/usr/share/doc/python-pyopencl-doc/html/misc.html is in python-pyopencl-doc 2017.2.2-1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>Installation — PyOpenCL 2017.2.2 documentation</title>
<link rel="stylesheet" href="_static/alabaster.css" type="text/css" />
<link rel="stylesheet" href="_static/pygments.css" type="text/css" />
<script type="text/javascript">
var DOCUMENTATION_OPTIONS = {
URL_ROOT: './',
VERSION: '2017.2.2',
COLLAPSE_INDEX: false,
FILE_SUFFIX: '.html',
HAS_SOURCE: false,
SOURCELINK_SUFFIX: '.txt'
};
</script>
<script type="text/javascript" src="_static/jquery.js"></script>
<script type="text/javascript" src="_static/underscore.js"></script>
<script type="text/javascript" src="_static/doctools.js"></script>
<link rel="index" title="Index" href="genindex.html" />
<link rel="search" title="Search" href="search.html" />
<link rel="prev" title="Built-in Utilities" href="tools.html" />
<link rel="stylesheet" href="_static/custom.css" type="text/css" />
<meta name="viewport" content="width=device-width, initial-scale=0.9, maximum-scale=0.9" />
</head>
<body>
<div class="document">
<div class="documentwrapper">
<div class="bodywrapper">
<div class="body" role="main">
<div class="section" id="installation">
<h1>Installation<a class="headerlink" href="#installation" title="Permalink to this headline">¶</a></h1>
<p>By far the easiest way to install PyOpenCL is to use the packages available in
<a class="reference external" href="https://conda-forge.org/">Conda Forge</a>. Conda Forge is a repository of
community-maintained packages for the <a class="reference external" href="https://conda.io/docs/">Conda</a>
package manager.</p>
<p>On Linux and OS X, the following set of instructions should work:</p>
<ol class="arabic">
<li><p class="first">Install a version of <a class="reference external" href="https://conda.io/miniconda.html">miniconda</a>
that fits your system. Both Python 2 and Python 3 work.
You can install these pieces of software in your user account and
do not need root/administrator privileges.</p>
<p>Note that if you already have Continuum Anaconda installed on your system,
you may just use that and do <em>not</em> need to install Miniconda.</p>
</li>
<li><p class="first"><code class="docutils literal"><span class="pre">source</span> <span class="pre">/WHERE/YOU/INSTALLED/MINICONDA/bin/activate</span> <span class="pre">root</span></code></p>
</li>
<li><p class="first"><code class="docutils literal"><span class="pre">conda</span> <span class="pre">config</span> <span class="pre">--add</span> <span class="pre">channels</span> <span class="pre">conda-forge</span></code></p>
</li>
<li><p class="first"><code class="docutils literal"><span class="pre">conda</span> <span class="pre">install</span> <span class="pre">pyopencl</span></code></p>
</li>
</ol>
<p>The analogous steps for Windows should also work.</p>
<p>Note that PyOpenCL is no fun (i.e. cannot run code) without an OpenCL device
driver (a so-called “ICD”, for “installable client driver”) that provides access
to hardware through OpenCL. If you get an error message like
<code class="docutils literal"><span class="pre">pyopencl.cffi_cl.LogicError:</span> <span class="pre">clGetPlatformIDs</span> <span class="pre">failed:</span> <span class="pre"><unknown</span> <span class="pre">error</span>
<span class="pre">-1001></span></code>, that means you have no OpenCL drivers installed.</p>
<p>Note that drivers (ICDs) are separate pieces of software from PyOpenCL. They
might be provided by your hardware vendor (e.g. for Nvidia or AMD GPUs). If
you have such hardware, see below for instructions on how to make those work
with PyOpenCL from Conda Forge.</p>
<p>It is important to note that OpenCL is not restricted to GPUs. In fact, no special
hardware is required to use OpenCL for computation–your existing CPU is enough.
On Linux, type:</p>
<ol class="arabic simple">
<li><code class="docutils literal"><span class="pre">conda</span> <span class="pre">install</span> <span class="pre">pocl</span></code></li>
</ol>
<p>to install a CPU-based OpenCL driver. On Windows, you may install e.g.
the <a class="reference external" href="https://software.intel.com/en-us/articles/opencl-drivers#latest_CPU_runtime">CPU OpenCL driver from Intel</a>.
OS X has support for OpenCL built into the operating system and does not need
additional software to run code based on PyOpenCL (but see below).</p>
<p>You are now ready to run code based on PyOpenCL, such as the <a class="reference external" href="https://github.com/inducer/pyopencl/tree/master/examples">code
examples</a>.</p>
<div class="section" id="using-vendor-supplied-opencl-drivers-linux">
<h2>Using vendor-supplied OpenCL drivers (Linux)<a class="headerlink" href="#using-vendor-supplied-opencl-drivers-linux" title="Permalink to this headline">¶</a></h2>
<p>The instructions above help you get a basic OpenCL environment going that
will work independently of whether you have specialized hardware (such as GPUs
or FPGAs) available. If you <em>do</em> have such hardware, read on for how to make
it work.</p>
<p>On Linux, PyOpenCL finds which drivers are installed by looking for files with
the extension <code class="docutils literal"><span class="pre">.icd</span></code> in a directory. PyOpenCL as installed from Conda will
look for these files in
<code class="file docutils literal"><span class="pre">/WHERE/YOU/INSTALLED/MINICONDA/etc/OpenCL/vendors</span></code>. They are just
simple text files containing either just the file names or the fully
qualified path names of the shared library providing the OpenCL driver.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p>If you ran the commands above in a
<a class="reference external" href="https://conda.io/docs/user-guide/tasks/manage-environments.html">Conda environment</a>
(i.e. if the environment indicator on your command line prompt says anything other
than <code class="docutils literal"><span class="pre">(root)</span></code>), then you may need to use a path like the following instead:</p>
<p><code class="file docutils literal"><span class="pre">/WHERE/YOU/INSTALLED/MINICONDA/envs/ENVIRONMENTNAME/etc/OpenCL/vendors</span></code></p>
<p class="last">Note that you should replace <code class="docutils literal"><span class="pre">ENVIRONMENTNAME</span></code> with the name of your environment,
shown between parentheses on your command line prompt.</p>
</div>
<p>If you have other OpenCL drivers installed (such as for your GPU), those will be
in <code class="file docutils literal"><span class="pre">/etc/OpenCL/vendors</span></code>. You can make them work with PyOpenCL from Conda Forge
by simply copying them to the above folder.</p>
<p>If you are looking for more information, see <a class="reference external" href="https://github.com/OCL-dev/ocl-icd">ocl-icd</a> and its documentation. Ocl-icd is the
“ICD loader” used by PyOpenCL when installed from Conda Forge. It represents the
code behind <code class="file docutils literal"><span class="pre">libOpenCL.so</span></code>.</p>
</div>
<div class="section" id="getting-a-better-cpu-based-opencl-driver-os-x">
<h2>Getting a better CPU-based OpenCL driver (OS X)<a class="headerlink" href="#getting-a-better-cpu-based-opencl-driver-os-x" title="Permalink to this headline">¶</a></h2>
<p>OS X has support for both CPU- and GPU-based OpenCL built in. Unfortunately,
the built-in drivers can be temperamental, and they have not advanced as quickly
as one might like. To make PyOpenCL use a more up-to-date (and open-source)
CPU-based OpenCL driver, type the following:</p>
<p><code class="docutils literal"><span class="pre">conda</span> <span class="pre">install</span> <span class="pre">osx-pocl-opencl</span> <span class="pre">pocl</span></code> (OS X)</p>
<p>Note that, by installing <code class="docutils literal"><span class="pre">osx-pocl-opencl</span></code>, you will no longer be able to
use PyOpenCL to talk to the system-wide Apple OpenCL drivers. To regain access
to those drivers, simply uninstall <code class="docutils literal"><span class="pre">osx-pocl-opencl</span></code> and reinstall <code class="docutils literal"><span class="pre">pyopencl</span></code>
afterwards.</p>
</div>
<div class="section" id="installing-from-source">
<h2>Installing from source<a class="headerlink" href="#installing-from-source" title="Permalink to this headline">¶</a></h2>
<p>Information on how to install PyOpenCL <em>from source</em> is maintained collaboratively on the
<a class="reference external" href="http://wiki.tiker.net/PyOpenCL/Installation">PyOpenCL Wiki</a>, but that should
mostly not be necessary unless you have very specific needs or would like to modify
PyOpenCL yourself.</p>
</div>
</div>
<div class="section" id="tips">
<h1>Tips<a class="headerlink" href="#tips" title="Permalink to this headline">¶</a></h1>
<div class="section" id="syntax-highlighting">
<h2>Syntax highlighting<a class="headerlink" href="#syntax-highlighting" title="Permalink to this headline">¶</a></h2>
<p>You can obtain Vim syntax highlighting for OpenCL C inlined in Python by
checking <a class="reference external" href="https://github.com/inducer/pyopencl/blob/master/contrib/pyopencl.vim">this file</a>.</p>
<p>Note that the triple-quoted strings containing the source must start with
<cite>“”“//CL// …”“”</cite>.</p>
</div>
<div class="section" id="ipython-integration">
<h2>IPython integration<a class="headerlink" href="#ipython-integration" title="Permalink to this headline">¶</a></h2>
<p>PyOpenCL comes with IPython integration, which lets you seamlessly integrate
PyOpenCL kernels into your IPython notebooks. Simply load the PyOpenCL
IPython extension using:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">%</span><span class="n">load_ext</span> <span class="n">pyopencl</span><span class="o">.</span><span class="n">ipython_ext</span>
</pre></div>
</div>
<p>and then use the <code class="docutils literal"><span class="pre">%%cl_kernel</span></code> ‘cell-magic’ command. See <a class="reference external" href="http://nbviewer.ipython.org/urls/raw.githubusercontent.com/inducer/pyopencl/master/examples/ipython-demo.ipynb">this notebook</a>
(which ships with PyOpenCL) for a demonstration.</p>
<p>You can pass build options to be used for building the program executable by using the <code class="docutils literal"><span class="pre">-o</span></code> flag on the first line of the cell (next to the <code class="docutils literal"><span class="pre">%%cl_kernel</span></code> directive). For example: <cite>%%cl_kernel -o “-cl-fast-relaxed-math”`</cite>.</p>
<p>There are also line magics: <code class="docutils literal"><span class="pre">cl_load_edit_kernel</span></code> which will load a file into the next cell (adding <code class="docutils literal"><span class="pre">cl_kernel</span></code> to the first line) and <code class="docutils literal"><span class="pre">cl_kernel_from_file</span></code> which will compile kernels from a file (as if you copy-and-pasted the contents of the file to a cell with <code class="docutils literal"><span class="pre">cl_kernel</span></code>). Both of these magics take options <code class="docutils literal"><span class="pre">-f</span></code> to specify the file and optionally <code class="docutils literal"><span class="pre">-o</span></code> for build options.</p>
<div class="versionadded">
<p><span class="versionmodified">New in version 2014.1.</span></p>
</div>
</div>
</div>
<div class="section" id="guidelines">
<h1>Guidelines<a class="headerlink" href="#guidelines" title="Permalink to this headline">¶</a></h1>
<div class="section" id="api-stability">
<span id="api-compatibility"></span><h2>API Stability<a class="headerlink" href="#api-stability" title="Permalink to this headline">¶</a></h2>
<p>I consider PyOpenCL’s API “stable”. That doesn’t mean it can’t
change. But if it does, your code will generally continue to run. It
may however start spewing warnings about things you need to change to
stay compatible with future versions.</p>
<p>Deprecation warnings will be around for a whole year, as identified by the
first number in the release name. (the “2014” in “2014.1”) I.e. a function
that was deprecated in 2014.n will generally be removed in 2015.n (or perhaps
later). Further, the stability promise applies for any code that’s part of a
released version. It doesn’t apply to undocumented bits of the API, and it
doesn’t apply to unreleased code downloaded from git.</p>
</div>
<div class="section" id="relation-with-opencl-s-c-bindings">
<span id="versus-c"></span><h2>Relation with OpenCL’s C Bindings<a class="headerlink" href="#relation-with-opencl-s-c-bindings" title="Permalink to this headline">¶</a></h2>
<p>We’ve tried to follow these guidelines when binding the OpenCL’s
C interface to Python:</p>
<ul class="simple">
<li>Remove the <cite>cl_</cite>, <cite>CL_</cite> and <cite>cl</cite> prefix from data types, macros and
function names.</li>
<li>Follow <span class="target" id="index-0"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0008"><strong>PEP 8</strong></a>, i.e.<ul>
<li>Make function names lowercase.</li>
<li>If a data type or function name is composed of more than one word,
separate the words with a single underscore.</li>
</ul>
</li>
<li><cite>get_info</cite> functions become attributes.</li>
<li>Object creation is done by constructors, to the extent possible.
(i.e. minimize use of “factory functions”)</li>
<li>If an operation involves two or more “complex” objects (like e.g. a
kernel enqueue involves a kernel and a queue), refuse the temptation
to guess which one should get a method for the operation.
Instead, simply leave that command to be a function.</li>
</ul>
</div>
<div class="section" id="interoperability-with-other-opencl-software">
<span id="interoperability"></span><h2>Interoperability with other OpenCL software<a class="headerlink" href="#interoperability-with-other-opencl-software" title="Permalink to this headline">¶</a></h2>
<p>Just about every object in <a class="reference internal" href="runtime.html#module-pyopencl" title="pyopencl"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl</span></code></a> supports the following
interface (here shown as an example for <a class="reference internal" href="runtime_memory.html#pyopencl.MemoryObject" title="pyopencl.MemoryObject"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.MemoryObject</span></code></a>,
from which <a class="reference internal" href="runtime_memory.html#pyopencl.Buffer" title="pyopencl.Buffer"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Buffer</span></code></a> and <code class="xref py py-class docutils literal"><span class="pre">pyopencl.Image</span></code> inherit):</p>
<ul class="simple">
<li><a class="reference internal" href="runtime_memory.html#pyopencl.MemoryObject.from_int_ptr" title="pyopencl.MemoryObject.from_int_ptr"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.MemoryObject.from_int_ptr()</span></code></a></li>
<li><a class="reference internal" href="runtime_memory.html#pyopencl.MemoryObject.int_ptr" title="pyopencl.MemoryObject.int_ptr"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.MemoryObject.int_ptr</span></code></a></li>
</ul>
<p>This allows retrieving the C-level pointer to an OpenCL object as a Python
integer, which may then be passed to other C libraries whose interfaces expose
OpenCL objects. It also allows turning C-level OpenCL objects obtained from
other software to be turned into the corresponding <a class="reference internal" href="runtime.html#module-pyopencl" title="pyopencl"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl</span></code></a> objects.</p>
<div class="versionadded">
<p><span class="versionmodified">New in version 2013.2.</span></p>
</div>
</div>
</div>
<div class="section" id="user-visible-changes">
<h1>User-visible Changes<a class="headerlink" href="#user-visible-changes" title="Permalink to this headline">¶</a></h1>
<div class="section" id="version-2016-3">
<h2>Version 2016.3<a class="headerlink" href="#version-2016-3" title="Permalink to this headline">¶</a></h2>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">This version is currently under development. You can get snapshots from
PyOpenCL’s <a class="reference external" href="https://github.com/inducer/pyopencl">git repository</a></p>
</div>
</div>
<div class="section" id="version-2016-2">
<h2>Version 2016.2<a class="headerlink" href="#version-2016-2" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Deprecate RANLUXCL. It will be removed in the 2018.x series of PyOpenCL.</li>
<li>Introduce Random123 random number generators. See <a class="reference internal" href="array.html#module-pyopencl.clrandom" title="pyopencl.clrandom"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl.clrandom</span></code></a>
for more information.</li>
<li>Add support for <strong>range</strong> and <strong>slice</strong> kwargs and data-less reductions
to <a class="reference internal" href="algorithm.html#pyopencl.reduction.ReductionKernel" title="pyopencl.reduction.ReductionKernel"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.reduction.ReductionKernel</span></code></a>.</li>
<li>Add support for SPIR-V. (See <a class="reference internal" href="runtime_program.html#pyopencl.Program" title="pyopencl.Program"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Program</span></code></a>.)</li>
<li>Add support for <a class="reference internal" href="runtime_memory.html#svm"><span class="std std-ref">Shared Virtual Memory (SVM)</span></a>.</li>
<li><a class="reference internal" href="runtime_memory.html#pyopencl.MemoryMap" title="pyopencl.MemoryMap"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.MemoryMap</span></code></a> is usable as a context manager.</li>
</ul>
</div>
<div class="section" id="version-2016-1">
<h2>Version 2016.1<a class="headerlink" href="#version-2016-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>The <code class="docutils literal"><span class="pre">from_int_ptr</span></code> methods now take a <em>retain</em> parameter for more convenient
ownership management.</li>
<li>Kernel build options (if passed as a list) are now properly quoted.
(This is a potentially compatibility-breaking change.)</li>
<li>Many bug fixes. (GL interop, Windows, event callbacks and more)</li>
</ul>
</div>
<div class="section" id="version-2015-2-4">
<h2>Version 2015.2.4<a class="headerlink" href="#version-2015-2-4" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fix building on Windows, using mingwpy and VS 2015.</li>
</ul>
</div>
<div class="section" id="version-2015-2-3">
<h2>Version 2015.2.3<a class="headerlink" href="#version-2015-2-3" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fix one more Ubuntu 14.x build issue.</li>
</ul>
</div>
<div class="section" id="version-2015-2-2">
<h2>Version 2015.2.2<a class="headerlink" href="#version-2015-2-2" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fix compatibility with CL 1.1</li>
<li>Fix compatibility with Ubuntu 14.x.</li>
<li>Various bug fixes</li>
</ul>
</div>
<div class="section" id="version-2015-2-1">
<h2>Version 2015.2.1<a class="headerlink" href="#version-2015-2-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fix global_offset kernel launch parameter</li>
</ul>
</div>
<div class="section" id="version-2015-2">
<h2>Version 2015.2<a class="headerlink" href="#version-2015-2" title="Permalink to this headline">¶</a></h2>
<ul>
<li><p class="first"><strong>[INCOMPATIBLE]</strong> Changed PyOpenCL’s complex numbers from <code class="docutils literal"><span class="pre">float2</span></code> and
<code class="docutils literal"><span class="pre">double2</span></code> OpenCL vector types to custom <code class="docutils literal"><span class="pre">struct</span></code>. This was changed
because it very easily introduced bugs where</p>
<ul class="simple">
<li>complex*complex</li>
<li>real+complex</li>
</ul>
<p><em>look</em> like they may do the right thing, but silently do the wrong thing.</p>
</li>
<li><p class="first">Rewrite of the wrapper layer to be based on CFFI</p>
</li>
<li><p class="first">Pypy compatibility</p>
</li>
<li><p class="first">Faster kernel invocation through Python launcher code generation</p>
</li>
<li><p class="first">POCL compatibility</p>
</li>
</ul>
</div>
<div class="section" id="version-2015-1">
<h2>Version 2015.1<a class="headerlink" href="#version-2015-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Support for new-style buffer protocol</li>
<li>Numerous fixes</li>
</ul>
</div>
<div class="section" id="version-2014-1">
<h2>Version 2014.1<a class="headerlink" href="#version-2014-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li><span class="xref std std-ref">ipython-integration</span></li>
<li>Bug fixes</li>
</ul>
</div>
<div class="section" id="version-2013-2">
<h2>Version 2013.2<a class="headerlink" href="#version-2013-2" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Add <a class="reference internal" href="array.html#pyopencl.array.Array.map_to_host" title="pyopencl.array.Array.map_to_host"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.array.Array.map_to_host()</span></code></a>.</li>
<li>Support <em>strides</em> on <a class="reference internal" href="runtime_memory.html#pyopencl.enqueue_map_buffer" title="pyopencl.enqueue_map_buffer"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_map_buffer()</span></code></a> and
<a class="reference internal" href="runtime_memory.html#pyopencl.enqueue_map_image" title="pyopencl.enqueue_map_image"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_map_image()</span></code></a>.</li>
<li><a class="reference internal" href="runtime_memory.html#pyopencl.ImageFormat" title="pyopencl.ImageFormat"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.ImageFormat</span></code></a> was made comparable and hashable.</li>
<li><a class="reference internal" href="algorithm.html#module-pyopencl.reduction" title="pyopencl.reduction"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl.reduction</span></code></a> supports slicing (contributed by Alex Nitz)</li>
<li>Added <a class="reference internal" href="#interoperability"><span class="std std-ref">Interoperability with other OpenCL software</span></a></li>
<li>Bug fixes</li>
</ul>
</div>
<div class="section" id="version-2013-1">
<h2>Version 2013.1<a class="headerlink" href="#version-2013-1" title="Permalink to this headline">¶</a></h2>
<ul>
<li><p class="first">Vastly improved <a class="reference internal" href="algorithm.html#custom-scan"><span class="std std-ref">Prefix Sums (“scan”)</span></a>.</p>
</li>
<li><p class="first">Add <a class="reference internal" href="array.html#pyopencl.tools.match_dtype_to_c_struct" title="pyopencl.tools.match_dtype_to_c_struct"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.tools.match_dtype_to_c_struct()</span></code></a>,
for better integration of the CL and <a class="reference external" href="file:///usr/share/doc/python-numpy-doc/html/reference/index.html#module-numpy" title="(in NumPy v1.13)"><code class="xref py py-mod docutils literal"><span class="pre">numpy</span></code></a> type systems.</p>
</li>
<li><p class="first">More/improved Bessel functions.
See <a class="reference external" href="https://github.com/inducer/pyopencl/tree/master/src/cl">the source</a>.</p>
</li>
<li><p class="first">Add <span class="target" id="index-1"></span><code class="xref std std-envvar docutils literal"><span class="pre">PYOPENCL_NO_CACHE</span></code> environment variable to aid debugging.
(e.g. with AMD’s CPU implementation, see
<a class="reference external" href="http://developer.amd.com/sdks/AMDAPPSDK/assets/AMD_Accelerated_Parallel_Processing_OpenCL_Programming_Guide.pdf">their programming guide</a>)</p>
</li>
<li><p class="first">Deprecated <a class="reference internal" href="array.html#pyopencl.tools.register_dtype" title="pyopencl.tools.register_dtype"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.tools.register_dtype()</span></code></a> in favor of
<a class="reference internal" href="array.html#pyopencl.tools.get_or_register_dtype" title="pyopencl.tools.get_or_register_dtype"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.tools.get_or_register_dtype()</span></code></a>.</p>
</li>
<li><p class="first">Clean up the <a class="reference internal" href="array.html#pyopencl.array.Array" title="pyopencl.array.Array"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.array.Array</span></code></a> constructor interface.</p>
</li>
<li><p class="first">Deprecate <code class="xref py py-class docutils literal"><span class="pre">pyopencl.array.DefaultAllocator</span></code>.</p>
</li>
<li><p class="first">Deprecate <code class="xref py py-class docutils literal"><span class="pre">pyopencl.tools.CLAllocator</span></code>.</p>
</li>
<li><p class="first">Introduce <a class="reference internal" href="tools.html#pyopencl.tools.DeferredAllocator" title="pyopencl.tools.DeferredAllocator"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.tools.DeferredAllocator</span></code></a>, <a class="reference internal" href="tools.html#pyopencl.tools.ImmediateAllocator" title="pyopencl.tools.ImmediateAllocator"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.tools.ImmediateAllocator</span></code></a>.</p>
</li>
<li><p class="first">Allow arrays whose beginning does not coincide with the beginning of their
<a class="reference internal" href="array.html#pyopencl.array.Array.data" title="pyopencl.array.Array.data"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.array.Array.data</span></code></a> <a class="reference internal" href="runtime_memory.html#pyopencl.Buffer" title="pyopencl.Buffer"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Buffer</span></code></a>.
See <a class="reference internal" href="array.html#pyopencl.array.Array.base_data" title="pyopencl.array.Array.base_data"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.array.Array.base_data</span></code></a> and <a class="reference internal" href="array.html#pyopencl.array.Array.offset" title="pyopencl.array.Array.offset"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.array.Array.offset</span></code></a>.
Note that not all functions in PyOpenCL support such arrays just yet. These
will fail with <a class="reference internal" href="array.html#pyopencl.array.ArrayHasOffsetError" title="pyopencl.array.ArrayHasOffsetError"><code class="xref py py-exc docutils literal"><span class="pre">pyopencl.array.ArrayHasOffsetError</span></code></a>.</p>
</li>
<li><p class="first">Add <a class="reference internal" href="array.html#pyopencl.array.Array.__getitem__" title="pyopencl.array.Array.__getitem__"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.array.Array.__getitem__()</span></code></a> and <a class="reference internal" href="array.html#pyopencl.array.Array.__setitem__" title="pyopencl.array.Array.__setitem__"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.array.Array.__setitem__()</span></code></a>,
supporting generic slicing.</p>
<p>It is <em>possible</em> to create non-contiguous arrays using this functionality.
Most operations (elementwise etc.) will not work on such arrays.</p>
<p>Note also that some operations (specifically, reductions and scans) on sliced
arrays that start past the beginning of the original array will fail for now.
This will be fixed in a future release.</p>
</li>
<li><p class="first"><a class="reference internal" href="runtime_queue.html#pyopencl.CommandQueue" title="pyopencl.CommandQueue"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.CommandQueue</span></code></a> may be used as a context manager (in a <code class="docutils literal"><span class="pre">with</span></code> statement)</p>
</li>
<li><p class="first">Add <a class="reference internal" href="array.html#pyopencl.clmath.atan2" title="pyopencl.clmath.atan2"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.clmath.atan2()</span></code></a>, <a class="reference internal" href="array.html#pyopencl.clmath.atan2pi" title="pyopencl.clmath.atan2pi"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.clmath.atan2pi()</span></code></a>.</p>
</li>
<li><p class="first">Add <a class="reference internal" href="array.html#pyopencl.array.concatenate" title="pyopencl.array.concatenate"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.array.concatenate()</span></code></a>.</p>
</li>
<li><p class="first">Add <a class="reference internal" href="runtime_program.html#pyopencl.Kernel.capture_call" title="pyopencl.Kernel.capture_call"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.Kernel.capture_call()</span></code></a>.</p>
</li>
</ul>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p>The addition of <a class="reference internal" href="array.html#pyopencl.array.Array.__getitem__" title="pyopencl.array.Array.__getitem__"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.array.Array.__getitem__()</span></code></a> has an unintended
consequence due to <a class="reference external" href="https://github.com/numpy/numpy/issues/3375">numpy bug 3375</a>. For instance, this
expression:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">numpy</span><span class="o">.</span><span class="n">float32</span><span class="p">(</span><span class="mi">5</span><span class="p">)</span> <span class="o">*</span> <span class="n">some_pyopencl_array</span>
</pre></div>
</div>
<p>may take a very long time to execute. This is because <a class="reference external" href="file:///usr/share/doc/python-numpy-doc/html/reference/index.html#module-numpy" title="(in NumPy v1.13)"><code class="xref py py-mod docutils literal"><span class="pre">numpy</span></code></a> first
builds an object array of (compute-device) scalars (!) before it decides that
that’s probably not such a bright idea and finally calls
<code class="xref py py-meth docutils literal"><span class="pre">pyopencl.array.Array.__rmul__()</span></code>.</p>
<p>Note that only left arithmetic operations of <a class="reference internal" href="array.html#pyopencl.array.Array" title="pyopencl.array.Array"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.array.Array</span></code></a>
by <a class="reference external" href="file:///usr/share/doc/python-numpy-doc/html/reference/index.html#module-numpy" title="(in NumPy v1.13)"><code class="xref py py-mod docutils literal"><span class="pre">numpy</span></code></a> scalars are affected. Python’s number types (<a class="reference external" href="file:///usr/share/doc/python3-doc/html/library/functions.html#float" title="(in Python v3.6)"><code class="xref py py-class docutils literal"><span class="pre">float</span></code></a> etc.)
are unaffected, as are right multiplications.</p>
<p>If a program that used to run fast suddenly runs extremely slowly, it is
likely that this bug is to blame.</p>
<p>Here’s what you can do:</p>
<ul class="last simple">
<li>Use Python scalars instead of <a class="reference external" href="file:///usr/share/doc/python-numpy-doc/html/reference/index.html#module-numpy" title="(in NumPy v1.13)"><code class="xref py py-mod docutils literal"><span class="pre">numpy</span></code></a> scalars.</li>
<li>Switch to right multiplications if possible.</li>
<li>Use a patched <a class="reference external" href="file:///usr/share/doc/python-numpy-doc/html/reference/index.html#module-numpy" title="(in NumPy v1.13)"><code class="xref py py-mod docutils literal"><span class="pre">numpy</span></code></a>. See the bug report linked above for a pull
request with a fix.</li>
<li>Switch to a fixed version of <a class="reference external" href="file:///usr/share/doc/python-numpy-doc/html/reference/index.html#module-numpy" title="(in NumPy v1.13)"><code class="xref py py-mod docutils literal"><span class="pre">numpy</span></code></a> when available.</li>
</ul>
</div>
</div>
<div class="section" id="version-2012-1">
<h2>Version 2012.1<a class="headerlink" href="#version-2012-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Support for complex numbers.</li>
<li>Support for Bessel functions. (experimental)</li>
<li>Numerous fixes.</li>
</ul>
</div>
<div class="section" id="version-2011-2">
<h2>Version 2011.2<a class="headerlink" href="#version-2011-2" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Add <code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_migrate_mem_object()</span></code>.</li>
<li>Add <a class="reference internal" href="runtime_memory.html#pyopencl.image_from_array" title="pyopencl.image_from_array"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.image_from_array()</span></code></a>.</li>
<li>IMPORTANT BUGFIX: Kernel caching was broken for all the 2011.1.x releases, with
severe consequences on the execution time of <a class="reference internal" href="array.html#pyopencl.array.Array" title="pyopencl.array.Array"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.array.Array</span></code></a>
operations.
Henrik Andresen at a <a class="reference external" href="http://gpulab.imm.dtu.dk/courses.html">PyOpenCL workshop at DTU</a>
first noticed the strange timings.</li>
<li>All comparable PyOpenCL objects are now also hashable.</li>
<li>Add <code class="xref py py-func docutils literal"><span class="pre">pyopencl.tools.context_dependent_memoize()</span></code> to the documented
functionality.</li>
<li>Base <a class="reference internal" href="array.html#module-pyopencl.clrandom" title="pyopencl.clrandom"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl.clrandom</span></code></a> on <a class="reference external" href="https://bitbucket.org/ivarun/ranluxcl">RANLUXCL</a>,
add functionality.</li>
<li>Add <a class="reference internal" href="runtime_queue.html#pyopencl.NannyEvent" title="pyopencl.NannyEvent"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.NannyEvent</span></code></a> objects.</li>
<li>Add <a class="reference internal" href="tools.html#module-pyopencl.characterize" title="pyopencl.characterize"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl.characterize</span></code></a>.</li>
<li>Ensure compatibility with OS X Lion.</li>
<li>Add <a class="reference internal" href="array.html#pyopencl.tools.register_dtype" title="pyopencl.tools.register_dtype"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.tools.register_dtype()</span></code></a> to enable scan/reduction on struct types.</li>
<li><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_migrate_mem_object()</span></code> was renamed
<a class="reference internal" href="runtime_memory.html#pyopencl.enqueue_migrate_mem_object_ext" title="pyopencl.enqueue_migrate_mem_object_ext"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_migrate_mem_object_ext()</span></code></a>.
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_migrate_mem_object()</span></code> now refers to the OpenCL 1.2 function
of this name, if available.</li>
<li><code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_sub_devices()</span></code> was renamed
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_sub_devices_ext()</span></code>.
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_sub_devices()</span></code> now refers to the OpenCL 1.2 function
of this name, if available.</li>
<li>Alpha support for OpenCL 1.2.</li>
</ul>
</div>
<div class="section" id="version-2011-1-2">
<h2>Version 2011.1.2<a class="headerlink" href="#version-2011-1-2" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>More bug fixes.</li>
</ul>
</div>
<div class="section" id="version-2011-1-1">
<h2>Version 2011.1.1<a class="headerlink" href="#version-2011-1-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fixes for Python 3 compatibility. (with work by Christoph Gohlke)</li>
</ul>
</div>
<div class="section" id="version-2011-1">
<h2>Version 2011.1<a class="headerlink" href="#version-2011-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>All <em>is_blocking</em> parameters now default to <em>True</em> to avoid
crashy-by-default behavior. (suggested by Jan Meinke)
In particular, this change affects
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_read_buffer()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_write_buffer()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_read_buffer_rect()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_write_buffer_rect()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_read_image()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_write_image()</span></code>,
<a class="reference internal" href="runtime_memory.html#pyopencl.enqueue_map_buffer" title="pyopencl.enqueue_map_buffer"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_map_buffer()</span></code></a>,
<a class="reference internal" href="runtime_memory.html#pyopencl.enqueue_map_image" title="pyopencl.enqueue_map_image"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_map_image()</span></code></a>.</li>
<li>Add <a class="reference internal" href="algorithm.html#module-pyopencl.reduction" title="pyopencl.reduction"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl.reduction</span></code></a>.</li>
<li>Add <a class="reference internal" href="array.html#reductions"><span class="std std-ref">Reductions</span></a>.</li>
<li>Add <a class="reference internal" href="algorithm.html#module-pyopencl.scan" title="pyopencl.scan"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl.scan</span></code></a>.</li>
<li>Add <a class="reference internal" href="runtime_memory.html#pyopencl.MemoryObject.get_host_array" title="pyopencl.MemoryObject.get_host_array"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.MemoryObject.get_host_array()</span></code></a>.</li>
<li>Deprecate context arguments of
<a class="reference internal" href="array.html#pyopencl.array.to_device" title="pyopencl.array.to_device"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.array.to_device()</span></code></a>,
<a class="reference internal" href="array.html#pyopencl.array.zeros" title="pyopencl.array.zeros"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.array.zeros()</span></code></a>,
<a class="reference internal" href="array.html#pyopencl.array.arange" title="pyopencl.array.arange"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.array.arange()</span></code></a>.</li>
<li>Make construction of <a class="reference internal" href="array.html#pyopencl.array.Array" title="pyopencl.array.Array"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.array.Array</span></code></a> more flexible (<em>cqa</em> argument.)</li>
<li>Add <a class="reference internal" href="tools.html#memory-pools"><span class="std std-ref">Memory Pools</span></a>.</li>
<li>Add vector types, see <a class="reference internal" href="array.html#pyopencl.array.vec" title="pyopencl.array.vec"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.array.vec</span></code></a>.</li>
<li>Add <a class="reference internal" href="array.html#pyopencl.array.Array.strides" title="pyopencl.array.Array.strides"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.array.Array.strides</span></code></a>, <a class="reference internal" href="array.html#pyopencl.array.Array.flags" title="pyopencl.array.Array.flags"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.array.Array.flags</span></code></a>.
Allow the creation of arrays in C and Fortran order.</li>
<li>Add <a class="reference internal" href="runtime_memory.html#pyopencl.enqueue_copy" title="pyopencl.enqueue_copy"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_copy()</span></code></a>. Deprecate all other transfer functions.</li>
<li>Add support for numerous extensions, among them device fission.</li>
<li>Add a compiler cache.</li>
<li>Add the ‘g_times_l’ keyword arg to kernel execution.</li>
</ul>
</div>
<div class="section" id="version-0-92">
<h2>Version 0.92<a class="headerlink" href="#version-0-92" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Add support for OpenCL 1.1.</li>
<li>Add support for the
<a class="reference external" href="ghttp://www.khronos.org/registry/cl/extensions/khr/cl_khr_gl_sharing.txt">cl_khr_gl_sharing</a>
extension, leading to working GL interoperability.</li>
<li>Add <a class="reference internal" href="runtime_program.html#pyopencl.Kernel.set_args" title="pyopencl.Kernel.set_args"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.Kernel.set_args()</span></code></a>.</li>
<li>The call signature of <a class="reference internal" href="runtime_program.html#pyopencl.Kernel.__call__" title="pyopencl.Kernel.__call__"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.Kernel.__call__()</span></code></a> changed to
emphasize the importance of <em>local_size</em>.</li>
<li>Add <a class="reference internal" href="runtime_program.html#pyopencl.Kernel.set_scalar_arg_dtypes" title="pyopencl.Kernel.set_scalar_arg_dtypes"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.Kernel.set_scalar_arg_dtypes()</span></code></a>.</li>
<li>Add support for the
<a class="reference external" href="http://www.khronos.org/registry/cl/extensions/khr/cl_nv_device_attribute_query.txt">cl_nv_device_attribute_query</a>
extension.</li>
<li>Add <a class="reference internal" href="array.html#pyopencl.array.Array" title="pyopencl.array.Array"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.array.Array()</span></code></a> and related functionality.</li>
<li>Make build not depend on Boost C++.</li>
</ul>
</div>
<div class="section" id="version-0-91-5">
<h2>Version 0.91.5<a class="headerlink" href="#version-0-91-5" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Add <a class="reference internal" href="runtime_memory.html#pyopencl.ImageFormat.channel_count" title="pyopencl.ImageFormat.channel_count"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.ImageFormat.channel_count</span></code></a>,
<a class="reference internal" href="runtime_memory.html#pyopencl.ImageFormat.dtype_size" title="pyopencl.ImageFormat.dtype_size"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.ImageFormat.dtype_size</span></code></a>,
<a class="reference internal" href="runtime_memory.html#pyopencl.ImageFormat.itemsize" title="pyopencl.ImageFormat.itemsize"><code class="xref py py-attr docutils literal"><span class="pre">pyopencl.ImageFormat.itemsize</span></code></a>.</li>
<li>Add missing <a class="reference internal" href="runtime_memory.html#pyopencl.enqueue_copy_buffer" title="pyopencl.enqueue_copy_buffer"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_copy_buffer()</span></code></a>.</li>
<li>Add <a class="reference internal" href="runtime_platform.html#pyopencl.create_some_context" title="pyopencl.create_some_context"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_some_context()</span></code></a>.</li>
<li>Add <a class="reference internal" href="runtime_queue.html#pyopencl.enqueue_barrier" title="pyopencl.enqueue_barrier"><code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_barrier()</span></code></a>, which was previously missing.</li>
</ul>
</div>
<div class="section" id="version-0-91-4">
<h2>Version 0.91.4<a class="headerlink" href="#version-0-91-4" title="Permalink to this headline">¶</a></h2>
<p>A bugfix release. No user-visible changes.</p>
</div>
<div class="section" id="version-0-91-3">
<h2>Version 0.91.3<a class="headerlink" href="#version-0-91-3" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>All parameters named <em>host_buffer</em> were renamed <em>hostbuf</em> for consistency
with the <a class="reference internal" href="runtime_memory.html#pyopencl.Buffer" title="pyopencl.Buffer"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Buffer</span></code></a> constructor introduced in 0.91.
Compatibility code is in place.</li>
<li>The <code class="xref py py-class docutils literal"><span class="pre">pyopencl.Image</span></code> constructor does not need a <em>shape</em> parameter if the
given <em>hostbuf</em> has <em>hostbuf.shape</em>.</li>
<li>The <a class="reference internal" href="runtime_platform.html#pyopencl.Context" title="pyopencl.Context"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Context</span></code></a> constructor can now be called without parameters.</li>
</ul>
</div>
<div class="section" id="version-0-91-2">
<h2>Version 0.91.2<a class="headerlink" href="#version-0-91-2" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li><a class="reference internal" href="runtime_program.html#pyopencl.Program.build" title="pyopencl.Program.build"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.Program.build()</span></code></a> now captures build logs and adds them
to the exception text.</li>
<li>Deprecate <code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_context_from_type()</span></code> in favor of second
form of <a class="reference internal" href="runtime_platform.html#pyopencl.Context" title="pyopencl.Context"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Context</span></code></a> constructor</li>
<li>Introduce <a class="reference internal" href="runtime_program.html#pyopencl.LocalMemory" title="pyopencl.LocalMemory"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.LocalMemory</span></code></a>.</li>
<li>Document kernel invocation and <a class="reference internal" href="runtime_program.html#pyopencl.Kernel.set_arg" title="pyopencl.Kernel.set_arg"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.Kernel.set_arg()</span></code></a>.</li>
</ul>
</div>
<div class="section" id="version-0-91-1">
<h2>Version 0.91.1<a class="headerlink" href="#version-0-91-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fixed a number of bugs, notably involving <a class="reference internal" href="runtime_memory.html#pyopencl.Sampler" title="pyopencl.Sampler"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Sampler</span></code></a>.</li>
<li><a class="reference internal" href="runtime_platform.html#pyopencl.Device" title="pyopencl.Device"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Device</span></code></a>, <a class="reference internal" href="runtime_platform.html#pyopencl.Platform" title="pyopencl.Platform"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Platform</span></code></a>,
<a class="reference internal" href="runtime_platform.html#pyopencl.Context" title="pyopencl.Context"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Context</span></code></a> now have nicer string representations.</li>
<li>Add <code class="xref py py-attr docutils literal"><span class="pre">Image.shape</span></code>. (suggested by David Garcia)</li>
</ul>
</div>
<div class="section" id="version-0-91">
<h2>Version 0.91<a class="headerlink" href="#version-0-91" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Add <a class="reference internal" href="runtime_gl.html#gl-interop"><span class="std std-ref">OpenCL Runtime: OpenGL Interoperability</span></a>.</li>
<li>Add a test suite.</li>
<li>Fix numerous <cite>get_info</cite> bugs. (reports by David Garcia and the test suite)</li>
<li>Add <a class="reference internal" href="runtime_memory.html#pyopencl.ImageFormat.__repr__" title="pyopencl.ImageFormat.__repr__"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.ImageFormat.__repr__()</span></code></a>.</li>
<li>Add <code class="xref py py-meth docutils literal"><span class="pre">pyopencl.addressing_mode.to_string()</span></code> and colleagues.</li>
<li>The <cite>pitch</cite> arguments to
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_image_2d()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_image_3d()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_read_image()</span></code>, and
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.enqueue_write_image()</span></code>
are now defaulted to zero. The argument order of <cite>enqueue_{read,write}_image</cite>
has changed for this reason.</li>
<li>Deprecate
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_image_2d()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_image_3d()</span></code>
in favor of the <code class="xref py py-class docutils literal"><span class="pre">pyopencl.Image</span></code> constructor.</li>
<li>Deprecate
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_program_with_source()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_program_with_binary()</span></code>
in favor of the <a class="reference internal" href="runtime_program.html#pyopencl.Program" title="pyopencl.Program"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Program</span></code></a> constructor.</li>
<li>Deprecate
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_buffer()</span></code>,
<code class="xref py py-func docutils literal"><span class="pre">pyopencl.create_host_buffer()</span></code>
in favor of the <a class="reference internal" href="runtime_memory.html#pyopencl.Buffer" title="pyopencl.Buffer"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.Buffer</span></code></a> constructor.</li>
<li><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.MemoryObject.get_image_info()</span></code> now actually exists.</li>
<li>Add <code class="xref py py-attr docutils literal"><span class="pre">pyopencl.MemoryObject.image.info</span></code>.</li>
<li>Fix API tracing.</li>
<li>Add constructor arguments to <a class="reference internal" href="runtime_memory.html#pyopencl.ImageFormat" title="pyopencl.ImageFormat"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.ImageFormat</span></code></a>. (suggested by David Garcia)</li>
</ul>
</div>
<div class="section" id="version-0-90-4">
<h2>Version 0.90.4<a class="headerlink" href="#version-0-90-4" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Add build fixes for Windows and OS X.</li>
</ul>
</div>
<div class="section" id="version-0-90-3">
<h2>Version 0.90.3<a class="headerlink" href="#version-0-90-3" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fix a GNU-ism in the C++ code of the wrapper.</li>
</ul>
</div>
<div class="section" id="version-0-90-2">
<h2>Version 0.90.2<a class="headerlink" href="#version-0-90-2" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fix <a class="reference internal" href="runtime_platform.html#pyopencl.Platform.get_info" title="pyopencl.Platform.get_info"><code class="xref py py-meth docutils literal"><span class="pre">pyopencl.Platform.get_info()</span></code></a>.</li>
<li>Fix passing properties to <a class="reference internal" href="runtime_queue.html#pyopencl.CommandQueue" title="pyopencl.CommandQueue"><code class="xref py py-class docutils literal"><span class="pre">pyopencl.CommandQueue</span></code></a>.
Also fix related documentation.</li>
</ul>
</div>
<div class="section" id="version-0-90-1">
<h2>Version 0.90.1<a class="headerlink" href="#version-0-90-1" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Fix building on the Mac.</li>
</ul>
</div>
<div class="section" id="version-0-90">
<h2>Version 0.90<a class="headerlink" href="#version-0-90" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Initial release.</li>
</ul>
</div>
</div>
<div class="section" id="license">
<span id="id1"></span><h1>License<a class="headerlink" href="#license" title="Permalink to this headline">¶</a></h1>
<p>PyOpenCL is licensed to you under the MIT/X Consortium license:</p>
<p>Copyright (c) 2009-13 Andreas Klöckner and Contributors.</p>
<p>Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation
files (the “Software”), to deal in the Software without
restriction, including without limitation the rights to use,
copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following
conditions:</p>
<p>The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.</p>
<p>THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
OTHER DEALINGS IN THE SOFTWARE.</p>
<p>PyOpenCL includes derivatives of parts of the <a class="reference external" href="https://github.com/thrust/thrust/">Thrust</a> computing package (in particular the scan
implementation). These parts are licensed as follows:</p>
<blockquote>
<div><p>Copyright 2008-2011 NVIDIA Corporation</p>
<p>Licensed under the Apache License, Version 2.0 (the “License”);
you may not use this file except in compliance with the License.
You may obtain a copy of the License at</p>
<blockquote>
<div><<a class="reference external" href="http://www.apache.org/licenses/LICENSE-2.0">http://www.apache.org/licenses/LICENSE-2.0</a>></div></blockquote>
<p>Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an “AS IS” BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.</p>
</div></blockquote>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">If you use Apache-licensed parts, be aware that these may be incompatible
with software licensed exclusively under GPL2. (Most software is licensed
as GPL2 or later, in which case this is not an issue.)</p>
</div>
<p>PyOpenCL includes parts of the Random123 suite of random number generators:</p>
<blockquote>
<div><p>Copyright 2010-2012, D. E. Shaw Research.
All rights reserved.</p>
<p>Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:</p>
<ul class="simple">
<li>Redistributions of source code must retain the above copyright
notice, this list of conditions, and the following disclaimer.</li>
<li>Redistributions in binary form must reproduce the above copyright
notice, this list of conditions, and the following disclaimer in the
documentation and/or other materials provided with the distribution.</li>
<li>Neither the name of D. E. Shaw Research nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.</li>
</ul>
<p>THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
“AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.</p>
</div></blockquote>
<p>PyOpenCL includes the RANLUXCL random number generator:</p>
<blockquote>
<div><p>Copyright (c) 2011 Ivar Ursin Nikolaisen</p>
<p>Permission is hereby granted, free of charge, to any person obtaining a copy of this
software and associated documentation files (the “Software”), to deal in the Software
without restriction, including without limitation the rights to use, copy, modify,
merge, publish, distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to the following
conditions:</p>
<p>The above copyright notice and this permission notice shall be included in all copies
or substantial portions of the Software.</p>
<p>THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE
OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.</p>
</div></blockquote>
</div>
<div class="section" id="frequently-asked-questions">
<h1>Frequently Asked Questions<a class="headerlink" href="#frequently-asked-questions" title="Permalink to this headline">¶</a></h1>
<p>The FAQ is maintained collaboratively on the
<a class="reference external" href="http://wiki.tiker.net/PyOpenCL/FrequentlyAskedQuestions">Wiki FAQ page</a>.</p>
</div>
<div class="section" id="citing-pyopencl">
<h1>Citing PyOpenCL<a class="headerlink" href="#citing-pyopencl" title="Permalink to this headline">¶</a></h1>
<p>We are not asking you to gratuitously cite PyOpenCL in work that is otherwise
unrelated to software. That said, if you do discuss some of the development
aspects of your code and would like to highlight a few of the ideas behind
PyOpenCL, feel free to cite <a class="reference external" href="http://dx.doi.org/10.1016/j.parco.2011.09.001">this article</a>:</p>
<blockquote>
<div>Andreas Klöckner, Nicolas Pinto, Yunsup Lee, Bryan Catanzaro, Paul Ivanov,
Ahmed Fasih, PyCUDA and PyOpenCL: A scripting-based approach to GPU
run-time code generation, Parallel Computing, Volume 38, Issue 3, March
2012, Pages 157-174.</div></blockquote>
<p>Here’s a Bibtex entry for your convenience:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="nd">@article</span><span class="p">{</span><span class="n">kloeckner_pycuda_2012</span><span class="p">,</span>
<span class="n">author</span> <span class="o">=</span> <span class="p">{{</span><span class="n">Kl</span><span class="p">{</span>\<span class="s2">"o}ckner}, Andreas</span>
<span class="ow">and</span> <span class="p">{</span><span class="n">Pinto</span><span class="p">},</span> <span class="n">Nicolas</span>
<span class="ow">and</span> <span class="p">{</span><span class="n">Lee</span><span class="p">},</span> <span class="n">Yunsup</span>
<span class="ow">and</span> <span class="p">{</span><span class="n">Catanzaro</span><span class="p">},</span> <span class="n">B</span><span class="o">.</span>
<span class="ow">and</span> <span class="p">{</span><span class="n">Ivanov</span><span class="p">},</span> <span class="n">Paul</span>
<span class="ow">and</span> <span class="p">{</span><span class="n">Fasih</span><span class="p">},</span> <span class="n">Ahmed</span> <span class="p">},</span>
<span class="n">title</span> <span class="o">=</span> <span class="s2">"{PyCUDA and PyOpenCL: A Scripting-Based Approach to GPU Run-Time Code Generation}"</span><span class="p">,</span>
<span class="n">journal</span> <span class="o">=</span> <span class="s2">"Parallel Computing"</span><span class="p">,</span>
<span class="n">volume</span> <span class="o">=</span> <span class="s2">"38"</span><span class="p">,</span>
<span class="n">number</span> <span class="o">=</span> <span class="s2">"3"</span><span class="p">,</span>
<span class="n">pages</span> <span class="o">=</span> <span class="s2">"157--174"</span><span class="p">,</span>
<span class="n">year</span> <span class="o">=</span> <span class="s2">"2012"</span><span class="p">,</span>
<span class="n">issn</span> <span class="o">=</span> <span class="s2">"0167-8191"</span><span class="p">,</span>
<span class="n">doi</span> <span class="o">=</span> <span class="s2">"10.1016/j.parco.2011.09.001"</span><span class="p">,</span>
<span class="p">}</span>
</pre></div>
</div>
</div>
<div class="section" id="acknowledgments">
<h1>Acknowledgments<a class="headerlink" href="#acknowledgments" title="Permalink to this headline">¶</a></h1>
<div class="section" id="contributors">
<h2>Contributors<a class="headerlink" href="#contributors" title="Permalink to this headline">¶</a></h2>
<p>Too many to list. Please see the
<a class="reference external" href="https://github.com/inducer/pyopencl/commits/master">commit log</a>
for detailed acknowledgments.</p>
</div>
<div class="section" id="funding">
<h2>Funding<a class="headerlink" href="#funding" title="Permalink to this headline">¶</a></h2>
<p>Andreas Klöckner’s work on <a class="reference internal" href="runtime.html#module-pyopencl" title="pyopencl"><code class="xref py py-mod docutils literal"><span class="pre">pyopencl</span></code></a> was supported in part by</p>
<ul class="simple">
<li>US Navy ONR grant number N00014-14-1-0117</li>
<li>the US National Science Foundation under grant numbers DMS-1418961 and CCF-1524433.</li>
</ul>
<p>AK also gratefully acknowledges a hardware gift from Nvidia Corporation. The
views and opinions expressed herein do not necessarily reflect those of the
funding agencies.</p>
</div>
</div>
</div>
</div>
</div>
<div class="sphinxsidebar" role="navigation" aria-label="main navigation">
<div class="sphinxsidebarwrapper">
<h1 class="logo"><a href="index.html">PyOpenCL</a></h1>
<h3>Navigation</h3>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="runtime.html">OpenCL Runtime: Basics</a></li>
<li class="toctree-l1"><a class="reference internal" href="runtime_const.html">OpenCL Runtime: Constants</a></li>
<li class="toctree-l1"><a class="reference internal" href="runtime_platform.html">OpenCL Runtime: Platforms, Devices and Contexts</a></li>
<li class="toctree-l1"><a class="reference internal" href="runtime_queue.html">OpenCL Runtime: Command Queues and Events</a></li>
<li class="toctree-l1"><a class="reference internal" href="runtime_memory.html">OpenCL Runtime: Memory</a></li>
<li class="toctree-l1"><a class="reference internal" href="runtime_program.html">OpenCL Runtime: Programs and Kernels</a></li>
<li class="toctree-l1"><a class="reference internal" href="runtime_gl.html">OpenCL Runtime: OpenGL Interoperability</a></li>
<li class="toctree-l1"><a class="reference internal" href="array.html">Multi-dimensional arrays</a></li>
<li class="toctree-l1"><a class="reference internal" href="algorithm.html">Parallel Algorithms</a></li>
<li class="toctree-l1"><a class="reference internal" href="howto.html">How-tos</a></li>
<li class="toctree-l1"><a class="reference internal" href="tools.html">Built-in Utilities</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">Installation</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#using-vendor-supplied-opencl-drivers-linux">Using vendor-supplied OpenCL drivers (Linux)</a></li>
<li class="toctree-l2"><a class="reference internal" href="#getting-a-better-cpu-based-opencl-driver-os-x">Getting a better CPU-based OpenCL driver (OS X)</a></li>
<li class="toctree-l2"><a class="reference internal" href="#installing-from-source">Installing from source</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="#tips">Tips</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#syntax-highlighting">Syntax highlighting</a></li>
<li class="toctree-l2"><a class="reference internal" href="#ipython-integration">IPython integration</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="#guidelines">Guidelines</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#api-stability">API Stability</a></li>
<li class="toctree-l2"><a class="reference internal" href="#relation-with-opencl-s-c-bindings">Relation with OpenCL’s C Bindings</a></li>
<li class="toctree-l2"><a class="reference internal" href="#interoperability-with-other-opencl-software">Interoperability with other OpenCL software</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="#user-visible-changes">User-visible Changes</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#version-2016-3">Version 2016.3</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2016-2">Version 2016.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2016-1">Version 2016.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2015-2-4">Version 2015.2.4</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2015-2-3">Version 2015.2.3</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2015-2-2">Version 2015.2.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2015-2-1">Version 2015.2.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2015-2">Version 2015.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2015-1">Version 2015.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2014-1">Version 2014.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2013-2">Version 2013.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2013-1">Version 2013.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2012-1">Version 2012.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2011-2">Version 2011.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2011-1-2">Version 2011.1.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2011-1-1">Version 2011.1.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-2011-1">Version 2011.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-92">Version 0.92</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-91-5">Version 0.91.5</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-91-4">Version 0.91.4</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-91-3">Version 0.91.3</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-91-2">Version 0.91.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-91-1">Version 0.91.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-91">Version 0.91</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-90-4">Version 0.90.4</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-90-3">Version 0.90.3</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-90-2">Version 0.90.2</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-90-1">Version 0.90.1</a></li>
<li class="toctree-l2"><a class="reference internal" href="#version-0-90">Version 0.90</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="#license">License</a></li>
<li class="toctree-l1"><a class="reference internal" href="#frequently-asked-questions">Frequently Asked Questions</a></li>
<li class="toctree-l1"><a class="reference internal" href="#citing-pyopencl">Citing PyOpenCL</a></li>
<li class="toctree-l1"><a class="reference internal" href="#acknowledgments">Acknowledgments</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#contributors">Contributors</a></li>
<li class="toctree-l2"><a class="reference internal" href="#funding">Funding</a></li>
</ul>
</li>
</ul>
<hr />
<ul>
<li class="toctree-l1"><a href="https://github.com/inducer/pyopencl">🚀 Github</a></li>
<li class="toctree-l1"><a href="https://pypi.python.org/pypi/pyopencl">💾 Download Releases</a></li>
</ul>
<div class="relations">
<h3>Related Topics</h3>
<ul>
<li><a href="index.html">Documentation overview</a><ul>
<li>Previous: <a href="tools.html" title="previous chapter">Built-in Utilities</a></li>
</ul></li>
</ul>
</div>
<div id="searchbox" style="display: none" role="search">
<h3>Quick search</h3>
<form class="search" action="search.html" method="get">
<div><input type="text" name="q" /></div>
<div><input type="submit" value="Go" /></div>
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
</div>
</div>
<div class="clearer"></div>
</div>
<div class="footer">
©2017, Andreas Kloeckner.
|
Powered by <a href="http://sphinx-doc.org/">Sphinx 1.6.5</a>
& <a href="https://github.com/bitprophet/alabaster">Alabaster 0.7.8</a>
</div>
</body>
</html>
|