HACKING 21 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596
  1. Hacking Tor: An Incomplete Guide
  2. ================================
  3. Getting started
  4. ---------------
  5. For full information on how Tor is supposed to work, look at the files in
  6. https://gitweb.torproject.org/torspec.git/tree
  7. For an explanation of how to change Tor's design to work differently, look at
  8. https://gitweb.torproject.org/torspec.git/blob_plain/HEAD:/proposals/001-process.txt
  9. For the latest version of the code, get a copy of git, and
  10. git clone https://git.torproject.org/git/tor
  11. We talk about Tor on the tor-talk mailing list. Design proposals and
  12. discussion belong on the tor-dev mailing list. We hang around on
  13. irc.oftc.net, with general discussion happening on #tor and development
  14. happening on #tor-dev.
  15. How we use Git branches
  16. -----------------------
  17. Each main development series (like 0.2.1, 0.2.2, etc) has its main work
  18. applied to a single branch. At most one series can be the development series
  19. at a time; all other series are maintenance series that get bug-fixes only.
  20. The development series is built in a git branch called "master"; the
  21. maintenance series are built in branches called "maint-0.2.0", "maint-0.2.1",
  22. and so on. We regularly merge the active maint branches forward.
  23. For all series except the development series, we also have a "release" branch
  24. (as in "release-0.2.1"). The release series is based on the corresponding
  25. maintenance series, except that it deliberately lags the maint series for
  26. most of its patches, so that bugfix patches are not typically included in a
  27. maintenance release until they've been tested for a while in a development
  28. release. Occasionally, we'll merge an urgent bugfix into the release branch
  29. before it gets merged into maint, but that's rare.
  30. If you're working on a bugfix for a bug that occurs in a particular version,
  31. base your bugfix branch on the "maint" branch for the first supported series
  32. that has that bug. (As of June 2013, we're supporting 0.2.3 and later.) If
  33. you're working on a new feature, base it on the master branch.
  34. How we log changes
  35. ------------------
  36. When you do a commit that needs a ChangeLog entry, add a new file to
  37. the "changes" toplevel subdirectory. It should have the format of a
  38. one-entry changelog section from the current ChangeLog file, as in
  39. o Major bugfixes:
  40. - Fix a potential buffer overflow. Fixes bug 99999; bugfix on
  41. 0.3.1.4-beta.
  42. To write a changes file, first categorize the change. Some common categories
  43. are: Minor bugfixes, Major bugfixes, Minor features, Major features, Code
  44. simplifications and refactoring. Then say what the change does. If
  45. it's a bugfix, mention what bug it fixes and when the bug was
  46. introduced. To find out which Git tag the change was introduced in,
  47. you can use "git describe --contains <sha1 of commit>".
  48. If at all possible, try to create this file in the same commit where you are
  49. making the change. Please give it a distinctive name that no other branch will
  50. use for the lifetime of your change. To verify the format of the changes file,
  51. you can use "make check-changes".
  52. When we go to make a release, we will concatenate all the entries
  53. in changes to make a draft changelog, and clear the directory. We'll
  54. then edit the draft changelog into a nice readable format.
  55. What needs a changes file?::
  56. A not-exhaustive list: Anything that might change user-visible
  57. behavior. Anything that changes internals, documentation, or the build
  58. system enough that somebody could notice. Big or interesting code
  59. rewrites. Anything about which somebody might plausibly wonder "when
  60. did that happen, and/or why did we do that" 6 months down the line.
  61. Why use changes files instead of Git commit messages?::
  62. Git commit messages are written for developers, not users, and they
  63. are nigh-impossible to revise after the fact.
  64. Why use changes files instead of entries in the ChangeLog?::
  65. Having every single commit touch the ChangeLog file tended to create
  66. zillions of merge conflicts.
  67. Useful tools
  68. ------------
  69. These aren't strictly necessary for hacking on Tor, but they can help track
  70. down bugs.
  71. Jenkins
  72. ~~~~~~~
  73. https://jenkins.torproject.org
  74. Dmalloc
  75. ~~~~~~~
  76. The dmalloc library will keep track of memory allocation, so you can find out
  77. if we're leaking memory, doing any double-frees, or so on.
  78. dmalloc -l ~/dmalloc.log
  79. (run the commands it tells you)
  80. ./configure --with-dmalloc
  81. Valgrind
  82. ~~~~~~~~
  83. valgrind --leak-check=yes --error-limit=no --show-reachable=yes src/or/tor
  84. (Note that if you get a zillion openssl warnings, you will also need to
  85. pass --undef-value-errors=no to valgrind, or rebuild your openssl
  86. with -DPURIFY.)
  87. Running lcov for unit test coverage
  88. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  89. Lcov is a utility that generates pretty HTML reports of test code coverage.
  90. To generate such a report:
  91. -----
  92. ./configure --enable-coverage
  93. make
  94. make coverage-html
  95. $BROWSER ./coverage_html/index.html
  96. -----
  97. This will run the tor unit test suite `./src/test/test` and generate the HTML
  98. coverage code report under the directory ./coverage_html/. To change the
  99. output directory, use `make coverage-html HTML_COVER_DIR=./funky_new_cov_dir`.
  100. Coverage diffs using lcov are not currently implemented, but are being
  101. investigated (as of July 2014).
  102. Running the unit tests
  103. ~~~~~~~~~~~~~~~~~~~~~~
  104. To quickly run all tests:
  105. -----
  106. make check
  107. -----
  108. To run unit tests only:
  109. -----
  110. make test
  111. -----
  112. To selectively run just some tests (the following can be combined
  113. arbitrarily):
  114. -----
  115. ./src/test/test <name_of_test> [<name of test 2>] ...
  116. ./src/test/test <prefix_of_name_of_test>.. [<prefix_of_name_of_test2>..] ...
  117. ./src/test/test :<name_of_excluded_test> [:<name_of_excluded_test2]...
  118. -----
  119. Running gcov for unit test coverage
  120. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  121. -----
  122. ./configure --enable-coverage
  123. make
  124. make check
  125. mkdir coverage-output
  126. ./scripts/test/coverage coverage-output
  127. -----
  128. (On OSX, you'll need to start with "--enable-coverage CC=clang".)
  129. Then, look at the .gcov files in coverage-output. '-' before a line means
  130. that the compiler generated no code for that line. '######' means that the
  131. line was never reached. Lines with numbers were called that number of times.
  132. If that doesn't work:
  133. * Try configuring Tor with --disable-gcc-hardening
  134. * You might need to run 'make clean' after you run './configure'.
  135. If you make changes to Tor and want to get another set of coverage results,
  136. you can run "make reset-gcov" to clear the intermediary gcov output.
  137. If you have two different "coverage-output" directories, and you want to see
  138. a meaningful diff between them, you can run:
  139. -----
  140. ./scripts/test/cov-diff coverage-output1 coverage-output2 | less
  141. -----
  142. In this diff, any lines that were visited at least once will have coverage
  143. "1". This lets you inspect what you (probably) really want to know: which
  144. untested lines were changed? Are there any new untested lines?
  145. Running integration tests
  146. ~~~~~~~~~~~~~~~~~~~~~~~~~
  147. We have the beginnings of a set of scripts to run integration tests using
  148. Chutney. To try them, set CHUTNEY_PATH to your chutney source directory, and
  149. run "make test-network".
  150. Profiling Tor with oprofile
  151. ~~~~~~~~~~~~~~~~~~~~~~~~~~~
  152. The oprofile tool runs (on Linux only!) to tell you what functions Tor is
  153. spending its CPU time in, so we can identify berformance pottlenecks.
  154. Here are some basic instructions
  155. - Build tor with debugging symbols (you probably already have, unless
  156. you messed with CFLAGS during the build process).
  157. - Build all the libraries you care about with debugging symbols
  158. (probably you only care about libssl, maybe zlib and Libevent).
  159. - Copy this tor to a new directory
  160. - Copy all the libraries it uses to that dir too (ldd ./tor will
  161. tell you)
  162. - Set LD_LIBRARY_PATH to include that dir. ldd ./tor should now
  163. show you it's using the libs in that dir
  164. - Run that tor
  165. - Reset oprofiles counters/start it
  166. * "opcontrol --reset; opcontrol --start", if Nick remembers right.
  167. - After a while, have it dump the stats on tor and all the libs
  168. in that dir you created.
  169. * "opcontrol --dump;"
  170. * "opreport -l that_dir/*"
  171. - Profit
  172. Coding conventions
  173. ------------------
  174. Patch checklist
  175. ~~~~~~~~~~~~~~~
  176. If possible, send your patch as one of these (in descending order of
  177. preference)
  178. - A git branch we can pull from
  179. - Patches generated by git format-patch
  180. - A unified diff
  181. Did you remember...
  182. - To build your code while configured with --enable-gcc-warnings?
  183. - To run "make check-spaces" on your code?
  184. - To run "make check-docs" to see whether all new options are on
  185. the manpage?
  186. - To write unit tests, as possible?
  187. - To base your code on the appropriate branch?
  188. - To include a file in the "changes" directory as appropriate?
  189. Whitespace and C conformance
  190. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  191. Invoke "make check-spaces" from time to time, so it can tell you about
  192. deviations from our C whitespace style. Generally, we use:
  193. - Unix-style line endings
  194. - K&R-style indentation
  195. - No space before newlines
  196. - A blank line at the end of each file
  197. - Never more than one blank line in a row
  198. - Always spaces, never tabs
  199. - No more than 79-columns per line.
  200. - Two spaces per indent.
  201. - A space between control keywords and their corresponding paren
  202. "if (x)", "while (x)", and "switch (x)", never "if(x)", "while(x)", or
  203. "switch(x)".
  204. - A space between anything and an open brace.
  205. - No space between a function name and an opening paren. "puts(x)", not
  206. "puts (x)".
  207. - Function declarations at the start of the line.
  208. We try hard to build without warnings everywhere. In particular, if you're
  209. using gcc, you should invoke the configure script with the option
  210. "--enable-gcc-warnings". This will give a bunch of extra warning flags to
  211. the compiler, and help us find divergences from our preferred C style.
  212. Getting emacs to edit Tor source properly
  213. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  214. Nick likes to put the following snippet in his .emacs file:
  215. -----
  216. (add-hook 'c-mode-hook
  217. (lambda ()
  218. (font-lock-mode 1)
  219. (set-variable 'show-trailing-whitespace t)
  220. (let ((fname (expand-file-name (buffer-file-name))))
  221. (cond
  222. ((string-match "^/home/nickm/src/libevent" fname)
  223. (set-variable 'indent-tabs-mode t)
  224. (set-variable 'c-basic-offset 4)
  225. (set-variable 'tab-width 4))
  226. ((string-match "^/home/nickm/src/tor" fname)
  227. (set-variable 'indent-tabs-mode nil)
  228. (set-variable 'c-basic-offset 2))
  229. ((string-match "^/home/nickm/src/openssl" fname)
  230. (set-variable 'indent-tabs-mode t)
  231. (set-variable 'c-basic-offset 8)
  232. (set-variable 'tab-width 8))
  233. ))))
  234. -----
  235. You'll note that it defaults to showing all trailing whitespace. The "cond"
  236. test detects whether the file is one of a few C free software projects that I
  237. often edit, and sets up the indentation level and tab preferences to match
  238. what they want.
  239. If you want to try this out, you'll need to change the filename regex
  240. patterns to match where you keep your Tor files.
  241. If you use emacs for editing Tor and nothing else, you could always just say:
  242. -----
  243. (add-hook 'c-mode-hook
  244. (lambda ()
  245. (font-lock-mode 1)
  246. (set-variable 'show-trailing-whitespace t)
  247. (set-variable 'indent-tabs-mode nil)
  248. (set-variable 'c-basic-offset 2)))
  249. -----
  250. There is probably a better way to do this. No, we are probably not going
  251. to clutter the files with emacs stuff.
  252. Functions to use
  253. ~~~~~~~~~~~~~~~~
  254. We have some wrapper functions like tor_malloc, tor_free, tor_strdup, and
  255. tor_gettimeofday; use them instead of their generic equivalents. (They
  256. always succeed or exit.)
  257. You can get a full list of the compatibility functions that Tor provides by
  258. looking through src/common/util.h and src/common/compat.h. You can see the
  259. available containers in src/common/containers.h. You should probably
  260. familiarize yourself with these modules before you write too much code, or
  261. else you'll wind up reinventing the wheel.
  262. Use 'INLINE' instead of 'inline', so that we work properly on Windows.
  263. Calling and naming conventions
  264. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  265. Whenever possible, functions should return -1 on error and 0 on success.
  266. For multi-word identifiers, use lowercase words combined with
  267. underscores. (e.g., "multi_word_identifier"). Use ALL_CAPS for macros and
  268. constants.
  269. Typenames should end with "_t".
  270. Function names should be prefixed with a module name or object name. (In
  271. general, code to manipulate an object should be a module with the same name
  272. as the object, so it's hard to tell which convention is used.)
  273. Functions that do things should have imperative-verb names
  274. (e.g. buffer_clear, buffer_resize); functions that return booleans should
  275. have predicate names (e.g. buffer_is_empty, buffer_needs_resizing).
  276. If you find that you have four or more possible return code values, it's
  277. probably time to create an enum. If you find that you are passing three or
  278. more flags to a function, it's probably time to create a flags argument that
  279. takes a bitfield.
  280. What To Optimize
  281. ~~~~~~~~~~~~~~~~
  282. Don't optimize anything if it's not in the critical path. Right now, the
  283. critical path seems to be AES, logging, and the network itself. Feel free to
  284. do your own profiling to determine otherwise.
  285. Log conventions
  286. ~~~~~~~~~~~~~~~
  287. https://www.torproject.org/docs/faq#LogLevel
  288. No error or warning messages should be expected during normal OR or OP
  289. operation.
  290. If a library function is currently called such that failure always means ERR,
  291. then the library function should log WARN and let the caller log ERR.
  292. Every message of severity INFO or higher should either (A) be intelligible
  293. to end-users who don't know the Tor source; or (B) somehow inform the
  294. end-users that they aren't expected to understand the message (perhaps
  295. with a string like "internal error"). Option (A) is to be preferred to
  296. option (B).
  297. Doxygen
  298. ~~~~~~~~
  299. We use the 'doxygen' utility to generate documentation from our
  300. source code. Here's how to use it:
  301. 1. Begin every file that should be documented with
  302. /**
  303. * \file filename.c
  304. * \brief Short description of the file.
  305. **/
  306. (Doxygen will recognize any comment beginning with /** as special.)
  307. 2. Before any function, structure, #define, or variable you want to
  308. document, add a comment of the form:
  309. /** Describe the function's actions in imperative sentences.
  310. *
  311. * Use blank lines for paragraph breaks
  312. * - and
  313. * - hyphens
  314. * - for
  315. * - lists.
  316. *
  317. * Write <b>argument_names</b> in boldface.
  318. *
  319. * \code
  320. * place_example_code();
  321. * between_code_and_endcode_commands();
  322. * \endcode
  323. */
  324. 3. Make sure to escape the characters "<", ">", "\", "%" and "#" as "\<",
  325. "\>", "\\", "\%", and "\#".
  326. 4. To document structure members, you can use two forms:
  327. struct foo {
  328. /** You can put the comment before an element; */
  329. int a;
  330. int b; /**< Or use the less-than symbol to put the comment
  331. * after the element. */
  332. };
  333. 5. To generate documentation from the Tor source code, type:
  334. $ doxygen -g
  335. To generate a file called 'Doxyfile'. Edit that file and run
  336. 'doxygen' to generate the API documentation.
  337. 6. See the Doxygen manual for more information; this summary just
  338. scratches the surface.
  339. Doxygen comment conventions
  340. ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  341. Say what functions do as a series of one or more imperative sentences, as
  342. though you were telling somebody how to be the function. In other words, DO
  343. NOT say:
  344. /** The strtol function parses a number.
  345. *
  346. * nptr -- the string to parse. It can include whitespace.
  347. * endptr -- a string pointer to hold the first thing that is not part
  348. * of the number, if present.
  349. * base -- the numeric base.
  350. * returns: the resulting number.
  351. */
  352. long strtol(const char *nptr, char **nptr, int base);
  353. Instead, please DO say:
  354. /** Parse a number in radix <b>base</b> from the string <b>nptr</b>,
  355. * and return the result. Skip all leading whitespace. If
  356. * <b>endptr</b> is not NULL, set *<b>endptr</b> to the first character
  357. * after the number parsed.
  358. **/
  359. long strtol(const char *nptr, char **nptr, int base);
  360. Doxygen comments are the contract in our abstraction-by-contract world: if
  361. the functions that call your function rely on it doing something, then your
  362. function should mention that it does that something in the documentation. If
  363. you rely on a function doing something beyond what is in its documentation,
  364. then you should watch out, or it might do something else later.
  365. Putting out a new release
  366. -------------------------
  367. Here are the steps Roger takes when putting out a new Tor release:
  368. 1) Use it for a while, as a client, as a relay, as a hidden service,
  369. and as a directory authority. See if it has any obvious bugs, and
  370. resolve those.
  371. 1.5) As applicable, merge the maint-X branch into the release-X branch.
  372. 2) Gather the changes/* files into a changelog entry, rewriting many
  373. of them and reordering to focus on what users and funders would find
  374. interesting and understandable.
  375. 2.1) Make sure that everything that wants a bug number has one.
  376. Make sure that everything which is a bugfix says what version
  377. it was a bugfix on.
  378. 2.2) Concatenate them.
  379. 2.3) Sort them by section. Within each section, sort by "version it's
  380. a bugfix on", else by numerical ticket order.
  381. 2.4) Clean them up:
  382. Standard idioms:
  383. "Fixes bug 9999; bugfix on 0.3.3.3-alpha."
  384. One space after a period.
  385. Make stuff very terse
  386. Make sure each section name ends with a colon
  387. Describe the user-visible problem right away
  388. Mention relevant config options by name. If they're rare or unusual,
  389. remind people what they're for
  390. Avoid starting lines with open-paren
  391. Present and imperative tense: not past.
  392. 'Relays', not 'servers' or 'nodes' or 'Tor relays'.
  393. "Stop FOOing", not "Fix a bug where we would FOO".
  394. Try not to let any given section be longer than about a page. Break up
  395. long sections into subsections by some sort of common subtopic. This
  396. guideline is especially important when organizing Release Notes for
  397. new stable releases.
  398. If a given changes stanza showed up in a different release (e.g.
  399. maint-0.2.1), be sure to make the stanzas identical (so people can
  400. distinguish if these are the same change).
  401. 2.5) Merge them in.
  402. 2.6) Clean everything one last time.
  403. 2.7) Run ./scripts/maint/format_changelog.py to make it prettier.
  404. 3) Compose a short release blurb to highlight the user-facing
  405. changes. Insert said release blurb into the ChangeLog stanza. If it's
  406. a stable release, add it to the ReleaseNotes file too. If we're adding
  407. to a release-0.2.x branch, manually commit the changelogs to the later
  408. git branches too.
  409. 4) Bump the version number in configure.ac and rebuild. Then run
  410. "make update-versions".
  411. 5) Make dist, put the tarball up somewhere, and tell #tor about it. Wait
  412. a while to see if anybody has problems building it. Try to get Sebastian
  413. or somebody to try building it on Windows.
  414. 6) Get at least two of weasel/arma/sebastian to put the new version number
  415. in their approved versions list.
  416. 7) Sign the tarball, then sign and push the git tag:
  417. gpg -ba <the_tarball>
  418. git tag -u <keyid> tor-0.2.x.y-status
  419. git push origin tag tor-0.2.x.y-status
  420. 8a) scp the tarball and its sig to the dist website, i.e.
  421. /srv/dist-master.torproject.org/htdocs/ on dist-master. When you want
  422. it to go live, you run "static-update-component dist.torproject.org"
  423. on dist-master.
  424. 8b) Edit "include/versions.wmi" and "Makefile" to note the new version.
  425. 9) Email the packagers (cc'ing tor-assistants) that a new tarball is up.
  426. The current list of packagers is:
  427. {weasel,gk,mikeperry} at torproject dot org
  428. {blueness} at gentoo dot org
  429. {paul} at invizbox dot io
  430. {ondrej.mikle} at gmail dot com
  431. {lfleischer} at archlinux dot org
  432. 10) Add the version number to Trac. To do this, go to Trac, log in,
  433. select "Admin" near the top of the screen, then select "Versions" from
  434. the menu on the left. At the right, there will be an "Add version"
  435. box. By convention, we enter the version in the form "Tor:
  436. 0.2.2.23-alpha" (or whatever the version is), and we select the date as
  437. the date in the ChangeLog.
  438. 11) Forward-port the ChangeLog.
  439. 12) Wait up to a day or two (for a development release), or until most
  440. packages are up (for a stable release), and mail the release blurb and
  441. changelog to tor-talk or tor-announce.
  442. (We might be moving to faster announcements, but don't announce until
  443. the website is at least updated.)
  444. 13) If it's a stable release, bump the version number in the maint-x.y.z
  445. branch to "newversion-dev", and do a "merge -s ours" merge to avoid
  446. taking that change into master. Do a similar 'merge -s theirs'
  447. merge to get the change (and only that change) into release. (Some
  448. of the build scripts require that maint merge cleanly into release.)