docs/devinfo.html

   1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
   2 <html lang="en">
   3 <head>
   4   <meta http-equiv="content-type" content="text/html; charset=utf-8">
   5   <title>Development Notes</title>
   6   <link rel="stylesheet" type="text/css" href="mesa.css">
   7 </head>
   8 <body>
   9
  10 <div class="header">
  11   <h1>The Mesa 3D Graphics Library</h1>
  12 </div>
  13
  14 <iframe src="contents.html"></iframe>
  15 <div class="content">
  16
  17 <h1>Development Notes</h1>
  18
  19
  20 <ul>
  21 <li><a href="#style">Coding Style</a>
  22 <li><a href="#submitting">Submitting Patches</a>
  23 <li><a href="#release">Making a New Mesa Release</a>
  24 <li><a href="#extensions">Adding Extensions</a>
  25 </ul>
  26
  27
  28 <h2 id="style">Coding Style</h2>
  29
  30 <p>
  31 Mesa is over 20 years old and the coding style has evolved over time.
  32 Some old parts use a style that's a bit out of date.
  33 If the guidelines below don't cover something, try following the format of
  34 existing, neighboring code.
  35 </p>
  36
  37 <p>
  38 Basic formatting guidelines
  39 </p>
  40
  41 <ul>
  42 <li>3-space indentation, no tabs.
  43 <li>Limit lines to 78 or fewer characters.  The idea is to prevent line
  44 wrapping in 80-column editors and terminals.  There are exceptions, such
  45 as if you're defining a large, static table of information.
  46 <li>Opening braces go on the same line as the if/for/while statement.
  47 For example:
  48 <pre>
  49    if (condition) {
  50       foo;
  51    } else {
  52       bar;
  53    }
  54 </pre>
  55
  56 <li>Put a space before/after operators.  For example, <tt>a = b + c;</tt>
  57 and not <tt>a=b+c;</tt>
  58
  59 <li>This GNU indent command generally does the right thing for formatting:
  60 <pre>
  61    indent -br -i3 -npcs --no-tabs infile.c -o outfile.c
  62 </pre>
  63
  64 <li>Use comments wherever you think it would be helpful for other developers.
  65 Several specific cases and style examples follow.  Note that we roughly
  66 follow <a href="http://www.stack.nl/~dimitri/doxygen/">Doxygen</a> conventions.
  67 <br>
  68 <br>
  69 Single-line comments:
  70 <pre>
  71    /* null-out pointer to prevent dangling reference below */
  72    bufferObj = NULL;
  73 </pre>
  74 Or,
  75 <pre>
  76    bufferObj = NULL;  /* prevent dangling reference below */
  77 </pre>
  78 Multi-line comment:
  79 <pre>
  80    /* If this is a new buffer object id, or one which was generated but
  81     * never used before, allocate a buffer object now.
  82     */
  83 </pre>
  84 We try to quote the OpenGL specification where prudent:
  85 <pre>
  86    /* Page 38 of the PDF of the OpenGL ES 3.0 spec says:
  87     *
  88     *     "An INVALID_OPERATION error is generated for any of the following
  89     *     conditions:
  90     *
  91     *     * <length> is zero."
  92     *
  93     * Additionally, page 94 of the PDF of the OpenGL 4.5 core spec
  94     * (30.10.2014) also says this, so it's no longer allowed for desktop GL,
  95     * either.
  96     */
  97 </pre>
  98 Function comment example:
  99 <pre>
 100    /**
 101     * Create and initialize a new buffer object.  Called via the
 102     * ctx->Driver.CreateObject() driver callback function.
 103     * \param  name  integer name of the object
 104     * \param  type  one of GL_FOO, GL_BAR, etc.
 105     * \return  pointer to new object or NULL if error
 106     */
 107    struct gl_object *
 108    _mesa_create_object(GLuint name, GLenum type)
 109    {
 110       /* function body */
 111    }
 112 </pre>
 113
 114 <li>Put the function return type and qualifiers on one line and the function
 115 name and parameters on the next, as seen above.  This makes it easy to use
 116 <code>grep ^function_name dir/*</code> to find function definitions.  Also,
 117 the opening brace goes on the next line by itself (see above.)
 118
 119 <li>Function names follow various conventions depending on the type of function:
 120 <pre>
 121    glFooBar()       - a public GL entry point (in glapi_dispatch.c)
 122    _mesa_FooBar()   - the internal immediate mode function
 123    save_FooBar()    - retained mode (display list) function in dlist.c
 124    foo_bar()        - a static (private) function
 125    _mesa_foo_bar()  - an internal non-static Mesa function
 126 </pre>
 127
 128 <li>Constants, macros and enumerant names are ALL_UPPERCASE, with _ between
 129 words.
 130 <li>Mesa usually uses camel case for local variables (Ex: "localVarname")
 131 while gallium typically uses underscores (Ex: "local_var_name").
 132 <li>Global variables are almost never used because Mesa should be thread-safe.
 133
 134 <li>Booleans.  Places that are not directly visible to the GL API
 135 should prefer the use of <tt>bool</tt>, <tt>true</tt>, and
 136 <tt>false</tt> over <tt>GLboolean</tt>, <tt>GL_TRUE</tt>, and
 137 <tt>GL_FALSE</tt>.  In C code, this may mean that
 138 <tt>#include &lt;stdbool.h&gt;</tt> needs to be added.  The
 139 <tt>try_emit_</tt>* methods in src/mesa/program/ir_to_mesa.cpp and
 140 src/mesa/state_tracker/st_glsl_to_tgsi.cpp can serve as examples.
 141
 142 </ul>
 143
 144
 145 <h2 id="submitting">Submitting patches</h2>
 146
 147 <p>
 148 The basic guidelines for submitting patches are:
 149 </p>
 150
 151 <ul>
 152 <li>Patches should be sufficiently tested before submitting.
 153 <li>Code patches should follow Mesa coding conventions.
 154 <li>Whenever possible, patches should only effect individual Mesa/Gallium
 155 components.
 156 <li>Patches should never introduce build breaks and should be bisectable (see
 157 <code>git bisect</code>.)
 158 <li>Patches should be properly formatted (see below).
 159 <li>Patches should be submitted to mesa-dev for review using
 160 <code>git send-email</code>.
 161 <li>Patches should not mix code changes with code formatting changes (except,
 162 perhaps, in very trivial cases.)
 163 </ul>
 164
 165 <h3>Patch formatting</h3>
 166
 167 <p>
 168 The basic rules for patch formatting are:
 169 </p>
 170
 171 <ul>
 172 <li>Lines should be limited to 75 characters or less so that git logs
 173 displayed in 80-column terminals avoid line wrapping.  Note that git
 174 log uses 4 spaces of indentation (4 + 75 &lt; 80).
 175 <li>The first line should be a short, concise summary of the change prefixed
 176 with a module name.  Examples:
 177 <pre>
 178     mesa: Add support for querying GL_VERTEX_ATTRIB_ARRAY_LONG
 179
 180     gallium: add PIPE_CAP_DEVICE_RESET_STATUS_QUERY
 181
 182     i965: Fix missing type in local variable declaration.
 183 </pre>
 184 <li>Subsequent patch comments should describe the change in more detail,
 185 if needed.  For example:
 186 <pre>
 187     i965: Remove end-of-thread SEND alignment code.
 188
 189     This was present in Eric's initial implementation of the compaction code
 190     for Sandybridge (commit 077d01b6). There is no documentation saying this
 191     is necessary, and removing it causes no regressions in piglit on any
 192     platform.
 193 </pre>
 194 <li>A "Signed-off-by:" line is not required, but not discouraged either.
 195 <li>If a patch address a bugzilla issue, that should be noted in the
 196 patch comment.  For example:
 197 <pre>
 198    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89689
 199 </pre>
 200 <li>If there have been several revisions to a patch during the review
 201 process, they should be noted such as in this example:
 202 <pre>
 203     st/mesa: add ARB_texture_stencil8 support (v4)
 204
 205     if we support stencil texturing, enable texture_stencil8
 206     there is no requirement to support native S8 for this,
 207     the texture can be converted to x24s8 fine.
 208
 209     v2: fold fixes from Marek in:
 210        a) put S8 last in the list
 211        b) fix renderable to always test for d/s renderable
 212         fixup the texture case to use a stencil only format
 213         for picking the format for the texture view.
 214     v3: hit fallback for getteximage
 215     v4: put s8 back in front, it shouldn't get picked now (Ilia)
 216 </pre>
 217 <li>If someone tested your patch, document it with a line like this:
 218 <pre>
 219     Tested-by: Joe Hacker &lt;jhacker@foo.com&gt;
 220 </pre>
 221 <li>If the patch was reviewed (usually the case) or acked by someone,
 222 that should be documented with:
 223 <pre>
 224     Reviewed-by: Joe Hacker &lt;jhacker@foo.com&gt;
 225     Acked-by: Joe Hacker &lt;jhacker@foo.com&gt;
 226 </pre>
 227 </ul>
 228
 229
 230
 231 <h3>Testing Patches</h3>
 232
 233 <p>
 234 It should go without saying that patches must be tested.  In general,
 235 do whatever testing is prudent.
 236 </p>
 237
 238 <p>
 239 You should always run the Mesa test suite before submitting patches.
 240 The test suite can be run using the 'make check' command. All tests
 241 must pass before patches will be accepted, this may mean you have
 242 to update the tests themselves.
 243 </p>
 244
 245 <p>
 246 Whenever possible and applicable, test the patch with
 247 <a href="http://people.freedesktop.org/~nh/piglit/">Piglit</a> to
 248 check for regressions.
 249 </p>
 250
 251
 252 <h3>Mailing Patches</h3>
 253
 254 <p>
 255 Patches should be sent to the Mesa mailing list for review.
 256 When submitting a patch make sure to use git send-email rather than attaching
 257 patches to emails. Sending patches as attachments prevents people from being
 258 able to provide in-line review comments.
 259 </p>
 260
 261 <p>
 262 When submitting follow-up patches you can use --in-reply-to to make v2, v3,
 263 etc patches show up as replies to the originals. This usually works well
 264 when you're sending out updates to individual patches (as opposed to
 265 re-sending the whole series). Using --in-reply-to makes
 266 it harder for reviewers to accidentally review old patches.
 267 </p>
 268
 269 <h3>Reviewing Patches</h3>
 270
 271 <p>
 272 When you've reviewed a patch on the mailing list, please be unambiguous
 273 about your review.  That is, state either
 274 <pre>
 275     Reviewed-by: Joe Hacker &lt;jhacker@foo.com&gt;
 276 </pre>
 277 or
 278 <pre>
 279     Acked-by: Joe Hacker &lt;jhacker@foo.com&gt;
 280 </pre>
 281 Rather than saying just "LGTM" or "Seems OK".
 282 </p>
 283
 284 <p>
 285 If small changes are suggested, it's OK to say something like:
 286 <pre>
 287    With the above fixes, Reviewed-by: Joe Hacker &lt;jhacker@foo.com&gt;
 288 </pre>
 289 which tells the patch author that the patch can be committed, as long
 290 as the issues are resolved first.
 291 </p>
 292
 293
 294 <h3>Marking a commit as a candidate for a stable branch</h3>
 295
 296 <p>
 297 If you want a commit to be applied to a stable branch,
 298 you should add an appropriate note to the commit message.
 299 </p>
 300
 301 <p>
 302 Here are some examples of such a note:
 303 </p>
 304 <ul>
 305   <li>CC: &lt;mesa-stable@lists.freedesktop.org&gt;</li>
 306   <li>CC: "9.2 10.0" &lt;mesa-stable@lists.freedesktop.org&gt;</li>
 307   <li>CC: "10.0" &lt;mesa-stable@lists.freedesktop.org&gt;</li>
 308 </ul>
 309
 310 Simply adding the CC to the mesa-stable list address is adequate to nominate
 311 the commit for the most-recently-created stable branch. It is only necessary
 312 to specify a specific branch name, (such as "9.2 10.0" or "10.0" in the
 313 examples above), if you want to nominate the commit for an older stable
 314 branch. And, as in these examples, you can nominate the commit for the older
 315 branch in addition to the more recent branch, or nominate the commit
 316 exclusively for the older branch.
 317
 318 This "CC" syntax for patch nomination will cause patches to automatically be
 319 copied to the mesa-stable@ mailing list when you use "git send-email" to send
 320 patches to the mesa-dev@ mailing list. Also, if you realize that a commit
 321 should be nominated for the stable branch after it has already been committed,
 322 you can send a note directly to the mesa-stable@lists.freedesktop.org where
 323 the Mesa stable-branch maintainers will receive it. Be sure to mention the
 324 commit ID of the commit of interest (as it appears in the mesa master branch).
 325
 326 The latest set of patches that have been nominated, accepted, or rejected for
 327 the upcoming stable release can always be seen on the
 328 <a href="http://cworth.org/~cworth/mesa-stable-queue/">Mesa Stable Queue</a>
 329 page.
 330
 331 <h3>Criteria for accepting patches to the stable branch</h3>
 332
 333 Mesa has a designated release manager for each stable branch, and the release
 334 manager is the only developer that should be pushing changes to these
 335 branches. Everyone else should simply nominate patches using the mechanism
 336 described above.
 337
 338 The stable-release manager will work with the list of nominated patches, and
 339 for each patch that meets the crtieria below will cherry-pick the patch with:
 340 <code>git cherry-pick -x &lt;commit&gt;</code>. The <code>-x</code> option is
 341 important so that the picked patch references the comit ID of the original
 342 patch.
 343
 344 The stable-release manager may at times need to force-push changes to the
 345 stable branches, for example, to drop a previously-picked patch that was later
 346 identified as causing a regression). These force-pushes may cause changes to
 347 be lost from the stable branch if developers push things directly. Consider
 348 yourself warned.
 349
 350 The stable-release manager is also given broad discretion in rejecting patches
 351 that have been nominated for the stable branch. The most basic rule is that
 352 the stable branch is for bug fixes only, (no new features, no
 353 regressions). Here is a non-exhaustive list of some reasons that a patch may
 354 be rejected:
 355
 356 <ul>
 357   <li>Patch introduces a regression. Any reported build breakage or other
 358   regression caused by a particular patch, (game no longer work, piglit test
 359   changes from PASS to FAIL), is justification for rejecting a patch.</li>
 360
 361   <li>Patch is too large, (say, larger than 100 lines)</li>
 362
 363   <li>Patch is not a fix. For example, a commit that moves code around with no
 364   functional change should be rejected.</li>
 365
 366   <li>Patch fix is not clearly described. For example, a commit message
 367   of only a single line, no description of the bug, no mention of bugzilla,
 368   etc.</li>
 369
 370   <li>Patch has not obviously been reviewed, For example, the commit message
 371   has no Reviewed-by, Signed-off-by, nor Tested-by tags from anyone but the
 372   author.</li>
 373
 374   <li>Patch has not already been merged to the master branch. As a rule, bug
 375   fixes should never be applied first to a stable branch. Patches should land
 376   first on the master branch and then be cherry-picked to a stable
 377   branch. (This is to avoid future releases causing regressions if the patch
 378   is not also applied to master.) The only things that might look like
 379   exceptions would be backports of patches from master that happen to look
 380   significantly different.</li>
 381
 382   <li>Patch depends on too many other patches. Ideally, all stable-branch
 383   patches should be self-contained. It sometimes occurs that a single, logical
 384   bug-fix occurs as two separate patches on master, (such as an original
 385   patch, then a subsequent fix-up to that patch). In such a case, these two
 386   patches should be squashed into a single, self-contained patch for the
 387   stable branch. (Of course, if the squashing makes the patch too large, then
 388   that could be a reason to reject the patch.)</li>
 389
 390   <li>Patch includes new feature development, not bug fixes. New OpenGL
 391   features, extensions, etc. should be applied to Mesa master and included in
 392   the next major release. Stable releases are intended only for bug fixes.
 393
 394   Note: As an exception to this rule, the stable-release manager may accept
 395   hardware-enabling "features". For example, backports of new code to support
 396   a newly-developed hardware product can be accepted if they can be reasonably
 397   determined to not have effects on other hardware.</li>
 398
 399   <li>Patch is a performance optimization. As a rule, performance patches are
 400   not candidates for the stable branch. The only exception might be a case
 401   where an application's performance was recently severely impacted so as to
 402   become unusable. The fix for this performance regression could then be
 403   considered for a stable branch. The optimization must also be
 404   non-controversial and the patches still need to meet the other criteria of
 405   being simple and self-contained</li>
 406
 407   <li>Patch introduces a new failure mode (such as an assert). While the new
 408   assert might technically be correct, for example to make Mesa more
 409   conformant, this is not the kind of "bug fix" we want in a stable
 410   release. The potential problem here is that an OpenGL program that was
 411   previously working, (even if technically non-compliant with the
 412   specification), could stop working after this patch. So that would be a
 413   regression that is unaacceptable for the stable branch.</li>
 414 </ul>
 415
 416
 417 <h2 id="release">Making a New Mesa Release</h2>
 418
 419 <p>
 420 These are the instructions for making a new Mesa release.
 421 </p>
 422
 423 <h3>Get latest source files</h3>
 424 <p>
 425 Use git to get the latest Mesa files from the git repository, from whatever
 426 branch is relevant. This document uses the convention X.Y.Z for the release
 427 being created, which should be created from a branch named X.Y.
 428 </p>
 429
 430 <h3>Perform basic testing</h3>
 431 <p>
 432 The release manager should, at the very least, test the code by compiling it,
 433 installing it, and running the latest piglit to ensure that no piglit tests
 434 have regressed since the previous release.
 435 </p>
 436
 437 <p>
 438 The release manager should do this testing with at least one hardware driver,
 439 (say, whatever is contained in the local development machine), as well as on
 440 both Gallium and non-Gallium software drivers. The software testing can be
 441 performed by running piglit with the following environment-variable set:
 442 </p>
 443
 444 <pre>
 445 LIBGL_ALWAYS_SOFTWARE=1
 446 </pre>
 447
 448 And Gallium vs. non-Gallium software drivers can be obtained by using the
 449 following configure flags on separate builds:
 450
 451 <pre>
 452 --with-dri-drivers=swrast
 453 --with-gallium-drivers=swrast
 454 </pre>
 455
 456 <p>
 457 Note: If both options are given in one build, both swrast_dri.so drivers will
 458 be compiled, but only one will be installed. The following command can be used
 459 to ensure the correct driver is being tested:
 460 </p>
 461
 462 <pre>
 463 LIBGL_ALWAYS_SOFTWARE=1 glxinfo | grep "renderer string"
 464 </pre>
 465
 466 If any regressions are found in this testing with piglit, stop here, and do
 467 not perform a release until regressions are fixed.
 468
 469 <h3>Update version in file VERSION</h3>
 470
 471 <p>
 472 Increment the version contained in the file VERSION at Mesa's top-level, then
 473 commit this change.
 474 </p>
 475
 476 <h3>Create release notes for the new release</h3>
 477
 478 <p>
 479 Create a new file docs/relnotes/X.Y.Z.html, (follow the style of the previous
 480 release notes). Note that the sha256sums section of the release notes should
 481 be empty at this point.
 482 </p>
 483
 484 <p>
 485 Two scripts are available to help generate portions of the release notes:
 486
 487 <pre>
 488         ./bin/bugzilla_mesa.sh
 489         ./bin/shortlog_mesa.sh
 490 </pre>
 491
 492 <p>
 493 The first script identifies commits that reference bugzilla bugs and obtains
 494 the descriptions of those bugs from bugzilla. The second script generates a
 495 log of all commits. In both cases, HTML-formatted lists are printed to stdout
 496 to be included in the release notes.
 497 </p>
 498
 499 <p>
 500 Commit these changes
 501 </p>
 502
 503 <h3>Make the release archives, signatures, and the release tag</h3>
 504 <p>
 505 From inside the Mesa directory:
 506 <pre>
 507         ./autogen.sh
 508         make -j1 tarballs
 509 </pre>
 510
 511 <p>
 512 After the tarballs are created, the sha256 checksums for the files will
 513 be computed and printed. These will be used in a step below.
 514 </p>
 515
 516 <p>
 517 It's important at this point to also verify that the constructed tar file
 518 actually builds:
 519 </p>
 520
 521 <pre>
 522         tar xjf MesaLib-X.Y.Z.tar.bz2
 523         cd Mesa-X.Y.Z
 524         ./configure --enable-gallium-llvm
 525         make -j6
 526         make install
 527 </pre>
 528
 529 <p>
 530 Some touch testing should also be performed at this point, (run glxgears or
 531 more involved OpenGL programs against the installed Mesa).
 532 </p>
 533
 534 <p>
 535 Create detached GPG signatures for each of the archive files created above:
 536 </p>
 537
 538 <pre>
 539         gpg --sign --detach MesaLib-X.Y.Z.tar.gz
 540         gpg --sign --detach MesaLib-X.Y.Z.tar.bz2
 541         gpg --sign --detach MesaLib-X.Y.Z.zip
 542 </pre>
 543
 544 <p>
 545 Tag the commit used for the build:
 546 </p>
 547
 548 <pre>
 549         git tag -s mesa-X.Y.X -m "Mesa X.Y.Z release"
 550 </pre>
 551
 552 <p>
 553 Note: It would be nice to investigate and fix the issue that causes the
 554 tarballs target to fail with multiple build process, such as with "-j4". It
 555 would also be nice to incorporate all of the above commands into a single
 556 makefile target. And instead of a custom "tarballs" target, we should
 557 incorporate things into the standard "make dist" and "make distcheck" targets.
 558 </p>
 559
 560 <h3>Add the sha256sums to the release notes</h3>
 561
 562 <p>
 563 Edit docs/relnotes/X.Y.Z.html to add the sha256sums printed as part of "make
 564 tarballs" in the previous step. Commit this change.
 565 </p>
 566
 567 <h3>Push all commits and the tag creates above</h3>
 568
 569 <p>
 570 This is the first step that cannot easily be undone. The release is going
 571 forward from this point:
 572 </p>
 573
 574 <pre>
 575         git push origin X.Y --tags
 576 </pre>
 577
 578 <h3>Install the release files and signatures on the distribution server</h3>
 579
 580 <p>
 581 The following commands can be used to copy the release archive files and
 582 signatures to the freedesktop.org server:
 583 </p>
 584
 585 <pre>
 586         scp MesaLib-X.Y.Z* people.freedesktop.org:
 587         ssh people.freedesktop.org
 588         cd /srv/ftp.freedesktop.org/pub/mesa
 589         mkdir X.Y.Z
 590         cd X.Y.Z
 591         mv ~/MesaLib-X.Y.Z* .
 592 </pre>
 593
 594 <h3>Back on mesa master, andd the new release notes into the tree</h3>
 595
 596 <p>
 597 Something like the following steps will do the trick:
 598 </p>
 599
 600 <pre>
 601         cp docs/relnotes/X.Y.Z.html /tmp
 602         git checkout master
 603         cp /tmp/X.Y.Z.html docs/relnotes
 604         git add docs/relnotes/X.Y.Z.html
 605 </pre>
 606
 607 <p>
 608 Also, edit docs/relnotes.html to add a link to the new release notes, and edit
 609 docs/index.html to add a news entry. Then commit and push:
 610 </p>
 611
 612 <pre>
 613         git commit -a -m "docs: Import X.Y.Z release notes, add news item."
 614         git push origin
 615 </pre>
 616
 617 <h3>Update the mesa3d.org website</h3>
 618
 619 <p>
 620 NOTE: The recent release managers have not been performing this step
 621 themselves, but leaving this to Brian Paul, (who has access to the
 622 sourceforge.net hosting for mesa3d.org). Brian is more than willing to grant
 623 the permission necessary to future release managers to do this step on their
 624 own.
 625 </p>
 626
 627 <p>
 628 Update the web site by copying the docs/ directory's files to
 629 /home/users/b/br/brianp/mesa-www/htdocs/ with:
 630 <br>
 631 <code>
 632 sftp USERNAME,mesa3d@web.sourceforge.net
 633 </code>
 634 </p>
 635
 636
 637 <h3>Announce the release</h3>
 638 <p>
 639 Make an announcement on the mailing lists:
 640
 641 <em>mesa-dev@lists.freedesktop.org</em>,
 642 and
 643 <em>mesa-announce@lists.freedesktop.org</em>
 644
 645 Follow the template of previously-sent release announcements. The following
 646 command can be used to generate the log of changes to be included in the
 647 release announcement:
 648
 649 <pre>
 650         git shortlog mesa-X.Y.Z-1..mesa-X.Y.Z
 651 </pre>
 652 </p>
 653
 654
 655 <h2 id="extensions">Adding Extensions</h2>
 656
 657 <p>
 658 To add a new GL extension to Mesa you have to do at least the following.
 659
 660 <ul>
 661 <li>
 662    If glext.h doesn't define the extension, edit include/GL/gl.h and add
 663    code like this:
 664    <pre>
 665      #ifndef GL_EXT_the_extension_name
 666      #define GL_EXT_the_extension_name 1
 667      /* declare the new enum tokens */
 668      /* prototype the new functions */
 669      /* TYPEDEFS for the new functions */
 670      #endif
 671    </pre>
 672 </li>
 673 <li>
 674    In the src/mapi/glapi/gen/ directory, add the new extension functions and
 675    enums to the gl_API.xml file.
 676    Then, a bunch of source files must be regenerated by executing the
 677    corresponding Python scripts.
 678 </li>
 679 <li>
 680    Add a new entry to the <code>gl_extensions</code> struct in mtypes.h
 681 </li>
 682 <li>
 683    Update the <code>extensions.c</code> file.
 684 </li>
 685 <li>
 686    From this point, the best way to proceed is to find another extension,
 687    similar to the new one, that's already implemented in Mesa and use it
 688    as an example.
 689 </li>
 690 <li>
 691    If the new extension adds new GL state, the functions in get.c, enable.c
 692    and attrib.c will most likely require new code.
 693 </li>
 694 <li>
 695    The dispatch tests check_table.cpp and dispatch_sanity.cpp
 696    should be updated with details about the new extensions functions. These
 697    tests are run using 'make check'
 698 </li>
 699 </ul>
 700
 701
 702
 703
 704 </div>
 705 </body>
 706 </html>