[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Bug-gsl] Some GSL-1.15 Tests check fail with the Visual c++ 2010 Expres
From: |
T . Dagobert |
Subject: |
[Bug-gsl] Some GSL-1.15 Tests check fail with the Visual c++ 2010 Express IDE |
Date: |
Thu, 20 Dec 2012 16:27:00 +0000 (UTC) |
User-agent: |
Loom/3.14 (http://gmane.org/) |
Hello.
I try to validate the installation of GSL-1.15 on Microsoft
Visual C++ 2010 Express (on Windows 7) for a scientific project (www.ipol.im).
I have followed the instructions described
in the converter (gsl-1.15-vc10.zip) of Brian Gladman
(http://gladman.plushost.co.uk/oldsite/computing/gnu_scientific_library.php)
As recommended,
I used the Microsoft SDK 7.1 compiler instead of the native v100 compiler for
compiling
the static and dynamic versions of the GSL both on 32 and 64 bits. Except two
compilation errors which I have been encountered and that I have fixed from
explanations in
http://4fire.wordpress.com/2012/03/18/gsl-1-15-building-with-visual-studio-
2010/
both cblas.dll, gsl.dll (resp. cblas.lib, gsl.lib) seem to have been well
built in {Debug, Release} x {Win32, x64} versions (so eight differents
versions).
However when I run the tests check in the {Debug} x {Win32, x64} versions,
two errors occur : with testvector.exe and testmatrix.exe tests (see below).
These errors seem to come from file test_complex_source.c. I though this error
was the same reported in http://blog.gmane.org/gmane.comp.lib.gsl.bugs/page=2
however, Brian Gladman's converter is still fixed with that.
I precise that there is no error in the {Release} x {Win32, x64} versions. But
as I have read from differents MSVC forums that even if there is no Release
errors while Debug errors occured, the Release DLL's are not safe and could
crash.
As I am not an expert at all in Windows, I have three questions :
- do you have any fix and a possible explanation ?
- do you think I can validate the GSL-1.15 installation from the Release test
check even if the Debug test failed ?
- do you think such an error could come with the professional Visual Studio
2010 IDE ?
Best regards.
T. Dagobert
File testvector.out reports :
FAIL: gsl_vector_set traps index below lower bound stride=1, N=1 [28344]
FAIL: gsl_vector_set traps index above upper bound stride=1, N=1 [28345]
FAIL: gsl_vector_set traps index at upper bound stride=1, N=1 [28346]
FAIL: gsl_vector_get traps index below lower bound stride=1, N=1 [28347]
FAIL: gsl_vector_get traps index above upper bound stride=1, N=1 [28349]
FAIL: gsl_vector_get traps index at upper bound stride=1, N=1 [28351]
FAIL: gsl_vector_float_set traps index below lower bound stride=1, N=1 [28353]
FAIL: gsl_vector_float_set traps index above upper bound stride=1, N=1 [28354]
FAIL: gsl_vector_float_set traps index at upper bound stride=1, N=1 [28355]
FAIL: gsl_vector_float_get traps index below lower bound stride=1, N=1 [28356]
FAIL: gsl_vector_float_get traps index above upper bound stride=1, N=1 [28358]
FAIL: gsl_vector_float_get traps index at upper bound stride=1, N=1 [28360]
File testmatrix.out reports :
FAIL: gsl_matrix_set traps 1st index above upper bound [751]
FAIL: gsl_matrix_set traps 2nd index above upper bound [752]
FAIL: gsl_matrix_set traps 1st index at upper bound [753]
FAIL: gsl_matrix_set traps 2nd index at upper bound [754]
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [Bug-gsl] Some GSL-1.15 Tests check fail with the Visual c++ 2010 Express IDE,
T . Dagobert <=