Imported Upstream version 5.18.0.161

Former-commit-id: 4db48158d3a35497b8f118ab21b5f08ac3d86d98

external/llvm/tools/opt-viewer/optpmap.py

@@ -0,0 +1,54 @@
+import sys
+import multiprocessing
+
+
+_current = None
+_total = None
+
+
+def _init(current, total):
+    global _current
+    global _total
+    _current = current
+    _total = total
+
+
+def _wrapped_func(func_and_args):
+    func, argument, should_print_progress = func_and_args
+
+    if should_print_progress:
+        with _current.get_lock():
+            _current.value += 1
+        sys.stdout.write('\r\t{} of {}'.format(_current.value, _total.value))
+        sys.stdout.flush()
+
+    return func(argument)
+
+
+def pmap(func, iterable, processes, should_print_progress, *args, **kwargs):
+    """
+    A parallel map function that reports on its progress.
+
+    Applies `func` to every item of `iterable` and return a list of the
+    results. If `processes` is greater than one, a process pool is used to run
+    the functions in parallel. `should_print_progress` is a boolean value that
+    indicates whether a string 'N of M' should be printed to indicate how many
+    of the functions have finished being run.
+    """
+    global _current
+    global _total
+    _current = multiprocessing.Value('i', 0)
+    _total = multiprocessing.Value('i', len(iterable))
+
+    func_and_args = [(func, arg, should_print_progress,) for arg in iterable]
+    if processes <= 1:
+        result = map(_wrapped_func, func_and_args, *args, **kwargs)
+    else:
+        pool = multiprocessing.Pool(initializer=_init,
+                                    initargs=(_current, _total,),
+                                    processes=processes)
+        result = pool.map(_wrapped_func, func_and_args, *args, **kwargs)
+
+    if should_print_progress:
+        sys.stdout.write('\r')
+    return result