Skip to content

bpo-29587: Enable implicit exception chaining with gen.throw() #19823

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
May 2, 2020

Conversation

cjerdonek
Copy link
Member

@cjerdonek cjerdonek commented Apr 30, 2020

This is a new version of PR #19811 to sort out the buildbot failures.

It enables implicit exception chaining when calling generator.throw(exc) by setting exc.__context__.

https://bugs.python.org/issue29587

Before this commit, if an exception was active inside a generator
when calling gen.throw(), then that exception was lost (i.e. there
was no implicit exception chaining).  This commit fixes that.
Copy link
Member

@vstinner vstinner left a comment

Choose a reason for hiding this comment

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

test_unittest still crash on this PR. I tested on Linux (Fedora 32).

$ make clean
$ make
$ ./python -m test -v test_unittest 
(...)
test_patch_nested_autospec_repr (unittest.test.testmock.testpatch.PatchTest) ... ok
test_patch_object_keyword_args (unittest.test.testmock.testpatch.PatchTest) ... ok
test_patch_object_with_spec_as_boolean (unittest.test.testmock.testpatch.PatchTest) ... ok
test_patch_orderdict (unittest.test.testmock.testpatch.PatchTest) ... ok
test_patch_propagates_exc_on_exit (unittest.test.testmock.testpatch.PatchTest) ... Fatal Python error: Segmentation fault

Current thread 0x00007f886bd1c740 (most recent call first):
  File "/home/vstinner/python/master/Lib/unittest/mock.py", line 1524 in __exit__
  File "/home/vstinner/python/master/Lib/unittest/test/testmock/testpatch.py", line 1673 in __exit__
  File "/home/vstinner/python/master/Lib/contextlib.py", line 498 in __exit__
  File "/home/vstinner/python/master/Lib/unittest/mock.py", line 1323 in decoration_helper
  File "/home/vstinner/python/master/Lib/contextlib.py", line 135 in __exit__
  File "/home/vstinner/python/master/Lib/unittest/mock.py", line 1337 in patched
  File "/home/vstinner/python/master/Lib/unittest/case.py", line 201 in handle
  File "/home/vstinner/python/master/Lib/unittest/case.py", line 733 in assertRaises
  File "/home/vstinner/python/master/Lib/unittest/test/testmock/testpatch.py", line 1692 in test_patch_propagates_exc_on_exit
(...)

@bedevere-bot
Copy link

When you're done making the requested changes, leave the comment: I have made the requested changes; please review again.

@cjerdonek
Copy link
Member Author

Is there a way to mark this PR as "don't merge" for now? This is the same PR as before, so it has the same issue. The Azure CI job seems not to be running, which is what exhibited the failure before.

test_unittest still crash on this PR. I tested on Linux (Fedora 32).

Yes, this is the same failure as before. Everything passes locally on my Mac, so it will take me more time to look into this.

@vstinner vstinner changed the title bpo-29587: Enable implicit exception chaining with gen.throw() [WIP] bpo-29587: Enable implicit exception chaining with gen.throw() Apr 30, 2020
@vstinner vstinner marked this pull request as draft April 30, 2020 22:49
@vstinner
Copy link
Member

vstinner commented Apr 30, 2020

Is there a way to mark this PR as "don't merge" for now?

Hum, I wasn't sure, so I did the 3 tricks that I know:

  • Add "do no merge label"
  • Add "[WIP]" in the PR title
  • Convert the PR to a draft

It seems like only the conversion to a draft technically prevents someone to merge the PR my mistake.

@cjerdonek
Copy link
Member Author

Looks like that will cover it. ;)

@cjerdonek
Copy link
Member Author

cjerdonek commented May 1, 2020

Okay, I believe my latest changes fix the buildbot failures from before.

Also, for future reference, below is a script I came up with that isolates the behavior difference between Mac and Fedora, with my prior version of the PR:

import re
import sys


def f():
    exc, exc_value, tb = sys.exc_info()
    print(f'CLEARING FRAME: {tb.tb_frame!r}')
    tb.tb_frame.clear()

def g():
    # Uncommenting the following line caused the tb_frame.clear() line
    # above to exhibit the following platform-specific behavior:
    # 1) On Mac, this is logged to stderr
    #  > TypeError: print_exception(): Exception expected for value,
    #    NoneType found
    # 2) On Fedora 32, the following error happens:
    #  > Fatal Python error: Segmentation fault
    data = re.compile('xxx')

    try:
        yield
    except Exception:
        f()

gen = g()
gen.send(None)
gen.throw(ValueError)

Maybe this suggests an issue elsewhere in Python's code base.

Should this PR also get a What's New?

@cjerdonek cjerdonek marked this pull request as ready for review May 1, 2020 08:43
@cjerdonek cjerdonek changed the title [WIP] bpo-29587: Enable implicit exception chaining with gen.throw() bpo-29587: Enable implicit exception chaining with gen.throw() May 1, 2020
@cjerdonek
Copy link
Member Author

I have made the requested changes; please review again.

@bedevere-bot
Copy link

Thanks for making the requested changes!

@vstinner: please review the changes made to this pull request.

@bedevere-bot bedevere-bot requested a review from vstinner May 1, 2020 08:47
Copy link
Member

@vstinner vstinner left a comment

Choose a reason for hiding this comment

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

LGTM. "./python -m test -j0 -r" does no longer crash on my Fedora.

I understood that passing NULL value to _PyErr_ChainExceptions() created the bug? Maybe add an assertion to prevent the situation to happen again?

@cjerdonek
Copy link
Member Author

Thank you.

Regarding an assertion, the thing is that it looks like _PyErr_ChainExceptions() is designed to accept a NULL value. So maybe in the future it can be made to work. (That's why I added an XXX comment.) So I wouldn't want to add an assertion making it seem like it always has to be like that. It could just something about how things are set up right now that was causing the crash.

Would you still like me to add an assertion and/or perhaps a code comment explaining this possibility?

@cjerdonek cjerdonek merged commit 0204726 into python:master May 2, 2020
@bedevere-bot
Copy link

@cjerdonek: Please replace # with GH- in the commit message next time. Thanks!

@cjerdonek cjerdonek deleted the fix-issue-29587-2 branch May 2, 2020 01:16
cjerdonek added a commit that referenced this pull request May 3, 2020
This is a follow-up to GH-19823 that removes the check that the
exception value isn't NULL, prior to calling _PyErr_ChainExceptions().
This enables implicit exception chaining for gen.throw() in more
circumstances.

The commit also adds a test that a particular code snippet involving
gen.throw() doesn't crash.  The test shows why the new
`gi_exc_state.exc_type != Py_None` check that was added is necessary.
Without the new check, the code snippet (as well as a number of other
tests) crashes on certain platforms (e.g. Fedora but not Mac).
@cjerdonek
Copy link
Member Author

For future reference, I was able to remove the NULL check in a later PR (#19877): 21893fb

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants