src/fs/rock/RockSwapDir.cc

   1 /*
   2  * DEBUG: section 47    Store Directory Routines
   3  */
   4
   5 #include "squid.h"
   6 #include "cache_cf.h"
   7 #include "ConfigOption.h"
   8 #include "DiskIO/DiskIOModule.h"
   9 #include "DiskIO/DiskIOStrategy.h"
  10 #include "DiskIO/ReadRequest.h"
  11 #include "DiskIO/WriteRequest.h"
  12 #include "fs/rock/RockSwapDir.h"
  13 #include "fs/rock/RockIoState.h"
  14 #include "fs/rock/RockIoRequests.h"
  15 #include "fs/rock/RockRebuild.h"
  16 #include "globals.h"
  17 #include "ipc/mem/Pages.h"
  18 #include "MemObject.h"
  19 #include "Parsing.h"
  20 #include "SquidConfig.h"
  21 #include "SquidMath.h"
  22 #include "tools.h"
  23
  24 #include <cstdlib>
  25 #include <iomanip>
  26
  27 #if HAVE_SYS_STAT_H
  28 #include <sys/stat.h>
  29 #endif
  30
  31 const int64_t Rock::SwapDir::HeaderSize = 16*1024;
  32
  33 Rock::SwapDir::SwapDir(): ::SwapDir("rock"),
  34     slotSize(HeaderSize), filePath(NULL), map(NULL), io(NULL), allSlots(NULL),
  35     waitingForPage(NULL)
  36 {
  37 }
  38
  39 Rock::SwapDir::~SwapDir()
  40 {
  41     delete io;
  42     delete map;
  43     safe_free(filePath);
  44 }
  45
  46 StoreSearch *
  47 Rock::SwapDir::search(String const url, HttpRequest *)
  48 {
  49     assert(false);
  50     return NULL; // XXX: implement
  51 }
  52
  53 void
  54 Rock::SwapDir::get(String const key, STOREGETCLIENT cb, void *data)
  55 {
  56     ::SwapDir::get(key, cb, data);
  57 }
  58
  59 // called when Squid core needs a StoreEntry with a given key
  60 StoreEntry *
  61 Rock::SwapDir::get(const cache_key *key)
  62 {
  63     if (!map || !theFile || !theFile->canRead())
  64         return NULL;
  65
  66     sfileno filen;
  67     const Ipc::StoreMapAnchor *const slot = map->openForReading(key, filen);
  68     if (!slot)
  69         return NULL;
  70
  71     const Ipc::StoreMapAnchor::Basics &basics = slot->basics;
  72
  73     // create a brand new store entry and initialize it with stored basics
  74     StoreEntry *e = new StoreEntry();
  75     e->lock_count = 0;
  76     e->swap_dirn = index;
  77     e->swap_filen = filen;
  78     e->swap_file_sz = basics.swap_file_sz;
  79     e->lastref = basics.lastref;
  80     e->timestamp = basics.timestamp;
  81     e->expires = basics.expires;
  82     e->lastmod = basics.lastmod;
  83     e->refcount = basics.refcount;
  84     e->flags = basics.flags;
  85     e->store_status = STORE_OK;
  86     e->setMemStatus(NOT_IN_MEMORY);
  87     e->swap_status = SWAPOUT_DONE;
  88     e->ping_status = PING_NONE;
  89     EBIT_SET(e->flags, ENTRY_CACHABLE);
  90     EBIT_CLR(e->flags, RELEASE_REQUEST);
  91     EBIT_CLR(e->flags, KEY_PRIVATE);
  92     EBIT_SET(e->flags, ENTRY_VALIDATED);
  93     e->hashInsert(key);
  94     trackReferences(*e);
  95
  96     return e;
  97     // the disk entry remains open for reading, protected from modifications
  98 }
  99
 100 void Rock::SwapDir::disconnect(StoreEntry &e)
 101 {
 102     assert(e.swap_dirn == index);
 103     assert(e.swap_filen >= 0);
 104     // cannot have SWAPOUT_NONE entry with swap_filen >= 0
 105     assert(e.swap_status != SWAPOUT_NONE);
 106
 107     // do not rely on e.swap_status here because there is an async delay
 108     // before it switches from SWAPOUT_WRITING to SWAPOUT_DONE.
 109
 110     // since e has swap_filen, its slot is locked for either reading or writing
 111     map->abortIo(e.swap_filen);
 112     e.swap_dirn = -1;
 113     e.swap_filen = -1;
 114     e.swap_status = SWAPOUT_NONE;
 115 }
 116
 117 uint64_t
 118 Rock::SwapDir::currentSize() const
 119 {
 120     const uint64_t spaceSize = !freeSlots ?
 121         maxSize() : (slotSize * freeSlots->size());
 122     // everything that is not free is in use
 123     return maxSize() - spaceSize;
 124 }
 125
 126 uint64_t
 127 Rock::SwapDir::currentCount() const
 128 {
 129     return map ? map->entryCount() : 0;
 130 }
 131
 132 /// In SMP mode only the disker process reports stats to avoid
 133 /// counting the same stats by multiple processes.
 134 bool
 135 Rock::SwapDir::doReportStat() const
 136 {
 137     return ::SwapDir::doReportStat() && (!UsingSmp() || IamDiskProcess());
 138 }
 139
 140 void
 141 Rock::SwapDir::swappedOut(const StoreEntry &)
 142 {
 143     // stats are not stored but computed when needed
 144 }
 145
 146 int64_t
 147 Rock::SwapDir::entryLimitAllowed() const
 148 {
 149     const int64_t eLimitLo = map ? map->entryLimit() : 0; // dynamic shrinking unsupported
 150     const int64_t eWanted = (maxSize() - HeaderSize)/slotSize;
 151     return min(max(eLimitLo, eWanted), entryLimitHigh());
 152 }
 153
 154 // TODO: encapsulate as a tool; identical to CossSwapDir::create()
 155 void
 156 Rock::SwapDir::create()
 157 {
 158     assert(path);
 159     assert(filePath);
 160
 161     if (UsingSmp() && !IamDiskProcess()) {
 162         debugs (47,3, HERE << "disker will create in " << path);
 163         return;
 164     }
 165
 166     debugs (47,3, HERE << "creating in " << path);
 167
 168     struct stat swap_sb;
 169     if (::stat(path, &swap_sb) < 0) {
 170         debugs (47, DBG_IMPORTANT, "Creating Rock db directory: " << path);
 171         const int res = mkdir(path, 0700);
 172         if (res != 0)
 173             createError("mkdir");
 174     }
 175
 176     const int swap = open(filePath, O_WRONLY|O_CREAT|O_TRUNC|O_BINARY, 0600);
 177     if (swap < 0)
 178         createError("create");
 179
 180 #if SLOWLY_FILL_WITH_ZEROS
 181     char block[1024];
 182     Must(maxSize() % sizeof(block) == 0);
 183     memset(block, '\0', sizeof(block));
 184
 185     for (off_t offset = 0; offset < maxSize(); offset += sizeof(block)) {
 186         if (write(swap, block, sizeof(block)) != sizeof(block))
 187             createError("write");
 188     }
 189 #else
 190     if (ftruncate(swap, maxSize()) != 0)
 191         createError("truncate");
 192
 193     char header[HeaderSize];
 194     memset(header, '\0', sizeof(header));
 195     if (write(swap, header, sizeof(header)) != sizeof(header))
 196         createError("write");
 197 #endif
 198
 199     close(swap);
 200 }
 201
 202 // report Rock DB creation error and exit
 203 void
 204 Rock::SwapDir::createError(const char *const msg) {
 205     debugs(47, DBG_CRITICAL, "ERROR: Failed to initialize Rock Store db in " <<
 206            filePath << "; " << msg << " error: " << xstrerror());
 207     fatal("Rock Store db creation error");
 208 }
 209
 210 void
 211 Rock::SwapDir::init()
 212 {
 213     debugs(47,2, HERE);
 214
 215     // XXX: SwapDirs aren't refcounted. We make IORequestor calls, which
 216     // are refcounted. We up our count once to avoid implicit delete's.
 217     lock();
 218
 219     Must(!map);
 220     map = new DirMap(inodeMapPath());
 221     map->cleaner = this;
 222
 223     const char *ioModule = needsDiskStrand() ? "IpcIo" : "Blocking";
 224     if (DiskIOModule *m = DiskIOModule::Find(ioModule)) {
 225         debugs(47,2, HERE << "Using DiskIO module: " << ioModule);
 226         io = m->createStrategy();
 227         io->init();
 228     } else {
 229         debugs(47, DBG_CRITICAL, "FATAL: Rock store is missing DiskIO module: " <<
 230                ioModule);
 231         fatal("Rock Store missing a required DiskIO module");
 232     }
 233
 234     theFile = io->newFile(filePath);
 235     theFile->configure(fileConfig);
 236     theFile->open(O_RDWR, 0644, this);
 237
 238     freeSlots = shm_old(Ipc::Mem::PageStack)(freeSlotsPath());
 239
 240     // Increment early. Otherwise, if one SwapDir finishes rebuild before
 241     // others start, storeRebuildComplete() will think the rebuild is over!
 242     // TODO: move store_dirs_rebuilding hack to store modules that need it.
 243     ++StoreController::store_dirs_rebuilding;
 244 }
 245
 246 bool
 247 Rock::SwapDir::needsDiskStrand() const
 248 {
 249     const bool wontEvenWorkWithoutDisker = Config.workers > 1;
 250     const bool wouldWorkBetterWithDisker = DiskIOModule::Find("IpcIo");
 251     return InDaemonMode() && (wontEvenWorkWithoutDisker ||
 252                               wouldWorkBetterWithDisker);
 253 }
 254
 255 void
 256 Rock::SwapDir::parse(int anIndex, char *aPath)
 257 {
 258     index = anIndex;
 259
 260     path = xstrdup(aPath);
 261
 262     // cache store is located at path/db
 263     String fname(path);
 264     fname.append("/rock");
 265     filePath = xstrdup(fname.termedBuf());
 266
 267     parseSize(false);
 268     parseOptions(0);
 269
 270     // Current openForWriting() code overwrites the old slot if needed
 271     // and possible, so proactively removing old slots is probably useless.
 272     assert(!repl); // repl = createRemovalPolicy(Config.replPolicy);
 273
 274     validateOptions();
 275 }
 276
 277 void
 278 Rock::SwapDir::reconfigure()
 279 {
 280     parseSize(true);
 281     parseOptions(1);
 282     // TODO: can we reconfigure the replacement policy (repl)?
 283     validateOptions();
 284 }
 285
 286 /// parse maximum db disk size
 287 void
 288 Rock::SwapDir::parseSize(const bool reconfiguring)
 289 {
 290     const int i = GetInteger();
 291     if (i < 0)
 292         fatal("negative Rock cache_dir size value");
 293     const uint64_t new_max_size =
 294         static_cast<uint64_t>(i) << 20; // MBytes to Bytes
 295     if (!reconfiguring)
 296         max_size = new_max_size;
 297     else if (new_max_size != max_size) {
 298         debugs(3, DBG_IMPORTANT, "WARNING: cache_dir '" << path << "' size "
 299                "cannot be changed dynamically, value left unchanged (" <<
 300                (max_size >> 20) << " MB)");
 301     }
 302 }
 303
 304 ConfigOption *
 305 Rock::SwapDir::getOptionTree() const
 306 {
 307     ConfigOptionVector *vector = dynamic_cast<ConfigOptionVector*>(::SwapDir::getOptionTree());
 308     assert(vector);
 309     vector->options.push_back(new ConfigOptionAdapter<SwapDir>(*const_cast<SwapDir *>(this), &SwapDir::parseSizeOption, &SwapDir::dumpSizeOption));
 310     vector->options.push_back(new ConfigOptionAdapter<SwapDir>(*const_cast<SwapDir *>(this), &SwapDir::parseTimeOption, &SwapDir::dumpTimeOption));
 311     vector->options.push_back(new ConfigOptionAdapter<SwapDir>(*const_cast<SwapDir *>(this), &SwapDir::parseRateOption, &SwapDir::dumpRateOption));
 312     return vector;
 313 }
 314
 315 bool
 316 Rock::SwapDir::allowOptionReconfigure(const char *const option) const
 317 {
 318     return strcmp(option, "slot-size") != 0 &&
 319            ::SwapDir::allowOptionReconfigure(option);
 320 }
 321
 322 /// parses time-specific options; mimics ::SwapDir::optionObjectSizeParse()
 323 bool
 324 Rock::SwapDir::parseTimeOption(char const *option, const char *value, int reconfiguring)
 325 {
 326     // TODO: ::SwapDir or, better, Config should provide time-parsing routines,
 327     // including time unit handling. Same for size and rate.
 328
 329     time_msec_t *storedTime;
 330     if (strcmp(option, "swap-timeout") == 0)
 331         storedTime = &fileConfig.ioTimeout;
 332     else
 333         return false;
 334
 335     if (!value)
 336         self_destruct();
 337
 338     // TODO: handle time units and detect parsing errors better
 339     const int64_t parsedValue = strtoll(value, NULL, 10);
 340     if (parsedValue < 0) {
 341         debugs(3, DBG_CRITICAL, "FATAL: cache_dir " << path << ' ' << option << " must not be negative but is: " << parsedValue);
 342         self_destruct();
 343     }
 344
 345     const time_msec_t newTime = static_cast<time_msec_t>(parsedValue);
 346
 347     if (!reconfiguring)
 348         *storedTime = newTime;
 349     else if (*storedTime != newTime) {
 350         debugs(3, DBG_IMPORTANT, "WARNING: cache_dir " << path << ' ' << option
 351                << " cannot be changed dynamically, value left unchanged: " <<
 352                *storedTime);
 353     }
 354
 355     return true;
 356 }
 357
 358 /// reports time-specific options; mimics ::SwapDir::optionObjectSizeDump()
 359 void
 360 Rock::SwapDir::dumpTimeOption(StoreEntry * e) const
 361 {
 362     if (fileConfig.ioTimeout)
 363         storeAppendPrintf(e, " swap-timeout=%" PRId64,
 364                           static_cast<int64_t>(fileConfig.ioTimeout));
 365 }
 366
 367 /// parses rate-specific options; mimics ::SwapDir::optionObjectSizeParse()
 368 bool
 369 Rock::SwapDir::parseRateOption(char const *option, const char *value, int isaReconfig)
 370 {
 371     int *storedRate;
 372     if (strcmp(option, "max-swap-rate") == 0)
 373         storedRate = &fileConfig.ioRate;
 374     else
 375         return false;
 376
 377     if (!value)
 378         self_destruct();
 379
 380     // TODO: handle time units and detect parsing errors better
 381     const int64_t parsedValue = strtoll(value, NULL, 10);
 382     if (parsedValue < 0) {
 383         debugs(3, DBG_CRITICAL, "FATAL: cache_dir " << path << ' ' << option << " must not be negative but is: " << parsedValue);
 384         self_destruct();
 385     }
 386
 387     const int newRate = static_cast<int>(parsedValue);
 388
 389     if (newRate < 0) {
 390         debugs(3, DBG_CRITICAL, "FATAL: cache_dir " << path << ' ' << option << " must not be negative but is: " << newRate);
 391         self_destruct();
 392     }
 393
 394     if (!isaReconfig)
 395         *storedRate = newRate;
 396     else if (*storedRate != newRate) {
 397         debugs(3, DBG_IMPORTANT, "WARNING: cache_dir " << path << ' ' << option
 398                << " cannot be changed dynamically, value left unchanged: " <<
 399                *storedRate);
 400     }
 401
 402     return true;
 403 }
 404
 405 /// reports rate-specific options; mimics ::SwapDir::optionObjectSizeDump()
 406 void
 407 Rock::SwapDir::dumpRateOption(StoreEntry * e) const
 408 {
 409     if (fileConfig.ioRate >= 0)
 410         storeAppendPrintf(e, " max-swap-rate=%d", fileConfig.ioRate);
 411 }
 412
 413 /// parses size-specific options; mimics ::SwapDir::optionObjectSizeParse()
 414 bool
 415 Rock::SwapDir::parseSizeOption(char const *option, const char *value, int reconfiguring)
 416 {
 417     uint64_t *storedSize;
 418     if (strcmp(option, "slot-size") == 0)
 419         storedSize = &slotSize;
 420     else
 421         return false;
 422
 423     if (!value)
 424         self_destruct();
 425
 426     // TODO: handle size units and detect parsing errors better
 427     const uint64_t newSize = strtoll(value, NULL, 10);
 428     if (newSize <= 0) {
 429         debugs(3, DBG_CRITICAL, "FATAL: cache_dir " << path << ' ' << option << " must be positive; got: " << newSize);
 430         self_destruct();
 431     }
 432
 433     if (newSize <= sizeof(DbCellHeader)) {
 434         debugs(3, DBG_CRITICAL, "FATAL: cache_dir " << path << ' ' << option << " must exceed " << sizeof(DbCellHeader) << "; got: " << newSize);
 435         self_destruct();
 436     }
 437
 438     if (!reconfiguring)
 439         *storedSize = newSize;
 440     else if (*storedSize != newSize) {
 441         debugs(3, DBG_IMPORTANT, "WARNING: cache_dir " << path << ' ' << option
 442                << " cannot be changed dynamically, value left unchanged: " <<
 443                *storedSize);
 444     }
 445
 446     return true;
 447 }
 448
 449 /// reports size-specific options; mimics ::SwapDir::optionObjectSizeDump()
 450 void
 451 Rock::SwapDir::dumpSizeOption(StoreEntry * e) const
 452 {
 453     storeAppendPrintf(e, " slot-size=%" PRId64, slotSize);
 454 }
 455
 456 /// check the results of the configuration; only level-0 debugging works here
 457 void
 458 Rock::SwapDir::validateOptions()
 459 {
 460     if (slotSize <= 0)
 461         fatal("Rock store requires a positive slot-size");
 462
 463     const int64_t maxSizeRoundingWaste = 1024 * 1024; // size is configured in MB
 464     const int64_t slotSizeRoundingWaste = slotSize;
 465     const int64_t maxRoundingWaste =
 466         max(maxSizeRoundingWaste, slotSizeRoundingWaste);
 467     const int64_t usableDiskSize = diskOffset(entryLimitAllowed());
 468     const int64_t diskWasteSize = maxSize() - usableDiskSize;
 469     Must(diskWasteSize >= 0);
 470
 471     // warn if maximum db size is not reachable due to sfileno limit
 472     if (entryLimitAllowed() == entryLimitHigh() &&
 473             diskWasteSize >= maxRoundingWaste) {
 474         debugs(47, DBG_CRITICAL, "Rock store cache_dir[" << index << "] '" << path << "':");
 475         debugs(47, DBG_CRITICAL, "\tmaximum number of entries: " << entryLimitAllowed());
 476         debugs(47, DBG_CRITICAL, "\tdb slot size: " << slotSize << " Bytes");
 477         debugs(47, DBG_CRITICAL, "\tmaximum db size: " << maxSize() << " Bytes");
 478         debugs(47, DBG_CRITICAL, "\tusable db size:  " << usableDiskSize << " Bytes");
 479         debugs(47, DBG_CRITICAL, "\tdisk space waste: " << diskWasteSize << " Bytes");
 480         debugs(47, DBG_CRITICAL, "WARNING: Rock store config wastes space.");
 481     }
 482 }
 483
 484 void
 485 Rock::SwapDir::rebuild()
 486 {
 487     //++StoreController::store_dirs_rebuilding; // see Rock::SwapDir::init()
 488     AsyncJob::Start(new Rebuild(this));
 489 }
 490
 491 bool
 492 Rock::SwapDir::canStore(const StoreEntry &e, int64_t diskSpaceNeeded, int &load) const
 493 {
 494     if (!::SwapDir::canStore(e, sizeof(DbCellHeader)+diskSpaceNeeded, load))
 495         return false;
 496
 497     if (!theFile || !theFile->canWrite())
 498         return false;
 499
 500     if (!map)
 501         return false;
 502
 503     // Do not start I/O transaction if there are less than 10% free pages left.
 504     // TODO: reserve page instead
 505     if (needsDiskStrand() &&
 506             Ipc::Mem::PageLevel(Ipc::Mem::PageId::ioPage) >= 0.9 * Ipc::Mem::PageLimit(Ipc::Mem::PageId::ioPage)) {
 507         debugs(47, 5, HERE << "too few shared pages for IPC I/O left");
 508         return false;
 509     }
 510
 511     if (io->shedLoad())
 512         return false;
 513
 514     load = io->load();
 515     return true;
 516 }
 517
 518 StoreIOState::Pointer
 519 Rock::SwapDir::createStoreIO(StoreEntry &e, StoreIOState::STFNCB *cbFile, StoreIOState::STIOCB *cbIo, void *data)
 520 {
 521     if (!theFile || theFile->error()) {
 522         debugs(47,4, HERE << theFile);
 523         return NULL;
 524     }
 525
 526     sfileno filen;
 527     Ipc::StoreMapAnchor *const slot =
 528         map->openForWriting(reinterpret_cast<const cache_key *>(e.key), filen);
 529     if (!slot) {
 530         debugs(47, 5, HERE << "map->add failed");
 531         return NULL;
 532     }
 533
 534     assert(filen >= 0);
 535     slot->set(e);
 536
 537     // XXX: We rely on our caller, storeSwapOutStart(), to set e.fileno.
 538     // If that does not happen, the entry will not decrement the read level!
 539
 540     Rock::SwapDir::Pointer self(this);
 541     IoState *sio = new IoState(self, &e, cbFile, cbIo, data);
 542
 543     sio->swap_dirn = index;
 544     sio->swap_filen = filen;
 545     sio->writeableAnchor_ = slot;
 546
 547     debugs(47,5, HERE << "dir " << index << " created new filen " <<
 548            std::setfill('0') << std::hex << std::uppercase << std::setw(8) <<
 549            sio->swap_filen << std::dec << " starting at " <<
 550            diskOffset(sio->swap_filen));
 551
 552     sio->file(theFile);
 553
 554     trackReferences(e);
 555     return sio;
 556 }
 557
 558 int64_t
 559 Rock::SwapDir::diskOffset(int filen) const
 560 {
 561      assert(filen >= 0);
 562      return HeaderSize + slotSize*filen;
 563 }
 564
 565 int64_t
 566 Rock::SwapDir::diskOffset(Ipc::Mem::PageId &pageId) const
 567 {
 568     assert(pageId);
 569     return diskOffset(pageId.number - 1);
 570 }
 571
 572 int64_t
 573 Rock::SwapDir::diskOffsetLimit() const
 574 {
 575     assert(map);
 576     return diskOffset(map->entryLimit());
 577 }
 578
 579 int
 580 Rock::SwapDir::entryMaxPayloadSize() const
 581 {
 582     return slotSize - sizeof(DbCellHeader);
 583 }
 584
 585 int
 586 Rock::SwapDir::entriesNeeded(const int64_t objSize) const
 587 {
 588     return (objSize + entryMaxPayloadSize() - 1) / entryMaxPayloadSize();
 589 }
 590
 591 bool
 592 Rock::SwapDir::useFreeSlot(Ipc::Mem::PageId &pageId)
 593 {
 594     if (freeSlots->pop(pageId)) {
 595         debugs(47, 5, "got a previously free slot: " << pageId);
 596         return true;
 597     }
 598
 599     // catch free slots delivered to noteFreeMapSlice()
 600     assert(!waitingForPage);
 601     waitingForPage = &pageId;
 602     if (map->purgeOne()) {
 603         assert(!waitingForPage); // noteFreeMapSlice() should have cleared it
 604         assert(pageId.set());
 605         debugs(47, 5, "got a previously busy slot: " << pageId);
 606         return true;
 607     }
 608     assert(waitingForPage == &pageId);
 609     waitingForPage = NULL;
 610
 611     debugs(47, 3, "cannot get a slot; entries: " << map->entryCount());
 612     return false;
 613 }
 614
 615 bool
 616 Rock::SwapDir::validSlotId(const SlotId slotId) const
 617 {
 618     return 0 <= slotId && slotId < entryLimitAllowed();
 619 }
 620
 621 void
 622 Rock::SwapDir::noteFreeMapSlice(const sfileno sliceId)
 623 {
 624     Ipc::Mem::PageId pageId;
 625     pageId.pool = index+1;
 626     pageId.number = sliceId+1;
 627     if (waitingForPage) {
 628         *waitingForPage = pageId;
 629         waitingForPage = NULL;
 630     } else {
 631         freeSlots->push(pageId);
 632     }
 633 }
 634
 635 // tries to open an old entry with swap_filen for reading
 636 StoreIOState::Pointer
 637 Rock::SwapDir::openStoreIO(StoreEntry &e, StoreIOState::STFNCB *cbFile, StoreIOState::STIOCB *cbIo, void *data)
 638 {
 639     if (!theFile || theFile->error()) {
 640         debugs(47,4, HERE << theFile);
 641         return NULL;
 642     }
 643
 644     if (e.swap_filen < 0) {
 645         debugs(47,4, HERE << e);
 646         return NULL;
 647     }
 648
 649     // Do not start I/O transaction if there are less than 10% free pages left.
 650     // TODO: reserve page instead
 651     if (needsDiskStrand() &&
 652             Ipc::Mem::PageLevel(Ipc::Mem::PageId::ioPage) >= 0.9 * Ipc::Mem::PageLimit(Ipc::Mem::PageId::ioPage)) {
 653         debugs(47, 5, HERE << "too few shared pages for IPC I/O left");
 654         return NULL;
 655     }
 656
 657     // The are two ways an entry can get swap_filen: our get() locked it for
 658     // reading or our storeSwapOutStart() locked it for writing. Peeking at our
 659     // locked entry is safe, but no support for reading a filling entry.
 660     const Ipc::StoreMapAnchor *slot = map->peekAtReader(e.swap_filen);
 661     if (!slot)
 662         return NULL; // we were writing afterall
 663
 664     Rock::SwapDir::Pointer self(this);
 665     IoState *sio = new IoState(self, &e, cbFile, cbIo, data);
 666
 667     sio->swap_dirn = index;
 668     sio->swap_filen = e.swap_filen;
 669     sio->readableAnchor_ = slot;
 670     sio->file(theFile);
 671
 672     debugs(47,5, HERE << "dir " << index << " has old filen: " <<
 673            std::setfill('0') << std::hex << std::uppercase << std::setw(8) <<
 674            sio->swap_filen);
 675
 676     assert(slot->sameKey(static_cast<const cache_key*>(e.key)));
 677     assert(slot->basics.swap_file_sz > 0);
 678     assert(slot->basics.swap_file_sz == e.swap_file_sz);
 679
 680     return sio;
 681 }
 682
 683 void
 684 Rock::SwapDir::ioCompletedNotification()
 685 {
 686     if (!theFile)
 687         fatalf("Rock cache_dir failed to initialize db file: %s", filePath);
 688
 689     if (theFile->error())
 690         fatalf("Rock cache_dir at %s failed to open db file: %s", filePath,
 691                xstrerror());
 692
 693     debugs(47, 2, "Rock cache_dir[" << index << "] limits: " <<
 694            std::setw(12) << maxSize() << " disk bytes and " <<
 695            std::setw(7) << map->entryLimit() << " entries");
 696
 697     rebuild();
 698 }
 699
 700 void
 701 Rock::SwapDir::closeCompleted()
 702 {
 703     theFile = NULL;
 704 }
 705
 706 void
 707 Rock::SwapDir::readCompleted(const char *buf, int rlen, int errflag, RefCount< ::ReadRequest> r)
 708 {
 709     ReadRequest *request = dynamic_cast<Rock::ReadRequest*>(r.getRaw());
 710     assert(request);
 711     IoState::Pointer sio = request->sio;
 712
 713     if (errflag == DISK_OK && rlen > 0)
 714         sio->offset_ += rlen;
 715
 716     StoreIOState::STRCB *callback = sio->read.callback;
 717     assert(callback);
 718     sio->read.callback = NULL;
 719     void *cbdata;
 720     if (cbdataReferenceValidDone(sio->read.callback_data, &cbdata))
 721         callback(cbdata, r->buf, rlen, sio.getRaw());
 722 }
 723
 724 void
 725 Rock::SwapDir::writeCompleted(int errflag, size_t rlen, RefCount< ::WriteRequest> r)
 726 {
 727     Rock::WriteRequest *request = dynamic_cast<Rock::WriteRequest*>(r.getRaw());
 728     assert(request);
 729     assert(request->sio !=  NULL);
 730     IoState &sio = *request->sio;
 731
 732     // quit if somebody called IoState::close() while we were waiting
 733     if (!sio.stillWaiting()) {
 734         debugs(79, 3, "ignoring closed entry " << sio.swap_filen);
 735         return;
 736     }
 737
 738     if (errflag == DISK_OK) {
 739         // do not increment sio.offset_ because we do it in sio->write()
 740         if (request->isLast) {
 741             // close, the entry gets the read lock
 742             map->closeForWriting(sio.swap_filen, true);
 743             sio.finishedWriting(errflag);
 744         }
 745     } else {
 746         writeError(sio.swap_filen);
 747         sio.finishedWriting(errflag);
 748         // and hope that Core will call disconnect() to close the map entry
 749     }
 750 }
 751
 752 void
 753 Rock::SwapDir::writeError(const sfileno fileno)
 754 {
 755     // Do not abortWriting here. The entry should keep the write lock
 756     // instead of losing association with the store and confusing core.
 757     map->freeEntry(fileno); // will mark as unusable, just in case
 758     // All callers must also call IoState callback, to propagate the error.
 759 }
 760
 761 bool
 762 Rock::SwapDir::full() const
 763 {
 764     return freeSlots != NULL && !freeSlots->size();
 765 }
 766
 767 // storeSwapOutFileClosed calls this nethod on DISK_NO_SPACE_LEFT,
 768 // but it should not happen for us
 769 void
 770 Rock::SwapDir::diskFull()
 771 {
 772     debugs(20, DBG_IMPORTANT, "BUG: No space left with rock cache_dir: " <<
 773            filePath);
 774 }
 775
 776 /// purge while full(); it should be sufficient to purge just one
 777 void
 778 Rock::SwapDir::maintain()
 779 {
 780     // The Store calls this to free some db space, but there is nothing wrong
 781     // with a full() db, except when db has to shrink after reconfigure, and
 782     // we do not support shrinking yet (it would have to purge specific slots).
 783     // TODO: Disable maintain() requests when they are pointless.
 784 }
 785
 786 void
 787 Rock::SwapDir::reference(StoreEntry &e)
 788 {
 789     debugs(47, 5, HERE << &e << ' ' << e.swap_dirn << ' ' << e.swap_filen);
 790     if (repl && repl->Referenced)
 791         repl->Referenced(repl, &e, &e.repl);
 792 }
 793
 794 bool
 795 Rock::SwapDir::dereference(StoreEntry &e, bool)
 796 {
 797     debugs(47, 5, HERE << &e << ' ' << e.swap_dirn << ' ' << e.swap_filen);
 798     if (repl && repl->Dereferenced)
 799         repl->Dereferenced(repl, &e, &e.repl);
 800
 801     // no need to keep e in the global store_table for us; we have our own map
 802     return false;
 803 }
 804
 805 bool
 806 Rock::SwapDir::unlinkdUseful() const
 807 {
 808     // no entry-specific files to unlink
 809     return false;
 810 }
 811
 812 void
 813 Rock::SwapDir::unlink(StoreEntry &e)
 814 {
 815     debugs(47, 5, HERE << e);
 816     ignoreReferences(e);
 817     map->freeEntry(e.swap_filen);
 818     disconnect(e);
 819 }
 820
 821 void
 822 Rock::SwapDir::trackReferences(StoreEntry &e)
 823 {
 824     debugs(47, 5, HERE << e);
 825     if (repl)
 826         repl->Add(repl, &e, &e.repl);
 827 }
 828
 829 void
 830 Rock::SwapDir::ignoreReferences(StoreEntry &e)
 831 {
 832     debugs(47, 5, HERE << e);
 833     if (repl)
 834         repl->Remove(repl, &e, &e.repl);
 835 }
 836
 837 void
 838 Rock::SwapDir::statfs(StoreEntry &e) const
 839 {
 840     storeAppendPrintf(&e, "\n");
 841     storeAppendPrintf(&e, "Maximum Size: %" PRIu64 " KB\n", maxSize() >> 10);
 842     storeAppendPrintf(&e, "Current Size: %.2f KB %.2f%%\n",
 843                       currentSize() / 1024.0,
 844                       Math::doublePercent(currentSize(), maxSize()));
 845
 846     if (map) {
 847         const int limit = map->entryLimit();
 848         storeAppendPrintf(&e, "Maximum entries: %9d\n", limit);
 849         if (limit > 0) {
 850             const int entryCount = map->entryCount();
 851             storeAppendPrintf(&e, "Current entries: %9d %.2f%%\n",
 852                               entryCount, (100.0 * entryCount / limit));
 853
 854             const unsigned int slotsFree = !freeSlots ? 0 : freeSlots->size();
 855             if (slotsFree <= static_cast<const unsigned int>(limit)) {
 856                 const int usedSlots = limit - static_cast<const int>(slotsFree);
 857                 storeAppendPrintf(&e, "Used slots:      %9d %.2f%%\n",
 858                                   usedSlots, (100.0 * usedSlots / limit));
 859             }
 860             if (limit < 100) { // XXX: otherwise too expensive to count
 861                 Ipc::ReadWriteLockStats stats;
 862                 map->updateStats(stats);
 863                 stats.dump(e);
 864             }
 865         }
 866     }
 867
 868     storeAppendPrintf(&e, "Pending operations: %d out of %d\n",
 869                       store_open_disk_fd, Config.max_open_disk_fds);
 870
 871     storeAppendPrintf(&e, "Flags:");
 872
 873     if (flags.selected)
 874         storeAppendPrintf(&e, " SELECTED");
 875
 876     if (flags.read_only)
 877         storeAppendPrintf(&e, " READ-ONLY");
 878
 879     storeAppendPrintf(&e, "\n");
 880
 881 }
 882
 883 const char *
 884 Rock::SwapDir::inodeMapPath() const {
 885     static String inodesPath;
 886     inodesPath = path;
 887     inodesPath.append("_inodes");
 888     return inodesPath.termedBuf();
 889 }
 890
 891 const char *
 892 Rock::SwapDir::freeSlotsPath() const {
 893     static String spacesPath;
 894     spacesPath = path;
 895     spacesPath.append("_spaces");
 896     return spacesPath.termedBuf();
 897 }
 898
 899 namespace Rock
 900 {
 901 RunnerRegistrationEntry(rrAfterConfig, SwapDirRr);
 902 }
 903
 904 void Rock::SwapDirRr::create(const RunnerRegistry &)
 905 {
 906     Must(mapOwners.empty() && freeSlotsOwners.empty());
 907     for (int i = 0; i < Config.cacheSwap.n_configured; ++i) {
 908         if (const Rock::SwapDir *const sd = dynamic_cast<Rock::SwapDir *>(INDEXSD(i))) {
 909             const int64_t capacity = sd->entryLimitAllowed();
 910
 911             SwapDir::DirMap::Owner *const mapOwner =
 912                 SwapDir::DirMap::Init(sd->inodeMapPath(), capacity);
 913             mapOwners.push_back(mapOwner);
 914
 915             // XXX: remove pool id and counters from PageStack
 916             Ipc::Mem::Owner<Ipc::Mem::PageStack> *const freeSlotsOwner =
 917                 shm_new(Ipc::Mem::PageStack)(sd->freeSlotsPath(),
 918                                              i+1, capacity,
 919                                              sizeof(DbCellHeader));
 920             freeSlotsOwners.push_back(freeSlotsOwner);
 921
 922             // XXX: add method to initialize PageStack with no free pages
 923             while (true) {
 924                 Ipc::Mem::PageId pageId;
 925                 if (!freeSlotsOwner->object()->pop(pageId))
 926                     break;
 927             }
 928         }
 929     }
 930 }
 931
 932 Rock::SwapDirRr::~SwapDirRr()
 933 {
 934     for (size_t i = 0; i < mapOwners.size(); ++i) {
 935         delete mapOwners[i];
 936         delete freeSlotsOwners[i];
 937     }
 938 }