src/gallium/drivers/llvmpipe/README

   1 LLVMPIPE -- a fork of softpipe that employs LLVM for code generation.
   2
   3
   4 Status
   5 ======
   6
   7 Done so far is:
   8
   9 - TGSI -> LLVM fragment shader translation
  10   - same level of support as the TGSI SSE2 exec machine
  11   - texture sampling via an intrinsic call
  12   - done in SoA
  13   - input interpolation also code generated
  14
  15 - blend -> LLVM (including logic ops)
  16   - SoA and AoS, but only the former used
  17
  18 - code is generic
  19   - intermediates can be vectors of floats, ubytes, fixed point, etc, and of
  20     any width and length
  21   - not all operations are implemented for these types yet though
  22
  23 Most mesa/progs/demos/* work. Speed is on par with Keith's softpipe-opt branch,
  24 which includes hand written fast implementations for common cases.
  25
  26 To do (probably by this order):
  27 - code generate the rest of the fragment pipeline, namely the
  28   depth/alpha/stencil state
  29 - concatenate the fragment pipeline (shader + depth/stencil/alpha + blend) in a
  30   single function
  31 - code generate texture sampling
  32 - translate TGSI control flow instructions
  33 - code generate the triangle setup and rasterization
  34
  35
  36 Requirements
  37 ============
  38
  39  - Linux
  40
  41  - udis86, http://udis86.sourceforge.net/
  42
  43      git clone git://udis86.git.sourceforge.net/gitroot/udis86
  44      cd udis86
  45      ./configure --with-pic
  46      make
  47      sudo make install
  48
  49  - LLVM 2.5. On Debian based distributions do:
  50
  51      aptitude install llvm-dev
  52
  53    There is a typo in one of the llvm-dev 2.5 headers, that causes compilation
  54    errors in the debug build:
  55
  56      --- /usr/include/llvm-c/Core.h.orig        2009-08-10 15:38:54.000000000 +0100
  57      +++ /usr/include/llvm-c/Core.h     2009-08-10 15:38:25.000000000 +0100
  58      @@ -831,7 +831,7 @@
  59         template<typename T>
  60         inline T **unwrap(LLVMValueRef *Vals, unsigned Length) {
  61           #if DEBUG
  62      -    for (LLVMValueRef *I = Vals, E = Vals + Length; I != E; ++I)
  63      +    for (LLVMValueRef *I = Vals, *E = Vals + Length; I != E; ++I)
  64             cast<T>(*I);
  65           #endif
  66           return reinterpret_cast<T**>(Vals);
  67
  68  - A x86 or amd64 processor with support for sse2, sse3, and sse4.1 SIMD
  69    instructions. This is necessary because we emit several SSE intrinsics for
  70    convenience. See /proc/cpuinfo to know what your CPU supports.
  71
  72  - scons (although it should be straightforward to fix the Makefiles as well)
  73
  74
  75 Building
  76 ========
  77
  78 To build everything invoke scons as:
  79
  80   scons debug=yes statetrackers=mesa drivers=llvmpipe winsys=xlib dri=false -k
  81
  82
  83 Using
  84 =====
  85
  86 Building will create a drop-in alternative for libGL.so. To use it set the
  87 environment variables:
  88
  89   export LD_LIBRARY_PATH=$PWD/build/linux-x86-debug/lib:$LD_LIBRARY_PATH
  90   export LD_LIBRARY_PATH=$PWD/build/linux-x86_64-debug/lib:$LD_LIBRARY_PATH
  91
  92
  93 Unit testing
  94 ============
  95
  96 Building will also create several unit tests in
  97 build/linux-???-debug/gallium/drivers/llvmpipe:
  98
  99  - lp_test_blend: blending
 100  - lp_test_conv: SIMD vector conversion
 101  - lp_test_format: pixel unpacking/packing
 102
 103