Golly (from 'GPU' and 'Polly' [1]) is an LLVM-based static race detector for detecting races in affine CUDA kernels, based on the integer linear programming solver in isl [2]. Race freedom is not guaranteed, but as far as possible, we aim to minimize false positives in error reporting to prevent leading users on wild goose chases.
Golly was developed on Ubuntu 22.04 Jammy Jellyfish, which ships with Python 3.10. The required packages for Ubuntu 22.04 are as follows:
$ sudo apt install -y cmake llvm clang pkg-config libclang-dev libisl-devTo be completely certain, please refer to the install script.
With the dependencies installed, Golly can be built with the following commands:
$ git clone --recurse-submodules https://github.com/hhkit/golly
$ mkdir build && cd build
$ cmake ..
$ cmake --build .
For a quick demo without worrying about dependencies, try opening this repository in a codespace. Note that the container is rather large, so remember to destroy your codespace after use.
Once built, golly can be executed from the base directory on any .cu file.
$ ./golly tests/simple.cu Block and grid dimensions can be specified with --blockDim and --gridDim flags.
Note: Block/grid specification for individual files are currently broken due to development focus on bulk testing.
Command, once fixed
```bash $ ./golly tests/simple.cu --blockDim 256 --gridDim 1 ```Two benchmark suites have been provided, PolyBench/GPU [3] and the Gklee [4] test suite. To access these benchmarks, golly must be cloned with submodules. If this has not been done, execute the following command:
$ git submodule update --initOtherwise, a test suite can be tested with a .yaml configuration. Either of the following commands should work:
$ ./golly tests/polybenchGpu.yaml
$ ./golly tests/GkleeTests.yaml[1] T. Grosser, A. Groesslinger, C. Lengauer, "Polly - Performing polyhedral optimizations on a low-level intermediate representation" Parallel Processing Letters 2012 22:04, doi: 10.1142/S0129626412500107
[2] Verdoolaege, S. (2010). isl: An Integer Set Library for the Polyhedral Model. In: Fukuda, K., Hoeven, J.v.d., Joswig, M., Takayama, N. (eds) Mathematical Software – ICMS 2010. ICMS 2010. Lecture Notes in Computer Science, vol 6327. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-15582-6_49
[3] S. Grauer-Gray, L. Xu, R. Searles, S. Ayalasomayajula and J. Cavazos, "Auto-tuning a high-level language targeted to GPU codes," 2012 Innovative Parallel Computing (InPar), San Jose, CA, USA, 2012, pp. 1-10, doi: 10.1109/InPar.2012.6339595.
[4] Guodong Li, Peng Li, Geof Sawaya, Ganesh Gopalakrishnan, Indradeep Ghosh, and Sreeranga P. Rajan. 2012. GKLEE: concolic verification and test generation for GPUs. SIGPLAN Not. 47, 8 (August 2012), 215–224. https://doi.org/10.1145/2370036.2145844