[issue20876] python -m test test_pathlib fails

Roundup Robot added the comment: New changeset 40e8a8e83ed0 by Antoine Pitrou in branch '3.4': Issue #20876: correctly close temporary file in test.support.fs_is_case_insensitive() https://hg.python.org/cpython/rev/40e8a8e83ed0 New changeset 5406ed13bd6e by Antoine Pitrou in branch 'default': I

[issue20876] python -m test test_pathlib fails

Antoine Pitrou added the comment: I have applied the patch to 3.4 and default. Thank you! -- resolution: -> fixed stage: patch review -> resolved status: open -> closed ___ Python tracker _

[issue20876] python -m test test_pathlib fails

Changes by Antoine Pitrou : -- stage: -> patch review versions: +Python 3.5 ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue20876] python -m test test_pathlib fails

Isaac Schwabacher added the comment: Fixed a truncated line in the patch. -- Added file: http://bugs.python.org/file38347/test_support.patch ___ Python tracker ___ __

[issue20876] python -m test test_pathlib fails

Isaac Schwabacher added the comment: This behavior is caused by the way NFS clients implement unlinking open files: instead of unlinking an open file, the filesystem renames it to .nfs and unlinks it on close. (The search term you want is "silly rename".) The reason this problem appears

[issue20876] python -m test test_pathlib fails

R. David Murray added the comment: My guess about that is that it is an nfs work file in the directory, and nfs didn't delete it by the time the final testing-cleanup rmdir was happening. But I am *purely* guessing, since I haven't used NFS in probably 20 years. So, essentially, I'm suspectin

[issue20876] python -m test test_pathlib fails

Antoine Pitrou added the comment: NFS should certainly be supported as far as the stdlib goes, but I'm not sure the whole test suite is guaranteed to succeed. Also, in this case, the behaviour is weird, e.g.: == ERROR: test_gl

[issue20876] python -m test test_pathlib fails

R. David Murray added the comment: Hmm. I could have sworn that we had some tests that catered to NFS (either skips or other doges), but I can't find any searching for the string 'nfs', so perhaps not. Do we consider 'nfs' to be an 'unsupported platform'? That doesn't seem likely... --

[issue20876] python -m test test_pathlib fails

New submission from Jey Narasimhan: I was using Ubuntu 11.04 and Python 2.7.1+. The dev code was checkout from hg clone http://hg.python.org/cpython. Here is the full traceback of the crash report: == CPython 3.4.0rc1+ (default:3ae2cd85a908, Mar 9 2014, 10:38:52) [GCC 4.5.2] == Linux-2.6.38

[issue20876] python -m test test_pathlib fails

Ned Deily added the comment: Looks like the tests were being run on an NFS-based file system. It's not surprising that the some of the pathlib tests might fail when using NFS due to subtle differences in semantics and timing. Is it possible for you to run the tests again using a local file sy

[issue20876] python -m test test_pathlib fails

Changes by R. David Murray : -- type: crash -> behavior versions: +Python 3.4 -Python 2.7 ___ Python tracker ___ ___ Python-bugs-list