[Scilab-Dev] 5.3.0 built with Atlas but with_atlas() returns F
bv
bvl at btconnect.com
Tue Feb 8 00:08:06 CET 2011
On Monday 07 February 2011 20:46:24 Dean S. Messing wrote:
> I'm trying to build the Image Processing Devel package under linux
> (outside of ATOMS). It immedately returns with an error saying I need
> to install Atlas. Problem is that I built scilab (GIT version from 27th
> Jan. 11) with Atlas. Note the 5th from last line.
>
> In fact here are the relevant lines from the output of the ./configure
>
> checking for msgcat... /usr/bin/msgcat
> checking if BLAS, ATLAS or MKL is available...
> checking how to get verbose linking output from gfortran... -v
> checking for Fortran 77 libraries of gfortran... -L/usr/lib64/atlas
> -L/usr/lib/gcc/x86_64-redhat-linux/4.4.5
> -L/usr/lib/gcc/x86_64-redhat-linux/4.4.5/../../../../lib64 -L/lib/../lib64
> -L/usr/lib/../lib64 -L/usr/lib/gcc/x86_64-redhat-linux/4.4.5/../../..
> -lpthread -ldl -lcurses -lgfortranbegin -lgfortran -lm checking for dummy
> main to link with Fortran 77 libraries... none
> checking for Fortran 77 name-mangling scheme... lower case, underscore, no
> extra underscore checking for sgemm_... no
> checking for ATL_xerbla in -latlas... yes
> checking for sgemm_ in -lf77blas... yes
> checking for cblas_dgemm in -lcblas... yes
> Atlas found
> checking if LAPACK is available...
> checking for cheev_... no
> checking for cheev_ in -llapack... yes
> Library -llapack found
>
>
> And here are the relevant lines from ldd:
>
> =>ldd /usr/local/bin/scilab-bin | fgrep -i atl
> libcblas.so.3 => /usr/lib64/atlas/libcblas.so.3 (0x00007f64572a3000)
> libf77blas.so.3 => /usr/lib64/atlas/libf77blas.so.3 (0x00007f6457086000)
> libatlas.so.3 => /usr/lib64/atlas/libatlas.so.3 (0x00007f6456895000)
> liblapack.so.3 => /usr/lib64/atlas/liblapack.so.3 (0x00007f6455d29000)
>
> But:
>
> -->with_atlas()
> ans =
>
> F
>
> Clues will be appreciated. :-)
>
> Dean
Maybe you should try the latest versions of atlas and compile from sources. I
had a go with atlas3.9.33/scilab-5.3.0 and it seems to have worked.
More information about the dev
mailing list