You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

GC.java 50KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248124912501251125212531254125512561257125812591260126112621263126412651266126712681269127012711272127312741275127612771278127912801281128212831284128512861287128812891290129112921293129412951296129712981299130013011302130313041305130613071308130913101311131213131314131513161317131813191320132113221323132413251326132713281329133013311332133313341335133613371338133913401341134213431344134513461347134813491350135113521353135413551356135713581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386138713881389139013911392139313941395139613971398139914001401140214031404140514061407140814091410141114121413141414151416141714181419142014211422142314241425142614271428142914301431143214331434143514361437143814391440144114421443144414451446144714481449145014511452145314541455145614571458145914601461146214631464146514661467146814691470147114721473147414751476147714781479148014811482148314841485148614871488148914901491149214931494149514961497149814991500150115021503150415051506150715081509151015111512151315141515151615171518151915201521152215231524152515261527152815291530153115321533153415351536153715381539154015411542154315441545154615471548154915501551155215531554155515561557155815591560156115621563156415651566156715681569157015711572157315741575157615771578157915801581158215831584158515861587158815891590159115921593159415951596159715981599160016011602160316041605160616071608
  1. /*
  2. * Copyright (C) 2012, Christian Halstrick <christian.halstrick@sap.com>
  3. * Copyright (C) 2011, Shawn O. Pearce <spearce@spearce.org>
  4. * and other copyright owners as documented in the project's IP log.
  5. *
  6. * This program and the accompanying materials are made available
  7. * under the terms of the Eclipse Distribution License v1.0 which
  8. * accompanies this distribution, is reproduced below, and is
  9. * available at http://www.eclipse.org/org/documents/edl-v10.php
  10. *
  11. * All rights reserved.
  12. *
  13. * Redistribution and use in source and binary forms, with or
  14. * without modification, are permitted provided that the following
  15. * conditions are met:
  16. *
  17. * - Redistributions of source code must retain the above copyright
  18. * notice, this list of conditions and the following disclaimer.
  19. *
  20. * - Redistributions in binary form must reproduce the above
  21. * copyright notice, this list of conditions and the following
  22. * disclaimer in the documentation and/or other materials provided
  23. * with the distribution.
  24. *
  25. * - Neither the name of the Eclipse Foundation, Inc. nor the
  26. * names of its contributors may be used to endorse or promote
  27. * products derived from this software without specific prior
  28. * written permission.
  29. *
  30. * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
  31. * CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
  32. * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
  33. * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  34. * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
  35. * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
  36. * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
  37. * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
  38. * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  39. * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
  40. * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
  41. * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
  42. * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  43. */
  44. package org.eclipse.jgit.internal.storage.file;
  45. import static org.eclipse.jgit.internal.storage.pack.PackExt.BITMAP_INDEX;
  46. import static org.eclipse.jgit.internal.storage.pack.PackExt.INDEX;
  47. import java.io.File;
  48. import java.io.FileOutputStream;
  49. import java.io.IOException;
  50. import java.io.OutputStream;
  51. import java.io.PrintWriter;
  52. import java.io.StringWriter;
  53. import java.nio.channels.Channels;
  54. import java.nio.channels.FileChannel;
  55. import java.nio.file.DirectoryNotEmptyException;
  56. import java.nio.file.DirectoryStream;
  57. import java.nio.file.Files;
  58. import java.nio.file.Path;
  59. import java.nio.file.StandardCopyOption;
  60. import java.text.MessageFormat;
  61. import java.text.ParseException;
  62. import java.time.Instant;
  63. import java.time.temporal.ChronoUnit;
  64. import java.util.ArrayList;
  65. import java.util.Collection;
  66. import java.util.Collections;
  67. import java.util.Comparator;
  68. import java.util.Date;
  69. import java.util.HashMap;
  70. import java.util.HashSet;
  71. import java.util.Iterator;
  72. import java.util.LinkedList;
  73. import java.util.List;
  74. import java.util.Map;
  75. import java.util.Objects;
  76. import java.util.Set;
  77. import java.util.TreeMap;
  78. import java.util.concurrent.Callable;
  79. import java.util.concurrent.ExecutorService;
  80. import java.util.regex.Pattern;
  81. import java.util.stream.Collectors;
  82. import java.util.stream.Stream;
  83. import org.eclipse.jgit.annotations.NonNull;
  84. import org.eclipse.jgit.dircache.DirCacheIterator;
  85. import org.eclipse.jgit.errors.CancelledException;
  86. import org.eclipse.jgit.errors.CorruptObjectException;
  87. import org.eclipse.jgit.errors.IncorrectObjectTypeException;
  88. import org.eclipse.jgit.errors.MissingObjectException;
  89. import org.eclipse.jgit.errors.NoWorkTreeException;
  90. import org.eclipse.jgit.internal.JGitText;
  91. import org.eclipse.jgit.internal.storage.pack.PackExt;
  92. import org.eclipse.jgit.internal.storage.pack.PackWriter;
  93. import org.eclipse.jgit.internal.storage.reftree.RefTreeNames;
  94. import org.eclipse.jgit.lib.ConfigConstants;
  95. import org.eclipse.jgit.lib.Constants;
  96. import org.eclipse.jgit.lib.FileMode;
  97. import org.eclipse.jgit.lib.NullProgressMonitor;
  98. import org.eclipse.jgit.lib.ObjectId;
  99. import org.eclipse.jgit.lib.ObjectIdSet;
  100. import org.eclipse.jgit.lib.ObjectLoader;
  101. import org.eclipse.jgit.lib.ObjectReader;
  102. import org.eclipse.jgit.lib.ProgressMonitor;
  103. import org.eclipse.jgit.lib.Ref;
  104. import org.eclipse.jgit.lib.Ref.Storage;
  105. import org.eclipse.jgit.lib.RefDatabase;
  106. import org.eclipse.jgit.lib.ReflogEntry;
  107. import org.eclipse.jgit.lib.ReflogReader;
  108. import org.eclipse.jgit.lib.internal.WorkQueue;
  109. import org.eclipse.jgit.revwalk.ObjectWalk;
  110. import org.eclipse.jgit.revwalk.RevObject;
  111. import org.eclipse.jgit.revwalk.RevWalk;
  112. import org.eclipse.jgit.storage.pack.PackConfig;
  113. import org.eclipse.jgit.treewalk.TreeWalk;
  114. import org.eclipse.jgit.treewalk.filter.TreeFilter;
  115. import org.eclipse.jgit.util.FileUtils;
  116. import org.eclipse.jgit.util.GitDateParser;
  117. import org.eclipse.jgit.util.SystemReader;
  118. import org.slf4j.Logger;
  119. import org.slf4j.LoggerFactory;
  120. /**
  121. * A garbage collector for git
  122. * {@link org.eclipse.jgit.internal.storage.file.FileRepository}. Instances of
  123. * this class are not thread-safe. Don't use the same instance from multiple
  124. * threads.
  125. *
  126. * This class started as a copy of DfsGarbageCollector from Shawn O. Pearce
  127. * adapted to FileRepositories.
  128. */
  129. public class GC {
  130. private final static Logger LOG = LoggerFactory
  131. .getLogger(GC.class);
  132. private static final String PRUNE_EXPIRE_DEFAULT = "2.weeks.ago"; //$NON-NLS-1$
  133. private static final String PRUNE_PACK_EXPIRE_DEFAULT = "1.hour.ago"; //$NON-NLS-1$
  134. private static final Pattern PATTERN_LOOSE_OBJECT = Pattern
  135. .compile("[0-9a-fA-F]{38}"); //$NON-NLS-1$
  136. private static final String PACK_EXT = "." + PackExt.PACK.getExtension();//$NON-NLS-1$
  137. private static final String BITMAP_EXT = "." //$NON-NLS-1$
  138. + PackExt.BITMAP_INDEX.getExtension();
  139. private static final String INDEX_EXT = "." + PackExt.INDEX.getExtension(); //$NON-NLS-1$
  140. private static final int DEFAULT_AUTOPACKLIMIT = 50;
  141. private static final int DEFAULT_AUTOLIMIT = 6700;
  142. private static volatile ExecutorService executor;
  143. /**
  144. * Set the executor for running auto-gc in the background. If no executor is
  145. * set JGit's own WorkQueue will be used.
  146. *
  147. * @param e
  148. * the executor to be used for running auto-gc
  149. */
  150. public static void setExecutor(ExecutorService e) {
  151. executor = e;
  152. }
  153. private final FileRepository repo;
  154. private ProgressMonitor pm;
  155. private long expireAgeMillis = -1;
  156. private Date expire;
  157. private long packExpireAgeMillis = -1;
  158. private Date packExpire;
  159. private PackConfig pconfig;
  160. /**
  161. * the refs which existed during the last call to {@link #repack()}. This is
  162. * needed during {@link #prune(Set)} where we can optimize by looking at the
  163. * difference between the current refs and the refs which existed during
  164. * last {@link #repack()}.
  165. */
  166. private Collection<Ref> lastPackedRefs;
  167. /**
  168. * Holds the starting time of the last repack() execution. This is needed in
  169. * prune() to inspect only those reflog entries which have been added since
  170. * last repack().
  171. */
  172. private long lastRepackTime;
  173. /**
  174. * Whether gc should do automatic housekeeping
  175. */
  176. private boolean automatic;
  177. /**
  178. * Whether to run gc in a background thread
  179. */
  180. private boolean background;
  181. /**
  182. * Creates a new garbage collector with default values. An expirationTime of
  183. * two weeks and <code>null</code> as progress monitor will be used.
  184. *
  185. * @param repo
  186. * the repo to work on
  187. */
  188. public GC(FileRepository repo) {
  189. this.repo = repo;
  190. this.pconfig = new PackConfig(repo);
  191. this.pm = NullProgressMonitor.INSTANCE;
  192. }
  193. /**
  194. * Runs a garbage collector on a
  195. * {@link org.eclipse.jgit.internal.storage.file.FileRepository}. It will
  196. * <ul>
  197. * <li>pack loose references into packed-refs</li>
  198. * <li>repack all reachable objects into new pack files and delete the old
  199. * pack files</li>
  200. * <li>prune all loose objects which are now reachable by packs</li>
  201. * </ul>
  202. *
  203. * If {@link #setAuto(boolean)} was set to {@code true} {@code gc} will
  204. * first check whether any housekeeping is required; if not, it exits
  205. * without performing any work.
  206. *
  207. * If {@link #setBackground(boolean)} was set to {@code true}
  208. * {@code collectGarbage} will start the gc in the background, and then
  209. * return immediately. In this case, errors will not be reported except in
  210. * gc.log.
  211. *
  212. * @return the collection of
  213. * {@link org.eclipse.jgit.internal.storage.file.PackFile}'s which
  214. * are newly created
  215. * @throws java.io.IOException
  216. * @throws java.text.ParseException
  217. * If the configuration parameter "gc.pruneexpire" couldn't be
  218. * parsed
  219. */
  220. // TODO(ms): change signature and return Future<Collection<PackFile>>
  221. @SuppressWarnings("FutureReturnValueIgnored")
  222. public Collection<PackFile> gc() throws IOException, ParseException {
  223. if (!background) {
  224. return doGc();
  225. }
  226. final GcLog gcLog = new GcLog(repo);
  227. if (!gcLog.lock()) {
  228. // there is already a background gc running
  229. return Collections.emptyList();
  230. }
  231. Callable<Collection<PackFile>> gcTask = () -> {
  232. try {
  233. Collection<PackFile> newPacks = doGc();
  234. if (automatic && tooManyLooseObjects()) {
  235. String message = JGitText.get().gcTooManyUnpruned;
  236. gcLog.write(message);
  237. gcLog.commit();
  238. }
  239. return newPacks;
  240. } catch (IOException | ParseException e) {
  241. try {
  242. gcLog.write(e.getMessage());
  243. StringWriter sw = new StringWriter();
  244. e.printStackTrace(new PrintWriter(sw));
  245. gcLog.write(sw.toString());
  246. gcLog.commit();
  247. } catch (IOException e2) {
  248. e2.addSuppressed(e);
  249. LOG.error(e2.getMessage(), e2);
  250. }
  251. } finally {
  252. gcLog.unlock();
  253. }
  254. return Collections.emptyList();
  255. };
  256. // TODO(ms): change signature and return the Future
  257. executor().submit(gcTask);
  258. return Collections.emptyList();
  259. }
  260. private ExecutorService executor() {
  261. return (executor != null) ? executor : WorkQueue.getExecutor();
  262. }
  263. private Collection<PackFile> doGc() throws IOException, ParseException {
  264. if (automatic && !needGc()) {
  265. return Collections.emptyList();
  266. }
  267. pm.start(6 /* tasks */);
  268. packRefs();
  269. // TODO: implement reflog_expire(pm, repo);
  270. Collection<PackFile> newPacks = repack();
  271. prune(Collections.emptySet());
  272. // TODO: implement rerere_gc(pm);
  273. return newPacks;
  274. }
  275. /**
  276. * Loosen objects in a pack file which are not also in the newly-created
  277. * pack files.
  278. *
  279. * @param inserter
  280. * @param reader
  281. * @param pack
  282. * @param existing
  283. * @throws IOException
  284. */
  285. private void loosen(ObjectDirectoryInserter inserter, ObjectReader reader, PackFile pack, HashSet<ObjectId> existing)
  286. throws IOException {
  287. for (PackIndex.MutableEntry entry : pack) {
  288. ObjectId oid = entry.toObjectId();
  289. if (existing.contains(oid)) {
  290. continue;
  291. }
  292. existing.add(oid);
  293. ObjectLoader loader = reader.open(oid);
  294. inserter.insert(loader.getType(),
  295. loader.getSize(),
  296. loader.openStream(),
  297. true /* create this object even though it's a duplicate */);
  298. }
  299. }
  300. /**
  301. * Delete old pack files. What is 'old' is defined by specifying a set of
  302. * old pack files and a set of new pack files. Each pack file contained in
  303. * old pack files but not contained in new pack files will be deleted. If
  304. * preserveOldPacks is set, keep a copy of the pack file in the preserve
  305. * directory. If an expirationDate is set then pack files which are younger
  306. * than the expirationDate will not be deleted nor preserved.
  307. * <p>
  308. * If we're not immediately expiring loose objects, loosen any objects
  309. * in the old pack files which aren't in the new pack files.
  310. *
  311. * @param oldPacks
  312. * @param newPacks
  313. * @throws ParseException
  314. * @throws IOException
  315. */
  316. private void deleteOldPacks(Collection<PackFile> oldPacks,
  317. Collection<PackFile> newPacks) throws ParseException, IOException {
  318. HashSet<ObjectId> ids = new HashSet<>();
  319. for (PackFile pack : newPacks) {
  320. for (PackIndex.MutableEntry entry : pack) {
  321. ids.add(entry.toObjectId());
  322. }
  323. }
  324. ObjectReader reader = repo.newObjectReader();
  325. ObjectDirectory dir = repo.getObjectDatabase();
  326. ObjectDirectoryInserter inserter = dir.newInserter();
  327. boolean shouldLoosen = !"now".equals(getPruneExpireStr()) && //$NON-NLS-1$
  328. getExpireDate() < Long.MAX_VALUE;
  329. prunePreserved();
  330. long packExpireDate = getPackExpireDate();
  331. oldPackLoop: for (PackFile oldPack : oldPacks) {
  332. checkCancelled();
  333. String oldName = oldPack.getPackName();
  334. // check whether an old pack file is also among the list of new
  335. // pack files. Then we must not delete it.
  336. for (PackFile newPack : newPacks)
  337. if (oldName.equals(newPack.getPackName()))
  338. continue oldPackLoop;
  339. if (!oldPack.shouldBeKept()
  340. && repo.getFS()
  341. .lastModifiedInstant(oldPack.getPackFile())
  342. .toEpochMilli() < packExpireDate) {
  343. oldPack.close();
  344. if (shouldLoosen) {
  345. loosen(inserter, reader, oldPack, ids);
  346. }
  347. prunePack(oldName);
  348. }
  349. }
  350. // close the complete object database. That's my only chance to force
  351. // rescanning and to detect that certain pack files are now deleted.
  352. repo.getObjectDatabase().close();
  353. }
  354. /**
  355. * Deletes old pack file, unless 'preserve-oldpacks' is set, in which case it
  356. * moves the pack file to the preserved directory
  357. *
  358. * @param packFile
  359. * @param packName
  360. * @param ext
  361. * @param deleteOptions
  362. * @throws IOException
  363. */
  364. private void removeOldPack(File packFile, String packName, PackExt ext,
  365. int deleteOptions) throws IOException {
  366. if (pconfig.isPreserveOldPacks()) {
  367. File oldPackDir = repo.getObjectDatabase().getPreservedDirectory();
  368. FileUtils.mkdir(oldPackDir, true);
  369. String oldPackName = "pack-" + packName + ".old-" + ext.getExtension(); //$NON-NLS-1$ //$NON-NLS-2$
  370. File oldPackFile = new File(oldPackDir, oldPackName);
  371. FileUtils.rename(packFile, oldPackFile);
  372. } else {
  373. FileUtils.delete(packFile, deleteOptions);
  374. }
  375. }
  376. /**
  377. * Delete the preserved directory including all pack files within
  378. */
  379. private void prunePreserved() {
  380. if (pconfig.isPrunePreserved()) {
  381. try {
  382. FileUtils.delete(repo.getObjectDatabase().getPreservedDirectory(),
  383. FileUtils.RECURSIVE | FileUtils.RETRY | FileUtils.SKIP_MISSING);
  384. } catch (IOException e) {
  385. // Deletion of the preserved pack files failed. Silently return.
  386. }
  387. }
  388. }
  389. /**
  390. * Delete files associated with a single pack file. First try to delete the
  391. * ".pack" file because on some platforms the ".pack" file may be locked and
  392. * can't be deleted. In such a case it is better to detect this early and
  393. * give up on deleting files for this packfile. Otherwise we may delete the
  394. * ".index" file and when failing to delete the ".pack" file we are left
  395. * with a ".pack" file without a ".index" file.
  396. *
  397. * @param packName
  398. */
  399. private void prunePack(String packName) {
  400. PackExt[] extensions = PackExt.values();
  401. try {
  402. // Delete the .pack file first and if this fails give up on deleting
  403. // the other files
  404. int deleteOptions = FileUtils.RETRY | FileUtils.SKIP_MISSING;
  405. for (PackExt ext : extensions)
  406. if (PackExt.PACK.equals(ext)) {
  407. File f = nameFor(packName, "." + ext.getExtension()); //$NON-NLS-1$
  408. removeOldPack(f, packName, ext, deleteOptions);
  409. break;
  410. }
  411. // The .pack file has been deleted. Delete as many as the other
  412. // files as you can.
  413. deleteOptions |= FileUtils.IGNORE_ERRORS;
  414. for (PackExt ext : extensions) {
  415. if (!PackExt.PACK.equals(ext)) {
  416. File f = nameFor(packName, "." + ext.getExtension()); //$NON-NLS-1$
  417. removeOldPack(f, packName, ext, deleteOptions);
  418. }
  419. }
  420. } catch (IOException e) {
  421. // Deletion of the .pack file failed. Silently return.
  422. }
  423. }
  424. /**
  425. * Like "git prune-packed" this method tries to prune all loose objects
  426. * which can be found in packs. If certain objects can't be pruned (e.g.
  427. * because the filesystem delete operation fails) this is silently ignored.
  428. *
  429. * @throws java.io.IOException
  430. */
  431. public void prunePacked() throws IOException {
  432. ObjectDirectory objdb = repo.getObjectDatabase();
  433. Collection<PackFile> packs = objdb.getPacks();
  434. File objects = repo.getObjectsDirectory();
  435. String[] fanout = objects.list();
  436. if (fanout != null && fanout.length > 0) {
  437. pm.beginTask(JGitText.get().pruneLoosePackedObjects, fanout.length);
  438. try {
  439. for (String d : fanout) {
  440. checkCancelled();
  441. pm.update(1);
  442. if (d.length() != 2)
  443. continue;
  444. String[] entries = new File(objects, d).list();
  445. if (entries == null)
  446. continue;
  447. for (String e : entries) {
  448. checkCancelled();
  449. if (e.length() != Constants.OBJECT_ID_STRING_LENGTH - 2)
  450. continue;
  451. ObjectId id;
  452. try {
  453. id = ObjectId.fromString(d + e);
  454. } catch (IllegalArgumentException notAnObject) {
  455. // ignoring the file that does not represent loose
  456. // object
  457. continue;
  458. }
  459. boolean found = false;
  460. for (PackFile p : packs) {
  461. checkCancelled();
  462. if (p.hasObject(id)) {
  463. found = true;
  464. break;
  465. }
  466. }
  467. if (found)
  468. FileUtils.delete(objdb.fileFor(id), FileUtils.RETRY
  469. | FileUtils.SKIP_MISSING
  470. | FileUtils.IGNORE_ERRORS);
  471. }
  472. }
  473. } finally {
  474. pm.endTask();
  475. }
  476. }
  477. }
  478. /**
  479. * Like "git prune" this method tries to prune all loose objects which are
  480. * unreferenced. If certain objects can't be pruned (e.g. because the
  481. * filesystem delete operation fails) this is silently ignored.
  482. *
  483. * @param objectsToKeep
  484. * a set of objects which should explicitly not be pruned
  485. * @throws java.io.IOException
  486. * @throws java.text.ParseException
  487. * If the configuration parameter "gc.pruneexpire" couldn't be
  488. * parsed
  489. */
  490. public void prune(Set<ObjectId> objectsToKeep) throws IOException,
  491. ParseException {
  492. long expireDate = getExpireDate();
  493. // Collect all loose objects which are old enough, not referenced from
  494. // the index and not in objectsToKeep
  495. Map<ObjectId, File> deletionCandidates = new HashMap<>();
  496. Set<ObjectId> indexObjects = null;
  497. File objects = repo.getObjectsDirectory();
  498. String[] fanout = objects.list();
  499. if (fanout == null || fanout.length == 0) {
  500. return;
  501. }
  502. pm.beginTask(JGitText.get().pruneLooseUnreferencedObjects,
  503. fanout.length);
  504. try {
  505. for (String d : fanout) {
  506. checkCancelled();
  507. pm.update(1);
  508. if (d.length() != 2)
  509. continue;
  510. File dir = new File(objects, d);
  511. File[] entries = dir.listFiles();
  512. if (entries == null || entries.length == 0) {
  513. FileUtils.delete(dir, FileUtils.IGNORE_ERRORS);
  514. continue;
  515. }
  516. for (File f : entries) {
  517. checkCancelled();
  518. String fName = f.getName();
  519. if (fName.length() != Constants.OBJECT_ID_STRING_LENGTH - 2)
  520. continue;
  521. if (repo.getFS().lastModifiedInstant(f)
  522. .toEpochMilli() >= expireDate) {
  523. continue;
  524. }
  525. try {
  526. ObjectId id = ObjectId.fromString(d + fName);
  527. if (objectsToKeep.contains(id))
  528. continue;
  529. if (indexObjects == null)
  530. indexObjects = listNonHEADIndexObjects();
  531. if (indexObjects.contains(id))
  532. continue;
  533. deletionCandidates.put(id, f);
  534. } catch (IllegalArgumentException notAnObject) {
  535. // ignoring the file that does not represent loose
  536. // object
  537. }
  538. }
  539. }
  540. } finally {
  541. pm.endTask();
  542. }
  543. if (deletionCandidates.isEmpty()) {
  544. return;
  545. }
  546. checkCancelled();
  547. // From the set of current refs remove all those which have been handled
  548. // during last repack(). Only those refs will survive which have been
  549. // added or modified since the last repack. Only these can save existing
  550. // loose refs from being pruned.
  551. Collection<Ref> newRefs;
  552. if (lastPackedRefs == null || lastPackedRefs.isEmpty())
  553. newRefs = getAllRefs();
  554. else {
  555. Map<String, Ref> last = new HashMap<>();
  556. for (Ref r : lastPackedRefs) {
  557. last.put(r.getName(), r);
  558. }
  559. newRefs = new ArrayList<>();
  560. for (Ref r : getAllRefs()) {
  561. Ref old = last.get(r.getName());
  562. if (!equals(r, old)) {
  563. newRefs.add(r);
  564. }
  565. }
  566. }
  567. if (!newRefs.isEmpty()) {
  568. // There are new/modified refs! Check which loose objects are now
  569. // referenced by these modified refs (or their reflogentries).
  570. // Remove these loose objects
  571. // from the deletionCandidates. When the last candidate is removed
  572. // leave this method.
  573. ObjectWalk w = new ObjectWalk(repo);
  574. try {
  575. for (Ref cr : newRefs) {
  576. checkCancelled();
  577. w.markStart(w.parseAny(cr.getObjectId()));
  578. }
  579. if (lastPackedRefs != null)
  580. for (Ref lpr : lastPackedRefs) {
  581. w.markUninteresting(w.parseAny(lpr.getObjectId()));
  582. }
  583. removeReferenced(deletionCandidates, w);
  584. } finally {
  585. w.dispose();
  586. }
  587. }
  588. if (deletionCandidates.isEmpty())
  589. return;
  590. // Since we have not left the method yet there are still
  591. // deletionCandidates. Last chance for these objects not to be pruned is
  592. // that they are referenced by reflog entries. Even refs which currently
  593. // point to the same object as during last repack() may have
  594. // additional reflog entries not handled during last repack()
  595. ObjectWalk w = new ObjectWalk(repo);
  596. try {
  597. for (Ref ar : getAllRefs())
  598. for (ObjectId id : listRefLogObjects(ar, lastRepackTime)) {
  599. checkCancelled();
  600. w.markStart(w.parseAny(id));
  601. }
  602. if (lastPackedRefs != null)
  603. for (Ref lpr : lastPackedRefs) {
  604. checkCancelled();
  605. w.markUninteresting(w.parseAny(lpr.getObjectId()));
  606. }
  607. removeReferenced(deletionCandidates, w);
  608. } finally {
  609. w.dispose();
  610. }
  611. if (deletionCandidates.isEmpty())
  612. return;
  613. checkCancelled();
  614. // delete all candidates which have survived: these are unreferenced
  615. // loose objects. Make a last check, though, to avoid deleting objects
  616. // that could have been referenced while the candidates list was being
  617. // built (by an incoming push, for example).
  618. Set<File> touchedFanout = new HashSet<>();
  619. for (File f : deletionCandidates.values()) {
  620. if (f.lastModified() < expireDate) {
  621. f.delete();
  622. touchedFanout.add(f.getParentFile());
  623. }
  624. }
  625. for (File f : touchedFanout) {
  626. FileUtils.delete(f,
  627. FileUtils.EMPTY_DIRECTORIES_ONLY | FileUtils.IGNORE_ERRORS);
  628. }
  629. repo.getObjectDatabase().close();
  630. }
  631. private long getExpireDate() throws ParseException {
  632. long expireDate = Long.MAX_VALUE;
  633. if (expire == null && expireAgeMillis == -1) {
  634. String pruneExpireStr = getPruneExpireStr();
  635. if (pruneExpireStr == null)
  636. pruneExpireStr = PRUNE_EXPIRE_DEFAULT;
  637. expire = GitDateParser.parse(pruneExpireStr, null, SystemReader
  638. .getInstance().getLocale());
  639. expireAgeMillis = -1;
  640. }
  641. if (expire != null)
  642. expireDate = expire.getTime();
  643. if (expireAgeMillis != -1)
  644. expireDate = System.currentTimeMillis() - expireAgeMillis;
  645. return expireDate;
  646. }
  647. private String getPruneExpireStr() {
  648. return repo.getConfig().getString(
  649. ConfigConstants.CONFIG_GC_SECTION, null,
  650. ConfigConstants.CONFIG_KEY_PRUNEEXPIRE);
  651. }
  652. private long getPackExpireDate() throws ParseException {
  653. long packExpireDate = Long.MAX_VALUE;
  654. if (packExpire == null && packExpireAgeMillis == -1) {
  655. String prunePackExpireStr = repo.getConfig().getString(
  656. ConfigConstants.CONFIG_GC_SECTION, null,
  657. ConfigConstants.CONFIG_KEY_PRUNEPACKEXPIRE);
  658. if (prunePackExpireStr == null)
  659. prunePackExpireStr = PRUNE_PACK_EXPIRE_DEFAULT;
  660. packExpire = GitDateParser.parse(prunePackExpireStr, null,
  661. SystemReader.getInstance().getLocale());
  662. packExpireAgeMillis = -1;
  663. }
  664. if (packExpire != null)
  665. packExpireDate = packExpire.getTime();
  666. if (packExpireAgeMillis != -1)
  667. packExpireDate = System.currentTimeMillis() - packExpireAgeMillis;
  668. return packExpireDate;
  669. }
  670. /**
  671. * Remove all entries from a map which key is the id of an object referenced
  672. * by the given ObjectWalk
  673. *
  674. * @param id2File
  675. * @param w
  676. * @throws MissingObjectException
  677. * @throws IncorrectObjectTypeException
  678. * @throws IOException
  679. */
  680. private void removeReferenced(Map<ObjectId, File> id2File,
  681. ObjectWalk w) throws MissingObjectException,
  682. IncorrectObjectTypeException, IOException {
  683. RevObject ro = w.next();
  684. while (ro != null) {
  685. checkCancelled();
  686. if (id2File.remove(ro.getId()) != null && id2File.isEmpty()) {
  687. return;
  688. }
  689. ro = w.next();
  690. }
  691. ro = w.nextObject();
  692. while (ro != null) {
  693. checkCancelled();
  694. if (id2File.remove(ro.getId()) != null && id2File.isEmpty()) {
  695. return;
  696. }
  697. ro = w.nextObject();
  698. }
  699. }
  700. private static boolean equals(Ref r1, Ref r2) {
  701. if (r1 == null || r2 == null) {
  702. return false;
  703. }
  704. if (r1.isSymbolic()) {
  705. return r2.isSymbolic() && r1.getTarget().getName()
  706. .equals(r2.getTarget().getName());
  707. }
  708. return !r2.isSymbolic()
  709. && Objects.equals(r1.getObjectId(), r2.getObjectId());
  710. }
  711. /**
  712. * Packs all non-symbolic, loose refs into packed-refs.
  713. *
  714. * @throws java.io.IOException
  715. */
  716. public void packRefs() throws IOException {
  717. Collection<Ref> refs = repo.getRefDatabase()
  718. .getRefsByPrefix(Constants.R_REFS);
  719. List<String> refsToBePacked = new ArrayList<>(refs.size());
  720. pm.beginTask(JGitText.get().packRefs, refs.size());
  721. try {
  722. for (Ref ref : refs) {
  723. checkCancelled();
  724. if (!ref.isSymbolic() && ref.getStorage().isLoose())
  725. refsToBePacked.add(ref.getName());
  726. pm.update(1);
  727. }
  728. ((RefDirectory) repo.getRefDatabase()).pack(refsToBePacked);
  729. } finally {
  730. pm.endTask();
  731. }
  732. }
  733. /**
  734. * Packs all objects which reachable from any of the heads into one pack
  735. * file. Additionally all objects which are not reachable from any head but
  736. * which are reachable from any of the other refs (e.g. tags), special refs
  737. * (e.g. FETCH_HEAD) or index are packed into a separate pack file. Objects
  738. * included in pack files which have a .keep file associated are never
  739. * repacked. All old pack files which existed before are deleted.
  740. *
  741. * @return a collection of the newly created pack files
  742. * @throws java.io.IOException
  743. * when during reading of refs, index, packfiles, objects,
  744. * reflog-entries or during writing to the packfiles
  745. * {@link java.io.IOException} occurs
  746. */
  747. public Collection<PackFile> repack() throws IOException {
  748. Collection<PackFile> toBeDeleted = repo.getObjectDatabase().getPacks();
  749. long time = System.currentTimeMillis();
  750. Collection<Ref> refsBefore = getAllRefs();
  751. Set<ObjectId> allHeadsAndTags = new HashSet<>();
  752. Set<ObjectId> allHeads = new HashSet<>();
  753. Set<ObjectId> allTags = new HashSet<>();
  754. Set<ObjectId> nonHeads = new HashSet<>();
  755. Set<ObjectId> txnHeads = new HashSet<>();
  756. Set<ObjectId> tagTargets = new HashSet<>();
  757. Set<ObjectId> indexObjects = listNonHEADIndexObjects();
  758. RefDatabase refdb = repo.getRefDatabase();
  759. for (Ref ref : refsBefore) {
  760. checkCancelled();
  761. nonHeads.addAll(listRefLogObjects(ref, 0));
  762. if (ref.isSymbolic() || ref.getObjectId() == null) {
  763. continue;
  764. }
  765. if (isHead(ref)) {
  766. allHeads.add(ref.getObjectId());
  767. } else if (isTag(ref)) {
  768. allTags.add(ref.getObjectId());
  769. } else if (RefTreeNames.isRefTree(refdb, ref.getName())) {
  770. txnHeads.add(ref.getObjectId());
  771. } else {
  772. nonHeads.add(ref.getObjectId());
  773. }
  774. if (ref.getPeeledObjectId() != null) {
  775. tagTargets.add(ref.getPeeledObjectId());
  776. }
  777. }
  778. List<ObjectIdSet> excluded = new LinkedList<>();
  779. for (PackFile f : repo.getObjectDatabase().getPacks()) {
  780. checkCancelled();
  781. if (f.shouldBeKept())
  782. excluded.add(f.getIndex());
  783. }
  784. // Don't exclude tags that are also branch tips
  785. allTags.removeAll(allHeads);
  786. allHeadsAndTags.addAll(allHeads);
  787. allHeadsAndTags.addAll(allTags);
  788. // Hoist all branch tips and tags earlier in the pack file
  789. tagTargets.addAll(allHeadsAndTags);
  790. nonHeads.addAll(indexObjects);
  791. // Combine the GC_REST objects into the GC pack if requested
  792. if (pconfig.getSinglePack()) {
  793. allHeadsAndTags.addAll(nonHeads);
  794. nonHeads.clear();
  795. }
  796. List<PackFile> ret = new ArrayList<>(2);
  797. PackFile heads = null;
  798. if (!allHeadsAndTags.isEmpty()) {
  799. heads = writePack(allHeadsAndTags, PackWriter.NONE, allTags,
  800. tagTargets, excluded);
  801. if (heads != null) {
  802. ret.add(heads);
  803. excluded.add(0, heads.getIndex());
  804. }
  805. }
  806. if (!nonHeads.isEmpty()) {
  807. PackFile rest = writePack(nonHeads, allHeadsAndTags, PackWriter.NONE,
  808. tagTargets, excluded);
  809. if (rest != null)
  810. ret.add(rest);
  811. }
  812. if (!txnHeads.isEmpty()) {
  813. PackFile txn = writePack(txnHeads, PackWriter.NONE, PackWriter.NONE,
  814. null, excluded);
  815. if (txn != null)
  816. ret.add(txn);
  817. }
  818. try {
  819. deleteOldPacks(toBeDeleted, ret);
  820. } catch (ParseException e) {
  821. // TODO: the exception has to be wrapped into an IOException because
  822. // throwing the ParseException directly would break the API, instead
  823. // we should throw a ConfigInvalidException
  824. throw new IOException(e);
  825. }
  826. prunePacked();
  827. deleteEmptyRefsFolders();
  828. deleteOrphans();
  829. deleteTempPacksIdx();
  830. lastPackedRefs = refsBefore;
  831. lastRepackTime = time;
  832. return ret;
  833. }
  834. private static boolean isHead(Ref ref) {
  835. return ref.getName().startsWith(Constants.R_HEADS);
  836. }
  837. private static boolean isTag(Ref ref) {
  838. return ref.getName().startsWith(Constants.R_TAGS);
  839. }
  840. private void deleteEmptyRefsFolders() throws IOException {
  841. Path refs = repo.getDirectory().toPath().resolve(Constants.R_REFS);
  842. // Avoid deleting a folder that was created after the threshold so that concurrent
  843. // operations trying to create a reference are not impacted
  844. Instant threshold = Instant.now().minus(30, ChronoUnit.SECONDS);
  845. try (Stream<Path> entries = Files.list(refs)
  846. .filter(Files::isDirectory)) {
  847. Iterator<Path> iterator = entries.iterator();
  848. while (iterator.hasNext()) {
  849. try (Stream<Path> s = Files.list(iterator.next())) {
  850. s.filter(path -> canBeSafelyDeleted(path, threshold)).forEach(this::deleteDir);
  851. }
  852. }
  853. }
  854. }
  855. private boolean canBeSafelyDeleted(Path path, Instant threshold) {
  856. try {
  857. return Files.getLastModifiedTime(path).toInstant().isBefore(threshold);
  858. }
  859. catch (IOException e) {
  860. LOG.warn(MessageFormat.format(
  861. JGitText.get().cannotAccessLastModifiedForSafeDeletion,
  862. path), e);
  863. return false;
  864. }
  865. }
  866. private void deleteDir(Path dir) {
  867. try (Stream<Path> dirs = Files.walk(dir)) {
  868. dirs.filter(this::isDirectory).sorted(Comparator.reverseOrder())
  869. .forEach(this::delete);
  870. } catch (IOException e) {
  871. LOG.error(e.getMessage(), e);
  872. }
  873. }
  874. private boolean isDirectory(Path p) {
  875. return p.toFile().isDirectory();
  876. }
  877. private void delete(Path d) {
  878. try {
  879. Files.delete(d);
  880. } catch (DirectoryNotEmptyException e) {
  881. // Don't log
  882. } catch (IOException e) {
  883. LOG.error(MessageFormat.format(JGitText.get().cannotDeleteFile, d),
  884. e);
  885. }
  886. }
  887. /**
  888. * Deletes orphans
  889. * <p>
  890. * A file is considered an orphan if it is either a "bitmap" or an index
  891. * file, and its corresponding pack file is missing in the list.
  892. * </p>
  893. */
  894. private void deleteOrphans() {
  895. Path packDir = repo.getObjectDatabase().getPackDirectory().toPath();
  896. List<String> fileNames = null;
  897. try (Stream<Path> files = Files.list(packDir)) {
  898. fileNames = files.map(path -> path.getFileName().toString())
  899. .filter(name -> (name.endsWith(PACK_EXT)
  900. || name.endsWith(BITMAP_EXT)
  901. || name.endsWith(INDEX_EXT)))
  902. .sorted(Collections.reverseOrder())
  903. .collect(Collectors.toList());
  904. } catch (IOException e1) {
  905. // ignore
  906. }
  907. if (fileNames == null) {
  908. return;
  909. }
  910. String base = null;
  911. for (String n : fileNames) {
  912. if (n.endsWith(PACK_EXT)) {
  913. base = n.substring(0, n.lastIndexOf('.'));
  914. } else {
  915. if (base == null || !n.startsWith(base)) {
  916. try {
  917. Files.delete(packDir.resolve(n));
  918. } catch (IOException e) {
  919. LOG.error(e.getMessage(), e);
  920. }
  921. }
  922. }
  923. }
  924. }
  925. private void deleteTempPacksIdx() {
  926. Path packDir = repo.getObjectDatabase().getPackDirectory().toPath();
  927. Instant threshold = Instant.now().minus(1, ChronoUnit.DAYS);
  928. if (!Files.exists(packDir)) {
  929. return;
  930. }
  931. try (DirectoryStream<Path> stream =
  932. Files.newDirectoryStream(packDir, "gc_*_tmp")) { //$NON-NLS-1$
  933. stream.forEach(t -> {
  934. try {
  935. Instant lastModified = Files.getLastModifiedTime(t)
  936. .toInstant();
  937. if (lastModified.isBefore(threshold)) {
  938. Files.deleteIfExists(t);
  939. }
  940. } catch (IOException e) {
  941. LOG.error(e.getMessage(), e);
  942. }
  943. });
  944. } catch (IOException e) {
  945. LOG.error(e.getMessage(), e);
  946. }
  947. }
  948. /**
  949. * @param ref
  950. * the ref which log should be inspected
  951. * @param minTime only reflog entries not older then this time are processed
  952. * @return the {@link ObjectId}s contained in the reflog
  953. * @throws IOException
  954. */
  955. private Set<ObjectId> listRefLogObjects(Ref ref, long minTime) throws IOException {
  956. ReflogReader reflogReader = repo.getReflogReader(ref.getName());
  957. if (reflogReader == null) {
  958. return Collections.emptySet();
  959. }
  960. List<ReflogEntry> rlEntries = reflogReader
  961. .getReverseEntries();
  962. if (rlEntries == null || rlEntries.isEmpty())
  963. return Collections.emptySet();
  964. Set<ObjectId> ret = new HashSet<>();
  965. for (ReflogEntry e : rlEntries) {
  966. if (e.getWho().getWhen().getTime() < minTime)
  967. break;
  968. ObjectId newId = e.getNewId();
  969. if (newId != null && !ObjectId.zeroId().equals(newId))
  970. ret.add(newId);
  971. ObjectId oldId = e.getOldId();
  972. if (oldId != null && !ObjectId.zeroId().equals(oldId))
  973. ret.add(oldId);
  974. }
  975. return ret;
  976. }
  977. /**
  978. * Returns a collection of all refs and additional refs.
  979. *
  980. * Additional refs which don't start with "refs/" are not returned because
  981. * they should not save objects from being garbage collected. Examples for
  982. * such references are ORIG_HEAD, MERGE_HEAD, FETCH_HEAD and
  983. * CHERRY_PICK_HEAD.
  984. *
  985. * @return a collection of refs pointing to live objects.
  986. * @throws IOException
  987. */
  988. private Collection<Ref> getAllRefs() throws IOException {
  989. RefDatabase refdb = repo.getRefDatabase();
  990. Collection<Ref> refs = refdb.getRefs();
  991. List<Ref> addl = refdb.getAdditionalRefs();
  992. if (!addl.isEmpty()) {
  993. List<Ref> all = new ArrayList<>(refs.size() + addl.size());
  994. all.addAll(refs);
  995. // add additional refs which start with refs/
  996. for (Ref r : addl) {
  997. checkCancelled();
  998. if (r.getName().startsWith(Constants.R_REFS)) {
  999. all.add(r);
  1000. }
  1001. }
  1002. return all;
  1003. }
  1004. return refs;
  1005. }
  1006. /**
  1007. * Return a list of those objects in the index which differ from whats in
  1008. * HEAD
  1009. *
  1010. * @return a set of ObjectIds of changed objects in the index
  1011. * @throws IOException
  1012. * @throws CorruptObjectException
  1013. * @throws NoWorkTreeException
  1014. */
  1015. private Set<ObjectId> listNonHEADIndexObjects()
  1016. throws CorruptObjectException, IOException {
  1017. if (repo.isBare()) {
  1018. return Collections.emptySet();
  1019. }
  1020. try (TreeWalk treeWalk = new TreeWalk(repo)) {
  1021. treeWalk.addTree(new DirCacheIterator(repo.readDirCache()));
  1022. ObjectId headID = repo.resolve(Constants.HEAD);
  1023. if (headID != null) {
  1024. try (RevWalk revWalk = new RevWalk(repo)) {
  1025. treeWalk.addTree(revWalk.parseTree(headID));
  1026. }
  1027. }
  1028. treeWalk.setFilter(TreeFilter.ANY_DIFF);
  1029. treeWalk.setRecursive(true);
  1030. Set<ObjectId> ret = new HashSet<>();
  1031. while (treeWalk.next()) {
  1032. checkCancelled();
  1033. ObjectId objectId = treeWalk.getObjectId(0);
  1034. switch (treeWalk.getRawMode(0) & FileMode.TYPE_MASK) {
  1035. case FileMode.TYPE_MISSING:
  1036. case FileMode.TYPE_GITLINK:
  1037. continue;
  1038. case FileMode.TYPE_TREE:
  1039. case FileMode.TYPE_FILE:
  1040. case FileMode.TYPE_SYMLINK:
  1041. ret.add(objectId);
  1042. continue;
  1043. default:
  1044. throw new IOException(MessageFormat.format(
  1045. JGitText.get().corruptObjectInvalidMode3,
  1046. String.format("%o", //$NON-NLS-1$
  1047. Integer.valueOf(treeWalk.getRawMode(0))),
  1048. (objectId == null) ? "null" : objectId.name(), //$NON-NLS-1$
  1049. treeWalk.getPathString(), //
  1050. repo.getIndexFile()));
  1051. }
  1052. }
  1053. return ret;
  1054. }
  1055. }
  1056. private PackFile writePack(@NonNull Set<? extends ObjectId> want,
  1057. @NonNull Set<? extends ObjectId> have, @NonNull Set<ObjectId> tags,
  1058. Set<ObjectId> tagTargets, List<ObjectIdSet> excludeObjects)
  1059. throws IOException {
  1060. checkCancelled();
  1061. File tmpPack = null;
  1062. Map<PackExt, File> tmpExts = new TreeMap<>((o1, o2) -> {
  1063. // INDEX entries must be returned last, so the pack
  1064. // scanner does pick up the new pack until all the
  1065. // PackExt entries have been written.
  1066. if (o1 == o2) {
  1067. return 0;
  1068. }
  1069. if (o1 == PackExt.INDEX) {
  1070. return 1;
  1071. }
  1072. if (o2 == PackExt.INDEX) {
  1073. return -1;
  1074. }
  1075. return Integer.signum(o1.hashCode() - o2.hashCode());
  1076. });
  1077. try (PackWriter pw = new PackWriter(
  1078. pconfig,
  1079. repo.newObjectReader())) {
  1080. // prepare the PackWriter
  1081. pw.setDeltaBaseAsOffset(true);
  1082. pw.setReuseDeltaCommits(false);
  1083. if (tagTargets != null) {
  1084. pw.setTagTargets(tagTargets);
  1085. }
  1086. if (excludeObjects != null)
  1087. for (ObjectIdSet idx : excludeObjects)
  1088. pw.excludeObjects(idx);
  1089. pw.preparePack(pm, want, have, PackWriter.NONE, tags);
  1090. if (pw.getObjectCount() == 0)
  1091. return null;
  1092. checkCancelled();
  1093. // create temporary files
  1094. String id = pw.computeName().getName();
  1095. File packdir = repo.getObjectDatabase().getPackDirectory();
  1096. tmpPack = File.createTempFile("gc_", ".pack_tmp", packdir); //$NON-NLS-1$ //$NON-NLS-2$
  1097. final String tmpBase = tmpPack.getName()
  1098. .substring(0, tmpPack.getName().lastIndexOf('.'));
  1099. File tmpIdx = new File(packdir, tmpBase + ".idx_tmp"); //$NON-NLS-1$
  1100. tmpExts.put(INDEX, tmpIdx);
  1101. if (!tmpIdx.createNewFile())
  1102. throw new IOException(MessageFormat.format(
  1103. JGitText.get().cannotCreateIndexfile, tmpIdx.getPath()));
  1104. // write the packfile
  1105. try (FileOutputStream fos = new FileOutputStream(tmpPack);
  1106. FileChannel channel = fos.getChannel();
  1107. OutputStream channelStream = Channels
  1108. .newOutputStream(channel)) {
  1109. pw.writePack(pm, pm, channelStream);
  1110. channel.force(true);
  1111. }
  1112. // write the packindex
  1113. try (FileOutputStream fos = new FileOutputStream(tmpIdx);
  1114. FileChannel idxChannel = fos.getChannel();
  1115. OutputStream idxStream = Channels
  1116. .newOutputStream(idxChannel)) {
  1117. pw.writeIndex(idxStream);
  1118. idxChannel.force(true);
  1119. }
  1120. if (pw.prepareBitmapIndex(pm)) {
  1121. File tmpBitmapIdx = new File(packdir, tmpBase + ".bitmap_tmp"); //$NON-NLS-1$
  1122. tmpExts.put(BITMAP_INDEX, tmpBitmapIdx);
  1123. if (!tmpBitmapIdx.createNewFile())
  1124. throw new IOException(MessageFormat.format(
  1125. JGitText.get().cannotCreateIndexfile,
  1126. tmpBitmapIdx.getPath()));
  1127. try (FileOutputStream fos = new FileOutputStream(tmpBitmapIdx);
  1128. FileChannel idxChannel = fos.getChannel();
  1129. OutputStream idxStream = Channels
  1130. .newOutputStream(idxChannel)) {
  1131. pw.writeBitmapIndex(idxStream);
  1132. idxChannel.force(true);
  1133. }
  1134. }
  1135. // rename the temporary files to real files
  1136. File realPack = nameFor(id, ".pack"); //$NON-NLS-1$
  1137. repo.getObjectDatabase().closeAllPackHandles(realPack);
  1138. tmpPack.setReadOnly();
  1139. FileUtils.rename(tmpPack, realPack, StandardCopyOption.ATOMIC_MOVE);
  1140. for (Map.Entry<PackExt, File> tmpEntry : tmpExts.entrySet()) {
  1141. File tmpExt = tmpEntry.getValue();
  1142. tmpExt.setReadOnly();
  1143. File realExt = nameFor(id,
  1144. "." + tmpEntry.getKey().getExtension()); //$NON-NLS-1$
  1145. try {
  1146. FileUtils.rename(tmpExt, realExt,
  1147. StandardCopyOption.ATOMIC_MOVE);
  1148. } catch (IOException e) {
  1149. File newExt = new File(realExt.getParentFile(),
  1150. realExt.getName() + ".new"); //$NON-NLS-1$
  1151. try {
  1152. FileUtils.rename(tmpExt, newExt,
  1153. StandardCopyOption.ATOMIC_MOVE);
  1154. } catch (IOException e2) {
  1155. newExt = tmpExt;
  1156. e = e2;
  1157. }
  1158. throw new IOException(MessageFormat.format(
  1159. JGitText.get().panicCantRenameIndexFile, newExt,
  1160. realExt), e);
  1161. }
  1162. }
  1163. boolean interrupted = false;
  1164. try {
  1165. FileSnapshot snapshot = FileSnapshot.save(realPack);
  1166. if (pconfig.doWaitPreventRacyPack(snapshot.size())) {
  1167. snapshot.waitUntilNotRacy();
  1168. }
  1169. } catch (InterruptedException e) {
  1170. interrupted = true;
  1171. }
  1172. try {
  1173. return repo.getObjectDatabase().openPack(realPack);
  1174. } finally {
  1175. if (interrupted) {
  1176. // Re-set interrupted flag
  1177. Thread.currentThread().interrupt();
  1178. }
  1179. }
  1180. } finally {
  1181. if (tmpPack != null && tmpPack.exists())
  1182. tmpPack.delete();
  1183. for (File tmpExt : tmpExts.values()) {
  1184. if (tmpExt.exists())
  1185. tmpExt.delete();
  1186. }
  1187. }
  1188. }
  1189. private File nameFor(String name, String ext) {
  1190. File packdir = repo.getObjectDatabase().getPackDirectory();
  1191. return new File(packdir, "pack-" + name + ext); //$NON-NLS-1$
  1192. }
  1193. private void checkCancelled() throws CancelledException {
  1194. if (pm.isCancelled() || Thread.currentThread().isInterrupted()) {
  1195. throw new CancelledException(JGitText.get().operationCanceled);
  1196. }
  1197. }
  1198. /**
  1199. * A class holding statistical data for a FileRepository regarding how many
  1200. * objects are stored as loose or packed objects
  1201. */
  1202. public static class RepoStatistics {
  1203. /**
  1204. * The number of objects stored in pack files. If the same object is
  1205. * stored in multiple pack files then it is counted as often as it
  1206. * occurs in pack files.
  1207. */
  1208. public long numberOfPackedObjects;
  1209. /**
  1210. * The number of pack files
  1211. */
  1212. public long numberOfPackFiles;
  1213. /**
  1214. * The number of objects stored as loose objects.
  1215. */
  1216. public long numberOfLooseObjects;
  1217. /**
  1218. * The sum of the sizes of all files used to persist loose objects.
  1219. */
  1220. public long sizeOfLooseObjects;
  1221. /**
  1222. * The sum of the sizes of all pack files.
  1223. */
  1224. public long sizeOfPackedObjects;
  1225. /**
  1226. * The number of loose refs.
  1227. */
  1228. public long numberOfLooseRefs;
  1229. /**
  1230. * The number of refs stored in pack files.
  1231. */
  1232. public long numberOfPackedRefs;
  1233. /**
  1234. * The number of bitmaps in the bitmap indices.
  1235. */
  1236. public long numberOfBitmaps;
  1237. @Override
  1238. public String toString() {
  1239. final StringBuilder b = new StringBuilder();
  1240. b.append("numberOfPackedObjects=").append(numberOfPackedObjects); //$NON-NLS-1$
  1241. b.append(", numberOfPackFiles=").append(numberOfPackFiles); //$NON-NLS-1$
  1242. b.append(", numberOfLooseObjects=").append(numberOfLooseObjects); //$NON-NLS-1$
  1243. b.append(", numberOfLooseRefs=").append(numberOfLooseRefs); //$NON-NLS-1$
  1244. b.append(", numberOfPackedRefs=").append(numberOfPackedRefs); //$NON-NLS-1$
  1245. b.append(", sizeOfLooseObjects=").append(sizeOfLooseObjects); //$NON-NLS-1$
  1246. b.append(", sizeOfPackedObjects=").append(sizeOfPackedObjects); //$NON-NLS-1$
  1247. b.append(", numberOfBitmaps=").append(numberOfBitmaps); //$NON-NLS-1$
  1248. return b.toString();
  1249. }
  1250. }
  1251. /**
  1252. * Returns information about objects and pack files for a FileRepository.
  1253. *
  1254. * @return information about objects and pack files for a FileRepository
  1255. * @throws java.io.IOException
  1256. */
  1257. public RepoStatistics getStatistics() throws IOException {
  1258. RepoStatistics ret = new RepoStatistics();
  1259. Collection<PackFile> packs = repo.getObjectDatabase().getPacks();
  1260. for (PackFile f : packs) {
  1261. ret.numberOfPackedObjects += f.getIndex().getObjectCount();
  1262. ret.numberOfPackFiles++;
  1263. ret.sizeOfPackedObjects += f.getPackFile().length();
  1264. if (f.getBitmapIndex() != null)
  1265. ret.numberOfBitmaps += f.getBitmapIndex().getBitmapCount();
  1266. }
  1267. File objDir = repo.getObjectsDirectory();
  1268. String[] fanout = objDir.list();
  1269. if (fanout != null && fanout.length > 0) {
  1270. for (String d : fanout) {
  1271. if (d.length() != 2)
  1272. continue;
  1273. File[] entries = new File(objDir, d).listFiles();
  1274. if (entries == null)
  1275. continue;
  1276. for (File f : entries) {
  1277. if (f.getName().length() != Constants.OBJECT_ID_STRING_LENGTH - 2)
  1278. continue;
  1279. ret.numberOfLooseObjects++;
  1280. ret.sizeOfLooseObjects += f.length();
  1281. }
  1282. }
  1283. }
  1284. RefDatabase refDb = repo.getRefDatabase();
  1285. for (Ref r : refDb.getRefs()) {
  1286. Storage storage = r.getStorage();
  1287. if (storage == Storage.LOOSE || storage == Storage.LOOSE_PACKED)
  1288. ret.numberOfLooseRefs++;
  1289. if (storage == Storage.PACKED || storage == Storage.LOOSE_PACKED)
  1290. ret.numberOfPackedRefs++;
  1291. }
  1292. return ret;
  1293. }
  1294. /**
  1295. * Set the progress monitor used for garbage collection methods.
  1296. *
  1297. * @param pm a {@link org.eclipse.jgit.lib.ProgressMonitor} object.
  1298. * @return this
  1299. */
  1300. public GC setProgressMonitor(ProgressMonitor pm) {
  1301. this.pm = (pm == null) ? NullProgressMonitor.INSTANCE : pm;
  1302. return this;
  1303. }
  1304. /**
  1305. * During gc() or prune() each unreferenced, loose object which has been
  1306. * created or modified in the last <code>expireAgeMillis</code> milliseconds
  1307. * will not be pruned. Only older objects may be pruned. If set to 0 then
  1308. * every object is a candidate for pruning.
  1309. *
  1310. * @param expireAgeMillis
  1311. * minimal age of objects to be pruned in milliseconds.
  1312. */
  1313. public void setExpireAgeMillis(long expireAgeMillis) {
  1314. this.expireAgeMillis = expireAgeMillis;
  1315. expire = null;
  1316. }
  1317. /**
  1318. * During gc() or prune() packfiles which are created or modified in the
  1319. * last <code>packExpireAgeMillis</code> milliseconds will not be deleted.
  1320. * Only older packfiles may be deleted. If set to 0 then every packfile is a
  1321. * candidate for deletion.
  1322. *
  1323. * @param packExpireAgeMillis
  1324. * minimal age of packfiles to be deleted in milliseconds.
  1325. */
  1326. public void setPackExpireAgeMillis(long packExpireAgeMillis) {
  1327. this.packExpireAgeMillis = packExpireAgeMillis;
  1328. expire = null;
  1329. }
  1330. /**
  1331. * Set the PackConfig used when (re-)writing packfiles. This allows to
  1332. * influence how packs are written and to implement something similar to
  1333. * "git gc --aggressive"
  1334. *
  1335. * @param pconfig
  1336. * the {@link org.eclipse.jgit.storage.pack.PackConfig} used when
  1337. * writing packs
  1338. */
  1339. public void setPackConfig(@NonNull PackConfig pconfig) {
  1340. this.pconfig = pconfig;
  1341. }
  1342. /**
  1343. * During gc() or prune() each unreferenced, loose object which has been
  1344. * created or modified after or at <code>expire</code> will not be pruned.
  1345. * Only older objects may be pruned. If set to null then every object is a
  1346. * candidate for pruning.
  1347. *
  1348. * @param expire
  1349. * instant in time which defines object expiration
  1350. * objects with modification time before this instant are expired
  1351. * objects with modification time newer or equal to this instant
  1352. * are not expired
  1353. */
  1354. public void setExpire(Date expire) {
  1355. this.expire = expire;
  1356. expireAgeMillis = -1;
  1357. }
  1358. /**
  1359. * During gc() or prune() packfiles which are created or modified after or
  1360. * at <code>packExpire</code> will not be deleted. Only older packfiles may
  1361. * be deleted. If set to null then every packfile is a candidate for
  1362. * deletion.
  1363. *
  1364. * @param packExpire
  1365. * instant in time which defines packfile expiration
  1366. */
  1367. public void setPackExpire(Date packExpire) {
  1368. this.packExpire = packExpire;
  1369. packExpireAgeMillis = -1;
  1370. }
  1371. /**
  1372. * Set the {@code gc --auto} option.
  1373. *
  1374. * With this option, gc checks whether any housekeeping is required; if not,
  1375. * it exits without performing any work. Some JGit commands run
  1376. * {@code gc --auto} after performing operations that could create many
  1377. * loose objects.
  1378. * <p>
  1379. * Housekeeping is required if there are too many loose objects or too many
  1380. * packs in the repository. If the number of loose objects exceeds the value
  1381. * of the gc.auto option JGit GC consolidates all existing packs into a
  1382. * single pack (equivalent to {@code -A} option), whereas git-core would
  1383. * combine all loose objects into a single pack using {@code repack -d -l}.
  1384. * Setting the value of {@code gc.auto} to 0 disables automatic packing of
  1385. * loose objects.
  1386. * <p>
  1387. * If the number of packs exceeds the value of {@code gc.autoPackLimit},
  1388. * then existing packs (except those marked with a .keep file) are
  1389. * consolidated into a single pack by using the {@code -A} option of repack.
  1390. * Setting {@code gc.autoPackLimit} to 0 disables automatic consolidation of
  1391. * packs.
  1392. * <p>
  1393. * Like git the following jgit commands run auto gc:
  1394. * <ul>
  1395. * <li>fetch</li>
  1396. * <li>merge</li>
  1397. * <li>rebase</li>
  1398. * <li>receive-pack</li>
  1399. * </ul>
  1400. * The auto gc for receive-pack can be suppressed by setting the config
  1401. * option {@code receive.autogc = false}
  1402. *
  1403. * @param auto
  1404. * defines whether gc should do automatic housekeeping
  1405. */
  1406. public void setAuto(boolean auto) {
  1407. this.automatic = auto;
  1408. }
  1409. /**
  1410. * @param background
  1411. * whether to run the gc in a background thread.
  1412. */
  1413. void setBackground(boolean background) {
  1414. this.background = background;
  1415. }
  1416. private boolean needGc() {
  1417. if (tooManyPacks()) {
  1418. addRepackAllOption();
  1419. } else {
  1420. return tooManyLooseObjects();
  1421. }
  1422. // TODO run pre-auto-gc hook, if it fails return false
  1423. return true;
  1424. }
  1425. private void addRepackAllOption() {
  1426. // TODO: if JGit GC is enhanced to support repack's option -l this
  1427. // method needs to be implemented
  1428. }
  1429. /**
  1430. * @return {@code true} if number of packs > gc.autopacklimit (default 50)
  1431. */
  1432. boolean tooManyPacks() {
  1433. int autopacklimit = repo.getConfig().getInt(
  1434. ConfigConstants.CONFIG_GC_SECTION,
  1435. ConfigConstants.CONFIG_KEY_AUTOPACKLIMIT,
  1436. DEFAULT_AUTOPACKLIMIT);
  1437. if (autopacklimit <= 0) {
  1438. return false;
  1439. }
  1440. // JGit always creates two packfiles, one for the objects reachable from
  1441. // branches, and another one for the rest
  1442. return repo.getObjectDatabase().getPacks().size() > (autopacklimit + 1);
  1443. }
  1444. /**
  1445. * Quickly estimate number of loose objects, SHA1 is distributed evenly so
  1446. * counting objects in one directory (bucket 17) is sufficient
  1447. *
  1448. * @return {@code true} if number of loose objects > gc.auto (default 6700)
  1449. */
  1450. boolean tooManyLooseObjects() {
  1451. int auto = getLooseObjectLimit();
  1452. if (auto <= 0) {
  1453. return false;
  1454. }
  1455. int n = 0;
  1456. int threshold = (auto + 255) / 256;
  1457. Path dir = repo.getObjectsDirectory().toPath().resolve("17"); //$NON-NLS-1$
  1458. if (!dir.toFile().exists()) {
  1459. return false;
  1460. }
  1461. try (DirectoryStream<Path> stream = Files.newDirectoryStream(dir, file -> {
  1462. Path fileName = file.getFileName();
  1463. return file.toFile().isFile() && fileName != null
  1464. && PATTERN_LOOSE_OBJECT.matcher(fileName.toString())
  1465. .matches();
  1466. })) {
  1467. for (Iterator<Path> iter = stream.iterator(); iter.hasNext(); iter
  1468. .next()) {
  1469. if (++n > threshold) {
  1470. return true;
  1471. }
  1472. }
  1473. } catch (IOException e) {
  1474. LOG.error(e.getMessage(), e);
  1475. }
  1476. return false;
  1477. }
  1478. private int getLooseObjectLimit() {
  1479. return repo.getConfig().getInt(ConfigConstants.CONFIG_GC_SECTION,
  1480. ConfigConstants.CONFIG_KEY_AUTO, DEFAULT_AUTOLIMIT);
  1481. }
  1482. }