GeezerVison

The World Wide Web, restored to 800 pixels of glory.

GeezerVison is showing github.com — gh-114058: The Tier2 Optimizer by Fidget-Spinner · Pull Request #114059 · python/cpython

Conversation

Fidget-Spinner commented Jan 14, 2024 •

edited

Loading

Uh oh!

There was an error while loading. Please reload this page.

Copy link

Copy Markdown

Member

Closes #114058

This PR turns on the optimizer for all uops. The tier 2 uops optimizer contains a few parts: the abstract interpreter, the IR, and the codegen.

The abstract interpreter does the following:

Value numbering for types (not on all expressions though, that is too expensive)
Type propagation
Constant propagation
Guard elimination
Poor man's loop invariant code motion for guards/loop duplication

Function inlining is left out for a future PR, as it's the most complex.

After analysis of the bytecode and doing all of the above, it emits optimized uops, and passes that to the executor.

When uops is enabled, **this passes the entire CPython test suite **. The significant milestone is that this is able to analyse and abstract interpret all CPython uops that we currently support. The other significant milestone is that this generates code that passes CPython's test suite.

Refleak tests will fail as well, as they need a design overhaul.

The design of this PR is here https://github.com/Fidget-Spinner/cpython_optimization_notes/blob/main/3.13/uops_optimizer.md

High level discussion here faster-cpython/ideas#648.

0-2% faster on Linux, 3% faster on macOS ARM64

Issue: Tier 2 optimizer's abstract interpreter #114058

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

bluss reacted with hooray emoji

erlend-aasland reacted with eyes emoji

All reactions

🎉 1 reaction
👀 1 reaction

Fidget-Spinner added 8 commits

January 14, 2024 07:20


          abstract interp

29db898


          the abstract interpreter

c1332cc


          cleanup

9d85c35


          run black

76cee0c


          the optimizer

b71aa06


          fix a whole bunch of bugs

f0e5dec


          properly handle runtime self_or_null

52e368f


          fix faulty assertion

a273a2f

Fidget-Spinner requested review from gvanrossum and markshannon as code owners

January 14, 2024 18:38

bedevere-app Bot added the awaiting core review label

Jan 14, 2024

Fidget-Spinner marked this pull request as draft

January 14, 2024 18:38

bedevere-app Bot removed the awaiting core review label

Jan 14, 2024

Fidget-Spinner changed the title ~~The Tier2 Optimizer's abstract interpreter~~ gh-114058: The Tier2 Optimizer's abstract interpreter

Jan 14, 2024

bedevere-app Bot mentioned this pull request

Jan 14, 2024

Tier 2 optimizer's abstract interpreter #114058

Closed

Fidget-Spinner commented Jan 14, 2024

Copy link

Copy Markdown

Member Author

Please give me some time to write out the proper docs explaining the abstract IR this uses.

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

Fidget-Spinner added the DO-NOT-MERGE label

Jan 14, 2024

Fidget-Spinner changed the title ~~gh-114058: The Tier2 Optimizer's abstract interpreter~~ gh-114058: The Tier2 Optimizer

Jan 14, 2024

Fidget-Spinner added 2 commits

January 15, 2024 12:07


          fix build

60a1d79


          fix all tests except test_capi and maybe test_ctypes

7077ad5

Fidget-Spinner commented Jan 16, 2024

Copy link

Copy Markdown

Member Author

All tests run with uops on now passes except for the following two:

test_capi - there's one test that counts memory allocation, which obviously fails because my optimizer allocates memory. I think I'm just going to skip this test when -Xuops is detected.
test_ctypes - this doesn't actually fail I just have no clue whether it passes because the linker on my system is broken.

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

Fidget-Spinner and others added 2 commits

January 16, 2024 22:39


          run black and re-enable tests

5169bf3


          📜🤖 Added by blurb_it.

0929bb8

brandtbucher self-requested a review

January 16, 2024 21:27

brandtbucher commented Jan 17, 2024

Copy link

Copy Markdown

Member

I think I'm just going to skip this test when -Xuops is detected.

Maybe we should just fix the test? This sort of seems to be kicking the can down the road, since eventually tier 2 will be on by default.

There's a without_optimizer helper in test.support. We also have other ways of getting and setting optimizers in _testinternalcapi if we're in a subprocess or something (check out test_opt.temporary_optimizer for an example).

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

brandtbucher commented Jan 17, 2024

Copy link

Copy Markdown

Member

1% slower on macOS (other platforms aren't building right now). 8% reduction in traces executed, but 3% increase in uops executed.

PGO failure on Windows:

Details

Running PGInstrument|x64 interpreter...
Using random seed: 1256678172
0:00:00 Run 44 tests sequentially
0:00:00 [ 1/44] test_array
0:00:05 [ 2/44] test_base64
0:00:07 load avg: 0.41 [ 3/44] test_binascii
0:00:07 load avg: 0.41 [ 4/44] test_binop
0:00:08 load avg: 0.41 [ 5/44] test_bisect
0:00:08 load avg: 0.42 [ 6/44] test_bytes
0:00:34 load avg: 0.50 [ 7/44] test_bz2
0:00:40 load avg: 0.49 [ 8/44] test_cmath
0:00:41 load avg: 0.48 [ 9/44] test_codecs
0:00:52 load avg: 0.46 [10/44] test_collections
0:01:03 load avg: 0.52 [11/44] test_complex
0:01:04 load avg: 0.53 [12/44] test_dataclasses
0:01:08 load avg: 0.56 [13/44] test_datetime
0:01:18 load avg: 0.56 [14/44] test_decimal
0:01:49 load avg: 0.61 [15/44] test_difflib -- test_decimal passed in 31.4 sec
0:01:59 load avg: 0.60 [16/44] test_embed
0:02:42 load avg: 0.33 [17/44] test_float -- test_embed passed in 42.7 sec
0:02:43 load avg: 0.34 [18/44] test_fstring
0:02:57 load avg: 0.34 [19/44] test_functools
Windows fatal exception: stack overflow

Thread 0x00002f90 (most recent call first):
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\libregrtest\win_utils.py", line 43 in _update_load

Current thread 0x000019dc (most recent call first):
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1[875](https://github.com/faster-cpython/benchmarking/actions/runs/7559538614/job/20583550792#step:10:876) in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  File "C:\actions-runner\_work\benchmarking\benchmarking\cpython\Lib\test\test_functools.py", line 1875 in fib
  ...

PGO failure on Linux:

Details

# Next, run the profile task to generate the profile information.
./python -m test --pgo --timeout=
Using random seed: [1366](https://github.com/faster-cpython/benchmarking/actions/runs/7559538614/job/20583550197#step:10:1367)[1376](https://github.com/faster-cpython/benchmarking/actions/runs/7559538614/job/20583550197#step:10:1377)73
0:00:00 load avg: 1.96 Run 44 tests sequentially
0:00:00 load avg: 1.96 [ 1/44] test_array
0:00:00 load avg: 1.89 [ 2/44] test_base64
0:00:01 load avg: 1.89 [ 3/44] test_binascii
0:00:01 load avg: 1.89 [ 4/44] test_binop
0:00:01 load avg: 1.89 [ 5/44] test_bisect
0:00:01 load avg: 1.89 [ 6/44] test_bytes
0:00:05 load avg: 1.89 [ 7/44] test_bz2
0:00:06 load avg: 1.82 [ 8/44] test_cmath
0:00:06 load avg: 1.82 [ 9/44] test_codecs
Fatal Python error: Segmentation fault

Current thread 0x00007fb926b78740 (most recent call first):
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/encodings/idna.py", line 37 in nameprep
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/encodings/idna.py", line 74 in ToASCII
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/encodings/idna.py", line 142 in ToUnicode
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/encodings/idna.py", line 222 in decode
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/test_codecs.py", line 1561 in test_builtin_decode_length_limit
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/unittest/case.py", line 589 in _callTestMethod
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/unittest/case.py", line 636 in run
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/unittest/case.py", line 692 in __call__
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/unittest/suite.py", line 122 in run
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/unittest/suite.py", line 84 in __call__
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/unittest/suite.py", line 122 in run
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/unittest/suite.py", line 84 in __call__
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/testresult.py", line 146 in run
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 57 in _run_suite
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 37 in run_unittest
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 132 in test_func
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 88 in regrtest_runner
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 135 in _load_run_test
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 178 in _runtest_env_changed_exc
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 278 in _runtest
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/single.py", line 306 in run_single_test
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/main.py", line 302 in run_test
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/main.py", line 336 in run_tests_sequentially
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/main.py", line 477 in _run_tests
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/main.py", line 509 in run_tests
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/main.py", line 672 in main
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/libregrtest/main.py", line 680 in main
  File "/home/ddfun/actions-runner/_work/benchmarking/benchmarking/cpython/Lib/test/__main__.py", line 2 in <module>
  File "<frozen runpy>", line 88 in _run_code
  File "<frozen runpy>", line 198 in _run_module_as_main

Extension modules: _testcapi, _testinternalcapi (total: 2)
Segmentation fault (core dumped)

Both look like recursion-related issues. The Windows one may be fixed on main as of yesterday.

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

Fidget-Spinner commented Jan 18, 2024

Copy link

Copy Markdown

Member Author

Thanks Brandt. Seems like the slowdown is due to bm_nbody, and there's some serious pessimization there for some reason. The only way such a massive slowdown could happen IMO, is if the executors are constantly being discarded and re-optimized again, since optimization is now not completely free (not too sure, will take a look).

Llike Mark said though, benchmark results aren't too important at the moment. Function inlining would be the most important optimization and that's missing from this PR, to be added in a future PR.

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.


          Merge remote-tracking branch 'upstream/main' into tier2_abstract_inte…

70ee73e

…rpreter

gvanrossum commented Jan 18, 2024

Copy link

Copy Markdown

Member

Nevertheless it would be wise to dig deeper into what goes on with bm_nbody. It may be an important canary. :-)

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.


          check for buffer overruns

7d66440

gvanrossum mentioned this pull request

Jan 31, 2024

Write about Tier 2 and JIT in "what's new 3.13" #114826

Merged

Fidget-Spinner added 23 commits

February 2, 2024 19:46


          Merge remote-tracking branch 'upstream/main' into tier2_abstract_inte…

07ba964

…rpreter


          Merge remote-tracking branch 'upstream/main' into tier2_abstract_inte…

50154ad

…rpreter


          Integrate the constant promoter

a0b58b1


          fix test

4573fca


          try fix memleak?

2096a8a


          Revert "try fix memleak?"

4ad2804

This reverts commit 2096a8a.


          fix memleak for real

224f6b4


          add more constant prop tests for promoted globals

b784617


          refactor


          Simplify codegen -- either types or consts, not both

6a6b11f


          cleanup, reduce memory usage even more

cfe1de0


          remove more memory allocations

5ad6211


          zero dynamic memory allocation

73daf91


          forgot return

cac3616


          streamline frame creation

e7df93a


          more cleanup

65fae9a


          fix bug in return code

3fec959


          fix on MSVC

dde7d12


          Add back constant evaluation

04c902b


          cleanup

05e93c2


          add type annotation

55f9bcb


          make const zapping more error-proof


          fix stack metadata

a8adada

gvanrossum reviewed

Feb 5, 2024

View reviewed changes

gvanrossum left a comment

Copy link

Copy Markdown

Member

There was a problem hiding this comment.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Here are some comments on the latest version which I (mostly) wrote before hearing about plans to merge with Mark's version. (I wonder if it might make sense to start a new PR for that, rather than carry the history of over 100 commits along?)

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

All reactions

Comment thread

Tools/cases_generator/tier2_abstract_generator.py

Comment on lines +131 to +135

		if var.name != MANGLED_NULL:
		out.emit(f"{var.name} = sym_init_unknown(ctx);\n")
		out.emit(f"if({var.name} == NULL) goto error;\n")
		if var.type_prop:
		out.emit(f"sym_set_type({var.name}, {var.type_prop[0]}, 0);\n")

gvanrossum Feb 4, 2024

Copy link

Copy Markdown

Member

There was a problem hiding this comment.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is an example of something that I think can be streamlined further. Instead of calling sym_init_unknown(ctx) followed by sym_set_type(...) there ought to be a single function you can call to create a symbol with a given type. Basically a bunch of convenience constructors.

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

All reactions

Comment thread

Include/internal/pycore_optimizer.h

Comment on lines +15 to +16

		// This the above + additional working space we need.
		#define UOP_MAX_TRACE_WORKING_LENGTH (UOP_MAX_TRACE_LENGTH * 2)

gvanrossum Feb 5, 2024

Copy link

Copy Markdown

Member

There was a problem hiding this comment.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't see this used in the code any more.

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

All reactions

Comment thread

Python/optimizer.c Resolved

Uh oh!

There was an error while loading. Please reload this page.

Comment thread

Python/optimizer_analysis.c Resolved

Uh oh!

There was an error while loading. Please reload this page.

Comment thread

Python/optimizer_analysis.c Resolved

Uh oh!

There was an error while loading. Please reload this page.

Comment thread

Python/optimizer_analysis.c Resolved

Uh oh!

There was an error while loading. Please reload this page.

Comment thread

Python/optimizer_analysis.c Resolved

Uh oh!

There was an error while loading. Please reload this page.

Comment thread

Python/optimizer_analysis.c Resolved

Uh oh!

There was an error while loading. Please reload this page.

Comment thread

Python/optimizer_analysis.c Resolved

Uh oh!

There was an error while loading. Please reload this page.

Comment thread

Python/optimizer_analysis.c

Comment on lines +381 to 389

		sym_set_type(sym, GUARD_TYPE_VERSION_TYPE, tp->tp_version_tag);
		}
		if (tp->tp_flags & Py_TPFLAGS_MANAGED_DICT) {
		PyDictOrValues dorv = *_PyObject_DictOrValuesPointer(obj);
		if(_PyDictOrValues_IsValues(dorv)) {
		sym_set_type(sym, GUARD_DORV_VALUES_TYPE, 0);
		}
		}
		if (tp == &PyLong_Type) {

gvanrossum Feb 5, 2024

Copy link

Copy Markdown

Member

There was a problem hiding this comment.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This could also use a longer explanation. What special cases are these two if blocks covering?

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

All reactions


          address reviews

ab60387

Fidget-Spinner commented Feb 6, 2024

Copy link

Copy Markdown

Member Author

Here are some comments on the latest version which I (mostly) wrote before hearing about plans to merge with Mark's version. (I wonder if it might make sense to start a new PR for that, rather than carry the history of over 100 commits along?)

Makes sense. I will close this PR and create a new one with the Mark's new DSL changes. Thanks for all your reviews!

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

Fidget-Spinner closed this

Feb 6, 2024

Fidget-Spinner mentioned this pull request

Feb 6, 2024

gh-114058: Foundations of the Tier2 redundancy eliminator #115085

Merged

Fidget-Spinner commented Feb 6, 2024

Copy link

Copy Markdown

Member Author

The successor PR with cleaned up history is at #115085.

All reactions

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

Reviewers

gvanrossum

gvanrossum left review comments

markshannon

Awaiting requested review from markshannon

markshannon is a code owner

brandtbucher

Awaiting requested review from brandtbucher

brandtbucher is a code owner

ericsnowcurrently

Awaiting requested review from ericsnowcurrently

ericsnowcurrently is a code owner

Uh oh!

There was an error while loading. Please reload this page.

Assignees

No one assigned

Labels

awaiting core review

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Tier 2 optimizer's abstract interpreter

Uh oh!

There was an error while loading. Please reload this page.

5 participants

Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.

You are visitor 01402503 — GeezerVison does not store the pages you look up.
Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required.

GeezerVison

Conversation

Fidget-Spinner commented Jan 14, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Fidget-Spinner commented Jan 14, 2024

Uh oh!

Fidget-Spinner commented Jan 16, 2024

Uh oh!

brandtbucher commented Jan 17, 2024

Uh oh!

brandtbucher commented Jan 17, 2024

Uh oh!

Fidget-Spinner commented Jan 18, 2024

Uh oh!

gvanrossum commented Jan 18, 2024

Uh oh!

gvanrossum left a comment

Choose a reason for hiding this comment

Uh oh!

gvanrossum Feb 4, 2024

Choose a reason for hiding this comment

Uh oh!

gvanrossum Feb 5, 2024

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

gvanrossum Feb 5, 2024

Choose a reason for hiding this comment

Uh oh!

Fidget-Spinner commented Feb 6, 2024

Uh oh!

Fidget-Spinner commented Feb 6, 2024

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Fidget-Spinner commented Jan 14, 2024 •

edited

Loading

Uh oh!

There was an error while loading. Please reload this page.