Some Multiprocessing Added #2
Conversation
|
Thank you for the PR! I think it would be great to have a |
| args_tuple_list.append([config_path, crashing_input, extra_args, True, True, results_queue]) | ||
| iterator += 1 | ||
|
|
||
| p = mp.Process(target=multi_run_target_manager, args=(args_tuple_list,)) |
There was a problem hiding this comment.
I think mp.Process does not limit the number of concurrent processes. We want to be able to limit this number so that other fuzzing runs are not impacted by the trace generation taking up all computation.
multiprocessing.Pool should be able to limit this number.
Given that, we can add a -n argument similar to the fuzzware pipeline -n <num_cores> argument which allows setting the max number of cores/parallel processes used by trace generation.
| job_args_multi.append((str(config_path), str(input_path), bbl_trace_path, ram_trace_path, mmio_trace_path, | ||
| bbl_set_path, mmio_set_path, extra_args, not verbose, bbl_hash_path, processed_queue)) | ||
| iteration += 1 | ||
| p = mp.Process(target=multi_proc_manager, args=(gen_traces,job_args_multi)) |
There was a problem hiding this comment.
Same here. We want to be able to limit the number of cores/parallel processes used
No description provided.