[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Felipe Rodrigues added the comment: Hi all! I'm seeing the same error that @pablogsal saw here, but I'm on Linux Mint 20.1 x86_64 (Kernel 5.4.0-77): ./python -m test test_os -R : -v == CPython 3.11.0a0 (heads/pr_26964:d375c08c75, Jul 3 2021, 07:47:01) [GCC 9.3.0] ==

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: I have not noticed it in a while, so I will close it for now unless I can reproduce it again. -- resolution: -> postponed stage: -> resolved status: open -> closed ___ Python tracker

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Ned Deily added the comment: @pablogsal, Is this still an issue for you? AFAIK, there have been no other reports of this behavior. -- ___ Python tracker ___

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: > WHat kind of file system is @test_28886_tmp/f1 created on? Apple_APFS -- ___ Python tracker ___

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: I will test this on other laptops to try to "bisect" configuration-wise, but for my investigation so far, it seems that the value in the (STRUCT_STAT) st->st_atime is directly wrong (is 1559932701 when the test fails and 1 when the test succeeds).

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Ned Deily added the comment: WHat kind of file system is @test_28886_tmp/f1 created on? -- ___ Python tracker ___ ___

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Carol Willing added the comment: @pablogsal My debug info looks similar to yours other than shell/terminal preferences. Similar specs on HW and OS. ./python.exe -m test test_os -R : -v Any other configs that you are frequently seeing the failures? --

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: Ok, this is the failure once I investigated with pdb: (Pdb) filename '@test_28886_tmp/f1' (Pdb) atime_ns 1002003000 (Pdb) mtime_ns 4005006000 (Pdb) set_time(filename, (atime_ns, mtime_ns)) (Pdb) st = os.stat(filename) (Pdb) st.st_atime, st.st_mtime

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: > For some reason, the files created for this test has a very weird inode > information. Ok, disregard what I am saying, that's actually part of the test. But I think this is related to the failure. I keep investigating. --

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: For some reason, the files created for this test has a very weird inode information: ❯ ll test_python_**/**/f* -rw-r--r-- 1 pgalindo3 staff 7B 1 Jan 1970 test_python_26370/@test_26370_tmp/f1 -- ___

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Change by Pablo Galindo Salgado : -- Removed message: https://bugs.python.org/msg344983 ___ Python tracker ___ ___ Python-bugs-list

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: The problem is that one of the filenames that the test creates has this inode information for some mysterious reason: ❯ ll @test_25508_tmp/f1 -rw-r--r-- 1 pgalindo3 staff 7B 1 Jan 1970 @test_25508_tmp/f1 --

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: The failure tends to be very reliable when running with -R, running the test in isolation or only one time normally succeeds. I will try to investigate more. The hardware is a MacBook Pro with Mojave 10.14.5, 2.2 GHz Intel Core i7 and 16 GB 2400 MHz

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Pablo Galindo Salgado added the comment: Python debug information CC.version: Configured with: --prefix=/Library/Developer/CommandLineTools/usr --with-gxx-include-dir=/usr/include/c++/4.2.1 Py_DEBUG: Yes (sys.gettotalrefcount() present) _decimal.__libmpdec_version__:

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Carol Willing added the comment: Confirming that they do pass on my Mojave system. test_utime (test.test_os.UtimeTests) ... ok test_utime_by_indexed (test.test_os.UtimeTests) ... ok ... test_utime_dir_fd (test.test_os.UtimeTests) ... ok Though I did have to go through some shenanigans

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Ned Deily added the comment: Also, what utime* results show up in ./configure output? I see checking utime.h usability... yes checking utime.h presence... yes checking for utime.h... yes -- ___ Python tracker

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Ned Deily added the comment: FWIW, it doesn't fail for me and I don't recall ever seeing a failure like this on macOS. What hardware? ./python.exe -m test.pythoninfo please! -- nosy: +ned.deily ___ Python tracker

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

Change by Karthikeyan Singaravelan : -- nosy: +p-ganssle ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue37195] test_utime fails on MacOS Mojave (Kernel Version 18.6.0:)

New submission from Pablo Galindo Salgado : ./python.exe -m test test_os -R : -v == FAIL: test_utime (test.test_os.UtimeTests) -- Traceback (most recent call