Re^2: Install Math::GSL fail

OK so per Anonymous, I deleted the existing Makefile. Per marto I started a new session. Per Khen1950fx I set the env variable, but it's at /usr/local/pkgconfig.

On trying to install again, I get

cpan[1]> install Math::GSL
Going to read '/root/.cpan/Metadata'
  Database was generated on Fri, 04 Mar 2011 10:27:20 GMT
Running install for module 'Math::GSL'
Running make for L/LE/LETO/Math-GSL-0.22.tar.gz
Checksum for /root/.cpan/sources/authors/id/L/LE/LETO/Math-GSL-0.22.ta
+r.gz ok
Scanning cache /root/.cpan/build for sizes
......................................................................
+......DONE

  CPAN.pm: Going to build L/LE/LETO/Math-GSL-0.22.tar.gz

Checking for GSL..Package gsl was not found in the pkg-config search p
+ath.
Perhaps you should add the directory containing `gsl.pc'
to the PKG_CONFIG_PATH environment variable
No package 'gsl' found
 at Build.PL line 108
*** can not find package gsl
*** check that it is properly installed and available in PKG_CONFIG_PA
+TH
 at Build.PL line 108
Warning: No success on command[/usr/bin/perl Build.PL ]
'YAML' not installed, will not store persistent state
  LETO/Math-GSL-0.22.tar.gz
  /usr/bin/perl Build.PL  -- NOT OK
Running Build test
  Make had some problems, won't test
Running Build install
  Make had some problems, won't install
Could not read '/root/.cpan/build/Math-GSL-0.22-pXritF/META.yml'. Fall
+ing back to other methods to determine prerequisites
Failed during this command:
 LETO/Math-GSL-0.22.tar.gz                    : writemakefile NO '/usr
+/bin/perl Build.PL ' returned status 256
[download]

GSL does not appear in /usr/local/pkgconfig, but it is installed

[root@stats corman]# yum info GSL
Loaded plugins: langpacks, presto, refresh-packagekit
Adding en_US to language list
Installed Packages
Name        : gsl
Arch        : i686
Version     : 1.14
Release     : 1.fc14
Size        : 2.3 M
Repo        : installed
From repo   : fedora
Summary     : The GNU Scientific Library for numerical analysis
URL         : http://www.gnu.org/software/gsl/
License     : GPLv3 and GFDL and BSD
Description : The GNU Scientific Library (GSL) is a collection of rout
+ines for
            : numerical analysis, written in C.
[download]

But locate gsl.pc returns nothing :-/ locate GSL only returns stuff in /home/corman/.cpan/build. And it strikes me that this could be part of the problem. When I first tried to install this I forgot to SU to root. Do these files need to be in /root/.cpan/build if I'm running CPAN as root?

But that still doesn't address the gsl.pc issue... UPDATE: I installed YAML. Now CPAN says:

cpan[3]> install Math::GSL
Running install for module 'Math::GSL'
Running Build for L/LE/LETO/Math-GSL-0.22.tar.gz
  Has already been unwrapped into directory /root/.cpan/build/Math-GSL
+-0.22-yrnKBo
  '/usr/bin/perl Build.PL ' returned status 256, won't make
Running Build test
  Make had some problems, won't test
Running Build install
  Make had some problems, won't install
[download]

Comment on Re^2: Install Math::GSL fail Select or Download Code

Replies are listed 'Best First'.
Re^3: Install Math::GSL fail by Eliya (Vicar) on Mar 04, 2011 at 18:28 UTC
GSL does not appear in /usr/local/pkgconfig, but it is installed Maybe you need to install the respective `*-dev` package for the GSL library (?) Not entirely sure about Fedora, but most distros split their library packages into two parts: one for the runtime libs (always needed), and one for the "development stuff" (with the package name usually ending in "-dev" or "-devel"), like header files, static libs, the package config (".pc") file, etc., which are only needed if you build software that links against the library in question.	[reply] [d/l]
Re^4: Install Math::GSL fail by cormanaz (Deacon) on Mar 04, 2011 at 22:47 UTC
Yes that was it. But now I am getting a bunch of warnings from GCC, and make test fails a couple. To wit: cpan[1]> install Math::GSL Going to read '/root/.cpan/Metadata' Database was generated on Fri, 04 Mar 2011 10:27:20 GMT Running install for module 'Math::GSL' Running make for L/LE/LETO/Math-GSL-0.22.tar.gz Checksum for /root/.cpan/sources/authors/id/L/LE/LETO/Math-GSL-0.22.ta +r.gz ok Scanning cache /root/.cpan/build for sizes ...................................................................... +......DONE CPAN.pm: Going to build L/LE/LETO/Math-GSL-0.22.tar.gz Checking for GSL..Found GSL version 1.14 Checking if gcc supports "-Wall"...yes Checking if gcc supports "-Wno-unused-function"...yes Checking if gcc supports "-Wno-unused-value"...yes Checking if gcc supports "-Wno-unused-function"...yes Checking if gcc supports "-Wno-unused-variable"...yes Checking if gcc supports "-g"...yes Creating new 'MYMETA.yml' with configuration results Creating new 'Build' script for 'Math-GSL' version '0.22' Have a great day! Building Math-GSL xs/Multiroots_wrap.c: In function ‘_wrap_gsl_multiroot_fdjacobian’: xs/Multiroots_wrap.c:2339:12: warning: assignment from incompatible po +inter type xs/Multiroots_wrap.c: In function ‘_wrap_gsl_multiroot_fsolver_functio +n_set’: xs/Multiroots_wrap.c:2902:12: warning: assignment from incompatible po +inter type xs/Multiroots_wrap.c: In function ‘_wrap_gsl_multiroot_fsolver_set’: xs/Multiroots_wrap.c:3337:12: warning: assignment from incompatible po +inter type xs/Roots_wrap.c: In function ‘_wrap_gsl_root_fdfsolver_fdf_set’: xs/Roots_wrap.c:3213:23: warning: character constant too long for its +type xs/Roots_wrap.c:3213:7: warning: passing argument 2 of ‘fprintf’ makes + pointer from integer without a cast /usr/include/stdio.h:355:12: note: expected ‘const char * __restrict__ +’ but argument is of type ‘int’ xs/Roots_wrap.c:3214:7: warning: ‘return’ with a value, in function re +turning void xs/Roots_wrap.c: In function ‘_wrap_gsl_root_fdfsolver_set’: xs/Roots_wrap.c:3776:23: warning: character constant too long for its +type xs/Roots_wrap.c:3776:7: warning: passing argument 2 of ‘fprintf’ makes + pointer from integer without a cast /usr/include/stdio.h:355:12: note: expected ‘const char * __restrict__ +’ but argument is of type ‘int’ xs/Roots_wrap.c:3777:7: warning: ‘return’ with a value, in function re +turning void LETO/Math-GSL-0.22.tar.gz ./Build -- OK Running Build test t/00-load.t ........ 1/52 # Testing Math::GSL 0.22 on i386-linux-threa +d-multi, Perl (/usr/bin/perl) 5.012003 t/00-load.t ........ ok t/01-pod.t ......... skipped: Test::Pod 1.00 required for testing POD t/BLAS.t ........... 1/109 Elements start differing at index 0, delta = 8.9263593745227e+195 $x->[0] = 8.9263593745227e+195 $y->[0] = 50 Elements start differing at index 0, delta = 7.43864376780752e+195 $x->[0] = 7.43864376780752e+195 $y->[0] = 34 Elements start differing at index 0, delta = 4.42423204080942e+214 $x->[0] = -4.42423204080942e+214 $y->[0] = 24 Elements start differing at index 0, delta = 4.22125628458722e+200 $x->[0] = 4.22125628458722e+200 $y->[0] = 46 Elements start differing at index 0, delta = 3.51771357048935e+200 $x->[0] = 3.51771357048935e+200 $y->[0] = 34 Elements start differing at index 0, delta = 24 $x->[0] = -8.0780671462807e-27 $y->[0] = 24 t/BLAS.t ........... ok t/BSpline.t ........ ok t/CBLAS.t .......... ok t/CDF.t ............ 1/794 # Failed test 'gsl_cdf_tdist_Qinv (3.18309885122757724e-5, 1.0) ?= 9 +999.99999999054, # res= +-9.45874489843845e-09, eps=2.3283064365387e-10' # at t/CDF.t line 834. # Failed test 'gsl_cdf_tdist_Pinv (3.18309885122757724e-5, 1.0) ?= - +9999.99999999054, # res= +-9.45874489843845e-09, eps=2.3283064365387e-10' # at t/CDF.t line 834. # Looks like you failed 2 tests of 794. t/CDF.t ............ Dubious, test returned 2 (wstat 512, 0x200) Failed 2/794 subtests t/Chebyshev.t ...... ok t/Combination.t .... ok t/Complex.t ........ ok t/Const.t .......... ok t/DHT.t ............ ok t/Deriv.t .......... ok t/Eigen.t .......... ok t/Errno.t .......... ok t/FFT.t ............ ok t/Fit.t ............ ok t/GSL.t ............ 1/18 Elements start differing at index 2, delta = 0.00099999999999989 $x->[2] = 3 $y->[2] = 3.001 Elements start differing at index 2, delta = 9.99999998363421e-08 $x->[2] = 3.0010001 $y->[2] = 3.0010002 t/GSL.t ............ ok t/Heapsort.t ....... ok t/Histogram.t ...... ok t/Histogram2D.t .... ok t/IEEEUtils.t ...... ok t/Integration.t .... ok t/Interp.t ......... ok t/Linalg.t ......... 1/70 Elements start differing at index 0, delta = 6 $x->[0] = -0 $y->[0] = -6 Elements start differing at index 0, delta = 8 $x->[0] = -4 $y->[0] = 4 t/Linalg.t ......... ok t/Machine.t ........ ok t/Matrix.t ......... 1/238 Elements start differing at index 1, delta = 0.707106781186548 $x->[1] = 4.32963728535968e-17 $y->[1] = -0.707106781186548 Elements start differing at index 1, delta = 0.707106781186548 $x->[1] = 4.32963728535968e-17 $y->[1] = 0.707106781186548 t/Matrix.t ......... 109/238 Elements start differing at index 1, delta = 1 $x->[1] = 3 $y->[1] = 2 t/Matrix.t ......... ok t/MatrixComplex.t .. 1/38 Elements start differing at index 1, delta = 1 $x->[1] = 3 $y->[1] = 4 Elements start differing at index 1, delta = 1 $x->[1] = 3 $y->[1] = 2 t/MatrixComplex.t .. ok t/Min.t ............ ok t/Monte.t .......... 1/15 Elements start differing at index 0, delta = 2.78505685129308e+16 $x->[0] = 0.333333333333333 $y->[0] = 2.78505685129308e+16 t/Monte.t .......... ok t/Multifit.t ....... ok t/Multimin.t ....... ok t/Multiroots.t ..... ok t/NTuple.t ......... 1/11 This is the end at t/NTuple.t line 14. t/NTuple.t ......... ok t/ODEIV.t .......... ok t/Permutation.t .... ok t/Poly.t ........... Elements start differing at index 0, delta = 4 $x->[0] = 0 $y->[0] = 4 t/Poly.t ........... 1/7 Elements start differing at index 0, delta = 0.674 $x->[0] = 0 $y->[0] = 0.674 t/Poly.t ........... ok t/PowInt.t ......... ok t/QRNG.t ........... ok t/RNG.t ............ ok t/Randist.t ........ ok t/Roots.t .......... ok t/SF.t ............. ok t/Siman.t .......... ok t/Sort.t ........... ok t/Spline.t ......... ok t/Statistics.t ..... ok t/Sum.t ............ ok t/Sys.t ............ ok t/Vector.t ......... 1/142 Elements start differing at index 0, delta = 9 $x->[0] = 1 $y->[0] = 10 Elements start differing at index 0, delta = 9 $x->[0] = 1 $y->[0] = 10 t/Vector.t ......... ok t/VectorComplex.t .. ok t/Wavelet.t ........ ok t/Wavelet2D.t ...... ok Test Summary Report ------------------- t/CDF.t (Wstat: 512 Tests: 794 Failed: 2) Failed tests: 481, 523 Non-zero exit status: 2 t/Matrix.t (Wstat: 0 Tests: 238 Failed: 0) TODO passed: 209-212 t/Vector.t (Wstat: 0 Tests: 142 Failed: 0) TODO passed: 141-142 Files=52, Tests=3578, 18 wallclock secs ( 1.53 usr 0.27 sys + 13.75 c +usr 1.58 csys = 17.13 CPU) Result: FAIL Failed 1/52 test programs. 2/3578 subtests failed. LETO/Math-GSL-0.22.tar.gz ./Build test -- NOT OK //hint// to see the cpan-testers results for installing this module, t +ry: reports LETO/Math-GSL-0.22.tar.gz Running Build install make test had returned bad status, won't install without force Failed during this command: LETO/Math-GSL-0.22.tar.gz : make_test NO [download]	[reply] [d/l]
Re^5: Install Math::GSL fail by syphilis (Archbishop) on Mar 05, 2011 at 06:28 UTC
Note that there's only 1 failing test script (namely t/CDF.t) and that it fails only 2 tests (tests 481 and 523). All the other stuff is in relation to 'TODO' tests, and they don't count as failures. On MS Windows, I get the same 2 failing tests as you get - I haven't yet investigated further. Looks to me that you've got a fairly good build of Math::GSL - why not just force install it and see how it goes for you ? Cheers, Rob	[reply]
Re^5: Install Math::GSL fail by aixtal (Sexton) on Jan 24, 2013 at 09:27 UTC
I get the same errors on Mac OSx 10.8.2 and gcc 4.6.3 : `xs/Roots_wrap.1.15.c:3283:7: error: void function '_wrap_gsl_root_fdfs +olver_fdf_set' should not return a value [-Wreturn-type] return (gsl_nan()); ^ ~~~~~~~~~~~ ... xs/Roots_wrap.1.15.c:3840:7: error: void function '_wrap_gsl_root_fdfs +olver_set' should not return a value [-Wreturn-type] return (gsl_nan()); ^ ~~~~~~~~~~~` [download] This thread is a bit old. I wonder if anybody has made any progress on this?	[reply] [d/l]
Re^6: Install Math::GSL fail by syphilis (Archbishop) on Jan 24, 2013 at 09:54 UTC
Re^7: Install Math::GSL fail by aixtal (Sexton) on Jan 24, 2013 at 10:14 UTC
Some notes below your chosen depth have not been shown here
Re^6: Install Math::GSL fail by Anonymous Monk on Jan 24, 2013 at 09:45 UTC
Re^7: Install Math::GSL fail by aixtal (Sexton) on Jan 24, 2013 at 09:56 UTC
Some notes below your chosen depth have not been shown here
Re^3: Install Math::GSL fail by Anonymous Monk on Mar 04, 2011 at 17:21 UTC
Why do you insist on using the cpan shell?	[reply]
Re^4: Install Math::GSL fail by cormanaz (Deacon) on Mar 04, 2011 at 22:50 UTC
Because that seems to be the generally accepted way of installing packages and because for a complex package like this I don't want to spend all day manually installing dependencies, etc. Why do you have a better way to do it?	[reply]
Re^5: Install Math::GSL fail by Anonymous Monk on Mar 05, 2011 at 12:57 UTC
Why do you have a better way to do it? Yes. By all means, use cpan to install all the dependencies Math::GSL has, but then use look to instll Math::GSL manually. Its easier than trying to filter all noise cpan makes other methods to determine prerequisites	[reply]


go ahead... be a heretic
	PerlMonks