[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Change by Mark Lawrence : -- nosy: -BreamoreBoy ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Mark Lawrence added the comment: Can this be reproduced on 3.4/5? -- nosy: +BreamoreBoy type: - behavior versions: +Python 3.4, Python 3.5 -Python 3.2 ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue12589

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: what information do you need to advance on this bug? It would be easier to debug if I had access to OpenBSD on a PowerPC host. -- ___ Python tracker rep...@bugs.python.org

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Remi Pointel pyt...@xiri.fr added the comment: Hi, what information do you need to advance on this bug? Cheers, Remi. -- ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue12589 ___

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Landry Breuil lan...@openbsd.org added the comment: floatobject.c is compiled with cc -pthread -c -fno-strict-aliasing -O2 -pipe -Wall -DTHREAD_STACK_SIZE=0x2 -fPIC -O2 -pipe -Wall -I. -IInclude -I./Include -fPIC -DPy_BUILD_CORE -o Objects/floatobject.o Objects/floatobject.c

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Changes by Landry Breuil lan...@openbsd.org: Added file: http://bugs.python.org/file22854/pyconfig.h ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue12589 ___

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Changes by Landry Breuil lan...@openbsd.org: Added file: http://bugs.python.org/file22855/math.h ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue12589 ___

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Landry Breuil lan...@openbsd.org added the comment: Interestingly, after an update of the system i can't build python 3.2.1 anymore : /usr/include/util.h:102: error: conflicting types for 'openpty' Include/pyport.h:634: error: previous declaration of 'openpty' was here /usr/include/util.h:105:

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Changes by Landry Breuil lan...@openbsd.org: Added file: http://bugs.python.org/file22857/build-macppc.log ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue12589 ___

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: The problem is in PyLong_FromDouble(): if (Py_IS_INFINITY(dval)) is evaluated as false, whereas dval *is* infinity. Possible causes: - Py_IS_INFINITY is not defined as # define Py_IS_INFINITY(X) isinf(X) (issue with the

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Landry Breuil lan...@openbsd.org added the comment: $grep -r '#define Py_IS_INF' . PC/pyconfig.h:#define Py_IS_INFINITY(X) (!_finite(X) !_isnan(X)) isinf(float('inf')) Breakpoint 2, __isinf (d=1.0604798301039825e-314) at /usr/src/lib/libc/gen/isinf.c:30 30 in

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Mark Dickinson dicki...@gmail.com added the comment: The '1.06...e-314' number in the gdb output is interesting: it indicates a byte-ordering issue, though maybe that issue is only pertinent to gdb itself. On a little-endian machine: struct.pack('d', 1.0604798301039825e-314)

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Mark Dickinson dicki...@gmail.com added the comment: Question: does this test also fail after configuring with the --with-pydebug flag? (Which I *think* should turn compiler optimizations off, amongst other things.) -- ___ Python tracker

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: $grep -r '#define Py_IS_INF' . PC/pyconfig.h:#define Py_IS_INFINITY(X) (!_finite(X) !_isnan(X)) The PC/ directory is specific to Windows. Py_IS_INFINITY should be defined in Include/pymath.h (at least, in the 3.2 branch)

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: If gdb has an endian issue, you may also try print /x d in the gdb shell. Oh, forget me: /x converts the argument to an integer... -- ___ Python tracker rep...@bugs.python.org

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Landry Breuil lan...@openbsd.org added the comment: Py_IS_INFINITY is defined as #ifndef Py_IS_INFINITY # if defined HAVE_DECL_ISINF HAVE_DECL_ISINF == 1 #define Py_IS_INFINITY(X) isinf(X) # else #define Py_IS_INFINITY(X) ((X)\

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: I'm not sure that your version of gdb understands macros. You may have to set a breakpoint on __isinf. Compile Python with -g -ggdb helps gdb. Py_IS_INFINITY is may not defined as # define Py_IS_INFINITY(X) isinf(X). To check

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Changes by STINNER Victor victor.stin...@haypocalc.com: -- nosy: +mark.dickinson ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue12589 ___ ___

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: Is HAVE_DECL_ISINF defined in pyconfig.h? PyLong_FromDouble() uses Py_IS_INFINITY(x): -- /* Py_IS_INFINITY(X) * Return 1 if float or double arg is an infinity, else 0. * Caution: *X is

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Mark Dickinson dicki...@gmail.com added the comment: The result of: struct.pack(d, float(inf)) would also be interesting. I'd expect to see: '\x7f\xf0\x00\x00\x00\x00\x00\x00' -- ___ Python tracker rep...@bugs.python.org

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Landry Breuil lan...@openbsd.org added the comment: import struct struct.pack(d, float(inf)) b'\x7f\xf0\x00\x00\x00\x00\x00\x00' struct.pack(f, float(inf)) b'\x7f\x80\x00\x00' And yes, HAVE_DECL_ISINF is defined to 1 in pyconfig.h -- ___ Python

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: You may try: $ ./python import ctypes import ctypes.util libc = ctypes.cdll.LoadLibrary(ctypes.util.find_library('c')) isinf = libc.isinf isinf.argtypes = (ctypes.c_double,) isinf(0.0) 0 isinf(float('inf')) 1 (try

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Landry Breuil lan...@openbsd.org added the comment: Python 3.2.1 (default, Jul 18 2011, 10:56:33) [GCC 4.2.1 20070719 ] on openbsd4 Type help, copyright, credits or license for more information. import ctypes import ctypes.util libc = ctypes.cdll.LoadLibrary(ctypes.util.find_library('c'))

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

Landry Breuil lan...@openbsd.org added the comment: $python3.2 Python 3.2.1 (default, Jul 18 2011, 10:56:33) [GCC 4.2.1 20070719 ] on openbsd4 int(float('inf')) 0 $sysctl hw hw.machine=macppc --

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

New submission from rpointel pyt...@xiri.fr: Hello, the test test_nan_inf failed on OpenBSD on powerpc architecture. It works fine on amd64 and sparc64. Don't hesitate if you need more informations. Details: Re-running test 'test_long' in verbose mode test__format__ (test.test_long.LongTest)

[issue12589] test_long.test_nan_inf() failed on OpenBSD (powerpc)

STINNER Victor victor.stin...@haypocalc.com added the comment: What is the result of int(float('inf')) ? -- nosy: +haypo ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue12589 ___