[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Help-gsl] Building gsl-1.15 under MinGW
From: |
Richard Haney |
Subject: |
[Help-gsl] Building gsl-1.15 under MinGW |
Date: |
Sat, 22 Sep 2012 22:24:03 -0700 (PDT) |
I have a similar problem to one described in the archives, and I have more
information about the problem.
This relates to a thread labeled:
Re: [Help-gsl] Building gsl-1.15 under MinGW
and can be found by searching the archives for 'HAVE_IEEE_COMPARISONS'.
The problem appears to be a failure of configure to modify the config.h file
(essentially copied from config.h.in) with updated macro definitions even
though configure seems otherwise to complete testing and running successfully.
I tried to build gsl-1.15 using cygwin bash for my command-line processor and
gcc.exe (tdm64-1) 4.6.1, but the header files subdirectory gsl all had nothing
but "!<symlink>ÿþ."
Evidently, the cygwin package was failing to handle symbolic links properly.
My command line for configure was
./configure CFLAGS="-g -O2 -march=native -fexceptions"
--prefix=/c/gsl-1.15_64_-g_O2 &> gsl-1.15_configure.out
So I installed MinGW with MSYS, and started with a fresh gsl-1.15 source
hierarchy.
I ran configure again, but this time using the MSYS command processor, and this
time I encountered the problem with the bad, non-updated config.h file
described in the above mentioned thread. But it turns out that the config.h
file created under cygwin appeared to be fully populated with a variety of
macro definitions, including the definition "#define HAVE_IEEE_COMPARISONS 1".
I copied this apparently good config.h to my gsl-1.15 build directory in place
of the bad config.h, and everything ran fine from then on, including the "make
test" and "make install".
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [Help-gsl] Building gsl-1.15 under MinGW,
Richard Haney <=