|
GPUOcelot
|
|
GPUOcelot
|
Able to run various optimization passes over PTX modules. More...
#include <PTXOptimizer.h>
Public Types | |
| enum | RegisterAllocationType { LinearScan, InvalidRegisterAllocationType } |
The type of register allocator to use. More... | |
| enum | PassType { InvalidPassType = 0x0, RemoveBarriers = 0x1, ReverseIfConversion = 0x2, SubkernelFormation = 0x4, StructuralTransform = 0x8, MIMDThreadScheduling = 0x10, DeadCodeElimination = 0x20, SplitBasicBlocks = 0x40, SyncElimination = 0x80 } |
The possible PTX to PTX passes. More... | |
Public Member Functions | |
| PTXOptimizer () | |
| The constructor sets the defaults. | |
| void | optimize () |
| Performs the optimizations. | |
Public Attributes | |
| std::string | input |
| The input file being optimized. | |
| std::string | output |
| The output file being generated. | |
| RegisterAllocationType | registerAllocationType |
| The type of register allocation to perform. | |
| int | passes |
| The set of passes to run. | |
| unsigned int | registerCount |
| The number of registers to allocate. | |
| unsigned int | subkernelSize |
| The target sub-kernel size. | |
| unsigned int | basicBlockSize |
| The target basic block size. | |
| bool | cfg |
| Print out the CFG of optimized kernels. | |
Able to run various optimization passes over PTX modules.
| tools::PTXOptimizer::PTXOptimizer | ( | ) |
The constructor sets the defaults.
| void tools::PTXOptimizer::optimize | ( | ) |
Performs the optimizations.
| unsigned int tools::PTXOptimizer::basicBlockSize |
The target basic block size.
Print out the CFG of optimized kernels.
| std::string tools::PTXOptimizer::input |
The input file being optimized.
| std::string tools::PTXOptimizer::output |
The output file being generated.
The set of passes to run.
The type of register allocation to perform.
| unsigned int tools::PTXOptimizer::registerCount |
The number of registers to allocate.
| unsigned int tools::PTXOptimizer::subkernelSize |
The target sub-kernel size.
1.7.3
