[issue44363] Address sanitizer (gcc version) is generating false positives

miss-islington added the comment: New changeset 0895e62c9b4f03b83120d44cb32587804084e0e2 by Miss Islington (bot) in branch '3.10': bpo-44363: Get test_capi passing with address sanitizer (GH-26639) https://github.com/python/cpython/commit/0895e62c9b4f03b83120d44cb32587804084e0e2 --

[issue44363] Address sanitizer (gcc version) is generating false positives

miss-islington added the comment: New changeset 865269458fde83f58f056d02a777e4328c763880 by Miss Islington (bot) in branch '3.9': bpo-44363: Get test_capi passing with address sanitizer (GH-26639) https://github.com/python/cpython/commit/865269458fde83f58f056d02a777e4328c763880 --

[issue44363] Address sanitizer (gcc version) is generating false positives

Change by miss-islington : -- pull_requests: +25228 pull_request: https://github.com/python/cpython/pull/26642 ___ Python tracker ___

[issue44363] Address sanitizer (gcc version) is generating false positives

Change by miss-islington : -- nosy: +miss-islington nosy_count: 3.0 -> 4.0 pull_requests: +25227 pull_request: https://github.com/python/cpython/pull/26641 ___ Python tracker

[issue44363] Address sanitizer (gcc version) is generating false positives

Pablo Galindo Salgado added the comment: New changeset 31aa0dbff4c1d39c9d77c6c8f4a61d0e46c1268b by Mark Shannon in branch 'main': bpo-44363: Get test_capi passing with address sanitizer (GH-26639) https://github.com/python/cpython/commit/31aa0dbff4c1d39c9d77c6c8f4a61d0e46c1268b --

[issue44363] Address sanitizer (gcc version) is generating false positives

Change by Mark Shannon : -- pull_requests: +25226 pull_request: https://github.com/python/cpython/pull/26639 ___ Python tracker ___

[issue44363] Address sanitizer (gcc version) is generating false positives

Mark Shannon added the comment: It looks like I've been a bit unfair to the address sanitizer. It does appear to produce incorrect locations sometimes, but that's not really a false positive and the reports are generally useful. -- ___ Python

[issue44363] Address sanitizer (gcc version) is generating false positives

Eric Snow added the comment: This was my bad. I only partially fixed original problem when un-reverting 3fa63e in gh-26609. I've merged the rest of the fix in gh-26626. Sorry about that. -- nosy: +eric.snow ___ Python tracker

[issue44363] Address sanitizer (gcc version) is generating false positives

Pablo Galindo Salgado added the comment: >From the gdb session looks to me that you are reading freed or invalid memory >from the dictionary keys. Well, I'm quite sure you are doing that, but I don't >know why they is happening. -- ___ Python

[issue44363] Address sanitizer (gcc version) is generating false positives

Mark Shannon added the comment: This line seems to be responsible for most of the failures: https://github.com/python/cpython/blob/main/Objects/frameobject.c#L985 Which does appear to be a true positive. -- ___ Python tracker

[issue44363] Address sanitizer (gcc version) is generating false positives

Pablo Galindo Salgado added the comment: commit e858ea15718709bc8ec3c13bd8451ff7d62cbe80 (HEAD -> specialize-load-attr) Author: Mark Shannon Date: Wed Jun 9 09:33:13 2021 +0100 Assert that address is in bounds (ASAN thinks it might not be). diff --git a/Python/ceval.c b/Python/ceval.c

[issue44363] Address sanitizer (gcc version) is generating false positives

Mark Shannon added the comment: What commit are you running that on? -- ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue44363] Address sanitizer (gcc version) is generating false positives

Pablo Galindo Salgado added the comment: This is not a false positive. I break into gdb at the moment the sanitizer makes the report and inspecting the values that apparently are wrong. I did that by breaking into __sanitizer::ColorizeReports which is called for making the report. THis is

[issue44363] Address sanitizer (gcc version) is generating false positives

Pablo Galindo Salgado added the comment: > It seems like ceval.c:1600 upsets the sanitizer, at least for gcc. Even if you don't trust the sanitizer, you can also use valgrind, it shows the same error: ❯ valgrind ./python -m test test_lib2to3 ==27010== Memcheck, a memory error detector

[issue44363] Address sanitizer (gcc version) is generating false positives

Pablo Galindo Salgado added the comment: > Do you think it would be it feasible to run the address sanitizer on all PRs, > so that we can keep it passing? That's a good idea,let me see how easy doing that would be -- ___ Python tracker

[issue44363] Address sanitizer (gcc version) is generating false positives

Mark Shannon added the comment: I still get quite a few failures on the main branch. It seems like ceval.c:1600 upsets the sanitizer, at least for gcc. There isn't anything wrong with that line, but as I plan to change it anyway I guess it doesn't matter: