brkho
/
mesa


			
							123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107
							<HTML>

<TITLE>Cell Driver</TITLE>

<link rel="stylesheet" type="text/css" href="mesa.css"></head>

<BODY>

<H1>Mesa Cell Driver</H1>

<p>
The Mesa
<a href="http://en.wikipedia.org/wiki/Cell_%28microprocessor%29" target="_parent">Cell</a>
driver is part of the 
<a href="http://www.tungstengraphics.com/wiki/index.php/Gallium3D" target="_parent">Gallium3D</a>
architecture.
</p>

<p>
<a href="http://www.tungstengraphics.com/" target="_parent">Tungsten Graphics</a>
is leading the project.
Two phases are planned.
First, to implement the framework for parallel rasterization using the Cell
SPEs, including texture mapping.
Second, to implement a full-featured OpenGL driver with support for GLSL, etc.
</p>


<H2>Source Code</H2>

<p>
The Cell driver source code is on the <code>gallium-0.1</code> branch of the
git repository.
After you've cloned the repository, check out the branch with:
</p>
<pre>
   git-checkout -b gallium-0.1 origin/gallium-0.1
</pre>
<p>
To build the driver you'll need the IBM Cell SDK (version 2.1 or 3.0).
To use the driver you'll need a Cell system, such as a PS3 running Linux,
or the Cell Simulator (untested, though).
</p>

<p>
If using Cell SDK 3.0, first edit configs/linux-cell and add
<code>-DSPU_MAIN_PARAM_LONG_LONG</code> to the SPU_CFLAGS.
</p>

<p>
To compile the code, run <code>make linux-cell</code>.
</p>

<p>
To use the library, make sure <code>LD_LIBRARY_PATH</code> points the Mesa/lib/
directory that contains <code>libGL.so</code>.
</p>

<p>
Verify that the Cell driver is being used by running <code>glxinfo</code>
and looking for:
<pre>
  OpenGL renderer string: Gallium 0.1, Cell on Xlib
</pre>


<H2>Driver Implementation Summary</H2>

<p>
Rasterization is parallelized across the SPUs in a tiled-based manner.
Batches of transformed triangles are sent to the SPUs (actually, pulled by from
main memory by the SPUs).
Each SPU loops over a set of 32x32-pixel screen tiles, rendering the triangles
into each tile.
Because of the limited SPU memory, framebuffer tiles are paged in/out of
SPU local store as needed.
Similarly, textures are tiled and brought into local store as needed.
</p>

<p>
More recently, vertex transformation has been parallelized across the SPUs
as well.
</p>


<H2>Status</H2>

<p>
As of February 2008 the driver supports smooth/flat shaded triangle rendering
with Z testing and simple texture mapping.
Simple demos like gears run successfully.
To test texture mapping, try progs/demos/texcyl (press right mouse button for
rendering options).
</p>


<H2>Contributing</H2>

<p>
If you're interested in contributing to the effort, familiarize yourself
with the code, join the <a href="lists.html">mesa3d-dev mailing list</a>,
and describe what you'd like to do.
</p>


</BODY>
</HTML>