[STYLE] run autopep8 and isort (#421)

Run:
```
isort ./python
autopep8 -i --ignore E501,E701,E731 $(find ./python/ -name '*.py')
```
with an `.isort.cfg` and then clean up a few warts. This PR should be a no-op; the idea is that this is all boring whitespace changes, and any config file changes will be in a different change to make it easier to review.

python/test/unit/runtime/test_cache.py

@@ -1,13 +1,16 @@
-import torch
-import triton
-from triton.code_gen import JITFunction
-import triton.language as tl
 import os
 import shutil
+
 import pytest
+import torch
+
+import triton
+import triton.language as tl
+from triton.code_gen import JITFunction
 
 tmpdir = ".tmp"
 
+
 @triton.jit
 def function_1(i):
     i = i + 1
@@ -20,18 +23,21 @@ def function_2(i):
     i = i + 1
     return i
 
+
 @triton.jit
 def kernel(X, i, BLOCK: tl.constexpr):
     i = i + 1
     i = function_1(i)
     tl.store(X, i)
 
+
 @triton.jit(do_not_specialize=["i"])
 def kernel_nospec(X, i, BLOCK: tl.constexpr):
     i = i + 1
     i = function_1(i)
     tl.store(X, i)
 
+
 def apply_src_change(target, old, new):
     delattr(kernel.fn, 'hash')
     delattr(function_1.fn, 'hash')
@@ -42,28 +48,34 @@ def apply_src_change(target, old, new):
     target.src = target.src.replace(new, old)
     return ret
 
+
 def test_nochange():
     baseline = kernel.cache_key
     updated = apply_src_change(kernel, 'i + 1', 'i + 1')
     assert baseline == updated
 
+
 def test_toplevel_change():
     baseline = kernel.cache_key
     updated = apply_src_change(kernel, 'i + 1', 'i + 2')
     assert baseline != updated
 
+
 def test_nested1_change():
     baseline = kernel.cache_key
     updated = apply_src_change(function_1, 'i + 1', 'i + 2')
     assert baseline != updated
 
+
 def reset_tmp_dir():
     os.environ["TRITON_CACHE_DIR"] = tmpdir
     if os.path.exists(tmpdir):
         shutil.rmtree(tmpdir)
 
+
 def test_reuse():
     counter = 0
+
     def inc_counter(key, binary, repr):
         nonlocal counter
         counter += 1
@@ -73,11 +85,12 @@ def test_reuse():
     for i in range(10):
         kernel[(1,)](x, 1, BLOCK=1024)
     assert counter == 1
-    
+
 
 @pytest.mark.parametrize('mode', ['enable', 'disable'])
 def test_specialize(mode):
     counter = 0
+
     def inc_counter(key, binary, repr):
         nonlocal counter
         counter += 1

python/test/unit/runtime/test_comm.py

@@ -1,9 +1,11 @@
-import torch
-import triton
-import pytest
 import subprocess
-import triton.language as tl
+
 import numpy as np
+import pytest
+import torch
+
+import triton
+import triton.language as tl
 
 
 def get_p2p_matrix():