A Disk-based Highly Parallel Interprocedural Static Analysis Engine.

DiskSpa is simple to use,with a very straight-forward complication procedure.

Ensure that you have a recent version of the Boost library installed in your system. You can obtain the library from here.

First, download the entire DiskSpa source code into your machine. Next, edit the src/makefile to set the paths to the Boost library include files and lib files in your machine. Do the same for the src/run file. Finally, run the makefile in the src folder using make. DiskSpa should now be compiled and ready to run.

cd src
./run <graph_file> <grammar_file> <number_partitions> <memory_budget> <num_threads>

1.根据内存大小设置划分数，内存不足进行重划分。
2.更高效的调度算法。
3.设计其他的数据结构，与当前数组效率进行对比。
4.优化多个有序数组合并去重算法。