Merge branch 'v2.3.3' of https://github.com/embree/embree into v2.3.3

doc/release.txt

@@ -7,6 +7,8 @@ Perform the following steps for each Embree release.
 [ ] Check Embree version in kernels/xeonphi/CMakeLists.txt
 [ ] Check Embree Library version in scripts/release_linux.sh
 [ ] Check Embree Library version in scripts/release_macos.sh
+[ ] Check Embree Library version in scripts/install_linux/install.sh
+[ ] Check Embree Library version in scripts/install_macos/install.sh
 
 [ ] Make sure all regression tests pass under Linux  : ./scripts/regression.py render linux   log models/embree/ | tee log/log.txt
 [ ] Make sure all regression tests pass under MacOSX : ./scripts/regression.py render macosx  log models/embree/ | tee log/log.txt

doc/src/compilation.html

@@ -55,7 +55,7 @@ <h2>Compiling Embree on Linux and Mac OS X</h2>
 <p>If you cannot install Embree on your system (e.g. when you don't
 have administrator rights) you need to add embree_root_directory/build
 to your LD_LIBRARY_PATH (and SINK_LD_LIBRARY_PATH in case you
-want to use Embree on Xeon Phi).</p>
+want to use Embree on Xeon Phi&trade;).</p>
 
 <p>The default configuration in the configuration dialog should be
 appropiate for most usages. The following table described all
@@ -79,10 +79,17 @@ <h2>Compiling Embree on Linux and Mac OS X</h2>
 </table>
 
 <p>You need at least Intel Compiler 11.1 or GCC 4.4 to enable AVX and
-Intel Compiler 12.1 or GCC 4.7 to enable AVX2. </p>
+Intel Compiler 12.1 or GCC 4.7 to enable AVX2.</p>
 
-<p>Enabling the buffer stride feature reduces performance for
-building spatial hierarchies on Xeon Phi&trade;.</p>
+<h2>Compiling Embree for Xeon Phi</h2>
+
+<p>Embree supports the Xeon Phi&trade; coprocessor under Linux. To
+compile Embree for Xeon Phi&trade; you need to enable the XEON_PHI_ISA
+option in CMake and have the Intel Compiler and the Intel&trade; Manycore
+Platform Software Stack (MPSS) installed.</p>
+
+<p>Enabling the buffer stride feature reduces performance for building
+spatial hierarchies on Xeon Phi&trade;.</p>
 
  <h2>Compiling Embree on Windows</h2>
 

doc/src/downloading.html

@@ -6,7 +6,7 @@ <h2>Downloading Embree</h2>
 
 <p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_linux.zip">embree-bin-2.3.3_linux.zip</a></p>
 <p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_win.zip">embree-bin-2.3.3_win.zip</a></p>
-<p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_macos.zip">embree-bin-2.3.3_macos.zip</a></p>
+<p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_macosx.zip">embree-bin-2.3.3_macosx.zip</a></p>
 
 <p>If you need to recompile Embree for your platform, please download
 the sources and follow the compilation instructions below:</p>

doc/src/new_features.html

@@ -1,6 +1,6 @@
 <h3>New features in Embree 2.3.3</h3>
 <ul>
-  <li> BVH builders more robust handling invalid input data (Xeon&trade;)</li>
+  <li> BVH builders more robustly handle invalid input data (Xeon&trade;)</li>
   <li> Motion blur support for hair geometry (Xeon&trade;)</li>
   <li> Improved motion blur performance for triangle geometry (Xeon&trade;)</li>
   <li> Improved robust ray tracing mode (Xeon&trade;)</li>

doc/src/readme.html

@@ -13,8 +13,8 @@ <h1>Embree: High Performance Ray Tracing Kernels
     <CONTENT_OVERVIEW>
     <CONTENT_PLATFORMS>
     <CONTENT_COMPILATION>
-    <CONTENT_API>
     <CONTENT_TUTORIALS>
+    <CONTENT_API>
     <CONTENT_SUPPORT>
   </body>
 </html>

doc/src/readme_bin.html

@@ -13,8 +13,8 @@ <h1>Embree: High Performance Ray Tracing Kernels
     <CONTENT_OVERVIEW>
     <CONTENT_PLATFORMS>
     <CONTENT_INSTALL>
-    <CONTENT_API>
     <CONTENT_TUTORIALS>
+    <CONTENT_API>
     <CONTENT_SUPPORT>
   </body>
 </html>

doc/src/tutorials.html

@@ -2,32 +2,40 @@
 <h2>Embree Tutorials</h2>
 
 <p>Embree comes with a set of tutorials aimed at helping users
-understand how embree can be used and extended. All tutorials exist in
+understand how Embree can be used and extended. All tutorials exist in
 an ISPC and C version to demonstrate the two versions of the API. Look
-for files names <code>tutorialXX_device.ispc</code> for the ISPC
+for files named <code>tutorialXX_device.ispc</code> for the ISPC
 implementation of the tutorial, and files named
-<code>tutorialXX_device.cpp</code> for the single ray C version of the
+<code>tutorialXX_device.cpp</code> for the single ray C++ version of the
 tutorial. To start the C++ version
 use the <code>tutorialXX</code> executables, to start the ISPC version
-use the <code>tutorialXX_ispc</code> executables.
+use the <code>tutorialXX_ispc</code> executables.</p>
 
-You can select an initial camera using the -vp (camera position), -vi
-(camera lookat point), -vu (camera up vector), and -fov (vertical
-field of view) command line parameters:
+<p>Under Linux Embree also comes with an ISPC version of all tutorials
+for the Xeon Phi&trade; Coprocessor. The executables of this version of
+the tutorials are named <code>tutorialXX_xeonphi</code> and only work
+if a Xeon Phi&trade; Coprocessor is present in the system. The Xeon
+Phi&trade; version of the tutorials get started on the host CPU, just
+like all other tutorials, and will connect automatically to one
+installed Xeon Phi&trade; Coprocessor in the system.</p>
+
+<p>For all tutorials, you can select an initial camera using the -vp
+(camera position), -vi (camera lookat point), -vu (camera up vector),
+and -fov (vertical field of view) command line parameters:</p>
 
 <p><pre><code>./tutorial00 -vp 10 10 10 -vi 0 0 0
 </code></pre></p>
 
-You can select the initial windows size using the -size command line
+<p>You can select the initial windows size using the -size command line
 parameter, or start the tutorials in fullscreen using the -fullscreen
-parameter:
+parameter:</p>
 
 <p><pre><code>./tutorial00 -size 1024 1024
 ./tutorial00 -fullscreen
 </code></pre></p>
 
-Implementation specific parameters can be passed to the ray tracing
-core through the -rtcore command line parameter, e.g.:
+<p>Implementation specific parameters can be passed to the ray tracing
+core through the -rtcore command line parameter, e.g.:</p>
 
 <p><pre><code>./tutorial00 -rtcore verbose=2,threads=1,accel=bvh4.triangle1
 </code></pre></p>

doc/www/downloads.html

@@ -60,7 +60,7 @@ <h2>Downloading Embree</h2>
 
 <p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_linux.zip">embree-bin-2.3.3_linux.zip</a></p>
 <p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_win.zip">embree-bin-2.3.3_win.zip</a></p>
-<p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_macos.zip">embree-bin-2.3.3_macos.zip</a></p>
+<p><a href="http://github.com/embree/embree-bin/archive/v2.3.3_macosx.zip">embree-bin-2.3.3_macosx.zip</a></p>
 
 <p>If you need to recompile Embree for your platform, please download
 the sources and follow the compilation instructions below:</p>
@@ -156,7 +156,7 @@ <h2>Compiling Embree on Linux and Mac OS X</h2>
 <p>If you cannot install Embree on your system (e.g. when you don't
 have administrator rights) you need to add embree_root_directory/build
 to your LD_LIBRARY_PATH (and SINK_LD_LIBRARY_PATH in case you
-want to use Embree on Xeon Phi).</p>
+want to use Embree on Xeon Phi&trade;).</p>
 
 <p>The default configuration in the configuration dialog should be
 appropiate for most usages. The following table described all
@@ -180,10 +180,17 @@ <h2>Compiling Embree on Linux and Mac OS X</h2>
 </table>
 
 <p>You need at least Intel Compiler 11.1 or GCC 4.4 to enable AVX and
-Intel Compiler 12.1 or GCC 4.7 to enable AVX2. </p>
+Intel Compiler 12.1 or GCC 4.7 to enable AVX2.</p>
 
-<p>Enabling the buffer stride feature reduces performance for
-building spatial hierarchies on Xeon Phi&trade;.</p>
+<h2>Compiling Embree for Xeon Phi</h2>
+
+<p>Embree supports the Xeon Phi&trade; coprocessor under Linux. To
+compile Embree for Xeon Phi&trade; you need to enable the XEON_PHI_ISA
+option in CMake and have the Intel Compiler and the Intel&trade; Manycore
+Platform Software Stack (MPSS) installed.</p>
+
+<p>Enabling the buffer stride feature reduces performance for building
+spatial hierarchies on Xeon Phi&trade;.</p>
 
  <h2>Compiling Embree on Windows</h2>
 

doc/www/index.html

@@ -86,7 +86,7 @@ <h2>Embree Overview</h2>
 
 <h3>New features in Embree 2.3.3</h3>
 <ul>
-  <li> BVH builders robust for bad input data. (Xeon&trade;)</li>
+  <li> BVH builders more robustly handle invalid input data (Xeon&trade;)</li>
   <li> Motion blur support for hair geometry (Xeon&trade;)</li>
   <li> Improved motion blur performance for triangle geometry (Xeon&trade;)</li>
   <li> Improved robust ray tracing mode (Xeon&trade;)</li>

doc/www/renderer.html

@@ -226,7 +226,8 @@ <h3>Example Models</h3>
  href="http://software.intel.com/sites/default/files/article/183323/crown0413.zip">Crown0413.zip
 (74MB)</a>.</p>
 
-<p><img src="images/crown.jpg" alt="Imperial Crown of Austria"> Model courtesy <a href="http://www.loramel.net">Martin Lubich</a>.</p>
+<p><img src="images/crown.jpg" alt="Imperial Crown of Austria" width="400" height="480"></p>
+<p>Model courtesy <a href="http://www.loramel.net">Martin Lubich</a></p>
 
 
       </div>

doc/www/tutorials.html

@@ -41,32 +41,40 @@ <h2 id="project_tagline">High Performance Ray Tracing Kernels</h2>
 <h2>Embree Tutorials</h2>
 
 <p>Embree comes with a set of tutorials aimed at helping users
-understand how embree can be used and extended. All tutorials exist in
+understand how Embree can be used and extended. All tutorials exist in
 an ISPC and C version to demonstrate the two versions of the API. Look
-for files names <code>tutorialXX_device.ispc</code> for the ISPC
+for files named <code>tutorialXX_device.ispc</code> for the ISPC
 implementation of the tutorial, and files named
-<code>tutorialXX_device.cpp</code> for the single ray C version of the
+<code>tutorialXX_device.cpp</code> for the single ray C++ version of the
 tutorial. To start the C++ version
 use the <code>tutorialXX</code> executables, to start the ISPC version
-use the <code>tutorialXX_ispc</code> executables.
+use the <code>tutorialXX_ispc</code> executables.</p>
 
-You can select an initial camera using the -vp (camera position), -vi
-(camera lookat point), -vu (camera up vector), and -fov (vertical
-field of view) command line parameters:
+<p>Under Linux Embree also comes with an ISPC version of all tutorials
+for the Xeon Phi&trade; Coprocessor. The executables of this version of
+the tutorials are named <code>tutorialXX_xeonphi</code> and only work
+if a Xeon Phi&trade; Coprocessor is present in the system. The Xeon
+Phi&trade; version of the tutorials get started on the host CPU, just
+like all other tutorials, and will connect automatically to one
+installed Xeon Phi&trade; Coprocessor in the system.</p>
+
+<p>For all tutorials, you can select an initial camera using the -vp
+(camera position), -vi (camera lookat point), -vu (camera up vector),
+and -fov (vertical field of view) command line parameters:</p>
 
 <p><pre><code>./tutorial00 -vp 10 10 10 -vi 0 0 0
 </code></pre></p>
 
-You can select the initial windows size using the -size command line
+<p>You can select the initial windows size using the -size command line
 parameter, or start the tutorials in fullscreen using the -fullscreen
-parameter:
+parameter:</p>
 
 <p><pre><code>./tutorial00 -size 1024 1024
 ./tutorial00 -fullscreen
 </code></pre></p>
 
-Implementation specific parameters can be passed to the ray tracing
-core through the -rtcore command line parameter, e.g.:
+<p>Implementation specific parameters can be passed to the ray tracing
+core through the -rtcore command line parameter, e.g.:</p>
 
 <p><pre><code>./tutorial00 -rtcore verbose=2,threads=1,accel=bvh4.triangle1
 </code></pre></p>