[Python-Dev] Accepting PEP 456 (Secure hash algorithm) (original) (raw)
Victor Stinner victor.stinner at gmail.com
Wed Nov 20 12:29:33 CET 2013
- Previous message: [Python-Dev] Accepting PEP 456 (Secure hash algorithm)
- Next message: [Python-Dev] Accepting PEP 456 (Secure hash algorithm)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
2013/11/20 Christian Heimes <christian at python.org>:
The PEP has landed in revision http://hg.python.org/cpython/rev/adb471b9cba1 . I don't expect any test failures as I have tested the PEP on a lot of platforms. The new code compiles and passes its tests on Linux, Windows, BSD, HUPX, Solaris with all supported CPUs (little and big endian, 32 bit and 64 bit).
It looks like dict, set and frozenset representation (repr(...)) now depends on the platform (probably 32 bit vs 64 bit), even if PYTHONHASHSEED is set. I don't know if it's an issue or not.
This is why test_gdb was failing. I fixed test_gdb in: http://hg.python.org/cpython/rev/11cb1c8faf11
The fix runs Python with a fixed PYTHONHASHSEED to get repr(value), and then compare it to the value from gdb.
Victor
- Previous message: [Python-Dev] Accepting PEP 456 (Secure hash algorithm)
- Next message: [Python-Dev] Accepting PEP 456 (Secure hash algorithm)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]