hanchenye-llvm-project/polly
Hongbin Zheng 7aee737062 IndependentBLocks: Do not visit the same instruction twice when moving the
operand tree.

This patch fix Bug 13491, and the original "FIXME" in IndependentBlocks.cpp.

Patched by Kevin Fan<kevin.fan@gmail.com>.

llvm-svn: 161105
2012-08-01 08:46:11 +00:00
..
autoconf Detect the cuda library available. 2012-06-06 12:16:10 +00:00
cmake Replace CUDA data types with Polly's GPGPU data types. 2012-07-04 21:45:03 +00:00
docs
include Revert "Add preliminary implementation for GPGPU code generation." 2012-07-13 07:44:56 +00:00
lib IndependentBLocks: Do not visit the same instruction twice when moving the 2012-08-01 08:46:11 +00:00
test IndependentBLocks: Do not visit the same instruction twice when moving the 2012-08-01 08:46:11 +00:00
tools Update libGPURuntime to be dual licensed under MIT and UIUC license. 2012-07-06 10:40:15 +00:00
utils codegen.intrinsic: Update testcase to work with NVPTX backend 2012-07-03 08:18:34 +00:00
www Create a new directory before running the polly script 2012-07-24 16:58:57 +00:00
CMakeLists.txt Replace CUDA data types with Polly's GPGPU data types. 2012-07-04 21:45:03 +00:00
CREDITS.txt (Test commit for polly) 2011-07-16 13:30:03 +00:00
LICENSE.txt Happy new year 2012! 2012-01-01 08:16:56 +00:00
Makefile Revert "Fix a bug introduced by r153739: We are not able to provide the correct" 2012-04-11 07:43:13 +00:00
Makefile.common.in
Makefile.config.in Detect the cuda library available. 2012-06-06 12:16:10 +00:00
README Remove some empty lines 2011-10-04 06:56:36 +00:00
configure Detect the cuda library available. 2012-06-06 12:16:10 +00:00

README

Polly - Polyhedral optimizations for LLVM

Polly uses a mathematical representation, the polyhedral model, to represent and
transform loops and other control flow structures. Using an abstract
representation it is possible to reason about transformations in a more general
way and to use highly optimized linear programming libraries to figure out the
optimal loop structure. These transformations can be used to do constant
propagation through arrays, remove dead loop iterations, optimize loops for
cache locality, optimize arrays, apply advanced automatic parallelization, drive
vectorization, or they can be used to do software pipelining.