https://brlcad.org/w/api.php?action=feedcontributions&user=85.246.127.87&feedformat=atomBRL-CAD - User contributions [en]2024-03-28T16:48:27ZUser contributionsMediaWiki 1.32.2https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8530User:Vasco.costa/GSoC15/logs2015-06-02T23:09:02Z<p>85.246.127.87: /* Development Phase */</p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:<code>do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()</code><br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.<br />
<br />
* Updated project proposal on Google Melange.<br />
<br />
* SVN r65153 fails to compile with a bogus error of an unused variable that's actually being used its just that GCC 4.9.1 is too dumb to figure that out.<br />
** https://sourceforge.net/p/brlcad/bugs/365/<br />
<br />
* Upgraded Ubuntu and GCC.<br />
<br />
* Made some simple ray generation code in C.</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8527User:Vasco.costa/GSoC15/logs2015-06-02T18:12:16Z<p>85.246.127.87: /* Development Phase */</p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:<code>do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()</code><br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.<br />
<br />
* Updated project proposal on Google Melange.<br />
<br />
* SVN r65153 fails to compile with a bogus error of an unused variable that's actually being used its just that GCC 4.9.1 is too dumb to figure that out.<br />
** https://sourceforge.net/p/brlcad/bugs/365/<br />
<br />
* Upgraded Ubuntu and GCC.</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8526User:Vasco.costa/GSoC15/logs2015-06-02T18:11:52Z<p>85.246.127.87: </p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:<code>do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()</code><br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.<br />
<br />
* Updated project proposal on Google Melange.<br />
<br />
* SVN r65153 fails to compile with a bogus error of an unused variable that's actually being used its just that GCC 4.9.1 is too dumb to figure that out.<br />
** https://sourceforge.net/p/brlcad/bugs/365/<br />
* Upgraded Ubuntu and GCC.</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8525User:Vasco.costa/GSoC15/logs2015-06-02T18:09:15Z<p>85.246.127.87: /* Development Phase */</p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:<code>do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()</code><br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.<br />
<br />
* Updated project proposal on Google Melange.<br />
<br />
* SVN r65153 fails to compile with a bogus error of an unused variable that's actually being used its just that GCC 4.9.1 is too dumb to figure that out.<br />
** https://sourceforge.net/p/brlcad/bugs/365/<br />
: Upgraded Ubuntu and GCC.</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8524User:Vasco.costa/GSoC15/logs2015-06-02T16:55:32Z<p>85.246.127.87: /* Development Phase */</p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:<code>do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()</code><br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.<br />
<br />
* Updated project proposal on Google Melange.<br />
<br />
* SVN r65153 fails to compile with a bogus error of an unused variable that's actually being used its just that GCC 4.9.1 is too dumb to figure that out.<br />
** https://sourceforge.net/p/brlcad/bugs/365/<br />
: Upgrading Ubuntu and GCC to see if the error goes away.</div>85.246.127.87https://brlcad.org/w/index.php?title=Google_Summer_of_Code/2015&diff=8523Google Summer of Code/20152015-06-02T16:54:39Z<p>85.246.127.87: /* Sigourney: a Boolean Weaver for BRL-CAD */</p>
<hr />
<div>= GSoC 2015 With BRL-CAD =<br />
<br />
'''BRL-CAD was accepted as a [http://www.google-melange.com/ mentoring organization] for the Google Summer of Code!'''<br />
<br />
We accepted '''12''' exceptional students that prepared quality proposals and demonstrated commitment through quality patches. Once again as an umbrella org, 4 of those students were selected for associated orgs LibreCAD, LinuxCNC, STEPcode, and OpenSCAD.<br />
<br />
= Accepted Projects =<br />
<br />
== GPU Accelerated Ray Trace Rendering for BRL-CAD ==<br />
* OpenCL-based accelerated ray trace rendering<br />
* Student: [[User:vasco.costa|Vasco Costa]]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/vasc/5657382461898752 Abstract]<br />
* [[User:vasco.costa/GSoC15/proposal|Project Proposal]]<br />
* [[User:vasco.costa/GSoC15/logs|Development Logs]]<br />
<br />
== Online Geometry Viewer (OGV) ==<br />
* Geometry Model Viewer on Web : Front-End Development<br />
* Student: [http://brlcad.org/wiki/User:Deekaysharma Deepak Kumar Sharma]<br />
* [http://www.google-melange.com/gsoc/project/details/google/gsoc2015/deepakkumarsharma/5764640680181760 Abstract]<br />
* [http://brlcad.org/wiki/User:Deekaysharma/GSoC2015/Proposal Project Proposal]<br />
* [http://brlcad.org/wiki/User:Deekaysharma/GSoc15/OGV_production_ready_plan Front-End : Production Ready Plan]<br />
* [[User:Deekaysharma/logs|Development Logs]]<br />
<br />
== Online Geometry Viewer (OGV): Back-end ==<br />
* Geometry Model Viewer on Web: Back-end Development<br />
* Student: [http://brlcad.org/wiki/User:MeShubham99 Shubham Chauhan]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/meshubham99/5639274879778816 Abstract]<br />
* [http://brlcad.org/wiki/User:MeShubham99/GSoc15/proposal Project Proposal]<br />
* [http://brlcad.org/wiki/User:MeShubham99/GSoc15/OGV_production_ready_plan Production Ready Plan]<br />
* [http://brlcad.org/w/index.php?title=User:MeShubham99/GSoc15/log_development Development Logs]<br />
<br />
== Synchronize Wiki with Docbook ==<br />
* merge the all docs with website and provide the editing option to user.<br />
* Student: [http://brlcad.org/w/index.php?title=User:Hiteshsofat Hitesh Sofat]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/hitesh123/5757334940811264 Abstract]<br />
* [http://brlcad.org/w/index.php?title=User:Hiteshsofat/GSoc15/proposal Project Proposal]<br />
* [http://brlcad.org/w/index.php?title=User:Hiteshsofat/GSoc15/log_development Development Logs]<br />
<br />
== Povray export ==<br />
* Converting BRL-CAD geometry into POV-Ray format<br />
* Student: [http://brlcad.org/w/index.php?title=User:Gurwinder_Singh Gurwinder Singh]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/gurwinder_singh_bains/5639274879778816 Abstract]<br />
* [http://brlcad.org/wiki/User:Gurwinder_Singh/GSoc15/proposal Project Proposal]<br />
* [http://brlcad.org/w/index.php?title=User:Gurwinder_Singh/GSoc15/log_development Development Logs]<br />
<br />
== OpenSCAD Importer ==<br />
* Converting OpenSCAD geometry into BRL-CAD format<br />
* Student: [[User:Andrei.ilinca24|Andrei Ilinca]]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/andrei_il/5676830073815040 Abstract]<br />
* [[User:Andrei.ilinca24/Proposal|Project Proposal]]<br />
* [[User:Andrei.ilinca24/logs|Development Logs]]<br />
<br />
== Object-Oriented C++ Geometry API ==<br />
* Adding Volume, Centroid and Surface Area functions for primitives (Enhancing GE)<br />
* Student : [[User:Dracarys983|Kalpit Thakkar]]<br />
* Mentor : Daniel Roßberg (d_rossberg)<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/dracarys983/5750085036015616 Abstract]<br />
* [[User:Dracarys983/Proposal|Project Proposal]]<br />
* [[User:Dracarys983/Development_Logs|Development Logs]]<br />
<br />
== STEP Geometry Viewer ==<br />
* A stand-alone STEP file viewer with great functionalities. <br />
* Student : [http://brlcad.org/wiki/User:MilindaFernando Milinda Fernando]<br />
* Mentor : Charlie Stirk <br />
* [[User:MilindaFernando/abstract|Abstract]]<br />
* [[User:MilindaFernando/gsoc2015_project_proposal|Project Proposal]]<br />
* [[User:MilindaFernando/gsoc2015_devlog|Development Logs]]<br />
<br />
== NMG / BoT Editing ==<br />
* Command-line interface for NMG / BoT editing. <br />
* Student : [[User:Bhollister|Brad Hollister]]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/brad_h/5697982787747840 Abstract]<br />
* [[User:Bhollister/Proposal|Project Proposal]]<br />
* [[User:Bhollister/DevLog|Development Logs]]<br />
<br />
== X3D Importer ==<br />
* A converter to convert from X3D file format to BRL-CAD format<br />
* Student: [http://brlcad.org/wiki/User:Konrado_DJ Konrad Djimeli]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/djimeli/5634387206995968 Abstract]<br />
* [http://brlcad.org/wiki/User:Konrado_DJ/GSoc2015/Proposal/X3D_Importer Project Proposal]<br />
* [http://brlcad.org/wiki/User:Konrado_DJ/GSoc2015/logs Development Logs]<br />
<br />
== ScadLexer for OpenSCAD ==<br />
* Lexer for QScintilla Editor. <br />
* Student: [http://brlcad.org/wiki/User:Shainasabarwal Shaina Sabarwal]<br />
* [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/shaina/5684049913839616 Abstract]<br />
* [http://brlcad.org/wiki/User:Shainasabarwal/ScadLexer_for_ScintillaEditor#Title:_SCAD_lexer_for_QScintilla_Editor Proposal]<br />
*[http://brlcad.org/wiki/User:Shainasabarwal/GSoC15/logs Development Logs]<br />
<br />
== Conic Curve Support for LibreCAD ==<br />
* Building Conic curve support (Hyperbola and Parabola) for LibreCAD.<br />
* Student: [[ User:Amalia | Ngassa Amalia ]]<br />
* [http://www.google-melange.com/gsoc/project/details/google/gsoc2015/amalia/5634387206995968 Abstract]<br />
* [http://brlcad.org/wiki/User:Amalia#Detailed_Project_Summary Project Proposal]<br />
* [http://brlcad.org/wiki/User:Amalia/Development_logs Development logs]<br />
<br />
<br />
<!-- MAKE A COPY OF THE TEMPLATE, DO NOT JUST FILL IN YOUR DETAILS...--><br />
<br />
==PROJECT TITLE (TEMPLATE) ==<br />
* 1-line brief project summary (<80 chars)<br />
* Student: [http://brlcad.org/wiki/User:whatever Your Name]<br />
* [http://www.google-melange.com/gsoc/project/google/gsoc2015/whatever/12345 Abstract]<br />
* [http://brlcad.org/wiki/User:whatever/GSoC15/proposal Project Proposal]<br />
* [http://brlcad.org/wiki/User:whatever/GSoC15/logs Development Logs]</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8521User:Vasco.costa/GSoC15/logs2015-06-02T16:23:06Z<p>85.246.127.87: </p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:<code>do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()</code><br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.<br />
<br />
* SVN r65153 fails to compile with a bogus error of an unused variable that's actually being used its just that GCC 4.9.1 is too dumb to figure that out.<br />
** https://sourceforge.net/p/brlcad/bugs/365/<br />
: Upgrading Ubuntu and GCC to see if the error goes away.</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8520User:Vasco.costa/GSoC15/logs2015-06-02T13:53:47Z<p>85.246.127.87: /* Development Phase */</p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:<code>do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()</code><br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8519User:Vasco.costa/GSoC15/logs2015-06-02T13:47:58Z<p>85.246.127.87: </p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
:do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()<br />
<br />
:The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.</div>85.246.127.87https://brlcad.org/w/index.php?title=User:Vasco.costa/GSoC15/logs&diff=8518User:Vasco.costa/GSoC15/logs2015-06-02T13:47:08Z<p>85.246.127.87: /* Development Phase */</p>
<hr />
<div>=Community Bonding Period=<br />
I have made two patches for OpenCL (OCL) shot code. One patch refactors the existing sph (Sphere) shot code, and the another patch implements ehy (Elliptical Hyperboloid) shot code.<br />
* https://sourceforge.net/p/brlcad/patches/341/<br />
* https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
=Background Research=<br />
* http://graphics.cs.kuleuven.be/publications/LD08CFRGRT/<br />
* http://research.microsoft.com/en-us/um/people/cloop/garanzhaloop2010.pdf<br />
<br />
=Development Phase=<br />
''Week 1 : 25-31 May''<br />
* Created some example .g files in mged for the primitives to be implemented this week. The Quick Reference Card proved to be quite useful.<br />
* Do the matrix ops for ehy (Elliptical Hyperboloid) in the OpenCL side.<br />
** https://sourceforge.net/p/brlcad/patches/346/<br />
<br />
* Made patch for ell (Generalized Ellipsoid) and arb8 (Arbitrary Polyhedron) OpenCL shots:<br />
** https://sourceforge.net/p/brlcad/patches/370/<br />
* ''M1 complete.''<br />
<br />
* Tried out a bunch of code browsing tools (cscope, LXR, doxygen, etc). The NetBeans IDE seems the most promising.<br />
<br />
''Week 2 : 1-7 Jun''<br />
* Read code to better understand the main rendering loop. It seems to be something like this:<br />
do_frame() &rarr; do_run() &rarr; worker()* &rarr; do_pixel()* &rarr; rt_shootray()* &rarr; rt_*_shot()<br />
<br />
The code is recursive (which is problematic for OpenCL). I'll work on a simplified version of the rendering loop which only does the primary rays in C as a first approach. After I get the non-recursive parallel friendly C code I'll work on the OpenCL port.</div>85.246.127.87