mesa3d/docs/llvmpipe.html

<HTML>

<TITLE>llvmpipe</TITLE>

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

<BODY>

<H1>Introduction</H1>

<p>
The Gallium llvmpipe driver is a software rasterizer that uses LLVM to
do runtime code generation.
Shaders, point/line/triangle rasterization and vertex processing are
implemented with LLVM IR which is translated to x86 or x86-64 machine
code.
Also, the driver is multithreaded to take advantage of multiple CPU cores
(up to 8 at this time).
It's the fastest software rasterizer for Mesa.
</p>


<h1>Requirements</h1>

<ul>
<li>
	 <p>An x86 or amd64 processor; 64-bit mode recommended.</p
   <p>
   Support for SSE2 is strongly encouraged.  Support for SSSE3, and SSE4.1 will
   yield the most efficient code.  The less features the CPU has the more
   likely is that you ran into underperforming, buggy, or incomplete code.
   </p>
   <p>
   See /proc/cpuinfo to know what your CPU supports.
   </p>
</li>
<li>
	 <p>LLVM: version 2.9 recommended; 2.6 or later required.</p>
   <b>NOTE</b>: LLVM 2.8 and earlier will not work on systems that support the
   Intel AVX extensions (e.g. Sandybridge).  LLVM's code generator will
   fail when trying to emit AVX instructions.  This was fixed in LLVM 2.9.
   </p>
   <p>
   For Linux, on a recent Debian based distribution do:
   </p>
<pre>
     aptitude install llvm-dev
</pre>
   For a RPM-based distribution do:
   </p>
<pre>
     yum install llvm-devel
</pre>

   <p>
	 For Windows you will need to build LLVM from source with MSVC or MINGW
	 (either natively or through cross compilers) and CMake, and set the LLVM
	 environment variable to the directory you installed it to.

   LLVM will be statically linked, so when building on MSVC it needs to be
   built with a matching CRT as Mesa, and you'll need to pass
   -DLLVM_USE_CRT_RELEASE=MTd for debug and checked builds,
   -DLLVM_USE_CRT_RELEASE=MTd for profile and release builds.

   You can build only the x86 target by passing -DLLVM_TARGETS_TO_BUILD=X86
   to cmake.
   </p>
</li>

<li>
   <p>scons (optional)</p>
</li>
</ul>


<h1>Building</h1>

To build everything on Linux invoke scons as:

<pre>
  scons build=debug libgl-xlib
</pre>

Alternatively, you can build it with GNU make, if you prefer, by invoking it as

<pre>
  make linux-llvm
</pre>

but the rest of these instructions assume that scons is used.

For Windows the procedure is similar except the target:

<pre>
  scons build=debug libgl-gdi
</pre>


<h1>Using</h1>

On Linux, building will create a drop-in alternative for libGL.so into

<pre>
  build/foo/gallium/targets/libgl-xlib/libGL.so
</pre>
or
<pre>
  lib/gallium/libGL.so
</pre>

To use it set the LD_LIBRARY_PATH environment variable accordingly.

For performance evaluation pass debug=no to scons, and use the corresponding
lib directory without the "-debug" suffix.

On Windows, building will create a drop-in alternative for opengl32.dll. To use
it put it in the same directory as the application. It can also be used by
replacing the native ICD driver, but it's quite an advanced usage, so if you
need to ask, don't even try it.


<h1>Profiling</h1>

To profile llvmpipe you should pass the options

<pre>
  scons build=profile <same-as-before>
</pre>

This will ensure that frame pointers are used both in C and JIT functions, and
that no tail call optimizations are done by gcc.

To better profile JIT code you'll need to build LLVM with oprofile integration.

<pre>
  /configure \
      --prefix=$install_dir \
      --enable-optimized \
      --disable-profiling \
      --enable-targets=host-only \
      --with-oprofile

  make -C "$build_dir"
  make -C "$build_dir" install

  find "$install_dir/lib" -iname '*.a' -print0 | xargs -0 strip --strip-debug
</pre>

The you should define

<pre>
  export LLVM=/path/to/llvm-2.6-profile
</pre>

and rebuild.


<h1>Unit testing</h1>

<p>
Building will also create several unit tests in
build/linux-???-debug/gallium/drivers/llvmpipe:
</p>

</ul>
<li> lp_test_blend: blending
<li> lp_test_conv: SIMD vector conversion
<li> lp_test_format: pixel unpacking/packing
</ul>

<p>
Some of this tests can output results and benchmarks to a tab-separated-file
for posterior analysis, e.g.:
</p>
<pre>
  build/linux-x86_64-debug/gallium/drivers/llvmpipe/lp_test_blend -o blend.tsv
</pre>


<h1>Development Notes</h1>

<ul>
<li>
  When looking to this code by the first time start in lp_state_fs.c, and
  then skim through the lp_bld_* functions called in there, and the comments
  at the top of the lp_bld_*.c functions.
</li>
<li>
  The driver-independent parts of the LLVM / Gallium code are found in
  src/gallium/auxiliary/gallivm/.  The filenames and function prefixes
  need to be renamed from "lp_bld_" to something else though.
</li>
<li>
  We use LLVM-C bindings for now. They are not documented, but follow the C++
  interfaces very closely, and appear to be complete enough for code
  generation. See
  http://npcontemplation.blogspot.com/2008/06/secret-of-llvm-c-bindings.html
  for a stand-alone example.  See the llvm-c/Core.h file for reference.
</li>
</ul>