty);
 882         }
 883
 884         EXIT;
 885 }
 886
 887 static unsigned long rpcs_in_flight(struct client_obd *cli)
 888 {
 889         return cli->cl_r_in_flight + cli->cl_w_in_flight;
 890 }
 891
 892 /* caller must hold loi_list_lock */
 893 void osc_wake_cache_waiters(struct client_obd *cli)
 894 {
 895         cfs_list_t *l, *tmp;
 896         struct osc_cache_waiter *ocw;
 897
 898         ENTRY;
 899         cfs_list_for_each_safe(l, tmp, &cli->cl_cache_waiters) {
 900                 /* if we can't dirty more, we must wait until some is written */
 901                 if ((cli->cl_dirty + CFS_PAGE_SIZE > cli->cl_dirty_max) ||
 902                    (cfs_atomic_read(&obd_dirty_pages) + 1 >
 903                     obd_max_dirty_pages)) {
 904                         CDEBUG(D_CACHE, "no dirty room: dirty: %ld "
 905                                "osc max %ld, sys max %d\n", cli->cl_dirty,
 906                                cli->cl_dirty_max, obd_max_dirty_pages);
 907                         return;
 908                 }
 909
 910                 /* if still dirty cache but no grant wait for pending RPCs that
 911                  * may yet return us some grant before doing sync writes */
 912                 if (cli->cl_w_in_flight && cli->cl_avail_grant < CFS_PAGE_SIZE) {
 913                         CDEBUG(D_CACHE, "%u BRW writes in flight, no grant\n",
 914                                cli->cl_w_in_flight);
 915                         return;
 916                 }
 917
 918                 ocw = cfs_list_entry(l, struct osc_cache_waiter, ocw_entry);
 919                 cfs_list_del_init(&ocw->ocw_entry);
 920                 if (cli->cl_avail_grant < CFS_PAGE_SIZE) {
 921                         /* no more RPCs in flight to return grant, do sync IO */
 922                         ocw->ocw_rc = -EDQUOT;
 923                         CDEBUG(D_INODE, "wake oap %p for sync\n", ocw->ocw_oap);
 924                 } else {
 925                         osc_consume_write_grant(cli,
 926                                                 &ocw->ocw_oap->oap_brw_page);
 927                 }
 928
 929                 cfs_waitq_signal(&ocw->ocw_waitq);
 930         }
 931
 932         EXIT;
 933 }
 934
 935 static void __osc_update_grant(struct client_obd *cli, obd_size grant)
 936 {
 937         client_obd_list_lock(&cli->cl_loi_list_lock);
 938         cli->cl_avail_grant += grant;
 939         client_obd_list_unlock(&cli->cl_loi_list_lock);
 940 }
 941
 942 static void osc_update_grant(struct client_obd *cli, struct ost_body *body)
 943 {
 944         if (body->oa.o_valid & OBD_MD_FLGRANT) {
 945                 CDEBUG(D_CACHE, "got "LPU64" extra grant\n", body->oa.o_grant);
 946                 __osc_update_grant(cli, body->oa.o_grant);
 947         }
 948 }
 949
 950 static int osc_set_info_async(struct obd_export *exp, obd_count keylen,
 951                               void *key, obd_count vallen, void *val,
 952                               struct ptlrpc_request_set *set);
 953
 954 static int osc_shrink_grant_interpret(const struct lu_env *env,
 955                                       struct ptlrpc_request *req,
 956                                       void *aa, int rc)
 957 {
 958         struct client_obd *cli = &req->rq_import->imp_obd->u.cli;
 959         struct obdo *oa = ((struct osc_grant_args *)aa)->aa_oa;
 960         struct ost_body *body;
 961
 962         if (rc != 0) {
 963                 __osc_update_grant(cli, oa->o_grant);
 964                 GOTO(out, rc);
 965         }
 966
 967         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
 968         LASSERT(body);
 969         osc_update_grant(cli, body);
 970 out:
 971         OBDO_FREE(oa);
 972         return rc;
 973 }
 974
 975 static void osc_shrink_grant_local(struct client_obd *cli, struct obdo *oa)
 976 {
 977         client_obd_list_lock(&cli->cl_loi_list_lock);
 978         oa->o_grant = cli->cl_avail_grant / 4;
 979         cli->cl_avail_grant -= oa->o_grant;
 980         client_obd_list_unlock(&cli->cl_loi_list_lock);
 981         if (!(oa->o_valid & OBD_MD_FLFLAGS)) {
 982                 oa->o_valid |= OBD_MD_FLFLAGS;
 983                 oa->o_flags = 0;
 984         }
 985         oa->o_flags |= OBD_FL_SHRINK_GRANT;
 986         osc_update_next_shrink(cli);
 987 }
 988
 989 /* Shrink the current grant, either from some large amount to enough for a
 990  * full set of in-flight RPCs, or if we have already shrunk to that limit
 991  * then to enough for a single RPC.  This avoids keeping more grant than
 992  * needed, and avoids shrinking the grant piecemeal. */
 993 static int osc_shrink_grant(struct client_obd *cli)
 994 {
 995         long target = (cli->cl_max_rpcs_in_flight + 1) *
 996                       cli->cl_max_pages_per_rpc;
 997
 998         client_obd_list_lock(&cli->cl_loi_list_lock);
 999         if (cli->cl_avail_grant <= target)
1000                 target = cli->cl_max_pages_per_rpc;
1001         client_obd_list_unlock(&cli->cl_loi_list_lock);
1002
1003         return osc_shrink_grant_to_target(cli, target);
1004 }
1005
1006 int osc_shrink_grant_to_target(struct client_obd *cli, long target)
1007 {
1008         int    rc = 0;
1009         struct ost_body     *body;
1010         ENTRY;
1011
1012         client_obd_list_lock(&cli->cl_loi_list_lock);
1013         /* Don't shrink if we are already above or below the desired limit
1014          * We don't want to shrink below a single RPC, as that will negatively
1015          * impact block allocation and long-term performance. */
1016         if (target < cli->cl_max_pages_per_rpc)
1017                 target = cli->cl_max_pages_per_rpc;
1018
1019         if (target >= cli->cl_avail_grant) {
1020                 client_obd_list_unlock(&cli->cl_loi_list_lock);
1021                 RETURN(0);
1022         }
1023         client_obd_list_unlock(&cli->cl_loi_list_lock);
1024
1025         OBD_ALLOC_PTR(body);
1026         if (!body)
1027                 RETURN(-ENOMEM);
1028
1029         osc_announce_cached(cli, &body->oa, 0);
1030
1031         client_obd_list_lock(&cli->cl_loi_list_lock);
1032         body->oa.o_grant = cli->cl_avail_grant - target;
1033         cli->cl_avail_grant = target;
1034         client_obd_list_unlock(&cli->cl_loi_list_lock);
1035         if (!(body->oa.o_valid & OBD_MD_FLFLAGS)) {
1036                 body->oa.o_valid |= OBD_MD_FLFLAGS;
1037                 body->oa.o_flags = 0;
1038         }
1039         body->oa.o_flags |= OBD_FL_SHRINK_GRANT;
1040         osc_update_next_shrink(cli);
1041
1042         rc = osc_set_info_async(cli->cl_import->imp_obd->obd_self_export,
1043                                 sizeof(KEY_GRANT_SHRINK), KEY_GRANT_SHRINK,
1044                                 sizeof(*body), body, NULL);
1045         if (rc != 0)
1046                 __osc_update_grant(cli, body->oa.o_grant);
1047         OBD_FREE_PTR(body);
1048         RETURN(rc);
1049 }
1050
1051 #define GRANT_SHRINK_LIMIT PTLRPC_MAX_BRW_SIZE
1052 static int osc_should_shrink_grant(struct client_obd *client)
1053 {
1054         cfs_time_t time = cfs_time_current();
1055         cfs_time_t next_shrink = client->cl_next_shrink_grant;
1056
1057         if ((client->cl_import->imp_connect_data.ocd_connect_flags &
1058              OBD_CONNECT_GRANT_SHRINK) == 0)
1059                 return 0;
1060
1061         if (cfs_time_aftereq(time, next_shrink - 5 * CFS_TICK)) {
1062                 if (client->cl_import->imp_state == LUSTRE_IMP_FULL &&
1063                     client->cl_avail_grant > GRANT_SHRINK_LIMIT)
1064                         return 1;
1065                 else
1066                         osc_update_next_shrink(client);
1067         }
1068         return 0;
1069 }
1070
1071 static int osc_grant_shrink_grant_cb(struct timeout_item *item, void *data)
1072 {
1073         struct client_obd *client;
1074
1075         cfs_list_for_each_entry(client, &item->ti_obd_list,
1076                                 cl_grant_shrink_list) {
1077                 if (osc_should_shrink_grant(client))
1078                         osc_shrink_grant(client);
1079         }
1080         return 0;
1081 }
1082
1083 static int osc_add_shrink_grant(struct client_obd *client)
1084 {
1085         int rc;
1086
1087         rc = ptlrpc_add_timeout_client(client->cl_grant_shrink_interval,
1088                                        TIMEOUT_GRANT,
1089                                        osc_grant_shrink_grant_cb, NULL,
1090                                        &client->cl_grant_shrink_list);
1091         if (rc) {
1092                 CERROR("add grant client %s error %d\n",
1093                         client->cl_import->imp_obd->obd_name, rc);
1094                 return rc;
1095         }
1096         CDEBUG(D_CACHE, "add grant client %s \n",
1097                client->cl_import->imp_obd->obd_name);
1098         osc_update_next_shrink(client);
1099         return 0;
1100 }
1101
1102 static int osc_del_shrink_grant(struct client_obd *client)
1103 {
1104         return ptlrpc_del_timeout_client(&client->cl_grant_shrink_list,
1105                                          TIMEOUT_GRANT);
1106 }
1107
1108 static void osc_init_grant(struct client_obd *cli, struct obd_connect_data *ocd)
1109 {
1110         /*
1111          * ocd_grant is the total grant amount we're expect to hold: if we've
1112          * been evicted, it's the new avail_grant amount, cl_dirty will drop
1113          * to 0 as inflight RPCs fail out; otherwise, it's avail_grant + dirty.
1114          *
1115          * race is tolerable here: if we're evicted, but imp_state already
1116          * left EVICTED state, then cl_dirty must be 0 already.
1117          */
1118         client_obd_list_lock(&cli->cl_loi_list_lock);
1119         if (cli->cl_import->imp_state == LUSTRE_IMP_EVICTED)
1120                 cli->cl_avail_grant = ocd->ocd_grant;
1121         else
1122                 cli->cl_avail_grant = ocd->ocd_grant - cli->cl_dirty;
1123
1124         if (cli->cl_avail_grant < 0) {
1125                 CWARN("%s: available grant < 0, the OSS is probably not running"
1126                       " with patch from bug20278 (%ld) \n",
1127                       cli->cl_import->imp_obd->obd_name, cli->cl_avail_grant);
1128                 /* workaround for 1.6 servers which do not have
1129                  * the patch from bug20278 */
1130                 cli->cl_avail_grant = ocd->ocd_grant;
1131         }
1132
1133         client_obd_list_unlock(&cli->cl_loi_list_lock);
1134
1135         CDEBUG(D_CACHE, "%s, setting cl_avail_grant: %ld cl_lost_grant: %ld \n",
1136                cli->cl_import->imp_obd->obd_name,
1137                cli->cl_avail_grant, cli->cl_lost_grant);
1138
1139         if (ocd->ocd_connect_flags & OBD_CONNECT_GRANT_SHRINK &&
1140             cfs_list_empty(&cli->cl_grant_shrink_list))
1141                 osc_add_shrink_grant(cli);
1142 }
1143
1144 /* We assume that the reason this OSC got a short read is because it read
1145  * beyond the end of a stripe file; i.e. lustre is reading a sparse file
1146  * via the LOV, and it _knows_ it's reading inside the file, it's just that
1147  * this stripe never got written at or beyond this stripe offset yet. */
1148 static void handle_short_read(int nob_read, obd_count page_count,
1149                               struct brw_page **pga)
1150 {
1151         char *ptr;
1152         int i = 0;
1153
1154         /* skip bytes read OK */
1155         while (nob_read > 0) {
1156                 LASSERT (page_count > 0);
1157
1158                 if (pga[i]->count > nob_read) {
1159                         /* EOF inside this page */
1160                         ptr = cfs_kmap(pga[i]->pg) +
1161                                 (pga[i]->off & ~CFS_PAGE_MASK);
1162                         memset(ptr + nob_read, 0, pga[i]->count - nob_read);
1163                         cfs_kunmap(pga[i]->pg);
1164                         page_count--;
1165                         i++;
1166                         break;
1167                 }
1168
1169                 nob_read -= pga[i]->count;
1170                 page_count--;
1171                 i++;
1172         }
1173
1174         /* zero remaining pages */
1175         while (page_count-- > 0) {
1176                 ptr = cfs_kmap(pga[i]->pg) + (pga[i]->off & ~CFS_PAGE_MASK);
1177                 memset(ptr, 0, pga[i]->count);
1178                 cfs_kunmap(pga[i]->pg);
1179                 i++;
1180         }
1181 }
1182
1183 static int check_write_rcs(struct ptlrpc_request *req,
1184                            int requested_nob, int niocount,
1185                            obd_count page_count, struct brw_page **pga)
1186 {
1187         int     i;
1188         __u32   *remote_rcs;
1189
1190         remote_rcs = req_capsule_server_sized_get(&req->rq_pill, &RMF_RCS,
1191                                                   sizeof(*remote_rcs) *
1192                                                   niocount);
1193         if (remote_rcs == NULL) {
1194                 CDEBUG(D_INFO, "Missing/short RC vector on BRW_WRITE reply\n");
1195                 return(-EPROTO);
1196         }
1197
1198         /* return error if any niobuf was in error */
1199         for (i = 0; i < niocount; i++) {
1200                 if ((int)remote_rcs[i] < 0)
1201                         return(remote_rcs[i]);
1202
1203                 if (remote_rcs[i] != 0) {
1204                         CDEBUG(D_INFO, "rc[%d] invalid (%d) req %p\n",
1205                                 i, remote_rcs[i], req);
1206                         return(-EPROTO);
1207                 }
1208         }
1209
1210         if (req->rq_bulk->bd_nob_transferred != requested_nob) {
1211                 CERROR("Unexpected # bytes transferred: %d (requested %d)\n",
1212                        req->rq_bulk->bd_nob_transferred, requested_nob);
1213                 return(-EPROTO);
1214         }
1215
1216         return (0);
1217 }
1218
1219 static inline int can_merge_pages(struct brw_page *p1, struct brw_page *p2)
1220 {
1221         if (p1->flag != p2->flag) {
1222                 unsigned mask = ~(OBD_BRW_FROM_GRANT|
1223                                   OBD_BRW_NOCACHE|OBD_BRW_SYNC|OBD_BRW_ASYNC);
1224
1225                 /* warn if we try to combine flags that we don't know to be
1226                  * safe to combine */
1227                 if ((p1->flag & mask) != (p2->flag & mask))
1228                         CERROR("is it ok to have flags 0x%x and 0x%x in the "
1229                                "same brw?\n", p1->flag, p2->flag);
1230                 return 0;
1231         }
1232
1233         return (p1->off + p1->count == p2->off);
1234 }
1235
1236 static obd_count osc_checksum_bulk(int nob, obd_count pg_count,
1237                                    struct brw_page **pga, int opc,
1238                                    cksum_type_t cksum_type)
1239 {
1240         __u32 cksum;
1241         int i = 0;
1242
1243         LASSERT (pg_count > 0);
1244         cksum = init_checksum(cksum_type);
1245         while (nob > 0 && pg_count > 0) {
1246                 unsigned char *ptr = cfs_kmap(pga[i]->pg);
1247                 int off = pga[i]->off & ~CFS_PAGE_MASK;
1248                 int count = pga[i]->count > nob ? nob : pga[i]->count;
1249
1250                 /* corrupt the data before we compute the checksum, to
1251                  * simulate an OST->client data error */
1252                 if (i == 0 && opc == OST_READ &&
1253                     OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_RECEIVE))
1254                         memcpy(ptr + off, "bad1", min(4, nob));
1255                 cksum = compute_checksum(cksum, ptr + off, count, cksum_type);
1256                 cfs_kunmap(pga[i]->pg);
1257                 LL_CDEBUG_PAGE(D_PAGE, pga[i]->pg, "off %d checksum %x\n",
1258                                off, cksum);
1259
1260                 nob -= pga[i]->count;
1261                 pg_count--;
1262                 i++;
1263         }
1264         /* For sending we only compute the wrong checksum instead
1265          * of corrupting the data so it is still correct on a redo */
1266         if (opc == OST_WRITE && OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_SEND))
1267                 cksum++;
1268
1269         return cksum;
1270 }
1271
1272 static int osc_brw_prep_request(int cmd, struct client_obd *cli,struct obdo *oa,
1273                                 struct lov_stripe_md *lsm, obd_count page_count,
1274                                 struct brw_page **pga,
1275                                 struct ptlrpc_request **reqp,
1276                                 struct obd_capa *ocapa, int reserve,
1277                                 int resend)
1278 {
1279         struct ptlrpc_request   *req;
1280         struct ptlrpc_bulk_desc *desc;
1281         struct ost_body         *body;
1282         struct obd_ioobj        *ioobj;
1283         struct niobuf_remote    *niobuf;
1284         int niocount, i, requested_nob, opc, rc;
1285         struct osc_brw_async_args *aa;
1286         struct req_capsule      *pill;
1287         struct brw_page *pg_prev;
1288
1289         ENTRY;
1290         if (OBD_FAIL_CHECK(OBD_FAIL_OSC_BRW_PREP_REQ))
1291                 RETURN(-ENOMEM); /* Recoverable */
1292         if (OBD_FAIL_CHECK(OBD_FAIL_OSC_BRW_PREP_REQ2))
1293                 RETURN(-EINVAL); /* Fatal */
1294
1295         if ((cmd & OBD_BRW_WRITE) != 0) {
1296                 opc = OST_WRITE;
1297                 req = ptlrpc_request_alloc_pool(cli->cl_import,
1298                                                 cli->cl_import->imp_rq_pool,
1299                                                 &RQF_OST_BRW_WRITE);
1300         } else {
1301                 opc = OST_READ;
1302                 req = ptlrpc_request_alloc(cli->cl_import, &RQF_OST_BRW_READ);
1303         }
1304         if (req == NULL)
1305                 RETURN(-ENOMEM);
1306
1307         for (niocount = i = 1; i < page_count; i++) {
1308                 if (!can_merge_pages(pga[i - 1], pga[i]))
1309                         niocount++;
1310         }
1311
1312         pill = &req->rq_pill;
1313         req_capsule_set_size(pill, &RMF_OBD_IOOBJ, RCL_CLIENT,
1314                              sizeof(*ioobj));
1315         req_capsule_set_size(pill, &RMF_NIOBUF_REMOTE, RCL_CLIENT,
1316                              niocount * sizeof(*niobuf));
1317         osc_set_capa_size(req, &RMF_CAPA1, ocapa);
1318
1319         rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, opc);
1320         if (rc) {
1321                 ptlrpc_request_free(req);
1322                 RETURN(rc);
1323         }
1324         req->rq_request_portal = OST_IO_PORTAL; /* bug 7198 */
1325         ptlrpc_at_set_req_timeout(req);
1326
1327         if (opc == OST_WRITE)
1328                 desc = ptlrpc_prep_bulk_imp(req, page_count,
1329                                             BULK_GET_SOURCE, OST_BULK_PORTAL);
1330         else
1331                 desc = ptlrpc_prep_bulk_imp(req, page_count,
1332                                             BULK_PUT_SINK, OST_BULK_PORTAL);
1333
1334         if (desc == NULL)
1335                 GOTO(out, rc = -ENOMEM);
1336         /* NB request now owns desc and will free it when it gets freed */
1337
1338         body = req_capsule_client_get(pill, &RMF_OST_BODY);
1339         ioobj = req_capsule_client_get(pill, &RMF_OBD_IOOBJ);
1340         niobuf = req_capsule_client_get(pill, &RMF_NIOBUF_REMOTE);
1341         LASSERT(body != NULL && ioobj != NULL && niobuf != NULL);
1342
1343         lustre_set_wire_obdo(&body->oa, oa);
1344
1345         obdo_to_ioobj(oa, ioobj);
1346         ioobj->ioo_bufcnt = niocount;
1347         osc_pack_capa(req, body, ocapa);
1348         LASSERT (page_count > 0);
1349         pg_prev = pga[0];
1350         for (requested_nob = i = 0; i < page_count; i++, niobuf++) {
1351                 struct brw_page *pg = pga[i];
1352                 int poff = pg->off & ~CFS_PAGE_MASK;
1353
1354                 LASSERT(pg->count > 0);
1355                 /* make sure there is no gap in the middle of page array */
1356                 LASSERTF(page_count == 1 ||
1357                          (ergo(i == 0, poff + pg->count == CFS_PAGE_SIZE) &&
1358                           ergo(i > 0 && i < page_count - 1,
1359                                poff == 0 && pg->count == CFS_PAGE_SIZE)   &&
1360                           ergo(i == page_count - 1, poff == 0)),
1361                          "i: %d/%d pg: %p off: "LPU64", count: %u\n",
1362                          i, page_count, pg, pg->off, pg->count);
1363 #ifdef __linux__
1364                 LASSERTF(i == 0 || pg->off > pg_prev->off,
1365                          "i %d p_c %u pg %p [pri %lu ind %lu] off "LPU64
1366                          " prev_pg %p [pri %lu ind %lu] off "LPU64"\n",
1367                          i, page_count,
1368                          pg->pg, page_private(pg->pg), pg->pg->index, pg->off,
1369                          pg_prev->pg, page_private(pg_prev->pg),
1370                          pg_prev->pg->index, pg_prev->off);
1371 #else
1372                 LASSERTF(i == 0 || pg->off > pg_prev->off,
1373                          "i %d p_c %u\n", i, page_count);
1374 #endif
1375                 LASSERT((pga[0]->flag & OBD_BRW_SRVLOCK) ==
1376                         (pg->flag & OBD_BRW_SRVLOCK));
1377
1378                 ptlrpc_prep_bulk_page(desc, pg->pg, poff, pg->count);
1379                 requested_nob += pg->count;
1380
1381                 if (i > 0 && can_merge_pages(pg_prev, pg)) {
1382                         niobuf--;
1383                         niobuf->len += pg->count;
1384                 } else {
1385                         niobuf->offset = pg->off;
1386                         niobuf->len    = pg->count;
1387                         niobuf->flags  = pg->flag;
1388                 }
1389                 pg_prev = pg;
1390         }
1391
1392         LASSERTF((void *)(niobuf - niocount) ==
1393                 req_capsule_client_get(&req->rq_pill, &RMF_NIOBUF_REMOTE),
1394                 "want %p - real %p\n", req_capsule_client_get(&req->rq_pill,
1395                 &RMF_NIOBUF_REMOTE), (void *)(niobuf - niocount));
1396
1397         osc_announce_cached(cli, &body->oa, opc == OST_WRITE ? requested_nob:0);
1398         if (resend) {
1399                 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0) {
1400                         body->oa.o_valid |= OBD_MD_FLFLAGS;
1401                         body->oa.o_flags = 0;
1402                 }
1403                 body->oa.o_flags |= OBD_FL_RECOV_RESEND;
1404         }
1405
1406         if (osc_should_shrink_grant(cli))
1407                 osc_shrink_grant_local(cli, &body->oa);
1408
1409         /* size[REQ_REC_OFF] still sizeof (*body) */
1410         if (opc == OST_WRITE) {
1411                 if (unlikely(cli->cl_checksum) &&
1412                     !sptlrpc_flavor_has_bulk(&req->rq_flvr)) {
1413                         /* store cl_cksum_type in a local variable since
1414                          * it can be changed via lprocfs */
1415                         cksum_type_t cksum_type = cli->cl_cksum_type;
1416
1417                         if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0) {
1418                                 oa->o_flags &= OBD_FL_LOCAL_MASK;
1419                                 body->oa.o_flags = 0;
1420                         }
1421                         body->oa.o_flags |= cksum_type_pack(cksum_type);
1422                         body->oa.o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1423                         body->oa.o_cksum = osc_checksum_bulk(requested_nob,
1424                                                              page_count, pga,
1425                                                              OST_WRITE,
1426                                                              cksum_type);
1427                         CDEBUG(D_PAGE, "checksum at write origin: %x\n",
1428                                body->oa.o_cksum);
1429                         /* save this in 'oa', too, for later checking */
1430                         oa->o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1431                         oa->o_flags |= cksum_type_pack(cksum_type);
1432                 } else {
1433                         /* clear out the checksum flag, in case this is a
1434                          * resend but cl_checksum is no longer set. b=11238 */
1435                         oa->o_valid &= ~OBD_MD_FLCKSUM;
1436                 }
1437                 oa->o_cksum = body->oa.o_cksum;
1438                 /* 1 RC per niobuf */
1439                 req_capsule_set_size(pill, &RMF_RCS, RCL_SERVER,
1440                                      sizeof(__u32) * niocount);
1441         } else {
1442                 if (unlikely(cli->cl_checksum) &&
1443                     !sptlrpc_flavor_has_bulk(&req->rq_flvr)) {
1444                         if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0)
1445                                 body->oa.o_flags = 0;
1446                         body->oa.o_flags |= cksum_type_pack(cli->cl_cksum_type);
1447                         body->oa.o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1448                 }
1449         }
1450         ptlrpc_request_set_replen(req);
1451
1452         CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
1453         aa = ptlrpc_req_async_args(req);
1454         aa->aa_oa = oa;
1455         aa->aa_requested_nob = requested_nob;
1456         aa->aa_nio_count = niocount;
1457         aa->aa_page_count = page_count;
1458         aa->aa_resends = 0;
1459         aa->aa_ppga = pga;
1460         aa->aa_cli = cli;
1461         CFS_INIT_LIST_HEAD(&aa->aa_oaps);
1462         if (ocapa && reserve)
1463                 aa->aa_ocapa = capa_get(ocapa);
1464
1465         *reqp = req;
1466         RETURN(0);
1467
1468  out:
1469         ptlrpc_req_finished(req);
1470         RETURN(rc);
1471 }
1472
1473 static int check_write_checksum(struct obdo *oa, const lnet_process_id_t *peer,
1474                                 __u32 client_cksum, __u32 server_cksum, int nob,
1475                                 obd_count page_count, struct brw_page **pga,
1476                                 cksum_type_t client_cksum_type)
1477 {
1478         __u32 new_cksum;
1479         char *msg;
1480         cksum_type_t cksum_type;
1481
1482         if (server_cksum == client_cksum) {
1483                 CDEBUG(D_PAGE, "checksum %x confirmed\n", client_cksum);
1484                 return 0;
1485         }
1486
1487         /* If this is mmaped file - it can be changed at any time */
1488         if (oa->o_valid & OBD_MD_FLFLAGS && oa->o_flags & OBD_FL_MMAP)
1489                 return 1;
1490
1491         if (oa->o_valid & OBD_MD_FLFLAGS)
1492                 cksum_type = cksum_type_unpack(oa->o_flags);
1493         else
1494                 cksum_type = OBD_CKSUM_CRC32;
1495
1496         new_cksum = osc_checksum_bulk(nob, page_count, pga, OST_WRITE,
1497                                       cksum_type);
1498
1499         if (cksum_type != client_cksum_type)
1500                 msg = "the server did not use the checksum type specified in "
1501                       "the original request - likely a protocol problem";
1502         else if (new_cksum == server_cksum)
1503                 msg = "changed on the client after we checksummed it - "
1504                       "likely false positive due to mmap IO (bug 11742)";
1505         else if (new_cksum == client_cksum)
1506                 msg = "changed in transit before arrival at OST";
1507         else
1508                 msg = "changed in transit AND doesn't match the original - "
1509                       "likely false positive due to mmap IO (bug 11742)";
1510
1511         LCONSOLE_ERROR_MSG(0x132, "BAD WRITE CHECKSUM: %s: from %s inode "DFID
1512                            " object "LPU64"/"LPU64" extent ["LPU64"-"LPU64"]\n",
1513                            msg, libcfs_nid2str(peer->nid),
1514                            oa->o_valid & OBD_MD_FLFID ? oa->o_parent_seq : (__u64)0,
1515                            oa->o_valid & OBD_MD_FLFID ? oa->o_parent_oid : 0,
1516                            oa->o_valid & OBD_MD_FLFID ? oa->o_parent_ver : 0,
1517                            oa->o_id,
1518                            oa->o_valid & OBD_MD_FLGROUP ? oa->o_seq : (__u64)0,
1519                            pga[0]->off,
1520                            pga[page_count-1]->off + pga[page_count-1]->count - 1);
1521         CERROR("original client csum %x (type %x), server csum %x (type %x), "
1522                "client csum now %x\n", client_cksum, client_cksum_type,
1523                server_cksum, cksum_type, new_cksum);
1524         return 1;
1525 }
1526
1527 /* Note rc enters this function as number of bytes transferred */
1528 static int osc_brw_fini_request(struct ptlrpc_request *req, int rc)
1529 {
1530         struct osc_brw_async_args *aa = (void *)&req->rq_async_args;
1531         const lnet_process_id_t *peer =
1532                         &req->rq_import->imp_connection->c_peer;
1533         struct client_obd *cli = aa->aa_cli;
1534         struct ost_body *body;
1535         __u32 client_cksum = 0;
1536         ENTRY;
1537
1538         if (rc < 0 && rc != -EDQUOT) {
1539                 DEBUG_REQ(D_INFO, req, "Failed request with rc = %d\n", rc);
1540                 RETURN(rc);
1541         }
1542
1543         LASSERTF(req->rq_repmsg != NULL, "rc = %d\n", rc);
1544         body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
1545         if (body == NULL) {
1546                 DEBUG_REQ(D_INFO, req, "Can't unpack body\n");
1547                 RETURN(-EPROTO);
1548         }
1549
1550 #ifdef HAVE_QUOTA_SUPPORT
1551         /* set/clear over quota flag for a uid/gid */
1552         if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE &&
1553             body->oa.o_valid & (OBD_MD_FLUSRQUOTA | OBD_MD_FLGRPQUOTA)) {
1554                 unsigned int qid[MAXQUOTAS] = { body->oa.o_uid, body->oa.o_gid };
1555
1556                 CDEBUG(D_QUOTA, "setdq for [%u %u] with valid "LPX64", flags %x\n",
1557                        body->oa.o_uid, body->oa.o_gid, body->oa.o_valid,
1558                        body->oa.o_flags);
1559                 lquota_setdq(quota_interface, cli, qid, body->oa.o_valid,
1560                              body->oa.o_flags);
1561         }
1562 #endif
1563
1564         osc_update_grant(cli, body);
1565
1566         if (rc < 0)
1567                 RETURN(rc);
1568
1569         if (aa->aa_oa->o_valid & OBD_MD_FLCKSUM)
1570                 client_cksum = aa->aa_oa->o_cksum; /* save for later */
1571
1572         if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE) {
1573                 if (rc > 0) {
1574                         CERROR("Unexpected +ve rc %d\n", rc);
1575                         RETURN(-EPROTO);
1576                 }
1577                 LASSERT(req->rq_bulk->bd_nob == aa->aa_requested_nob);
1578
1579                 if (sptlrpc_cli_unwrap_bulk_write(req, req->rq_bulk))
1580                         RETURN(-EAGAIN);
1581
1582                 if ((aa->aa_oa->o_valid & OBD_MD_FLCKSUM) && client_cksum &&
1583                     check_write_checksum(&body->oa, peer, client_cksum,
1584                                          body->oa.o_cksum, aa->aa_requested_nob,
1585                                          aa->aa_page_count, aa->aa_ppga,
1586                                          cksum_type_unpack(aa->aa_oa->o_flags)))
1587                         RETURN(-EAGAIN);
1588
1589                 rc = check_write_rcs(req, aa->aa_requested_nob,aa->aa_nio_count,
1590                                      aa->aa_page_count, aa->aa_ppga);
1591                 GOTO(out, rc);
1592         }
1593
1594         /* The rest of this function executes only for OST_READs */
1595
1596         /* if unwrap_bulk failed, return -EAGAIN to retry */
1597         rc = sptlrpc_cli_unwrap_bulk_read(req, req->rq_bulk, rc);
1598         if (rc < 0)
1599                 GOTO(out, rc = -EAGAIN);
1600
1601         if (rc > aa->aa_requested_nob) {
1602                 CERROR("Unexpected rc %d (%d requested)\n", rc,
1603                        aa->aa_requested_nob);
1604                 RETURN(-EPROTO);
1605         }
1606
1607         if (rc != req->rq_bulk->bd_nob_transferred) {
1608                 CERROR ("Unexpected rc %d (%d transferred)\n",
1609                         rc, req->rq_bulk->bd_nob_transferred);
1610                 return (-EPROTO);
1611         }
1612
1613         if (rc < aa->aa_requested_nob)
1614                 handle_short_read(rc, aa->aa_page_count, aa->aa_ppga);
1615
1616         if (body->oa.o_valid & OBD_MD_FLCKSUM) {
1617                 static int cksum_counter;
1618                 __u32      server_cksum = body->oa.o_cksum;
1619                 char      *via;
1620                 char      *router;
1621                 cksum_type_t cksum_type;
1622
1623                 if (body->oa.o_valid & OBD_MD_FLFLAGS)
1624                         cksum_type = cksum_type_unpack(body->oa.o_flags);
1625                 else
1626                         cksum_type = OBD_CKSUM_CRC32;
1627                 client_cksum = osc_checksum_bulk(rc, aa->aa_page_count,
1628                                                  aa->aa_ppga, OST_READ,
1629                                                  cksum_type);
1630
1631                 if (peer->nid == req->rq_bulk->bd_sender) {
1632                         via = router = "";
1633                 } else {
1634                         via = " via ";
1635                         router = libcfs_nid2str(req->rq_bulk->bd_sender);
1636                 }
1637
1638                 if (server_cksum == ~0 && rc > 0) {
1639                         CERROR("Protocol error: server %s set the 'checksum' "
1640                                "bit, but didn't send a checksum.  Not fatal, "
1641                                "but please notify on http://bugs.whamcloud.com/\n",
1642                                libcfs_nid2str(peer->nid));
1643                 } else if (server_cksum != client_cksum) {
1644                         LCONSOLE_ERROR_MSG(0x133, "%s: BAD READ CHECKSUM: from "
1645                                            "%s%s%s inode "DFID" object "
1646                                            LPU64"/"LPU64" extent "
1647                                            "["LPU64"-"LPU64"]\n",
1648                                            req->rq_import->imp_obd->obd_name,
1649                                            libcfs_nid2str(peer->nid),
1650                                            via, router,
1651                                            body->oa.o_valid & OBD_MD_FLFID ?
1652                                                 body->oa.o_parent_seq : (__u64)0,
1653                                            body->oa.o_valid & OBD_MD_FLFID ?
1654                                                 body->oa.o_parent_oid : 0,
1655                                            body->oa.o_valid & OBD_MD_FLFID ?
1656                                                 body->oa.o_parent_ver : 0,
1657                                            body->oa.o_id,
1658                                            body->oa.o_valid & OBD_MD_FLGROUP ?
1659                                                 body->oa.o_seq : (__u64)0,
1660                                            aa->aa_ppga[0]->off,
1661                                            aa->aa_ppga[aa->aa_page_count-1]->off +
1662                                            aa->aa_ppga[aa->aa_page_count-1]->count -
1663                                                                         1);
1664                         CERROR("client %x, server %x, cksum_type %x\n",
1665                                client_cksum, server_cksum, cksum_type);
1666                         cksum_counter = 0;
1667                         aa->aa_oa->o_cksum = client_cksum;
1668                         rc = -EAGAIN;
1669                 } else {
1670                         cksum_counter++;
1671                         CDEBUG(D_PAGE, "checksum %x confirmed\n", client_cksum);
1672                         rc = 0;
1673                 }
1674         } else if (unlikely(client_cksum)) {
1675                 static int cksum_missed;
1676
1677                 cksum_missed++;
1678                 if ((cksum_missed & (-cksum_missed)) == cksum_missed)
1679                         CERROR("Checksum %u requested from %s but not sent\n",
1680                                cksum_missed, libcfs_nid2str(peer->nid));
1681         } else {
1682                 rc = 0;
1683         }
1684 out:
1685         if (rc >= 0)
1686                 lustre_get_wire_obdo(aa->aa_oa, &body->oa);
1687
1688         RETURN(rc);
1689 }
1690
1691 static int osc_brw_internal(int cmd, struct obd_export *exp, struct obdo *oa,
1692                             struct lov_stripe_md *lsm,
1693                             obd_count page_count, struct brw_page **pga,
1694                             struct obd_capa *ocapa)
1695 {
1696         struct ptlrpc_request *req;
1697         int                    rc;
1698         cfs_waitq_t            waitq;
1699         int                    resends = 0;
1700         struct l_wait_info     lwi;
1701
1702         ENTRY;
1703
1704         cfs_waitq_init(&waitq);
1705
1706 restart_bulk:
1707         rc = osc_brw_prep_request(cmd, &exp->exp_obd->u.cli, oa, lsm,
1708                                   page_count, pga, &req, ocapa, 0, resends);
1709         if (rc != 0)
1710                 return (rc);
1711
1712         rc = ptlrpc_queue_wait(req);
1713
1714         if (rc == -ETIMEDOUT && req->rq_resend) {
1715                 DEBUG_REQ(D_HA, req,  "BULK TIMEOUT");
1716                 ptlrpc_req_finished(req);
1717                 goto restart_bulk;
1718         }
1719
1720         rc = osc_brw_fini_request(req, rc);
1721
1722         ptlrpc_req_finished(req);
1723         if (osc_recoverable_error(rc)) {
1724                 resends++;
1725                 if (!client_should_resend(resends, &exp->exp_obd->u.cli)) {
1726                         CERROR("too many resend retries, returning error\n");
1727                         RETURN(-EIO);
1728                 }
1729
1730                 lwi = LWI_TIMEOUT_INTR(cfs_time_seconds(resends), NULL, NULL, NULL);
1731                 l_wait_event(waitq, 0, &lwi);
1732
1733                 goto restart_bulk;
1734         }
1735
1736         RETURN (rc);
1737 }
1738
1739 int osc_brw_redo_request(struct ptlrpc_request *request,
1740                          struct osc_brw_async_args *aa)
1741 {
1742         struct ptlrpc_request *new_req;
1743         struct ptlrpc_request_set *set = request->rq_set;
1744         struct osc_brw_async_args *new_aa;
1745         struct osc_async_page *oap;
1746         int rc = 0;
1747         ENTRY;
1748
1749         if (!client_should_resend(aa->aa_resends, aa->aa_cli)) {
1750                 CERROR("too many resent retries, returning error\n");
1751                 RETURN(-EIO);
1752         }
1753
1754         DEBUG_REQ(D_ERROR, request, "redo for recoverable error");
1755
1756         rc = osc_brw_prep_request(lustre_msg_get_opc(request->rq_reqmsg) ==
1757                                         OST_WRITE ? OBD_BRW_WRITE :OBD_BRW_READ,
1758                                   aa->aa_cli, aa->aa_oa,
1759                                   NULL /* lsm unused by osc currently */,
1760                                   aa->aa_page_count, aa->aa_ppga,
1761                                   &new_req, aa->aa_ocapa, 0, 1);
1762         if (rc)
1763                 RETURN(rc);
1764
1765         client_obd_list_lock(&aa->aa_cli->cl_loi_list_lock);
1766
1767         cfs_list_for_each_entry(oap, &aa->aa_oaps, oap_rpc_item) {
1768                 if (oap->oap_request != NULL) {
1769                         LASSERTF(request == oap->oap_request,
1770                                  "request %p != oap_request %p\n",
1771                                  request, oap->oap_request);
1772                         if (oap->oap_interrupted) {
1773                                 client_obd_list_unlock(&aa->aa_cli->cl_loi_list_lock);
1774                                 ptlrpc_req_finished(new_req);
1775                                 RETURN(-EINTR);
1776                         }
1777                 }
1778         }
1779         /* New request takes over pga and oaps from old request.
1780          * Note that copying a list_head doesn't work, need to move it... */
1781         aa->aa_resends++;
1782         new_req->rq_interpret_reply = request->rq_interpret_reply;
1783         new_req->rq_async_args = request->rq_async_args;
1784         new_req->rq_sent = cfs_time_current_sec() + aa->aa_resends;
1785
1786         new_aa = ptlrpc_req_async_args(new_req);
1787
1788         CFS_INIT_LIST_HEAD(&new_aa->aa_oaps);
1789         cfs_list_splice(&aa->aa_oaps, &new_aa->aa_oaps);
1790         CFS_INIT_LIST_HEAD(&aa->aa_oaps);
1791
1792         cfs_list_for_each_entry(oap, &new_aa->aa_oaps, oap_rpc_item) {
1793                 if (oap->oap_request) {
1794                         ptlrpc_req_finished(oap->oap_request);
1795                         oap->oap_request = ptlrpc_request_addref(new_req);
1796                 }
1797         }
1798
1799         new_aa->aa_ocapa = aa->aa_ocapa;
1800         aa->aa_ocapa = NULL;
1801
1802         /* use ptlrpc_set_add_req is safe because interpret functions work
1803          * in check_set context. only one way exist with access to request
1804          * from different thread got -EINTR - this way protected with
1805          * cl_loi_list_lock */
1806         ptlrpc_set_add_req(set, new_req);
1807
1808         client_obd_list_unlock(&aa->aa_cli->cl_loi_list_lock);
1809
1810         DEBUG_REQ(D_INFO, new_req, "new request");
1811         RETURN(0);
1812 }
1813
1814 /*
1815  * ugh, we want disk allocation on the target to happen in offset order.  we'll
1816  * follow sedgewicks advice and stick to the dead simple shellsort -- it'll do
1817  * fine for our small page arrays and doesn't require allocation.  its an
1818  * insertion sort that swaps elements that are strides apart, shrinking the
1819  * stride down until its '1' and the array is sorted.
1820  */
1821 static void sort_brw_pages(struct brw_page **array, int num)
1822 {
1823         int stride, i, j;
1824         struct brw_page *tmp;
1825
1826         if (num == 1)
1827                 return;
1828         for (stride = 1; stride < num ; stride = (stride * 3) + 1)
1829                 ;
1830
1831         do {
1832                 stride /= 3;
1833                 for (i = stride ; i < num ; i++) {
1834                         tmp = array[i];
1835                         j = i;
1836                         while (j >= stride && array[j - stride]->off > tmp->off) {
1837                                 array[j] = array[j - stride];
1838                                 j -= stride;
1839                         }
1840                         array[j] = tmp;
1841                 }
1842         } while (stride > 1);
1843 }
1844
1845 static obd_count max_unfragmented_pages(struct brw_page **pg, obd_count pages)
1846 {
1847         int count = 1;
1848         int offset;
1849         int i = 0;
1850
1851         LASSERT (pages > 0);
1852         offset = pg[i]->off & ~CFS_PAGE_MASK;
1853
1854         for (;;) {
1855                 pages--;
1856                 if (pages == 0)         /* that's all */
1857                         return count;
1858
1859                 if (offset + pg[i]->count < CFS_PAGE_SIZE)
1860                         return count;   /* doesn't end on page boundary */
1861
1862                 i++;
1863                 offset = pg[i]->off & ~CFS_PAGE_MASK;
1864                 if (offset != 0)        /* doesn't start on page boundary */
1865                         return count;
1866
1867                 count++;
1868         }
1869 }
1870
1871 static struct brw_page **osc_build_ppga(struct brw_page *pga, obd_count count)
1872 {
1873         struct brw_page **ppga;
1874         int i;
1875
1876         OBD_ALLOC(ppga, sizeof(*ppga) * count);
1877         if (ppga == NULL)
1878                 return NULL;
1879
1880         for (i = 0; i < count; i++)
1881                 ppga[i] = pga + i;
1882         return ppga;
1883 }
1884
1885 static void osc_release_ppga(struct brw_page **ppga, obd_count count)
1886 {
1887         LASSERT(ppga != NULL);
1888         OBD_FREE(ppga, sizeof(*ppga) * count);
1889 }
1890
1891 static int osc_brw(int cmd, struct obd_export *exp, struct obd_info *oinfo,
1892                    obd_count page_count, struct brw_page *pga,
1893                    struct obd_trans_info *oti)
1894 {
1895         struct obdo *saved_oa = NULL;
1896         struct brw_page **ppga, **orig;
1897         struct obd_import *imp = class_exp2cliimp(exp);
1898         struct client_obd *cli;
1899         int rc, page_count_orig;
1900         ENTRY;
1901
1902         LASSERT((imp != NULL) && (imp->imp_obd != NULL));
1903         cli = &imp->imp_obd->u.cli;
1904
1905         if (cmd & OBD_BRW_CHECK) {
1906                 /* The caller just wants to know if there's a chance that this
1907                  * I/O can succeed */
1908
1909                 if (imp->imp_invalid)
1910                         RETURN(-EIO);
1911                 RETURN(0);
1912         }
1913
1914         /* test_brw with a failed create can trip this, maybe others. */
1915         LASSERT(cli->cl_max_pages_per_rpc);
1916
1917         rc = 0;
1918
1919         orig = ppga = osc_build_ppga(pga, page_count);
1920         if (ppga == NULL)
1921                 RETURN(-ENOMEM);
1922         page_count_orig = page_count;
1923
1924         sort_brw_pages(ppga, page_count);
1925         while (page_count) {
1926                 obd_count pages_per_brw;
1927
1928                 if (page_count > cli->cl_max_pages_per_rpc)
1929                         pages_per_brw = cli->cl_max_pages_per_rpc;
1930                 else
1931                         pages_per_brw = page_count;
1932
1933                 pages_per_brw = max_unfragmented_pages(ppga, pages_per_brw);
1934
1935                 if (saved_oa != NULL) {
1936                         /* restore previously saved oa */
1937                         *oinfo->oi_oa = *saved_oa;
1938                 } else if (page_count > pages_per_brw) {
1939                         /* save a copy of oa (brw will clobber it) */
1940                         OBDO_ALLOC(saved_oa);
1941                         if (saved_oa == NULL)
1942                                 GOTO(out, rc = -ENOMEM);
1943                         *saved_oa = *oinfo->oi_oa;
1944                 }
1945
1946                 rc = osc_brw_internal(cmd, exp, oinfo->oi_oa, oinfo->oi_md,
1947                                       pages_per_brw, ppga, oinfo->oi_capa);
1948
1949                 if (rc != 0)
1950                         break;
1951
1952                 page_count -= pages_per_brw;
1953                 ppga += pages_per_brw;
1954         }
1955
1956 out:
1957         osc_release_ppga(orig, page_count_orig);
1958
1959         if (saved_oa != NULL)
1960                 OBDO_FREE(saved_oa);
1961
1962         RETURN(rc);
1963 }
1964
1965 /* The companion to osc_enter_cache(), called when @oap is no longer part of
1966  * the dirty accounting.  Writeback completes or truncate happens before
1967  * writing starts.  Must be called with the loi lock held. */
1968 static void osc_exit_cache(struct client_obd *cli, struct osc_async_page *oap,
1969                            int sent)
1970 {
1971         osc_release_write_grant(cli, &oap->oap_brw_page, sent);
1972 }
1973
1974
1975 /* This maintains the lists of pending pages to read/write for a given object
1976  * (lop).  This is used by osc_check_rpcs->osc_next_loi() and loi_list_maint()
1977  * to quickly find objects that are ready to send an RPC. */
1978 static int lop_makes_rpc(struct client_obd *cli, struct loi_oap_pages *lop,
1979                          int cmd)
1980 {
1981         int optimal;
1982         ENTRY;
1983
1984         if (lop->lop_num_pending == 0)
1985                 RETURN(0);
1986
1987         /* if we have an invalid import we want to drain the queued pages
1988          * by forcing them through rpcs that immediately fail and complete
1989          * the pages.  recovery relies on this to empty the queued pages
1990          * before canceling the locks and evicting down the llite pages */
1991         if (cli->cl_import == NULL || cli->cl_import->imp_invalid)
1992                 RETURN(1);
1993
1994         /* stream rpcs in queue order as long as as there is an urgent page
1995          * queued.  this is our cheap solution for good batching in the case
1996          * where writepage marks some random page in the middle of the file
1997          * as urgent because of, say, memory pressure */
1998         if (!cfs_list_empty(&lop->lop_urgent)) {
1999                 CDEBUG(D_CACHE, "urgent request forcing RPC\n");
2000                 RETURN(1);
2001         }
2002         /* fire off rpcs when we have 'optimal' rpcs as tuned for the wire. */
2003         optimal = cli->cl_max_pages_per_rpc;
2004         if (cmd & OBD_BRW_WRITE) {
2005                 /* trigger a write rpc stream as long as there are dirtiers
2006                  * waiting for space.  as they're waiting, they're not going to
2007                  * create more pages to coalesce with what's waiting.. */
2008                 if (!cfs_list_empty(&cli->cl_cache_waiters)) {
2009                         CDEBUG(D_CACHE, "cache waiters forcing RPC\n");
2010                         RETURN(1);
2011                 }
2012                 /* +16 to avoid triggering rpcs that would want to include pages
2013                  * that are being queued but which can't be made ready until
2014                  * the queuer finishes with the page. this is a wart for
2015                  * llite::commit_write() */
2016                 optimal += 16;
2017         }
2018         if (lop->lop_num_pending >= optimal)
2019                 RETURN(1);
2020
2021         RETURN(0);
2022 }
2023
2024 static int lop_makes_hprpc(struct loi_oap_pages *lop)
2025 {
2026         struct osc_async_page *oap;
2027         ENTRY;
2028
2029         if (cfs_list_empty(&lop->lop_urgent))
2030                 RETURN(0);
2031
2032         oap = cfs_list_entry(lop->lop_urgent.next,
2033                          struct osc_async_page, oap_urgent_item);
2034
2035         if (oap->oap_async_flags & ASYNC_HP) {
2036                 CDEBUG(D_CACHE, "hp request forcing RPC\n");
2037                 RETURN(1);
2038         }
2039
2040         RETURN(0);
2041 }
2042
2043 static void on_list(cfs_list_t *item, cfs_list_t *list,
2044                     int should_be_on)
2045 {
2046         if (cfs_list_empty(item) && should_be_on)
2047                 cfs_list_add_tail(item, list);
2048         else if (!cfs_list_empty(item) && !should_be_on)
2049                 cfs_list_del_init(item);
2050 }
2051
2052 /* maintain the loi's cli list membership invariants so that osc_send_oap_rpc
2053  * can find pages to build into rpcs quickly */
2054 void loi_list_maint(struct client_obd *cli, struct lov_oinfo *loi)
2055 {
2056         if (lop_makes_hprpc(&loi->loi_write_lop) ||
2057             lop_makes_hprpc(&loi->loi_read_lop)) {
2058                 /* HP rpc */
2059                 on_list(&loi->loi_ready_item, &cli->cl_loi_ready_list, 0);
2060                 on_list(&loi->loi_hp_ready_item, &cli->cl_loi_hp_ready_list, 1);
2061         } else {
2062                 on_list(&loi->loi_hp_ready_item, &cli->cl_loi_hp_ready_list, 0);
2063                 on_list(&loi->loi_ready_item, &cli->cl_loi_ready_list,
2064                         lop_makes_rpc(cli, &loi->loi_write_lop, OBD_BRW_WRITE)||
2065                         lop_makes_rpc(cli, &loi->loi_read_lop, OBD_BRW_READ));
2066         }
2067
2068         on_list(&loi->loi_write_item, &cli->cl_loi_write_list,
2069                 loi->loi_write_lop.lop_num_pending);
2070
2071         on_list(&loi->loi_read_item, &cli->cl_loi_read_list,
2072                 loi->loi_read_lop.lop_num_pending);
2073 }
2074
2075 static void lop_update_pending(struct client_obd *cli,
2076                                struct loi_oap_pages *lop, int cmd, int delta)
2077 {
2078         lop->lop_num_pending += delta;
2079         if (cmd & OBD_BRW_WRITE)
2080                 cli->cl_pending_w_pages += delta;
2081         else
2082                 cli->cl_pending_r_pages += delta;
2083 }
2084
2085 /**
2086  * this is called when a sync waiter receives an interruption.  Its job is to
2087  * get the caller woken as soon as possible.  If its page hasn't been put in an
2088  * rpc yet it can dequeue immediately.  Otherwise it has to mark the rpc as
2089  * desiring interruption which will forcefully complete the rpc once the rpc
2090  * has timed out.
2091  */
2092 int osc_oap_interrupted(const struct lu_env *env, struct osc_async_page *oap)
2093 {
2094         struct loi_oap_pages *lop;
2095         struct lov_oinfo *loi;
2096         int rc = -EBUSY;
2097         ENTRY;
2098
2099         LASSERT(!oap->oap_interrupted);
2100         oap->oap_interrupted = 1;
2101
2102         /* ok, it's been put in an rpc. only one oap gets a request reference */
2103         if (oap->oap_request != NULL) {
2104                 ptlrpc_mark_interrupted(oap->oap_request);
2105                 ptlrpcd_wake(oap->oap_request);
2106                 ptlrpc_req_finished(oap->oap_request);
2107                 oap->oap_request = NULL;
2108         }
2109
2110         /*
2111          * page completion may be called only if ->cpo_prep() method was
2112          * executed by osc_io_submit(), that also adds page the to pending list
2113          */
2114         if (!cfs_list_empty(&oap->oap_pending_item)) {
2115                 cfs_list_del_init(&oap->oap_pending_item);
2116                 cfs_list_del_init(&oap->oap_urgent_item);
2117
2118                 loi = oap->oap_loi;
2119                 lop = (oap->oap_cmd & OBD_BRW_WRITE) ?
2120                         &loi->loi_write_lop : &loi->loi_read_lop;
2121                 lop_update_pending(oap->oap_cli, lop, oap->oap_cmd, -1);
2122                 loi_list_maint(oap->oap_cli, oap->oap_loi);
2123                 rc = oap->oap_caller_ops->ap_completion(env,
2124                                           oap->oap_caller_data,
2125                                           oap->oap_cmd, NULL, -EINTR);
2126         }
2127
2128         RETURN(rc);
2129 }
2130
2131 /* this is trying to propogate async writeback errors back up to the
2132  * application.  As an async write fails we record the error code for later if
2133  * the app does an fsync.  As long as errors persist we force future rpcs to be
2134  * sync so that the app can get a sync error and break the cycle of queueing
2135  * pages for which writeback will fail. */
2136 static void osc_process_ar(struct osc_async_rc *ar, __u64 xid,
2137                            int rc)
2138 {
2139         if (rc) {
2140                 if (!ar->ar_rc)
2141                         ar->ar_rc = rc;
2142
2143                 ar->ar_force_sync = 1;
2144                 ar->ar_min_xid = ptlrpc_sample_next_xid();
2145                 return;
2146
2147         }
2148
2149         if (ar->ar_force_sync && (xid >= ar->ar_min_xid))
2150                 ar->ar_force_sync = 0;
2151 }
2152
2153 void osc_oap_to_pending(struct osc_async_page *oap)
2154 {
2155         struct loi_oap_pages *lop;
2156
2157         if (oap->oap_cmd & OBD_BRW_WRITE)
2158                 lop = &oap->oap_loi->loi_write_lop;
2159         else
2160                 lop = &oap->oap_loi->loi_read_lop;
2161
2162         if (oap->oap_async_flags & ASYNC_HP)
2163                 cfs_list_add(&oap->oap_urgent_item, &lop->lop_urgent);
2164         else if (oap->oap_async_flags & ASYNC_URGENT)
2165                 cfs_list_add_tail(&oap->oap_urgent_item, &lop->lop_urgent);
2166         cfs_list_add_tail(&oap->oap_pending_item, &lop->lop_pending);
2167         lop_update_pending(oap->oap_cli, lop, oap->oap_cmd, 1);
2168 }
2169
2170 /* this must be called holding the loi list lock to give coverage to exit_cache,
2171  * async_flag maintenance, and oap_request */
2172 static void osc_ap_completion(const struct lu_env *env,
2173                               struct client_obd *cli, struct obdo *oa,
2174                               struct osc_async_page *oap, int sent, int rc)
2175 {
2176         __u64 xid = 0;
2177
2178         ENTRY;
2179         if (oap->oap_request != NULL) {
2180                 xid = ptlrpc_req_xid(oap->oap_request);
2181                 ptlrpc_req_finished(oap->oap_request);
2182                 oap->oap_request = NULL;
2183         }
2184
2185         cfs_spin_lock(&oap->oap_lock);
2186         oap->oap_async_flags = 0;
2187         cfs_spin_unlock(&oap->oap_lock);
2188         oap->oap_interrupted = 0;
2189
2190         if (oap->oap_cmd & OBD_BRW_WRITE) {
2191                 osc_process_ar(&cli->cl_ar, xid, rc);
2192                 osc_process_ar(&oap->oap_loi->loi_ar, xid, rc);
2193         }
2194
2195         if (rc == 0 && oa != NULL) {
2196                 if (oa->o_valid & OBD_MD_FLBLOCKS)
2197                         oap->oap_loi->loi_lvb.lvb_blocks = oa->o_blocks;
2198                 if (oa->o_valid & OBD_MD_FLMTIME)
2199                         oap->oap_loi->loi_lvb.lvb_mtime = oa->o_mtime;
2200                 if (oa->o_valid & OBD_MD_FLATIME)
2201                         oap->oap_loi->loi_lvb.lvb_atime = oa->o_atime;
2202                 if (oa->o_valid & OBD_MD_FLCTIME)
2203                         oap->oap_loi->loi_lvb.lvb_ctime = oa->o_ctime;
2204         }
2205
2206         rc = oap->oap_caller_ops->ap_completion(env, oap->oap_caller_data,
2207                                                 oap->oap_cmd, oa, rc);
2208
2209         /* ll_ap_completion (from llite) drops PG_locked. so, a new
2210          * I/O on the page could start, but OSC calls it under lock
2211          * and thus we can add oap back to pending safely */
2212         if (rc)
2213                 /* upper layer wants to leave the page on pending queue */
2214                 osc_oap_to_pending(oap);
2215         else
2216                 osc_exit_cache(cli, oap, sent);
2217         EXIT;
2218 }
2219
2220 static int brw_interpret(const struct lu_env *env,
2221                          struct ptlrpc_request *req, void *data, int rc)
2222 {
2223         struct osc_brw_async_args *aa = data;
2224         struct client_obd *cli;
2225         int async;
2226         ENTRY;
2227
2228         rc = osc_brw_fini_request(req, rc);
2229         CDEBUG(D_INODE, "request %p aa %p rc %d\n", req, aa, rc);
2230         if (osc_recoverable_error(rc)) {
2231                 /* Only retry once for mmaped files since the mmaped page
2232                  * might be modified at anytime. We have to retry at least
2233                  * once in case there WAS really a corruption of the page
2234                  * on the network, that was not caused by mmap() modifying
2235                  * the page. Bug11742 */
2236                 if ((rc == -EAGAIN) && (aa->aa_resends > 0) &&
2237                     aa->aa_oa->o_valid & OBD_MD_FLFLAGS &&
2238                     aa->aa_oa->o_flags & OBD_FL_MMAP) {
2239                         rc = 0;
2240                 } else {
2241                         rc = osc_brw_redo_request(req, aa);
2242                         if (rc == 0)
2243                                 RETURN(0);
2244                 }
2245         }
2246
2247         if (aa->aa_ocapa) {
2248                 capa_put(aa->aa_ocapa);
2249                 aa->aa_ocapa = NULL;
2250         }
2251
2252         cli = aa->aa_cli;
2253
2254         client_obd_list_lock(&cli->cl_loi_list_lock);
2255
2256         /* We need to decrement before osc_ap_completion->osc_wake_cache_waiters
2257          * is called so we know whether to go to sync BRWs or wait for more
2258          * RPCs to complete */
2259         if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE)
2260                 cli->cl_w_in_flight--;
2261         else
2262                 cli->cl_r_in_flight--;
2263
2264         async = cfs_list_empty(&aa->aa_oaps);
2265         if (!async) { /* from osc_send_oap_rpc() */
2266                 struct osc_async_page *oap, *tmp;
2267                 /* the caller may re-use the oap after the completion call so
2268                  * we need to clean it up a little */
2269                 cfs_list_for_each_entry_safe(oap, tmp, &aa->aa_oaps,
2270                                              oap_rpc_item) {
2271                         cfs_list_del_init(&oap->oap_rpc_item);
2272                         osc_ap_completion(env, cli, aa->aa_oa, oap, 1, rc);
2273                 }
2274                 OBDO_FREE(aa->aa_oa);
2275         } else { /* from async_internal() */
2276                 obd_count i;
2277                 for (i = 0; i < aa->aa_page_count; i++)
2278                         osc_release_write_grant(aa->aa_cli, aa->aa_ppga[i], 1);
2279         }
2280         osc_wake_cache_waiters(cli);
2281         osc_check_rpcs(env, cli);
2282         client_obd_list_unlock(&cli->cl_loi_list_lock);
2283         if (!async)
2284                 cl_req_completion(env, aa->aa_clerq, rc);
2285         osc_release_ppga(aa->aa_ppga, aa->aa_page_count);
2286
2287         RETURN(rc);
2288 }
2289
2290 static struct ptlrpc_request *osc_build_req(const struct lu_env *env,
2291                                             struct client_obd *cli,
2292                                             cfs_list_t *rpc_list,
2293                                             int page_count, int cmd)
2294 {
2295         struct ptlrpc_request *req;
2296         struct brw_page **pga = NULL;
2297         struct osc_brw_async_args *aa;
2298         struct obdo *oa = NULL;
2299         const struct obd_async_page_ops *ops = NULL;
2300         void *caller_data = NULL;
2301         struct osc_async_page *oap;
2302         struct osc_async_page *tmp;
2303         struct ost_body *body;
2304         struct cl_req *clerq = NULL;
2305         enum cl_req_type crt = (cmd & OBD_BRW_WRITE) ? CRT_WRITE : CRT_READ;
2306         struct ldlm_lock *lock = NULL;
2307         struct cl_req_attr crattr;
2308         int i, rc, mpflag = 0;
2309
2310         ENTRY;
2311         LASSERT(!cfs_list_empty(rpc_list));
2312
2313         if (cmd & OBD_BRW_MEMALLOC)
2314                 mpflag = cfs_memory_pressure_get_and_set();
2315
2316         memset(&crattr, 0, sizeof crattr);
2317         OBD_ALLOC(pga, sizeof(*pga) * page_count);
2318         if (pga == NULL)
2319                 GOTO(out, req = ERR_PTR(-ENOMEM));
2320
2321         OBDO_ALLOC(oa);
2322         if (oa == NULL)
2323                 GOTO(out, req = ERR_PTR(-ENOMEM));
2324
2325         i = 0;
2326         cfs_list_for_each_entry(oap, rpc_list, oap_rpc_item) {
2327                 struct cl_page *page = osc_oap2cl_page(oap);
2328                 if (ops == NULL) {
2329                         ops = oap->oap_caller_ops;
2330                         caller_data = oap->oap_caller_data;
2331
2332                         clerq = cl_req_alloc(env, page, crt,
2333                                              1 /* only 1-object rpcs for
2334                                                 * now */);
2335                         if (IS_ERR(clerq))
2336                                 GOTO(out, req = (void *)clerq);
2337                         lock = oap->oap_ldlm_lock;
2338                 }
2339                 pga[i] = &oap->oap_brw_page;
2340                 pga[i]->off = oap->oap_obj_off + oap->oap_page_off;
2341                 CDEBUG(0, "put page %p index %lu oap %p flg %x to pga\n",
2342                        pga[i]->pg, cfs_page_index(oap->oap_page), oap, pga[i]->flag);
2343                 i++;
2344                 cl_req_page_add(env, clerq, page);
2345         }
2346
2347         /* always get the data for the obdo for the rpc */
2348         LASSERT(ops != NULL);
2349         crattr.cra_oa = oa;
2350         crattr.cra_capa = NULL;
2351         cl_req_attr_set(env, clerq, &crattr, ~0ULL);
2352         if (lock) {
2353                 oa->o_handle = lock->l_remote_handle;
2354                 oa->o_valid |= OBD_MD_FLHANDLE;
2355         }
2356
2357         rc = cl_req_prep(env, clerq);
2358         if (rc != 0) {
2359                 CERROR("cl_req_prep failed: %d\n", rc);
2360                 GOTO(out, req = ERR_PTR(rc));
2361         }
2362
2363         sort_brw_pages(pga, page_count);
2364         rc = osc_brw_prep_request(cmd, cli, oa, NULL, page_count,
2365                                   pga, &req, crattr.cra_capa, 1, 0);
2366         if (rc != 0) {
2367                 CERROR("prep_req failed: %d\n", rc);
2368                 GOTO(out, req = ERR_PTR(rc));
2369         }
2370
2371         if (cmd & OBD_BRW_MEMALLOC)
2372                 req->rq_memalloc = 1;
2373
2374         /* Need to update the timestamps after the request is built in case
2375          * we race with setattr (locally or in queue at OST).  If OST gets
2376          * later setattr before earlier BRW (as determined by the request xid),
2377          * the OST will not use BRW timestamps.  Sadly, there is no obvious
2378          * way to do this in a single call.  bug 10150 */
2379         body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
2380         cl_req_attr_set(env, clerq, &crattr,
2381                         OBD_MD_FLMTIME|OBD_MD_FLCTIME|OBD_MD_FLATIME);
2382
2383         CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
2384         aa = ptlrpc_req_async_args(req);
2385         CFS_INIT_LIST_HEAD(&aa->aa_oaps);
2386         cfs_list_splice(rpc_list, &aa->aa_oaps);
2387         CFS_INIT_LIST_HEAD(rpc_list);
2388         aa->aa_clerq = clerq;
2389 out:
2390         if (cmd & OBD_BRW_MEMALLOC)
2391                 cfs_memory_pressure_restore(mpflag);
2392
2393         capa_put(crattr.cra_capa);
2394         if (IS_ERR(req)) {
2395                 if (oa)
2396                         OBDO_FREE(oa);
2397                 if (pga)
2398                         OBD_FREE(pga, sizeof(*pga) * page_count);
2399                 /* this should happen rarely and is pretty bad, it makes the
2400                  * pending list not follow the dirty order */
2401                 client_obd_list_lock(&cli->cl_loi_list_lock);
2402                 cfs_list_for_each_entry_safe(oap, tmp, rpc_list, oap_rpc_item) {
2403                         cfs_list_del_init(&oap->oap_rpc_item);
2404
2405                         /* queued sync pages can be torn down while the pages
2406                          * were between the pending list and the rpc */
2407                         if (oap->oap_interrupted) {
2408                                 CDEBUG(D_INODE, "oap %p interrupted\n", oap);
2409                                 osc_ap_completion(env, cli, NULL, oap, 0,
2410                                                   oap->oap_count);
2411                                 continue;
2412                         }
2413                         osc_ap_completion(env, cli, NULL, oap, 0, PTR_ERR(req));
2414                 }
2415                 if (clerq && !IS_ERR(clerq))
2416                         cl_req_completion(env, clerq, PTR_ERR(req));
2417         }
2418         RETURN(req);
2419 }
2420
2421 /**
2422  * prepare pages for ASYNC io and put pages in send queue.
2423  *
2424  * \param cmd OBD_BRW_* macroses
2425  * \param lop pending pages
2426  *
2427  * \return zero if no page added to send queue.
2428  * \return 1 if pages successfully added to send queue.
2429  * \return negative on errors.
2430  */
2431 static int
2432 osc_send_oap_rpc(const struct lu_env *env, struct client_obd *cli,
2433                  struct lov_oinfo *loi,
2434                  int cmd, struct loi_oap_pages *lop)
2435 {
2436         struct ptlrpc_request *req;
2437         obd_count page_count = 0;
2438         struct osc_async_page *oap = NULL, *tmp;
2439         struct osc_brw_async_args *aa;
2440         const struct obd_async_page_ops *ops;
2441         CFS_LIST_HEAD(rpc_list);
2442         int srvlock = 0, mem_tight = 0;
2443         struct cl_object *clob = NULL;
2444         obd_off starting_offset = OBD_OBJECT_EOF;
2445         unsigned int ending_offset;
2446         int starting_page_off = 0;
2447         ENTRY;
2448
2449         /* ASYNC_HP pages first. At present, when the lock the pages is
2450          * to be canceled, the pages covered by the lock will be sent out
2451          * with ASYNC_HP. We have to send out them as soon as possible. */
2452         cfs_list_for_each_entry_safe(oap, tmp, &lop->lop_urgent, oap_urgent_item) {
2453                 if (oap->oap_async_flags & ASYNC_HP)
2454                         cfs_list_move(&oap->oap_pending_item, &lop->lop_pending);
2455                 if (++page_count >= cli->cl_max_pages_per_rpc)
2456                         break;
2457         }
2458         page_count = 0;
2459
2460         /* first we find the pages we're allowed to work with */
2461         cfs_list_for_each_entry_safe(oap, tmp, &lop->lop_pending,
2462                                      oap_pending_item) {
2463                 ops = oap->oap_caller_ops;
2464
2465                 LASSERTF(oap->oap_magic == OAP_MAGIC, "Bad oap magic: oap %p, "
2466                          "magic 0x%x\n", oap, oap->oap_magic);
2467
2468                 if (clob == NULL) {
2469                         /* pin object in memory, so that completion call-backs
2470                          * can be safely called under client_obd_list lock. */
2471                         clob = osc_oap2cl_page(oap)->cp_obj;
2472                         cl_object_get(clob);
2473                 }
2474
2475                 if (page_count != 0 &&
2476                     srvlock != !!(oap->oap_brw_flags & OBD_BRW_SRVLOCK)) {
2477                         CDEBUG(D_PAGE, "SRVLOCK flag mismatch,"
2478                                " oap %p, page %p, srvlock %u\n",
2479                                oap, oap->oap_brw_page.pg, (unsigned)!srvlock);
2480                         break;
2481                 }
2482
2483                 /* If there is a gap at the start of this page, it can't merge
2484                  * with any previous page, so we'll hand the network a
2485                  * "fragmented" page array that it can't transfer in 1 RDMA */
2486                 if (oap->oap_obj_off < starting_offset) {
2487                         if (starting_page_off != 0)
2488                                 break;
2489
2490                         starting_page_off = oap->oap_page_off;
2491                         starting_offset = oap->oap_obj_off + starting_page_off;
2492                 } else if (oap->oap_page_off != 0)
2493                         break;
2494
2495                 /* in llite being 'ready' equates to the page being locked
2496                  * until completion unlocks it.  commit_write submits a page
2497                  * as not ready because its unlock will happen unconditionally
2498                  * as the call returns.  if we race with commit_write giving
2499                  * us that page we don't want to create a hole in the page
2500                  * stream, so we stop and leave the rpc to be fired by
2501                  * another dirtier or kupdated interval (the not ready page
2502                  * will still be on the dirty list).  we could call in
2503                  * at the end of ll_file_write to process the queue again. */
2504                 if (!(oap->oap_async_flags & ASYNC_READY)) {
2505                         int rc = ops->ap_make_ready(env, oap->oap_caller_data,
2506                                                     cmd);
2507                         if (rc < 0)
2508                                 CDEBUG(D_INODE, "oap %p page %p returned %d "
2509                                                 "instead of ready\n", oap,
2510                                                 oap->oap_page, rc);
2511                         switch (rc) {
2512                         case -EAGAIN:
2513                                 /* llite is telling us that the page is still
2514                                  * in commit_write and that we should try
2515                                  * and put it in an rpc again later.  we
2516                                  * break out of the loop so we don't create
2517                                  * a hole in the sequence of pages in the rpc
2518                                  * stream.*/
2519                                 oap = NULL;
2520                                 break;
2521                         case -EINTR:
2522                                 /* the io isn't needed.. tell the checks
2523                                  * below to complete the rpc with EINTR */
2524                                 cfs_spin_lock(&oap->oap_lock);
2525                                 oap->oap_async_flags |= ASYNC_COUNT_STABLE;
2526                                 cfs_spin_unlock(&oap->oap_lock);
2527                                 oap->oap_count = -EINTR;
2528                                 break;
2529                         case 0:
2530                                 cfs_spin_lock(&oap->oap_lock);
2531                                 oap->oap_async_flags |= ASYNC_READY;
2532                                 cfs_spin_unlock(&oap->oap_lock);
2533                                 break;
2534                         default:
2535                                 LASSERTF(0, "oap %p page %p returned %d "
2536                                             "from make_ready\n", oap,
2537                                             oap->oap_page, rc);
2538                                 break;
2539                         }
2540                 }
2541                 if (oap == NULL)
2542                         break;
2543                 /*
2544                  * Page submitted for IO has to be locked. Either by
2545                  * ->ap_make_ready() or by higher layers.
2546                  */
2547 #if defined(__KERNEL__) && defined(__linux__)
2548                 {
2549                         struct cl_page *page;
2550
2551                         page = osc_oap2cl_page(oap);
2552
2553                         if (page->cp_type == CPT_CACHEABLE &&
2554                             !(PageLocked(oap->oap_page) &&
2555                               (CheckWriteback(oap->oap_page, cmd)))) {
2556                                 CDEBUG(D_PAGE, "page %p lost wb %lx/%x\n",
2557                                        oap->oap_page,
2558                                        (long)oap->oap_page->flags,
2559                                        oap->oap_async_flags);
2560                                 LBUG();
2561                         }
2562                 }
2563 #endif
2564
2565                 /* take the page out of our book-keeping */
2566                 cfs_list_del_init(&oap->oap_pending_item);
2567                 lop_update_pending(cli, lop, cmd, -1);
2568                 cfs_list_del_init(&oap->oap_urgent_item);
2569
2570                 /* ask the caller for the size of the io as the rpc leaves. */
2571                 if (!(oap->oap_async_flags & ASYNC_COUNT_STABLE)) {
2572                         oap->oap_count =
2573                                 ops->ap_refresh_count(env, oap->oap_caller_data,
2574                                                       cmd);
2575                         LASSERT(oap->oap_page_off + oap->oap_count <= CFS_PAGE_SIZE);
2576                 }
2577                 if (oap->oap_count <= 0) {
2578                         CDEBUG(D_CACHE, "oap %p count %d, completing\n", oap,
2579                                oap->oap_count);
2580                         osc_ap_completion(env, cli, NULL,
2581                                           oap, 0, oap->oap_count);
2582                         continue;
2583                 }
2584
2585                 /* now put the page back in our accounting */
2586                 cfs_list_add_tail(&oap->oap_rpc_item, &rpc_list);
2587                 if (page_count++ == 0)
2588                         srvlock = !!(oap->oap_brw_flags & OBD_BRW_SRVLOCK);
2589
2590                 if (oap->oap_brw_flags & OBD_BRW_MEMALLOC)
2591                         mem_tight = 1;
2592
2593                 /* End on a PTLRPC_MAX_BRW_SIZE boundary.  We want full-sized
2594                  * RPCs aligned on PTLRPC_MAX_BRW_SIZE boundaries to help reads
2595                  * have the same alignment as the initial writes that allocated
2596                  * extents on the server. */
2597                 ending_offset = oap->oap_obj_off + oap->oap_page_off +
2598                                 oap->oap_count;
2599                 if (!(ending_offset & (PTLRPC_MAX_BRW_SIZE - 1)))
2600                         break;
2601
2602                 if (page_count >= cli->cl_max_pages_per_rpc)
2603                         break;
2604
2605                 /* If there is a gap at the end of this page, it can't merge
2606                  * with any subsequent pages, so we'll hand the network a
2607                  * "fragmented" page array that it can't transfer in 1 RDMA */
2608                 if (oap->oap_page_off + oap->oap_count < CFS_PAGE_SIZE)
2609                         break;
2610         }
2611
2612         osc_wake_cache_waiters(cli);
2613
2614         loi_list_maint(cli, loi);
2615
2616         client_obd_list_unlock(&cli->cl_loi_list_lock);
2617
2618         if (clob != NULL)
2619                 cl_object_put(env, clob);
2620
2621         if (page_count == 0) {
2622                 client_obd_list_lock(&cli->cl_loi_list_lock);
2623                 RETURN(0);
2624         }
2625
2626         req = osc_build_req(env, cli, &rpc_list, page_count,
2627                             mem_tight ? (cmd | OBD_BRW_MEMALLOC) : cmd);
2628         if (IS_ERR(req)) {
2629                 LASSERT(cfs_list_empty(&rpc_list));
2630                 loi_list_maint(cli, loi);
2631                 RETURN(PTR_ERR(req));
2632         }
2633
2634         aa = ptlrpc_req_async_args(req);
2635
2636         starting_offset &= PTLRPC_MAX_BRW_SIZE - 1;
2637         if (cmd == OBD_BRW_READ) {
2638                 lprocfs_oh_tally_log2(&cli->cl_read_page_hist, page_count);
2639                 lprocfs_oh_tally(&cli->cl_read_rpc_hist, cli->cl_r_in_flight);
2640                 lprocfs_oh_tally_log2(&cli->cl_read_offset_hist,
2641                                       (starting_offset >> CFS_PAGE_SHIFT) + 1);
2642         } else {
2643                 lprocfs_oh_tally_log2(&cli->cl_write_page_hist, page_count);
2644                 lprocfs_oh_tally(&cli->cl_write_rpc_hist,
2645                                  cli->cl_w_in_flight);
2646                 lprocfs_oh_tally_log2(&cli->cl_write_offset_hist,
2647                                       (starting_offset >> CFS_PAGE_SHIFT) + 1);
2648         }
2649         ptlrpc_lprocfs_brw(req, aa->aa_requested_nob);
2650
2651         client_obd_list_lock(&cli->cl_loi_list_lock);
2652
2653         if (cmd == OBD_BRW_READ)
2654                 cli->cl_r_in_flight++;
2655         else
2656                 cli->cl_w_in_flight++;
2657
2658         /* queued sync pages can be torn down while the pages
2659          * were between the pending list and the rpc */
2660         tmp = NULL;
2661         cfs_list_for_each_entry(oap, &aa->aa_oaps, oap_rpc_item) {
2662                 /* only one oap gets a request reference */
2663                 if (tmp == NULL)
2664                         tmp = oap;
2665                 if (oap->oap_interrupted && !req->rq_intr) {
2666                         CDEBUG(D_INODE, "oap %p in req %p interrupted\n",
2667                                oap, req);
2668                         ptlrpc_mark_interrupted(req);
2669                 }
2670         }
2671         if (tmp != NULL)
2672                 tmp->oap_request = ptlrpc_request_addref(req);
2673
2674         DEBUG_REQ(D_INODE, req, "%d pages, aa %p. now %dr/%dw in flight",
2675                   page_count, aa, cli->cl_r_in_flight, cli->cl_w_in_flight);
2676
2677         req->rq_interpret_reply = brw_interpret;
2678         ptlrpcd_add_req(req, PSCOPE_BRW);
2679         RETURN(1);
2680 }
2681
2682 #define LOI_DEBUG(LOI, STR, args...)                                     \
2683         CDEBUG(D_INODE, "loi ready %d wr %d:%d rd %d:%d " STR,           \
2684                !cfs_list_empty(&(LOI)->loi_ready_item) ||                \
2685                !cfs_list_empty(&(LOI)->loi_hp_ready_item),               \
2686                (LOI)->loi_write_lop.lop_num_pending,                     \
2687                !cfs_list_empty(&(LOI)->loi_write_lop.lop_urgent),        \
2688                (LOI)->loi_read_lop.lop_num_pending,                      \
2689                !cfs_list_empty(&(LOI)->loi_read_lop.lop_urgent),         \
2690                args)                                                     \
2691
2692 /* This is called by osc_check_rpcs() to find which objects have pages that
2693  * we could be sending.  These lists are maintained by lop_makes_rpc(). */
2694 struct lov_oinfo *osc_next_loi(struct client_obd *cli)
2695 {
2696         ENTRY;
2697
2698         /* First return objects that have blocked locks so that they
2699          * will be flushed quickly and other clients can get the lock,
2700          * then objects which have pages ready to be stuffed into RPCs */
2701         if (!cfs_list_empty(&cli->cl_loi_hp_ready_list))
2702                 RETURN(cfs_list_entry(cli->cl_loi_hp_ready_list.next,
2703                                       struct lov_oinfo, loi_hp_ready_item));
2704         if (!cfs_list_empty(&cli->cl_loi_ready_list))
2705                 RETURN(cfs_list_entry(cli->cl_loi_ready_list.next,
2706                                       struct lov_oinfo, loi_ready_item));
2707
2708         /* then if we have cache waiters, return all objects with queued
2709          * writes.  This is especially important when many small files
2710          * have filled up the cache and not been fired into rpcs because
2711          * they don't pass the nr_pending/object threshhold */
2712         if (!cfs_list_empty(&cli->cl_cache_waiters) &&
2713             !cfs_list_empty(&cli->cl_loi_write_list))
2714                 RETURN(cfs_list_entry(cli->cl_loi_write_list.next,
2715                                       struct lov_oinfo, loi_write_item));
2716
2717         /* then return all queued objects when we have an invalid import
2718          * so that they get flushed */
2719         if (cli->cl_import == NULL || cli->cl_import->imp_invalid) {
2720                 if (!cfs_list_empty(&cli->cl_loi_write_list))
2721                         RETURN(cfs_list_entry(cli->cl_loi_write_list.next,
2722                                               struct lov_oinfo,
2723                                               loi_write_item));
2724                 if (!cfs_list_empty(&cli->cl_loi_read_list))
2725                         RETURN(cfs_list_entry(cli->cl_loi_read_list.next,
2726                                               struct lov_oinfo, loi_read_item));
2727         }
2728         RETURN(NULL);
2729 }
2730
2731 static int osc_max_rpc_in_flight(struct client_obd *cli, struct lov_oinfo *loi)
2732 {
2733         struct osc_async_page *oap;
2734         int hprpc = 0;
2735
2736         if (!cfs_list_empty(&loi->loi_write_lop.lop_urgent)) {
2737                 oap = cfs_list_entry(loi->loi_write_lop.lop_urgent.next,
2738                                      struct osc_async_page, oap_urgent_item);
2739                 hprpc = !!(oap->oap_async_flags & ASYNC_HP);
2740         }
2741
2742         if (!hprpc && !cfs_list_empty(&loi->loi_read_lop.lop_urgent)) {
2743                 oap = cfs_list_entry(loi->loi_read_lop.lop_urgent.next,
2744                                      struct osc_async_page, oap_urgent_item);
2745                 hprpc = !!(oap->oap_async_flags & ASYNC_HP);
2746         }
2747
2748         return rpcs_in_flight(cli) >= cli->cl_max_rpcs_in_flight + hprpc;
2749 }
2750
2751 /* called with the loi list lock held */
2752 void osc_check_rpcs(const struct lu_env *env, struct client_obd *cli)
2753 {
2754         struct lov_oinfo *loi;
2755         int rc = 0, race_counter = 0;
2756         ENTRY;
2757
2758         while ((loi = osc_next_loi(cli)) != NULL) {
2759                 LOI_DEBUG(loi, "%lu in flight\n", rpcs_in_flight(cli));
2760
2761                 if (osc_max_rpc_in_flight(cli, loi))
2762                         break;
2763
2764                 /* attempt some read/write balancing by alternating between
2765                  * reads and writes in an object.  The makes_rpc checks here
2766                  * would be redundant if we were getting read/write work items
2767                  * instead of objects.  we don't want send_oap_rpc to drain a
2768                  * partial read pending queue when we're given this object to
2769                  * do io on writes while there are cache waiters */
2770                 if (lop_makes_rpc(cli, &loi->loi_write_lop, OBD_BRW_WRITE)) {
2771                         rc = osc_send_oap_rpc(env, cli, loi, OBD_BRW_WRITE,
2772                                               &loi->loi_write_lop);
2773                         if (rc < 0) {
2774                                 CERROR("Write request failed with %d\n", rc);
2775
2776                                 /* osc_send_oap_rpc failed, mostly because of
2777                                  * memory pressure.
2778                                  *
2779                                  * It can't break here, because if:
2780                                  *  - a page was submitted by osc_io_submit, so
2781                                  *    page locked;
2782                                  *  - no request in flight
2783                                  *  - no subsequent request
2784                                  * The system will be in live-lock state,
2785                                  * because there is no chance to call
2786                                  * osc_io_unplug() and osc_check_rpcs() any
2787                                  * more. pdflush can't help in this case,
2788                                  * because it might be blocked at grabbing
2789                                  * the page lock as we mentioned.
2790                                  *
2791                                  * Anyway, continue to drain pages. */
2792                                 /* break; */
2793                         }
2794
2795                         if (rc > 0)
2796                                 race_counter = 0;
2797                         else if (rc == 0)
2798                                 race_counter++;
2799                 }
2800                 if (lop_makes_rpc(cli, &loi->loi_read_lop, OBD_BRW_READ)) {
2801                         rc = osc_send_oap_rpc(env, cli, loi, OBD_BRW_READ,
2802                                               &loi->loi_read_lop);
2803                         if (rc < 0)
2804                                 CERROR("Read request failed with %d\n", rc);
2805
2806                         if (rc > 0)
2807                                 race_counter = 0;
2808                         else if (rc == 0)
2809                                 race_counter++;
2810                 }
2811
2812                 /* attempt some inter-object balancing by issuing rpcs
2813                  * for each object in turn */
2814                 if (!cfs_list_empty(&loi->loi_hp_ready_item))
2815                         cfs_list_del_init(&loi->loi_hp_ready_item);
2816                 if (!cfs_list_empty(&loi->loi_ready_item))
2817                         cfs_list_del_init(&loi->loi_ready_item);
2818                 if (!cfs_list_empty(&loi->loi_write_item))
2819                         cfs_list_del_init(&loi->loi_write_item);
2820                 if (!cfs_list_empty(&loi->loi_read_item))
2821                         cfs_list_del_init(&loi->loi_read_item);
2822
2823                 loi_list_maint(cli, loi);
2824
2825                 /* send_oap_rpc fails with 0 when make_ready tells it to
2826                  * back off.  llite's make_ready does this when it tries
2827                  * to lock a page queued for write that is already locked.
2828                  * we want to try sending rpcs from many objects, but we
2829                  * don't want to spin failing with 0.  */
2830                 if (race_counter == 10)
2831                         break;
2832         }
2833         EXIT;
2834 }
2835
2836 /* we're trying to queue a page in the osc so we're subject to the
2837  * 'cl_dirty_max' limit on the number of pages that can be queued in the osc.
2838  * If the osc's queued pages are already at that limit, then we want to sleep
2839  * until there is space in the osc's queue for us.  We also may be waiting for
2840  * write credits from the OST if there are RPCs in flight that may return some
2841  * before we fall back to sync writes.
2842  *
2843  * We need this know our allocation was granted in the presence of signals */
2844 static int ocw_granted(struct client_obd *cli, struct osc_cache_waiter *ocw)
2845 {
2846         int rc;
2847         ENTRY;
2848         client_obd_list_lock(&cli->cl_loi_list_lock);
2849         rc = cfs_list_empty(&ocw->ocw_entry) || rpcs_in_flight(cli) == 0;
2850         client_obd_list_unlock(&cli->cl_loi_list_lock);
2851         RETURN(rc);
2852 };
2853
2854 /**
2855  * Non-blocking version of osc_enter_cache() that consumes grant only when it
2856  * is available.
2857  */
2858 int osc_enter_cache_try(const struct lu_env *env,
2859                         struct client_obd *cli, struct lov_oinfo *loi,
2860                         struct osc_async_page *oap, int transient)
2861 {
2862         int has_grant;
2863
2864         has_grant = cli->cl_avail_grant >= CFS_PAGE_SIZE;
2865         if (has_grant) {
2866                 osc_consume_write_grant(cli, &oap->oap_brw_page);
2867                 if (transient) {
2868                         cli->cl_dirty_transit += CFS_PAGE_SIZE;
2869                         cfs_atomic_inc(&obd_dirty_transit_pages);
2870                         oap->oap_brw_flags |= OBD_BRW_NOCACHE;
2871                 }
2872         }
2873         return has_grant;
2874 }
2875
2876 /* Caller must hold loi_list_lock - we drop/regain it if we need to wait for
2877  * grant or cache space. */
2878 static int osc_enter_cache(const struct lu_env *env,
2879                            struct client_obd *cli, struct lov_oinfo *loi,
2880                            struct osc_async_page *oap)
2881 {
2882         struct osc_cache_waiter ocw;
2883         struct l_wait_info lwi = LWI_INTR(LWI_ON_SIGNAL_NOOP, NULL);
2884
2885         ENTRY;
2886
2887         CDEBUG(D_CACHE, "dirty: %ld/%d dirty_max: %ld/%d dropped: %lu "
2888                "grant: %lu\n", cli->cl_dirty, cfs_atomic_read(&obd_dirty_pages),
2889                cli->cl_dirty_max, obd_max_dirty_pages,
2890                cli->cl_lost_grant, cli->cl_avail_grant);
2891
2892         /* force the caller to try sync io.  this can jump the list
2893          * of queued writes and create a discontiguous rpc stream */
2894         if (OBD_FAIL_CHECK(OBD_FAIL_OSC_NO_GRANT) ||
2895             cli->cl_dirty_max < CFS_PAGE_SIZE     ||
2896             cli->cl_ar.ar_force_sync || loi->loi_ar.ar_force_sync)
2897                 RETURN(-EDQUOT);
2898
2899         /* Hopefully normal case - cache space and write credits available */
2900         if (cli->cl_dirty + CFS_PAGE_SIZE <= cli->cl_dirty_max &&
2901             cfs_atomic_read(&obd_dirty_pages) + 1 <= obd_max_dirty_pages &&
2902             osc_enter_cache_try(env, cli, loi, oap, 0))
2903                 RETURN(0);
2904
2905         /* It is safe to block as a cache waiter as long as there is grant
2906          * space available or the hope of additional grant being returned
2907          * when an in flight write completes.  Using the write back cache
2908          * if possible is preferable to sending the data synchronously
2909          * because write pages can then be merged in to large requests.
2910          * The addition of this cache waiter will causing pending write
2911          * pages to be sent immediately. */
2912         if (cli->cl_w_in_flight || cli->cl_avail_grant >= CFS_PAGE_SIZE) {
2913                 cfs_list_add_tail(&ocw.ocw_entry, &cli->cl_cache_waiters);
2914                 cfs_waitq_init(&ocw.ocw_waitq);
2915                 ocw.ocw_oap = oap;
2916                 ocw.ocw_rc = 0;
2917
2918                 loi_list_maint(cli, loi);
2919                 osc_check_rpcs(env, cli);
2920                 client_obd_list_unlock(&cli->cl_loi_list_lock);
2921
2922                 CDEBUG(D_CACHE, "sleeping for cache space\n");
2923                 l_wait_event(ocw.ocw_waitq, ocw_granted(cli, &ocw), &lwi);
2924
2925                 client_obd_list_lock(&cli->cl_loi_list_lock);
2926                 if (!cfs_list_empty(&ocw.ocw_entry)) {
2927                         cfs_list_del(&ocw.ocw_entry);
2928                         RETURN(-EINTR);
2929                 }
2930                 RETURN(ocw.ocw_rc);
2931         }
2932
2933         RETURN(-EDQUOT);
2934 }
2935
2936
2937 int osc_prep_async_page(struct obd_export *exp, struct lov_stripe_md *lsm,
2938                         struct lov_oinfo *loi, cfs_page_t *page,
2939                         obd_off offset, const struct obd_async_page_ops *ops,
2940                         void *data, void **res, int nocache,
2941                         struct lustre_handle *lockh)
2942 {
2943         struct osc_async_page *oap;
2944
2945         ENTRY;
2946
2947         if (!page)
2948                 return cfs_size_round(sizeof(*oap));
2949
2950         oap = *res;
2951         oap->oap_magic = OAP_MAGIC;
2952         oap->oap_cli = &exp->exp_obd->u.cli;
2953         oap->oap_loi = loi;
2954
2955         oap->oap_caller_ops = ops;
2956         oap->oap_caller_data = data;
2957
2958         oap->oap_page = page;
2959         oap->oap_obj_off = offset;
2960         if (!client_is_remote(exp) &&
2961             cfs_capable(CFS_CAP_SYS_RESOURCE))
2962                 oap->oap_brw_flags = OBD_BRW_NOQUOTA;
2963
2964         LASSERT(!(offset & ~CFS_PAGE_MASK));
2965
2966         CFS_INIT_LIST_HEAD(&oap->oap_pending_item);
2967         CFS_INIT_LIST_HEAD(&oap->oap_urgent_item);
2968         CFS_INIT_LIST_HEAD(&oap->oap_rpc_item);
2969         CFS_INIT_LIST_HEAD(&oap->oap_page_list);
2970
2971         cfs_spin_lock_init(&oap->oap_lock);
2972         CDEBUG(D_CACHE, "oap %p page %p obj off "LPU64"\n", oap, page, offset);
2973         RETURN(0);
2974 }
2975
2976 int osc_queue_async_io(const struct lu_env *env, struct obd_export *exp,
2977                        struct lov_stripe_md *lsm, struct lov_oinfo *loi,
2978                        struct osc_async_page *oap, int cmd, int off,
2979                        int count, obd_flag brw_flags, enum async_flags async_flags)
2980 {
2981         struct client_obd *cli = &exp->exp_obd->u.cli;
2982         int rc = 0;
2983         ENTRY;
2984
2985         if (oap->oap_magic != OAP_MAGIC)
2986                 RETURN(-EINVAL);
2987
2988         if (cli->cl_import == NULL || cli->cl_import->imp_invalid)
2989                 RETURN(-EIO);
2990
2991         if (!cfs_list_empty(&oap->oap_pending_item) ||
2992             !cfs_list_empty(&oap->oap_urgent_item) ||
2993             !cfs_list_empty(&oap->oap_rpc_item))
2994                 RETURN(-EBUSY);
2995
2996         /* check if the file's owner/group is over quota */
2997         if ((cmd & OBD_BRW_WRITE) && !(cmd & OBD_BRW_NOQUOTA)) {
2998                 struct cl_object *obj;
2999                 struct cl_attr    attr; /* XXX put attr into thread info */
3000                 unsigned int qid[MAXQUOTAS];
3001
3002                 obj = cl_object_top(osc_oap2cl_page(oap)->cp_obj);
3003
3004                 cl_object_attr_lock(obj);
3005                 rc = cl_object_attr_get(env, obj, &attr);
3006                 cl_object_attr_unlock(obj);
3007
3008                 qid[USRQUOTA] = attr.cat_uid;
3009                 qid[GRPQUOTA] = attr.cat_gid;
3010                 if (rc == 0 &&
3011                     lquota_chkdq(quota_interface, cli, qid) == NO_QUOTA)
3012                         rc = -EDQUOT;
3013                 if (rc)
3014                         RETURN(rc);
3015         }
3016
3017         if (loi == NULL)
3018                 loi = lsm->lsm_oinfo[0];
3019
3020         client_obd_list_lock(&cli->cl_loi_list_lock);
3021
3022         LASSERT(off + count <= CFS_PAGE_SIZE);
3023         oap->oap_cmd = cmd;
3024         oap->oap_page_off = off;
3025         oap->oap_count = count;
3026         oap->oap_brw_flags = brw_flags;
3027         /* Give a hint to OST that requests are coming from kswapd - bug19529 */
3028         if (cfs_memory_pressure_get())
3029                 oap->oap_brw_flags |= OBD_BRW_MEMALLOC;
3030         cfs_spin_lock(&oap->oap_lock);
3031         oap->oap_async_flags = async_flags;
3032         cfs_spin_unlock(&oap->oap_lock);
3033
3034         if (cmd & OBD_BRW_WRITE) {
3035                 rc = osc_enter_cache(env, cli, loi, oap);
3036                 if (rc) {
3037                         client_obd_list_unlock(&cli->cl_loi_list_lock);
3038                         RETURN(rc);
3039                 }
3040         }
3041
3042         osc_oap_to_pending(oap);
3043         loi_list_maint(cli, loi);
3044
3045         LOI_DEBUG(loi, "oap %p page %p added for cmd %d\n", oap, oap->oap_page,
3046                   cmd);
3047
3048         osc_check_rpcs(env, cli);
3049         client_obd_list_unlock(&cli->cl_loi_list_lock);
3050
3051         RETURN(0);
3052 }
3053
3054 /* aka (~was & now & flag), but this is more clear :) */
3055 #define SETTING(was, now, flag) (!(was & flag) && (now & flag))
3056
3057 int osc_set_async_flags_base(struct client_obd *cli,
3058                              struct lov_oinfo *loi, struct osc_async_page *oap,
3059                              obd_flag async_flags)
3060 {
3061         struct loi_oap_pages *lop;
3062         int flags = 0;
3063         ENTRY;
3064
3065         LASSERT(!cfs_list_empty(&oap->oap_pending_item));
3066
3067         if (oap->oap_cmd & OBD_BRW_WRITE) {
3068                 lop = &loi->loi_write_lop;
3069         } else {
3070                 lop = &loi->loi_read_lop;
3071         }
3072
3073         if ((oap->oap_async_flags & async_flags) == async_flags)
3074                 RETURN(0);
3075
3076         if (SETTING(oap->oap_async_flags, async_flags, ASYNC_READY))
3077                 flags |= ASYNC_READY;
3078
3079         if (SETTING(oap->oap_async_flags, async_flags, ASYNC_URGENT) &&
3080             cfs_list_empty(&oap->oap_rpc_item)) {
3081                 if (oap->oap_async_flags & ASYNC_HP)
3082                         cfs_list_add(&oap->oap_urgent_item, &lop->lop_urgent);
3083                 else
3084                         cfs_list_add_tail(&oap->oap_urgent_item,
3085                                           &lop->lop_urgent);
3086                 flags |= ASYNC_URGENT;
3087                 loi_list_maint(cli, loi);
3088         }
3089         cfs_spin_lock(&oap->oap_lock);
3090         oap->oap_async_flags |= flags;
3091         cfs_spin_unlock(&oap->oap_lock);
3092
3093         LOI_DEBUG(loi, "oap %p page %p has flags %x\n", oap, oap->oap_page,
3094                         oap->oap_async_flags);
3095         RETURN(0);
3096 }
3097
3098 int osc_teardown_async_page(struct obd_export *exp, struct lov_stripe_md *lsm,
3099                             struct lov_oinfo *loi, struct osc_async_page *oap)
3100 {
3101         struct client_obd *cli = &exp->exp_obd->u.cli;
3102         struct loi_oap_pages *lop;
3103         int rc = 0;
3104         ENTRY;
3105
3106         if (oap->oap_magic != OAP_MAGIC)
3107                 RETURN(-EINVAL);
3108
3109         if (loi == NULL)
3110                 loi = lsm->lsm_oinfo[0];
3111
3112         if (oap->oap_cmd & OBD_BRW_WRITE) {
3113                 lop = &loi->loi_write_lop;
3114         } else {
3115                 lop = &loi->loi_read_lop;
3116         }
3117
3118         client_obd_list_lock(&cli->cl_loi_list_lock);
3119
3120         if (!cfs_list_empty(&oap->oap_rpc_item))
3121                 GOTO(out, rc = -EBUSY);
3122
3123         osc_exit_cache(cli, oap, 0);
3124         osc_wake_cache_waiters(cli);
3125
3126         if (!cfs_list_empty(&oap->oap_urgent_item)) {
3127                 cfs_list_del_init(&oap->oap_urgent_item);
3128                 cfs_spin_lock(&oap->oap_lock);
3129                 oap->oap_async_flags &= ~(ASYNC_URGENT | ASYNC_HP);
3130                 cfs_spin_unlock(&oap->oap_lock);
3131         }
3132         if (!cfs_list_empty(&oap->oap_pending_item)) {
3133                 cfs_list_del_init(&oap->oap_pending_item);
3134                 lop_update_pending(cli, lop, oap->oap_cmd, -1);
3135         }
3136         loi_list_maint(cli, loi);
3137         LOI_DEBUG(loi, "oap %p page %p torn down\n", oap, oap->oap_page);
3138 out:
3139         client_obd_list_unlock(&cli->cl_loi_list_lock);
3140         RETURN(rc);
3141 }
3142
3143 static int osc_set_lock_data_with_check(struct ldlm_lock *lock,
3144                                         struct ldlm_enqueue_info *einfo)
3145 {
3146         void *data = einfo->ei_cbdata;
3147         int set = 0;
3148
3149         LASSERT(lock != NULL);
3150         LASSERT(lock->l_blocking_ast == einfo->ei_cb_bl);
3151         LASSERT(lock->l_resource->lr_type == einfo->ei_type);
3152         LASSERT(lock->l_completion_ast == einfo->ei_cb_cp);
3153         LASSERT(lock->l_glimpse_ast == einfo->ei_cb_gl);
3154
3155         lock_res_and_lock(lock);
3156         cfs_spin_lock(&osc_ast_guard);
3157
3158         if (lock->l_ast_data == NULL)
3159                 lock->l_ast_data = data;
3160         if (lock->l_ast_data == data)
3161                 set = 1;
3162
3163         cfs_spin_unlock(&osc_ast_guard);
3164         unlock_res_and_lock(lock);