Update ATLAS to stable version 3.10

[vbraun]

The new atlas release now builds netlib lapack itself, so the lapack tarball is now included in the ATLAS spkg.

• Updated to newest upstream source, various patches are no longer required
• SAGE_ATLAS_LIB=path now searches in path/libatlas.so instead of path/lib/libatlas.so so it works for people with atlas in /lib64, too.
• Threading is now enabled by default
• Flush before os.system (#13210)

Upstream has made some attempt at changing the layout of the shared libraries, which is now different from the static libraries. The atlas spkg contains a stub autoconf/libtools project that unpacks the static libraries and repacks them into equivalent shared libraries.

By default, ATLAS will now try twice to get timings and fail immediately if throttling is enabled. If auto-tuning fails build with SAGE_ATLAS_ARCH=fast, and if that fails with SAGE_ATLAS_ARCH=base. On x86, the fast and base targets are the new ATLAS generic targets x86SSE3 and x86SSE2/x86x87.

The updated cvxopt spkg fixes a blas path issue on Darwin.

The current spkg version is at

​http://www.stp.dias.ie/~vbraun/Sage/spkg/atlas-3.10.0.spkg ​http://www.stp.dias.ie/~vbraun/Sage/spkg/cvxopt-1.1.5.p1.spkg

Apply trac_10508_root_repo.patch to the SAGE_ROOT repository and 10508_doctest.patch, trac_10508_update_atlas_docs.patch​ to the Sage repository.

Remove the lapack and blas packages.

[vbraun]

Note that you cannot just use sage -f atlas-3.9.32.spkg to update atlas only. Many other spkgs use blas/lapack and must be rebuilt. The easiest way is to do a separate Sage installation...

The cvxopt spkg needs to be updated to link correctly with this atlas release, see #10509.

[dimpase]

Does it mean that LAPACK spkg can be removed, too?

[vbraun]

Yes, the lapack spkg can be removed.

I'm still trying to debug some issues with linbox...

[fbissey]

BLAS can also be removed if we go with this. f2c which was used before we got ATLAS to provide cblas by f2c-ing the BLAS package should also be removed (I think it listed in scipy's dependency only).

[dimpase]

on 32-bit x86 Linux (Debian squeeze) I get the following, when trying to install the spkg (applied the patches to a pristine Sage 4.6.1.alpha3):

...
make -j1 libatlas.so libptf77blas.so libf77blas.so \
                libptcblas.so libcblas.so liblapack.so
make[3]: Entering directory `/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/lib'
ld -melf_i386 -shared -soname /usr/local/src/sage/sage-4.6.1.alpha3/local/lib/libatlas.so -o libatlas.so \
           -rpath-link /usr/local/src/sage/sage-4.6.1.alpha3/local/lib \
           --whole-archive libatlas.a --no-whole-archive -lc -lm
make[3]: *** No rule to make target `libptf77blas.a', needed by `libptf77blas.so'.  Stop.
make[3]: Leaving directory `/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/at
las-3.9.32/ATLAS-build/lib'
make[2]: *** [ptshared] Error 2
make[2]: Leaving directory `/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/at
las-3.9.32/ATLAS-build/lib'
Configuration:
    SAGE_LOCAL: /usr/local/src/sage/sage-4.6.1.alpha3/local
    linker_Solaris?: False
    PPC?: False
    SPKG_DIR: /usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32
    linker_GNU?: True
    ld: GNU
    system: Linux
    Darwin?: False
    machine: i686
    fortran: gfortran
    Solaris?: False
    fortran_g95?: False
    bits: 32bit
    CYGWIN?: False
    SPARC?: False
    fortran_GNU?: True
    FreeBSD?: False
    32bit?: True
    Linux?: True
    64bit?: False
    Intel?: False
    processor: 

[dimpase]

Replying to dimpase:

on 32-bit x86 Linux (Debian squeeze) I get the following, when trying to install the spkg (applied the patches to a pristine Sage 4.6.1.alpha3):

complete install.log is here: ​http://boxen.math.washington.edu/home/dima/tmp/install-alt3.9.log.gz

[vbraun]

Hi Dima,

I think the problem is

make[6]: Entering directory `/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/tune/sysinfo'
gcc -c -DL2SIZE=4194304 -I/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/include -I/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/../src//include -I/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/../src//include/contrib -DAdd_ -DF77_INTEGER=int -DStringSunStyle -DATL_OS_Linux -DATL_ARCH_CoreDuo -DATL_CPUMHZ=800 -DATL_SSE3 -DATL_SSE2 -DATL_SSE1 -DATL_GAS_x8632  -fomit-frame-pointer -O3 -mfpmath=387 -fPIC -m32 /usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/../src//tune/sysinfo/findNT.c
gcc -DL2SIZE=4194304 -I/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/include -I/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/../src//include -I/usr/local/src/sage/sage-4.6.1.alpha3/spkg/build/atlas-3.9.32/ATLAS-build/../src//include/contrib -DAdd_ -DF77_INTEGER=int -DStringSunStyle -DATL_OS_Linux -DATL_ARCH_CoreDuo -DATL_CPUMHZ=800 -DATL_SSE3 -DATL_SSE2 -DATL_SSE1 -DATL_GAS_x8632  -fomit-frame-pointer -O3 -mfpmath=387 -fPIC -m32 -o xfindNT findNT.o ATL_walltime.o -lm
/usr/lib/crt1.o: In function `_start':
(.text+0x18): undefined reference to `main'
collect2: ld returned 1 exit status
make[6]: *** [xfindNT] Error 1

This fails because ATL_NCPU is not set. Do you have a single-core processor? I guess the threaded libraries are not built in that case.

[dimpase]

Replying to vbraun:

This fails because ATL_NCPU is not set. Do you have a single-core processor? I guess the threaded libraries are not built in that case.

yes, it's single-core. An old Pentium M. (Atlas 3.8 spkg does not build on it, at all)

[vbraun]

I changed the spkg-install to make only single-threaded shared libraries if necessary. Should work now.

For simplicity, I made spkgs for cvxopts and sage_scripts. So you need to add

​http://www.stp.dias.ie/~vbraun/Sage/spkg/atlas-3.9.32.spkg ​http://www.stp.dias.ie/~vbraun/Sage/spkg/cvxopt-1.1.3.p0.spkg ​http://www.stp.dias.ie/~vbraun/Sage/spkg/sage_scripts-4.6.1.alpha3.p0.spkg

to $SAGE_ROOT/spkg/standard and then

• replace spkg/install with the attached version (Note that sage_scripts overwrites this file during installation, aargh)
• replace spkg/standard/deps with the attached version.

I'm still having doctest errors with linbox...

[dimpase]

Replying to vbraun:

I changed the spkg-install to make only single-threaded shared libraries if necessary. Should work now.

It builds OK, but then testlong gives quite a bit of failures:

The following tests failed:
        sage -t  -long -force_lib "devel/sage/doc/en/bordeaux_2008/elliptic_curves.rst"
        sage -t  -long -force_lib "devel/sage/sage/modular/modsym/space.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/modsym/tests.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/modsym/subspace.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/modsym/ambient.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/modform/space.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/modform/element.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/modform/ambient.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/hecke/submodule.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/abvar/abvar.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/abvar/torsion_subgroup.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/abvar/cuspidal_subgroup.py"
        sage -t  -long -force_lib "devel/sage/sage/modular/abvar/finite_subgroup.py"
        sage -t  -long -force_lib "devel/sage/sage/matrix/matrix_integer_dense.pyx"
        sage -t  -long -force_lib "devel/sage/sage/matrix/matrix_integer_dense_hnf.py"
        sage -t  -long -force_lib "devel/sage/sage/rings/qqbar.py"
        sage -t  -long -force_lib "devel/sage/sage/finance/time_series.pyx"
        sage -t  -long -force_lib "devel/sage/sage/schemes/elliptic_curves/padic_lseries.py"
        sage -t  -long -force_lib "devel/sage/sage/schemes/elliptic_curves/ell_modular_symbols.py"
        sage -t  -long -force_lib "devel/sage/sage/schemes/elliptic_curves/ell_rational_field.py"
        sage -t  -long -force_lib "devel/sage/sage/schemes/elliptic_curves/sha_tate.py"
Total time for all tests: 30045.4 seconds

I do not know how many of them are Atlas related, though. The log is here: ​http://boxen.math.washington.edu/home/dima/tmp/testlong-atl3.9.log

[fbissey]

A lot of it is probably related. Was this a build from scratch? linbox calls seem to be particularly affected, there are a lot of failures in that path: sage.matrix.matrix_integer_dense.Matrix_integer_dense._charpoly_linbox stuff going through: sage.matrix.matrix_rational_dense.Matrix_rational_dense.right_kernel is also affected and that smells like atlas at work too.

I must say we have observed a number of failures related to ATLAS-3.9.23 in sage-on-gentoo: ​https://github.com/cschwan/sage-on-gentoo/issues#issue/3 you have failure there too but not due to iml/ATLAS as far as I can see. ​https://github.com/cschwan/sage-on-gentoo/issues#issue/6 more subtle. I see you have the failure with devel/sage/sage/finance/time_series.pyx you may want to look the comment I wrote about it in the section "known test failures for Sage on Gentoo": ​https://github.com/cschwan/sage-on-gentoo/wiki/Known-test-failures note that sage-on-gentoo can use (c)blas-reference, ATLAS or gsl-cblas or some combinations - in fact amd and intel libraries could in principle be used but no one I know has tried. ​http://www.gentoo.org/proj/en/science/blas-lapack.xml

Not sure about the SIGFPE, it could come from a result rounded to 0 or the like.

[dimpase]

Replying to fbissey:

A lot of it is probably related. Was this a build from scratch?

yes, it's a build from scratch, on the same machine (old Pentium M) as already discussed above.

linbox calls seem to be particularly affected, there are a lot of failures in that path:

I wonder if these are Atlas bugs, or Linbox bugs...

One should try an OSX build.

[fbissey]

why OS X in particular? I could do that but not before the 5th of January, when my university reopens. Ok I could do it before that but I'll enjoy the break a little bit more :)

[dimpase]

Replying to fbissey:

why OS X in particular? I could do that but not before the 5th of January, when my university reopens. Ok I could do it before that but I'll enjoy the break a little bit more :)

Or does OSX remain disabled for Atlas, i.e. it's not built?

[fbissey]

Replying to dimpase:

Replying to fbissey:

why OS X in particular? I could do that but not before the 5th of January, when my university reopens. Ok I could do it before that but I'll enjoy the break a little bit more :)

Or does OSX remain disabled for Atlas, i.e. it's not built?

I had forgotten about that. I checked the spkg and ATLAS is not built on cygwin and OS X. So it makes sense.

[vbraun]

I get the same doctest errors. Most of them are linbox related. The SIGFPE is from converting a NAN into a GMP integer, but I haven't gotten to the root of the NAN yet. In trying to debug this I've noticed that there are a bunch of valgrind warnings in the linbox code path we are using. I've asked some more specific questions on the linbox-use mailinglist:

​https://groups.google.com/d/topic/linbox-use/N3QNNOQuTAc/discussion

But so far no final conclusion.

[fbissey]

Replying to dimpase:

Replying to vbraun:

This fails because ATL_NCPU is not set. Do you have a single-core processor? I guess the threaded libraries are not built in that case.

yes, it's single-core. An old Pentium M. (Atlas 3.8 spkg does not build on it, at all)

I notice you say it is a Pentium M, yet ATLAS is compiling things with the assumption that it is a coreduo: -DATL_ARCH_CoreDuo -DATL_CPUMHZ=800 -DATL_SSE3 -DATL_SSE2 -DATL_SSE1 -DATL_GAS_x8632 I am guessing the speed is right but the rest is not. Was your successful build using these parameters? I had in one instance a non-working ATLAS because the cpu type was misdetected (wanted to use sse3 when it didn't have them). The build was curiously successful but the library was unusable.

[dimpase]

Replying to fbissey:

Replying to dimpase:

Replying to vbraun:

This fails because ATL_NCPU is not set. Do you have a single-core processor? I guess the threaded libraries are not built in that case.

yes, it's single-core. An old Pentium M. (Atlas 3.8 spkg does not build on it, at all)

I notice you say it is a Pentium M, yet ATLAS is compiling things with the assumption that it is a coreduo: -DATL_ARCH_CoreDuo -DATL_CPUMHZ=800 -DATL_SSE3 -DATL_SSE2 -DATL_SSE1 -DATL_GAS_x8632 I am guessing the speed is right but the rest is not. Was your successful build using these parameters? I had in one instance a non-working ATLAS because the cpu type was misdetected (wanted to use sse3 when it didn't have them). The build was curiously successful but the library was unusable.

The processor is Pentium M Banias 1.1GHz, ​http://ark.intel.com/Product.aspx?id=27600 (according to GotoBlas? installation procedure :-))

The Atlas built is not totally useless, it works for many doctests. By the way, 3.8 also thinks it's a CoreDuo?. And it works. I saw somewhere a note that it's OK.

[fbissey]

Replying to vbraun:

I get the same doctest errors. Most of them are linbox related. The SIGFPE is from converting a NAN into a GMP integer, but I haven't gotten to the root of the NAN yet. In trying to debug this I've noticed that there are a bunch of valgrind warnings in the linbox code path we are using. I've asked some more specific questions on the linbox-use mailinglist:

​https://groups.google.com/d/topic/linbox-use/N3QNNOQuTAc/discussion

But so far no final conclusion.

Converting a NAN into a GMP integer is exactly what was happening in ​https://github.com/cschwan/sage-on-gentoo/issues/3 and it didn't happen when using gslcblas. I will do a full build of sage-on-gentoo with 3.9.40 (or 41) and see if I can see anything.

[fbissey]

Yuck, still got problems leading to linbox, I got the ones leading to iml too. Note that using another cblas like gslcblas/openblas/reference(netlib) all these problems disappear which seem to indicate that there is something going on in ATLAS itself or that all the others gets it wrong (I realise that on a stock sage you may have trouble compiling iml with anything else than ATLAS, it requires some patching to the configure script to be able to do so).

[timdumol]

It may be worth noting that new released are available for both stable and development:

​http://sourceforge.net/projects/math-atlas/files/Stable/3.8.4/|Stable ​http://sourceforge.net/projects/math-atlas/files/Developer%20%28unstable%29/3.9.41/|Unstable

[fbissey]

I have 3.9.41 now, the only difference was it compiled faster because there was tuning for my cpu. But we could investigate 3.8.4. Thanks for pointing it is out, some of us (me at least) didn't think it would ever happen. We may have a stable ATLAS supporting newer CPUs at last, and it looks like I could test it quickly.

[fbissey]

OK 3.8.4 doesn't suffer from any of the drawbacks that are apparent in 3.9.23+.

The big drawback is that it doesn't build lapack nicely on its own, provided we point to the sources like the latest 3.9.xx do. That means spkg-install would need a little bit more work.

Opinions?

[vbraun]

Can we keep this ticket focused on the development version of atlas and discuss the stable ATLAS on #10226? I updated the spkg on the latter ticket to the new stable ATLAS release.

[fbissey]

Sure. My current opinion and that's something I am pushing to sage-on-gentoo users is to avoid ATLAS 3.9.xx for the time being. It is possible that ATLAS-3.9.xx is doing something permissible that iml and linbox are not ready to catch. My line of thought is that I remember that some algorithm in iml takes the inverse of a result returned by cblas. If the result is 0 instead of a small value we may have our NaN, more likely some result from ATLAS is a NaN.

[leif]

Note that you can copy updated spkgs into spkg/standard/ and then do

env SAGE_UPGRADING=yes make build

to rebuild all dependent packages.

(We should perhaps add make targets for that to the top-level Makefile and document them in the Developer's Guide, as this is currently just a side-effect of the effort to make upgrading work.)

If you at the same time need to apply patches to the Sage library, things get a bit more complicated, as e.g. Sage switches to the main branch before reinstalling the Sage library package. One safe way is to first apply the patches, create a new sage-x.y.z-whatever spkg (with devel/sage/spkg-dist) and replace the one in spkg/standard/ with that one (or at least make sure newest_version will pick up the right one).

Note that the extension modules' dependencies in module_list.py are currently far from complete. #8664 adds some in a generic way by adding them automatically in setup.py, i.e. lets modules also depend on the headers of the libraries they use (which [only] works if the headers' mtimes get modified / updated during installation of their corresponding libraries). The dumb alternative is to run sage -ba-force after an "upgrade" process.

[leif]

Ping.

[dimpase]

Replying to leif:

Ping.

just fired up my vintage MacOSX 32-bit Powerbbok PPC... Will know more in, like, 12 hours...

[fbissey]

Several issues:

• atlas is now at version 3.9.49 (last I checked)
• I have reports of it failling to build on some old opterons
• @dimpase: do you need atlas on OS X ppc? Don't you use vectorize like the other OS X?
• I haven't checked but I am quite sure from other reports that cblas_dgem{m,v} is still buggy (someone posted that R built against that Atlas lapack was giving them trouble and R upstream pointed the finger to Atlas).

[dimpase]

Replying to fbissey:

Several issues:

• atlas is now at version 3.9.49 (last I checked)
• I have reports of it failling to build on some old opterons

so what? I have a 5-year old 32-bit Intel laptop on which the sage-current Atlas does not build.

• @dimpase: do you need atlas on OS X ppc? Don't you use vectorize like the other OS X?

indeed. But that's for #10509. Oops, was posting on the wrong ticket...

[kcrisman]

With respect to comment:33:ticket:12011, should this be closed as a dup of #12011?

[dimpase]

propose to close this one and refer to #12011 the the continuation of the upgrade work.

[vbraun]

I would say #12011 is a duplicate of my ticket but oh well ;-)

[jdemeyer]

Seems like #12011 isn't a duplicate after all...

[jdemeyer]

Obviously, the patches to spkg/install and spkg/standard/deps must be rebased.

[jdemeyer]

It makes a lot more sense to me to put the LAPACK tarball at the top level of the spkg instead of in patches/.

patches/ATLAS-lib/autom4te.cache should be removed.

[jdemeyer]

I don't like using assert for control flow, because that's not what it's meant for.

Why not replace those by (see #13210)

if rc != 0:
    print "Error: foo"
    sys.exit(rc)

[jdemeyer]

There is something wrong with the history in SPKG.txt (atlas-3.8.4 is completely missing and there is atlas-3.9.68 for #12011 which never got merged)

[vbraun]

I don't have any strong opinion on where to put the lapack tarball, except that our convention of only allowing a single src/ directory is shortsighted.

Note that I'm not using assertions for control flow, i.e. I'm not using

  try:
    <command>
  except AssertionError:
    <alternative>

Note that you could theoretically also catch the SystemExit exception, so sys.exit() isn't different from assert in that regard. I'm only using assertions to ensure the following contract holds: spkg-install completes successfully only if the relevant atlas configure/make completed with rc==0

I'll replace the asserts by something that returns rc as exit code, though.

[jdemeyer]

Replying to vbraun:

I'm only using assertions to ensure the following contract holds: spkg-install completes successfully only if the relevant atlas configure/make completed with rc==0

I think (IMHO but I might be wrong) that assertions should be used only in a situation where an assertion being false indicates a bug in the program. If rc != 0 in spkg-install, that is an ordinary error condition, not a bug in the spkg-install script.

[jdemeyer]

Starting from sage-5.1, I get one doctest failure:

File "/release/merger/sage-5.1-atlas/devel/sage-main/sage/rings/polynomial/polynomial_element.pyx", line 1039:
    sage: parent(poly)([ 0.0 if abs(c)<=1e-14 else c for c in poly.coeffs() ])
Expected:
    1.0
Got:
    1.02140518266e-14*x^2 + 1.0

[dimpase]

Replying to jdemeyer:

Starting from sage-5.1, I get one doctest failure:

     sage: parent(poly)([ 0.0 if abs(c)<=1e-14 else c for c in poly.coeffs() ])

it's most probably not at Atlas problem, but rather that 1e-14 (any solid rationale behind this choice? I guess not.) needs to be adjusted.

[benjaminfjones]

Testing in sage-5.1.rc1 on x86_64 debian Linux:

$ uname -a
Linux sage 2.6.32 #1 SMP Fri Sep 2 21:08:57 CDT 2011 x86_64 GNU/Linux

$ head -18 /proc/cpuinfo 
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 44
model name      : Intel(R) Xeon(R) CPU           X5690  @ 3.47GHz
stepping        : 2
cpu MHz         : 3465.790
cache size      : 12288 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 11
wp              : yes
flags           : fpu de tsc msr pae cx8 sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good aperfmperf pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt aes hypervisor lahf_lm ida arat
bogomips        : 6931.58
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

BEFORE new ATLAS spkg

sage -t  "devel/sage-main/sage/modular/modsym/ambient.py"   
         [11.6 s]

sage -t  "devel/sage-main/sage/modular/hecke/ambient_module.py"
         [4.2 s]

sage -t  "devel/sage-main/sage/modular/hecke/hecke_operator.py"
         [3.0 s]

AFTER new spkg

1. untar'd fresh sage-5.1.rc1
2. replaced atlas-3.8.4.p1.spkg with atlas-3.9.85.spkg
3. make build

SPKG build log: ​http://sage.math.washington.edu/home/bjones/atlas-3.9.85.log

Build atlas-3.9.85

real    11m18.595s
user    11m15.906s
sys     2m56.099s
Successfully installed atlas-3.9.85

Tests

sage -t  "devel/sage-main/sage/modular/modsym/ambient.py"   
         [11.6 s]

sage -t  "devel/sage-main/sage/modular/hecke/ambient_module.py"
         [4.2 s]

sage -t  "devel/sage-main/sage/modular/hecke/hecke_operator.py"
         [3.1 s]

All tests

$ make ptestlong
...
All tests passed!
Total time for all tests: 1282.1 seconds

[fbissey]

3.10.0 has just been released a few hours ago. Do we try to go for it? Like Jeroen remarked there may be tolerance issues depending on the hardware. I know I have one when using openblas instead of ATLAS.

[vbraun]

The new spkg will attempt to build on OSX if SAGE_ATLAS_ARCH is set. Untested, but a start.

I also am planning to allow SAGE_ATLAS_LIB=system and use ldconfig -p to get the system atlas, if available. Though perhaps we should do that in a later version. This spkg seems to fix some build issues that people reported on the sage mailinglists so it might be good to get it in sooner rather than later.

[jdemeyer]

Replying to vbraun:

so it might be good to get it in sooner rather than later.

+1, let's not worry now about changing the working of SAGE_ATLAS_LIB.

[jdemeyer]

In spkg-install, I find it confusing that you have a function build() which calls configure() and make_atlas().

I would get rid of build() and simply call configure() and make_atlas() directly.

[jdemeyer]

About SAGE_FAT_BINARY: at one point you are checking

os.environ.has_key('SAGE_FAT_BINARY')

and at another you are checking

os.environ.get('SAGE_FAT_BINARY', 'no') == 'yes' and conf['Intel?']

I think the correct check is

os.environ.get('SAGE_FAT_BINARY', 'no') == 'yes'

"SAGE_FAT_BINARY" has evolved to mean "build a generic binary on any processor", so it's not Intel specific anymore. Just call configure_base().

[jdemeyer]

Somebody needs to review 10508_doctest.patch

[jdemeyer]

Detail: in system_with_flush(), your print command should not have a space after "Running". The space is automatically added by print.

[vbraun]

Updated spkg to remove the space and make the SAGE_FAT_BINARY check uniform.

I'm fine with the doctests patch...

[vbraun]

I'm always calling build() except in one place where I want to allow configure() to fail (because throttling is enabled).

[jdemeyer]

Is this expected? Compare the build time on arando (Ubuntu 12.04 i386):

real    104m44.021s
user    97m25.753s
sys     1m18.157s
Successfully installed atlas-3.8.4.p1

real    259m50.222s
user    233m21.879s
sys     7m38.405s
Successfully installed atlas-3.10.0

[vbraun]

I've noticed before that the build time for the "generic" binary is rather long. Its not entirely generic, it is still doing timing for cache edges. But the result will be a working library, it won't probe funky asm implementations that other CPUs might not support. I'll ask on the ATLAS mailinglist for clarification.

[jdemeyer]

Timing on x86_64 with SAGE_FAT_BINARY=yes on sage.math:

real    139m50.733s
user    134m30.710s
sys     7m12.390s
Successfully installed atlas-3.8.4.p1

real    350m5.662s
user    330m23.460s
sys     22m1.010s
Successfully installed atlas-3.10.0

Why does the new ATLAS take so much longer to build than the old one?

[leif]

Replying to jdemeyer:

Timing on x86_64 with SAGE_FAT_BINARY=yes on sage.math:

real    139m50.733s
user    134m30.710s
sys     7m12.390s
Successfully installed atlas-3.8.4.p1

real    350m5.662s
user    330m23.460s
sys     22m1.010s
Successfully installed atlas-3.10.0

Why does the new ATLAS take so much longer to build than the old one?

I can beat that:

Finished installing shared ATLAS library.

real	821m14.881s
user	739m58.560s
sys	58m23.440s
Successfully installed atlas-3.10.0

(That's with sage -f ... on an otherwise idle machine, an AMD Fusion E-450 running Ubuntu 10.04.4 LTS x86_64.)

It took far more time than building all of Sage [in parallel, on that dual-core CPU], including the old ATLAS spkg. I currently have no figures for a separate ATLAS 3.8.4 build, but the timings from previous parallel Sage builds on that machine vary between

real	214m52.096s
user	175m5.390s
sys	9m56.210s
Successfully installed atlas-3.8.4.p1

and

real	256m32.949s
user	217m30.760s
sys	10m31.050s
Successfully installed atlas-3.8.4.p1

(The LAPACK and BLAS spkg build times in these builds are a few minutes and less than one minute [wall time], respectively.)

I was actually hoping ATLAS 3.9.x / 3.10 meanwhile "knows" these AMD CPUs and therefore builds at least a bit faster...

[leif]

FWIW, ptestlong passed (after rebuilding all dependent packages) with Sage 5.2.beta0 (without applying the doctest patch).

[fbissey]

We have noticed the building time increase in Gentoo as well. At 3.9.82 I think. Apparently they have changed how they detect the compiler and that's what causing the spike. But it seem fixed in 3.10.0 unless we are carrying a specific patch in Gentoo

     Fri Jun 29 13:42:08 2012 >>> sci-libs/atlas-3.9.80
       merge time: 8 minutes and 4 seconds.

     Wed Jul  4 14:56:21 2012 >>> sci-libs/atlas-3.9.82
       merge time: 3 hours, 54 minutes and 30 seconds.

     Wed Jul 11 11:11:23 2012 >>> sci-libs/atlas-3.10.0
       merge time: 8 minutes and 34 seconds.

[vbraun]

If I set SAGE_ATLAS_ARCH=Corei2,AVX,SSE3,SSE2,SSE1 then I can also compile atlas-3.10.0 in about 10 minutes. The issue is only with the "generic" archdefs, which seem to be not sufficiently specialzied. I've raised this issue on the ATLAS mailinglist.

[vbraun]

The updated atlas spkg also installs a script atlas-config in $SAGE_LOCAL/bin which can be used to compute new architectural defaults. I need somebody on Linux i386 to first install the new spkg and then run

SAGE_FAT_BINARY=yes atlas-config --archdef

to grind out the archdefs for the i386 "generic" target. I'm currently doing this for x86_64, but I don't have a i386 machine. This will use sudo to turn off CPU throttling so you need to be a sudoer.

[vbraun@volker-desktop sage-5.2.beta1]$ ./local/bin/atlas-config --help
usage: atlas-config [-h] [--unthrottle PID] [--archdef]

(Re-)Build ATLAS (http://math-atlas.sourceforge.net) according to the
SAGE_ATLAS_ARCH environment variable

optional arguments:
  -h, --help        show this help message and exit
  --unthrottle PID  switch CPU throttling off until PID finishes
  --archdef         build archdef tarball and save it to the current directory

[vbraun]

I've updated the spkg with new 64-bit generic archdefs, this now builds in about 10 mins.

[leif]

Replying to vbraun:

I've updated the spkg with new 64-bit generic archdefs, this now builds in about 10 mins.

md5sum?

I incidentally just downloaded some new version (163f090f18bb8616e93617677a644cd8) and triggered a full build from scratch.

[vbraun]

The newest version is 8a16c9d39add1c6c3f37e13986e2a3cc and thats whats linked in the ticket.

[vbraun]

The new version d33e9114156d8373fa61f957b379e029 changes the "fast" 64-bit archdef to P4E64SSE3.

It turns out that there are no 64-bit generic archdefs, which might have been one reason for why ATLAS was slow to compile. The SPKG uses the existing 32-bit archdefs or the 64-bit one that I made. On x86 the spkg should produce a working ATLAS library within 10-30 mins, and only go through the tuning process if either

• CPU throttling is disabled (scaling_governor=performance, needs root), or
• SAGE_ATLAS_ARCH is explicitly set to something different from "base" / "fast".

[vbraun]

PS: Surprisingly enough, the new ATLAS spkg actually compiled on OSX (bsd.math)! If you want to try yourself just set SAGE_ATLAS_ARCH="base" to force building on OSX.

[benjaminfjones]

Replying to vbraun:

PS: Surprisingly enough, the new ATLAS spkg actually compiled on OSX (bsd.math)! If you want to try yourself just set SAGE_ATLAS_ARCH="base" to force building on OSX.

Very cool. I just got ATLAS to build on my OS X 10.6.8 machine setting SAGE_ATLAS_ARCH=Corei2. Running long tests now.

[benjaminfjones]

Update: ATLAS-3.10.0 built successfully on my OS X 10.6.8 machine with SAGE_ATLAS_ARCH=Corei2, the build took approx. 16 mins. The build log is at ​http://sage.math.washington.edu/home/bjones/atlas-3.10.0.log. Sage passes all make ptestlong tests. The spkg looks very good to me. I'd give this a positive review, but maybe it should be tested by a few other reviewers on other platforms first.

[dimpase]

Replying to benjaminfjones:

Update: ATLAS-3.10.0 built successfully on my OS X 10.6.8 machine with SAGE_ATLAS_ARCH=Corei2, the build took approx. 16 mins. The build log is at ​http://sage.math.washington.edu/home/bjones/atlas-3.10.0.log. Sage passes all make ptestlong tests. The spkg looks very good to me. I'd give this a positive review, but maybe it should be tested by a few other reviewers on other platforms first.

I've built it successfully on OS X 10.6.8 (with Core2 Duo) and setting SAGE_ATLAS_ARCH="base".

[leif]

Well, there's at least room for nitpicking (a couple of typos and some inconsistencies as well as superfluous code in spkg-install and probably configuration.py, don't recall)... I'll maybe take a look at it again tomorrow, and probably provide a patch (provided Volker doesn't plan to make further major changes to these files).

How about also installing a user script for convenience to save the built ATLAS libraries to another place (for later use with SAGE_ATLAS_LIB)?

Regarding the mentioned excessive tuning times, I also wonder whether we should use something like SAGE_ATLAS_ARCH=fast (or base) by default, i.e., only do self-tuning if the user explicitly asks for it in some way. I guess the Sage Installation Guide needs to get updated anyway w.r.t. ATLAS and environment variables.

[leif]

The root repo patch should get rebased for Sage 5.2.rc0.

[jhpalmieri]

On hawk (OpenSolaris):

DONE configure                                                                                     
Finished configuring ATLAS.                                                                        
Running make -j1                                                                                   
make[2]: Entering directory `/export/home/palmieri/testing/ATLAS/sage-5.2.rc0/spkg/build/atlas-3.1\
0.0/ATLAS-build'                                                                                   
make[2]: warning: -jN forced in submake: disabling jobserver mode.                                 
make -j1 -f Make.top build                                                                         
make[3]: Entering directory `/export/home/palmieri/testing/ATLAS/sage-5.2.rc0/spkg/build/atlas-3.1\
0.0/ATLAS-build'                                                                                   
Make.top:1: Make.inc: No such file or directory                                                    
Make.top:325: warning: overriding commands for target `/AtlasTest'                                 
Make.top:76: warning: ignoring old commands for target `/AtlasTest'                                
make[3]: *** No rule to make target `Make.inc'.  Stop.                                             
make[3]: Leaving directory `/export/home/palmieri/testing/ATLAS/sage-5.2.rc0/spkg/build/atlas-3.10\
.0/ATLAS-build'                                                                                    
make[2]: *** [build] Error 2                                                                       
make[2]: Leaving directory `/export/home/palmieri/testing/ATLAS/sage-5.2.rc0/spkg/build/atlas-3.10\
.0/ATLAS-build'                                                                                    
------------------------------------------------------------                                       
  File "./spkg-install", line 478, in <module>                                                     
    assert_success(rc, bad='Failed to build ATLAS.', good='Finished building ATLAS core.')         
  File "./spkg-install", line 74, in assert_success                                                
    traceback.print_stack(file=sys.stdout)                                                         
------------------------------------------------------------                                       
Error:  Failed to build ATLAS.                                                                     
                                                                                                   
real    4m10.778s                                                                                  
user    0m7.766s                                                                                   
sys     0m8.391s                                                                                   
Successfully installed atlas-3.10.0                                                                
Deleting temporary build directory                                                                 
/export/home/palmieri/testing/ATLAS/sage-5.2.rc0/spkg/build/atlas-3.10.0                           
Finished installing atlas-3.10.0.spkg 

I don't know why it's not building, but it shouldn't exit saying "Successfully installed atlas-3.10.0". I added a print statement, and "rc" is 512. The ​documentation for sys.exit says that for the argument, "Most systems require it to be in the range 0-127, and produce undefined results otherwise." We could instead do this:

spkg-install

@@ -74 +74 @@
     traceback.print_stack(file=sys.stdout)
     print '-'*60
     if bad is not None:
-        print 'Error: ', bad
-    sys.exit(rc)
+        sys.exit('Error: %s' % bad)
+    sys.exit(1)
 
 ######################################################################
 ### Skip building ATLAS on specific systems

[leif]

On Ubuntu 10.04.4 LTS x86_64 (AMD E-450), with Sage 5.2.rc0 and SAGE_ATLAS_ARCH=fast I get:

...

Building using specific architecture.
Fast configuration on Intel x86_64 compatible CPUs.
Running configure with arch = P4E64SSE3, isa extensions ('SSE3', 'SSE2', 'SSE1'), archdef dir None
Traceback (most recent call last):
  File "./spkg-install", line 454, in <module>
    rc = build()
  File "./spkg-install", line 447, in build
    rc = configure(arch, isa_ext, archdef_dir)
  File "./spkg-install", line 315, in configure
    cmd += ' -A '+str(ATLAS_MACHTYPE.index(arch))
ValueError: tuple.index(x): x not in tuple

real    0m0.701s
user    0m0.090s
sys     0m0.060s
************************************************************************
Error installing package atlas-3.10.0
************************************************************************

[fbissey]

Built successfully on power7. A few oddities in the log but I don't think they are important

make -j1 atlas_run atldir=/hpc/scratch/frb15/sandbox/sage-5.1.beta5/spkg/build/atlas-3.10.0/ATLAS-build exe=xprobe_comp redir=config1.out \
                args="-v 0 -o atlconf.txt -O 1 -A 7 -Si nof77 0 -V 6  -Fa ic '-fPIC' -C sm 'gcc' -Fa sm '-fPIC' -C dm 'gcc' -Fa dm '-fPIC' 
-C sk 'gcc' -Fa sk '-fPIC' -C dk 'gcc' -Fa dk '-fPIC' -C xc 'gcc' -Fa xc '-fPIC' -Fa gc '-fPIC' -C if 'sage_fortran' -Fa if '-fPIC' -b 64 -
d b /hpc/scratch/frb15/sandbox/sage-5.1.beta5/spkg/build/atlas-3.10.0/ATLAS-build"
make[1]: Entering directory `/hpc/scratch/frb15/sandbox/sage-5.1.beta5/spkg/build/atlas-3.10.0/ATLAS-build'
cd /hpc/scratch/frb15/sandbox/sage-5.1.beta5/spkg/build/atlas-3.10.0/ATLAS-build ; ./xprobe_comp -v 0 -o atlconf.txt -O 1 -A 7 -Si nof77 0 
-V 6  -Fa ic '-fPIC' -C sm 'gcc' -Fa sm '-fPIC' -C dm 'gcc' -Fa dm '-fPIC' -C sk 'gcc' -Fa sk '-fPIC' -C dk 'gcc' -Fa dk '-fPIC' -C xc 'gcc
' -Fa xc '-fPIC' -Fa gc '-fPIC' -C if 'sage_fortran' -Fa if '-fPIC' -b 64 -d b /hpc/scratch/frb15/sandbox/sage-5.1.beta5/spkg/build/atlas-3
.10.0/ATLAS-build > config1.out
sh: -c: line 0: unexpected EOF while looking for matching ``'
sh: -c: line 1: syntax error: unexpected end of file
sh: -c: line 0: unexpected EOF while looking for matching ``'
sh: -c: line 1: syntax error: unexpected end of file
sh: -c: line 0: unexpected EOF while looking for matching ``'
sh: -c: line 1: syntax error: unexpected end of file
sh: -c: line 0: unexpected EOF while looking for matching ``'
sh: -c: line 1: syntax error: unexpected end of file
sh: -c: line 0: unexpected EOF while looking for matching ``'
sh: -c: line 1: syntax error: unexpected end of file
probe_f2c.o: In function `ATL_tmpnam':
/hpc/scratch/frb15/sandbox/sage-5.1.beta5/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/include/atlas_sys.h:224: warning: the use of `
tmpnam' is dangerous, better use `mkstemp'

I noticed that libcblas.so is using rpath

ldd -r local/lib/libcblas.so.2.1.0 
        linux-vdso64.so.1 =>  (0x0000040000000000)
        libatlas.so.2 => /hpc/scratch/frb15/sandbox/sage-5.1.beta5/local/lib/libatlas.so.2 (0x0000040000060000)
        libpthread.so.0 => /lib64/power7/libpthread.so.0 (0x00000400008a0000)
        libm.so.6 => /lib64/power7/libm.so.6 (0x00000400008e0000)
        libc.so.6 => /lib64/power7/libc.so.6 (0x00000400009c0000)
        /lib64/ld64.so.1 (0x0000000024560000)

And that's from outside the sage shell. However for f77blas and presumably lapack, libgfortran is not rpath-ed (I am using sage gcc's spkg in this case).

That is important info if you want to put atlas libraries in another location.

I will run a few tests shortly.

[benjaminfjones]

Volker, do you still need architecture defaults for i386? I have access to an old laptop with a Centrino M processor that reports being i386 / i686 in uname -a and in /proc/cpuinfo:

cpu family 6
cpu model 13
Intel Pentium M

I haven't ever built Sage on it, I imagine it would take several years, but I'm willing to give it a try.

[vbraun]

• John, I'm against changing the exit codes, we report whatever the sub-process spat out. So even if its >127, thats just what we were handed so clearly its a supported exit code.

I don't have an account on hawk, but in any case that should not stop us from shipping an updated ATLAS that fixes build issues on modern hard/software. We can always clean up second-tier platforms later. But do post the whole log, the error is likely further up.

• Leif, are you working on a patch? In configure_fast(), we should set

  arch = 'P4E'
  

  instead of P4E64SSE3 is added later. This is what causes your build failure.

• Francois: the weird messages are normal. The rpaths are set by libtools and are what libtools likes to set. As long as we set LD_LIBRARY_PATH this doesn't matter except when we distribute the binaries. This needs to be fixed some day in all shared libs but not on this ticket.

• Benjamin: no I don't need 32-bit archdefs, this should be covered by what ATLAS already ships with. The atlas spkg should build relatively quickly now except for the cases in 72

[jhpalmieri]

Replying to vbraun:

• John, I'm against changing the exit codes, we report whatever the sub-process spat out. So even if its >127, thats just what we were handed so clearly its a supported exit code.

But after sys.exit(512), Python has return code of zero, on sage.math, OS X, and OpenSolaris. So sage-spkg thinks the spkg installed correctly, which it clearly didn't. From the ​documentation for os.system, it looks to me that its output should be divided by 512 to get a return code suitable for sys.exit, but I'm not sure about that. However you want to fix it, it has to be changed: it's not acceptable for the compilation to fail but for spkg-install to have a return code of zero.

I don't have an account on hawk, but in any case that should not stop us from shipping an updated ATLAS that fixes build issues on modern hard/software. We can always clean up second-tier platforms later. But do post the whole log, the error is likely further up.

The log is posted at ​http://sage.math.washington.edu/home/palmieri/misc/atlas-3.10.0.log.

[jdemeyer]

Something along the following lines should be used to handle rc, see ​http://docs.python.org/library/os.html

if os.WIFEXITED(rc):
    rc = os.WEXITSTATUS(rc)
elif os.WIFSIGNALED(rc):
    rc = 128 + os.WTERMSIG(rc)
else:
    raise SystemError("Unknown return value %i for os.system()"%rc)

[vbraun]

I've investigated and reported the Solaris build issue upstream at ​https://sourceforge.net/tracker/?func=detail&aid=3545418&group_id=23725&atid=379483

[vbraun]

The return value is actually the return value of os.system(), which is described in ​http://docs.python.org/library/os.html#os.wait

To extract the exit status, we should just do sys.exit((rc >> 8) & 0x7f). Leif, are you working on the spkg right now?

[vbraun]

The workaround for the Solaris build issue is to use the fqn for $CC and sage_fortran

[vbraun]

Since Leif apparently isn't around I implemented the fqn workaround for the Solaris build and the return status issues. Solaris build is still broken but now at a different place. Updated spkg at the same place, md5sum is 878695a26071cfe73a9977bd8413b748.

[jhpalmieri]

This version now builds on hawk: ​log file here.

[vbraun]

Sounds good! I updated the spkg with yet another SPARC Solaris fix, md5sum is 6dbcf22c920626380f2cba877cca4cb1. Though still doesn't work on mark/skynet, but at least makes it now into the compile phase. In any case SPARC solaris issues shouldn't delay this ticket.

[leif]

Replying to leif:

On Ubuntu 10.04.4 LTS x86_64 (AMD E-450), with Sage 5.2.rc0 and SAGE_ATLAS_ARCH=fast I get:

ValueError: tuple.index(x): x not in tuple

Using SAGE_ATLAS_ARCH=base in contrast worked (and ptestlong passed with Sage 5.2.rc0, FWIW):

real    34m12.187s
user    30m3.170s
sys     5m30.850s
Successfully installed atlas-3.10.0

Still not that fast, but approximately within your estimates...

[vbraun]

Replying to leif:

real 34m12.187s

Thats pretty good for 18W TDP. I take it compiling all of Sage takes 2+ hours on that machine?

[leif]

Replying to vbraun:

Replying to leif:

real 34m12.187s

Thats pretty good for 18W TDP. I take it compiling all of Sage takes 2+ hours on that machine?

Sure. Although ATLAS currently consumes only <= 9W ;-)

Unfortunately ATLAS is built quite late (due to its odd dependency on Sage's Python -- while your script is apparently designed to support Python 2.4 as well), so a fair amount of the time spent building Sage only one core is used (because the remaining packages directly or indirectly depend on ATLAS).

I reinstalled the updated spkg again with SAGE_ATLAS_ARCH=fast:

real    40m20.227s
user    36m26.220s
sys     6m9.500s
Successfully installed atlas-3.10.0

[jhpalmieri]

An update: on hawk, I unpacked a sage-5.2.rc0 tarball, replaced the old ATLAS spkg with this one, and built from scratch. There are a bunch of doctest failures:

The following tests failed:                                                                        
                                                                                                   
        sage -t  --long -force_lib devel/sage/sage/matrix/matrix2.pyx # 12 doctests failed
        sage -t  --long -force_lib devel/sage/sage/misc/functional.py # 1 doctests failed
        sage -t  --long -force_lib devel/sage/sage/finance/time_series.pyx # 6 doctests failed
        sage -t  --long -force_lib devel/sage/sage/numerical/test.py # Killed/crashed
        sage -t  --long -force_lib devel/sage/sage/modular/modform/numerical.py # 3 doctests failed
        sage -t  --long -force_lib devel/sage/sage/numerical/optimize.py # Killed/crashed
        sage -t  --long -force_lib devel/sage/sage/matrix/matrix_double_dense.pyx # 68 doctests failed
        sage -t  --long -force_lib devel/sage/doc/en/a_tour_of_sage/index.rst # Killed/crashed
        sage -t  --long -force_lib devel/sage/doc/en/numerical_sage/cvxopt.rst # Killed/crashed
        sage -t  --long -force_lib devel/sage/doc/fr/a_tour_of_sage/index.rst # Killed/crashed
        sage -t  --long -force_lib devel/sage/doc/tr/a_tour_of_sage/index.rst # Killed/crashed
        sage -t  --long -force_lib devel/sage/sage/combinat/e_one_star.py # Killed/crashed

For example:

File "/export/home/palmieri/testing/ATLAS/sage-5.2.rc0/devel/sage-main/sage/matrix/matrix2.pyx", line 8157:
    sage: (A - M*G).zero_at(10^-12)
Expected:
    [0.0 0.0 0.0]
    [0.0 0.0 0.0]
    [0.0 0.0 0.0]
Got:
    [                                 0.0                                  0.0                                  0.0]
    [  -0.10532733041 + 0.0950573490006*I   0.017805411596 - 0.0512258178986*I -0.0226596712913 + 0.0414519876977*I]
    [  0.100615400305 + 0.0962034401538*I -0.0779990660567 - 0.0543172822202*I   0.057608664751 + 0.0154619373789*I]

and

File "/export/home/palmieri/testing/ATLAS/sage-5.2.rc0/devel/sage-main/sage/misc/functional.py", line 1144:
    sage: norm(M)
Expected:
    10.6903311292
Got:
    10.4323182134

I'll try to build again in case something wrong the first time.

[leif]

Replying to jhpalmieri:

An update: on hawk, I unpacked a sage-5.2.rc0 tarball, replaced the old ATLAS spkg with this one, and built from scratch. There are a bunch of doctest failures:

Did you apply the root repo patch (to remove the BLAS and LAPACK spkgs)?

[jhpalmieri]

Oops, no, I forgot. One more time...

[jhpalmieri]

Now Sage doesn't build on hawk, I guess due to the problems noted on #10509: cvxopt doesn't build, because it says

ld: fatal: library -lblas: not found

I'll skip building cvxopt and continue with the rest of the build.

[vbraun]

I don't think the cvxopt problem is due to #10509. The cvxopt spkg explicitly links against blas, this is bad. From the cvxopt patches/setup.py.patch:

+ libraries = ['m','lapack','gsl','blas','gslcblas','cblas','gfortran','atlas']

this is wrong, it should be f77blas if the Fortran version is actually used or not there at all. Of course all modern systems have a libblas.so somewhere so the linker finds it, notes that it is not used, and proceeds. Except that on Hawk, I guess, there is no system-wide libblas. We should proceed removing blas in this ticket and then fix cvxopt on second-tier platforms later.

[jhpalmieri]

I meant that linking to blas was ​noted at #10509 as a possible problem, not that #10509 was causing this issue.

[vbraun]

You are right, that the patch on #10509 should have been applied to cvxopt a long time ago then it wouldn't break here.

[leif]

Replying to leif:

Replying to vbraun:

Replying to leif:

real 34m12.187s

Thats pretty good for 18W TDP. I take it compiling all of Sage takes 2+ hours on that machine?

[...]

I reinstalled the updated spkg again with SAGE_ATLAS_ARCH=fast:

real    40m20.227s
user    36m26.220s
sys     6m9.500s
Successfully installed atlas-3.10.0

ROFL, with SAGE_ATLAS_ARCH="AMD64K10h,SSE3,SSE2,SSE1,3DNow" (which involves self-tuning) it took

real	36m15.153s
user	31m23.290s
sys	5m56.960s
Successfully installed atlas-3.10.0

Also a bit strange is that the timing for ptestlong (all for Sage 5.2.rc0, GCC 4.4.3) was

base < fast < AMD64K10h

(i.e., fastest with SAGE_ATLAS_ARCH=base), although I think at least during the last run the machine was partially loaded with other stuff as well, and clearly ptestlong isn't very appropriate to benchmark ATLAS performance... ;-)

[Not going to use the ATLAS tools for comparison right now, perhaps later...]

[leif]

P.S.:

Another weird thing are (non-fatal w.r.t. the build) errors like

FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (armv7) for -march= switch
FlagCheck.c:1: error: bad value (armv7) for -mtune= switch
FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (ultrasparc) for -mtune= switch
FlagCheck.c:1: error: bad value (970) for -mtune= switch

even if one specifies the architecture (i.e., on x86).

[jhpalmieri]

On hawk: with the appropriate patches applied, I still get some test failures, but as far as I can tell, they're all due to cvxopt being broken. So it looks pretty good.

[leif]

Is it intentional that static libraries no longer get installed (although built)?

[fbissey]

On cvxopt I am doing a new spkg in #13160 I'll check what I have done there. My main issue with the current spkg is it is horribly overlinked.

[vbraun]

Replying to leif:

Is it intentional that static libraries no longer get installed (although built)?

Upstream really only builds static libraries. But static libraries suck for our purposes. So yet, it is intentional that the static libraries are not installed.

[dimpase]

Replying to vbraun:

I don't think the cvxopt problem is due to #10509. The cvxopt spkg explicitly links against blas, this is bad. From the cvxopt patches/setup.py.patch:

+ libraries = ['m','lapack','gsl','blas','gslcblas','cblas','gfortran','atlas']

this is wrong, it should be f77blas if the Fortran version is actually used or not there at all. Of course all modern systems have a libblas.so somewhere so the linker finds it, notes that it is not used, and proceeds. Except that on Hawk, I guess, there is no system-wide libblas. We should proceed removing blas in this ticket and then fix cvxopt on second-tier platforms later.

Perhaps this ticket should get #13160 as a dependence. One thing I checked is that it appears to work on OSX 10.6.8, both with native blas/lapack, and with Atlas 3.10 from this ticket. I imagine #13160 can get finalized quickly.

[leif]

Replying to leif:

[Not going to use the ATLAS tools for comparison right now, perhaps later...]

FWIW, while make time works, make atlvat2.pdf ... (to build ATLAS vs. ATLAS comparison charts) seems to be broken -- for me it always fails with a buffer overflow.

[leif]

Replying to vbraun:

Replying to leif:

Is it intentional that static libraries no longer get installed (although built)?

Upstream really only builds static libraries. But static libraries suck for our purposes. So yet, it is intentional that the static libraries are not installed.

Well, as long as also the shared libraries are present, they're (usually) preferred over the static ones (i.e., unless one explicitly asks for linking against the latter), so copying these into $SAGE_LOCAL/lib/ IMHO shouldn't hurt. (The static libraries are btw. needed to compare different ATLAS installations; the only way to "keep" them is to reinstall the ATLAS spkg with ./sage -f -s ... or to set SAGE_KEEP_BUILT_SPKGS, and manually copy them.)

Note that previously installed static ATLAS libraries currently don't get removed. Don't know whether that may cause trouble (e.g. with upgrading); see above.

[vbraun]

Its true that it doesn't hurt to have the static libraries as long as you don't use them. This is like saying that a knife doesn't hurt until you are stabbed with it. True, but why put a sharp blade under the couch pillow in hopes that nobody will sit on it?

It would be nice to have some system to compare different atlas versions and compile runs, but thats definitely for another ticket. Ideally the atlas-config python script could save the atlas libraries in a private directory, for example by setting a special environment variable while building atlas. And then have some way to tabulate the performance of different installs.

[vbraun]

I think the only remaining blocker is that it (or rather, cvxopt) doesn't build on hawk. Since I don't have an account, can someone test it (the spkg + patches from this ticket + the cvxopt spkg from #13160)? Everything else in this ticket has been reviewed already, we just have to check that the interaction with cvxopt is fixed on the last "fully supported" platform.

[jhpalmieri]

Cvxopt still doesn't build. I see the same error when using the ATLAS spkg here or when setting SAGE_ATLAS_LIB=/ATLAS32. ​Here is the log.

Does the ATLAS spkg here build on skynet/mark (and Solaris on sparc in general)?

[fbissey]

Why the heck is it not finding gsl? Oh I see the include line is actually wrong. I'll check the spkg but we should continue with cvxopt issues at #13160.

[vbraun]

John, now that you verified that it works on Hawk is there anything else that prevents you from pressing the positive review button? ;-)

[jhpalmieri]

I'm testing on a few skynet machines. Should I expect it to work on mark?

[vbraun]

It worked for me on mark (sparc solaris)

[jdemeyer]

Replying to vbraun:

It worked for me on mark (sparc solaris)

OMG cool, Skynet is back up. I was totally unaware of that!

[leif]

I'm still not happy with "discarding" the built static libraries; there should at least be some convenient way to save them (other than SAGE_KEEP_BUILT_SPKGS=yes or installing with sage (-i|-f) -s ....)

Another issue is the extremely increased build time on some machines if one doesn't set SAGE_ATLAS_ARCH. Don't know how we could handle that, but it certainly gives rise to a lot of user complaints.

[leif]

P.S.: W.r.t. the "knife": If you don't want to install static libraries (somewhere), at least previous ones should get removed (or moved somewhere else) upon a successful ATLAS build.

[vbraun]

Now that I added the generic 64-bit archdefs the default build time (without setting SAGE_ATLAS_ARCH) should be moderate on all x86 systems. I.e. less CPU time than building the rest of Sage.

Your suggestions about handling the static libraries are enhancement requests. By itself, its useless to keep a backup of the static libraries somewhere. I agree that one should keep them around and devise a way to benchmark them, but not on this ticket. Also I'm against attempting to delete stuff from previous installs unless it actively conflicts with the new spkg. Which it does not, the damage of statically linking is already done.

[jhpalmieri]

I'm willing to give this a positive review now. Leif, what about you? Can we defer your issues to a follow-up?

[leif]

Replying to vbraun:

Now that I added the generic 64-bit archdefs the default build time (without setting SAGE_ATLAS_ARCH) should be moderate on all x86 systems. I.e. less CPU time than building the rest of Sage.

Ok, hopefully...

Your suggestions about handling the static libraries are enhancement requests. By itself, its useless to keep a backup of the static libraries somewhere. I agree that one should keep them around and devise a way to benchmark them, but not on this ticket.

I'd rather say not installing them [anywhere] is a regression w.r.t. the previous spkg.

Also I'm against attempting to delete stuff from previous installs unless it actively conflicts with the new spkg. Which it does not, the damage of statically linking is already done.

If so, it shouldn't hurt to keep ATLAS installing them either... ;-)

[I'd expect "more damage" when having different .a and .so library versions.]

[leif]

I'm not wanting to hold up this ticket, but IMHO the Installation Guide should get updated, at least documenting the new atlas-config script.

[vbraun]

Replying to leif:

I'd rather say not installing them [anywhere] is a regression w.r.t. the previous spkg.

Its a major improvement, not a regression!

If so, it shouldn't hurt to keep ATLAS installing them either... ;-)

As I explained previously, thats not true. We have to change things to make them better. But we can't un-link the static linkage that has happened previously, so when you upgrade from an existing spkg you potentially keep the old code. And there is nothing a new atlas spkg can do about this. If you want to be sure that you don't have cruft statically linked you'll have to do a clean install. This is precisely why it was a bad idea to install static libraries previously.

[vbraun]

Initial patch

[vbraun]

I've added documentation to the installation guide for the atlas-config script and updated the environment variables.

[jhpalmieri]

I note that on OS X, the test suite doesn't run, even if ATLAS is built. This could perhaps be dealt with on a follow-up ticket? Maybe only skip self-tests if

"$UNAME" = "Darwin" -a -z "$SAGE_ATLAS_ARCH"

or something like that?

More seriously, I'm having problems with this on the skynet machine iras. I'm building Sage from scratch, using the spkgs from this ticket and #13160. The ATLAS build started about 15 hours ago, and nothing has been written to the log file for at least 6 hours. I'm going to quit and restart the build...

[jhpalmieri]

On skynet machine iras:

real    2251m2.189s
user    2175m58.212s
sys     16m29.112s

This is not really acceptable. And then:

RHS=74, nrm=184.678650
RHS=75, nrm=248.384750
RHS=76, nrm=291.507111
RHS=77, nrm=337.993896
RHS=78, nrm=335.039612
RHS=79, nrm=477.389954
RHS=80, nrm=334.815796
make[5]: *** [csanity_test] Error 1
make[5]: *** Waiting for unfinished jobs....
make[5]: Leaving directory `/home/palmieri/iras/sage-5.2.10508/spkg/build/atlas-\
3.10.0/ATLAS-build/bin'
make[4]: *** [sanity_test] Error 2
make[4]: Leaving directory `/home/palmieri/iras/sage-5.2.10508/spkg/build/atlas-\
3.10.0/ATLAS-build/bin'
make[3]: *** [sanity_test] Error 2
make[3]: Leaving directory `/home/palmieri/iras/sage-5.2.10508/spkg/build/atlas-\
3.10.0/ATLAS-build'
make[2]: *** [test] Error 2
make[2]: Leaving directory `/home/palmieri/iras/sage-5.2.10508/spkg/build/atlas-\
3.10.0/ATLAS-build'
An error occurred when running the ATLAS self-tests.

[leif]

Replying to jhpalmieri:

On skynet machine iras:

real    2251m2.189s
user    2175m58.212s
sys     16m29.112s

Ouch. How long did 3.8.4 take to build, and did you suspend the ECM background jobs? (They're back again...)

Presumably something to report upstream, but perhaps we can tweak it ourselves somehow.

[vbraun]

Clint Whaley (ATLAS developer) doesn't have access to Itanium any more, so its not surprising that it does a full search. If its too slow for you then get yourself a new computer ;-) I'll have a look at the self-test issue, but it seems that the build/install works. Right?

I'd also like to remind everyone that this ticket, even if it doesn't pass its own testsuite on some dead arch, fixes actual issues with modern CPUs as reported on the mailinglist.

[vbraun]

Upstream ticket at ​https://sourceforge.net/tracker/?func=detail&aid=3554553&group_id=23725&atid=379483

[dimpase]

Replying to vbraun:

Clint Whaley (ATLAS developer) doesn't have access to Itanium any more, so its not surprising that it does a full search. If its too slow for you then get yourself a new computer ;-) I'll have a look at the self-test issue, but it seems that the build/install works. Right?

Can't one use the current ATLAS spkg on Itaniums instead of the new one? Or there are incompatibility problems?

Itaniums are not a dying breed, at least AFAIK Intels still plans to release a new Itanium chip this year. So just hoping that they'd be gone soon is too optimistic. Should we try getting Clint Whaley access to an Itanium machine?

[vbraun]

Replying to dimpase:

Can't one use the current ATLAS spkg on Itaniums instead of the new one? Or there are incompatibility problems?

We can use a preinstalled ATLAS library on the buildbots. Given that this is by far the slowest part of the compile cycle, I think thats reasonable. We even do this officially on OSX where we always use the system-provided BLAS. But perhaps easier would be to update the Itanium archdef so that it doesn't have to tune. I just started a compile on iras, but don't expect the answer too soon ;-)

Itaniums are not a dying breed, at least AFAIK Intels still plans to release a new Itanium chip this year.

Well Intel made some unfortunate (for them) contractual commitments here... But ia64 is certainly not getting much love.

Should we try getting Clint Whaley access to an Itanium machine?

I've asked to that extend on the sage-skynet list.

[aginiewicz]

It worked on 2 machines I tested, but failed on a PIII machine - it turns out that Atlas 3.10.0 does not work on PIII but the fix is simple - see ​http://math-atlas.sourceforge.net/errata.html#piiisse2 for confirmation.

I've uploaded spkg containing fix for this to ​https://github.com/downloads/aginiewicz/spkgs/atlas-3.10.0.spkg

[leif]

Replying to vbraun:

I'd also like to remind everyone that this ticket, even if it doesn't pass its own testsuite on some dead arch, fixes actual issues with modern CPUs as reported on the mailinglist.

You mean the (reproducible) segfaults on some AVX CPUs? These can be solved by reinstalling the ATLAS spkg (one or two times) -- which doesn't take that long, as these CPUs are fast... ;-)

No, seriously, as a temporary "solution", we could at least put the old ATLAS, BLAS and LAPACK spkgs into the experimental branch, such that (e.g.) Itanium users can still access them (to be installed manually with ./sage -f ... of course).

W.r.t. the "dead archictecture": It's IMHO always good to test on a broad range of platforms, as that might expose real bugs that just incidentally don't show up on the "mainstream" (Intel, AMD, meanwhile perhaps also ARM); if one day GCC does no longer support Itanium, the situation will be different.

[leif]

Replying to vbraun:

Replying to dimpase:

Can't one use the current ATLAS spkg on Itaniums instead of the new one? Or there are incompatibility problems?

We can use a preinstalled ATLAS library on the buildbots. Given that this is by far the slowest part of the compile cycle, I think thats reasonable.

Therefore I was suggesting to also install some script to easily "save" the ATLAS libs somewhere for later use with SAGE_ATLAS_LIB (orthogonal to the static libraries issue).

[jhpalmieri]

Replying to leif:

Replying to jhpalmieri:

On skynet machine iras:

real    2251m2.189s
user    2175m58.212s
sys     16m29.112s

Ouch. How long did 3.8.4 take to build, and did you suspend the ECM background jobs? (They're back again...)

real    61m42.034s
user    52m55.672s
sys     1m54.940s

When compiling the new ATLAS spkg, I suspended two of the background processes (that is, I did touch /tmp/iras1 and touch /tmp/iras2). When compiling the previous version, I didn't bother to do this.

Replying to vbraun:

I'd also like to remind everyone that this ticket, even if it doesn't pass its own testsuite on some dead arch, fixes actual issues with modern CPUs as reported on the mailinglist.

I don't feel too strongly about it either way, but I would be a little surprised if Jeroen is willing to merge this if it takes so long to build on iras.

[vbraun]

Thats quite a speedup. I guess in your earlier test ATLAS failed to use the SAGE_ATLAS_ARCH=fast build because of system load (first attempt), then tried SAGE_ATLAS_ARCH=base which is the old Itanium (not Itanium 2), which then did a full search for timings. Something like that must have happened, the compile is not just 30x faster without background jobs running.

So I guess all problems are solved?

[jhpalmieri]

Ah, no, I think you misunderstand: it's quite a slow-down. The shorter time (~1 hour) is for 3.8.4 without suspending background processes, the longer time (~37 hours) is for 3.10.0 with two background processes suspended.

[jdemeyer]

I'm going to test this on the buildbot soon, that should give more information.

[jdemeyer]

This causes a build failure for cvxopt on OS X 10.4 PPC:

running build_ext
building 'gsl' extension
creating build/temp.macosx-10.4-ppc-2.7
creating build/temp.macosx-10.4-ppc-2.7/C
gcc -fno-strict-aliasing -fwrapv -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -I/Users/buildbot/build/sage/moufang-1/moufang_full/build/sage-5.3.beta3/local/include -I/Users/buildbot/build/sage/moufang-1/moufang_full/build/sage-5.3.beta3/local/include/python2.7 -c C/gsl.c -o build/temp.macosx-10.4-ppc-2.7/C/gsl.o
C/gsl.c: In function 'initgsl':
C/gsl.c:198:13: warning: variable 'm' set but not used [-Wunused-but-set-variable]
gcc -bundle -undefined dynamic_lookup -L/Users/buildbot/build/sage/moufang-1/moufang_full/build/sage-5.3.beta3/local/lib build/temp.macosx-10.4-ppc-2.7/C/gsl.o -L/Users/buildbot/build/sage/moufang-1/moufang_full/build/sage-5.3.beta3/local/lib -L/Users/buildbot/build/sage/moufang-1/moufang_full/build/sage-5.3.beta3/local/lib -lm -lgsl -lblas -lgfortran -lm -o build/lib.macosx-10.4-ppc-2.7/cvxopt/gsl.so
/usr/bin/ld: can't locate file for: -lblas
collect2: ld returned 1 exit status
error: command 'gcc' failed with exit status 1

[jdemeyer]

Also, ATLAS failed to build on various Buildbot systems: arando (Linux i386), lena (Linux x86_64), cicero (Linux i386), iras (Linux ia64), hawk (OpenSolaris? i386).

I just tried the build once, I will retry them. But in any case, the build is more flaky than the previous version of ATLAS.

[vbraun]

The OSX 10.4 issue is a cvxopt bug. Where is BLAS in the ancient vecLib and why didn't Apple call it libblas? In any case I don't have access to a museum Mac, so somebody else will have to figure that out.

[fbissey]

Technically if we want to use atlas on OS X a number of spkg including cvxopt needs to be updated. I am not sure I would call it a cvxopt bug because that's part of the configuration we introduced in spkg-install. Before the inclusion of this spkg cvxopt probably would have picked up libblas from the blas spkg. We can fix this in cvxopt of course.

[jhpalmieri]

I have had no problems building this on hawk (see /export/home/palmieri/testing/ATLAS/sage-5.3.beta2). On iras, as I said above, it built, but it took a long time. I haven't tried the others that you mentioned.

[jdemeyer]

More buildbot results:

First attempt failed for arando (Linux i386), lena (Linux x86_64), flavius (Linux x86_64), cicero (Linux i386), iras (Linux ia64), hawk (OpenSolaris i386).

Second attempt failed for arando (Linux i386), flavius (Linux x86_64), cicero (Linux i386), iras (Linux ia64), hawk (OpenSolaris i386).

[jhpalmieri]

On the first try, this built successfully for me on iras, lena, and hawk. (I've built it at least twice on iras and hawk, with no failures.) On the other hand, cicero (Linux x86) and flavius (Linux x86_64) both fail to build, with essentially the same error. ​Log file here. They also both took a long time (about 17 hours for cicero) trying to build the ATLAS spkg. I think we need some new different settings, because the build times for cicero, flavius (almost 19 hours), and iras (34.5 hours this time) are way too long.

[vbraun]

Oh I see, cicero/flavius have only a single CPU core. Talking about a blast from the past! In that case ATLAS doesn't build threaded static libraries, so we can't build threaded shared libraries. They are also so old that ATLAS can't figure out whether CPU throttling is enabled, so the full timings are collected instead of immediately falling back to a pre-defined archdef. Instead of trying to catch this in the ATLAS spkg and potentially not running the tuning in cases where we ought to, I think we should just configure the buildbot to set SAGE_ATLAS_ARCH=base on these ancient machines.

[jhpalmieri]

I thought one of the purposes of the buildbot was to test that Sage builds "out of the box" on a variety of platforms, so tuning environment variables for use with the buildbot machines doesn't seem like the right solution. But maybe I'm not thinking about the buildbot the right way.

Also, supporting "ancient" machines is important: if we want mathematicians in underdeveloped countries, for example, to be able to use Sage, they may not have access to anything better.

[vbraun]

Well nobody stops you from tuning ATLAS for a day if you have an ancient computer. Its just annoying on the buildbot if you want to test something else than the ATLAS build. I guess the question is: Do we want, by default, to build a state-of-the-art BLAS library on all computers or only on sufficiently modern ones?

[jdemeyer]

Replying to vbraun:

Oh I see, cicero/flavius have only a single CPU core. Talking about a blast from the past! In that case ATLAS doesn't build threaded static libraries, so we can't build threaded shared libraries.

Is that a bug in ATLAS or a bug in spkg-install? The version of ATLAS currently in Sage doesn't have this problem.

And I'm with John on trying the build "out of the box" on the buildbots as much as possible. And I don't consider single-core machines ancient.

[dimpase]

Replying to jdemeyer:

And I'm with John on trying the build "out of the box" on the buildbots as much as possible.

well, 37 hours for a single package sounds like buildbot abuse. And I think that the default for slow machines with a single CPUs should be a basic Atlas install, with as little tuning as possible. There should be however be good documentation available on how to improve Atlas performance on a host, by rebuilding the spkg.

And I don't consider single-core machines ancient.

One way or another, single-core machines do not need many features of the new Atlas. Perhaps it's time to think about versioning possibilities: a meta-package Atlas that installs an appropriate for the architecture version?

[vbraun]

Replying to jdemeyer:

And I'm with John on trying the build "out of the box" on the buildbots as much as possible.

So if you are fine with waiting 2 days for the slower buildbots to complete then that settles it.

[jhpalmieri]

Is there some sort of preliminary testing which could be done in spkg-install to try to catch machines which might take many hours to build ATLAS? For those machines, we could set the appropriate environment variable to skip as much tuning as possible, and then for quicker machines, we do more tuning. Something like

if [ -z "$SAGE_ATLAS_ARCH" && machine_is_too_old ]; then
   SAGE_ATLAS_ARCH='skip tuning' # or whatever
fi

(For 'machine_is_too_old', maybe just seeing if it only has one core would be enough.) Then the user can force tuning by setting SAGE_ATLAS_ARCH to something appropriate. I'm not sure this the the right variable, but anyway, I hope you understand the idea. Is this workable?

[vbraun]

The old machines are slow because they are old and because there are no up-to-date architectural defaults for ancient hardware. None of this can be fixed by a magic environment variable (except pointing SAGE_ATLAS_LIB to the system atlas libraries). Basically, you can't skip tuning sparc just because you have i386 archdefs. There is a lot of assembler code that will be unhappy.

Also, I don't think there is an easy way to know that atlas would take a long time without actually building it. We could just install a reference blas on all non-starte-of-the-art machines (say, less than 4 cores and 2GHz). But this will be a major rewrite of how Sage builds its BLAS libraries.

[jdemeyer]

Replying to vbraun:

So if you are fine with waiting 2 days for the slower buildbots to complete then that settles it.

It doesn't quite settle it, because the new spkg simply fails to build on some machines.

[vbraun]

I know, and I will make sure it builds on those machines. What I can't do is make ATLAS build fast on ancient hardware.

[jdemeyer]

I would like to know the answer to this question: do what extent are these issues upstream issues or issues with Sage? Because the current ATLAS-3.8.4 works fine.

Replying to jdemeyer:

Replying to vbraun:

Oh I see, cicero/flavius have only a single CPU core. Talking about a blast from the past! In that case ATLAS doesn't build threaded static libraries, so we can't build threaded shared libraries.

Is that a bug in ATLAS or a bug in spkg-install? The version of ATLAS currently in Sage doesn't have this problem.

[vbraun]

The current ATLAS does not work fine for some modern CPUs (search the mailinglist for reports), nor does it have AVX support. Realistically, if you want to have a fast BLAS then you need to track upstream pretty closely since a lot of linear algebra performance boils down to having a good (asm) implementation tuned to your particular hardware.

The build failures that I have seen logs for were just bugs in the Sage buildscript that I'm working on fixing.

[vbraun]

The updated spkg fixes the issue with single-core processors. Builds fine on cicero.

[jhpalmieri]

Builds on cicero and flavius. On mark (Solaris), I see a problem building scipy. Could this be related? ​Log here.

Edit: more info, from the numpy log:

atlas_threads_info:
Setting PTATLAS=ATLAS
  libraries lapack_atlas not found in /home/palmieri/mark/sage-5.3.beta1-ATLAS/local/lib
numpy.distutils.system_info.atlas_threads_info
Setting PTATLAS=ATLAS
/home/palmieri/mark/sage-5.3.beta1-ATLAS/spkg/build/numpy-1.5.1.p1/src/numpy/distutils/system_info.py:1010: UserWarning:
*********************************************************************
    Lapack library (from ATLAS) is probably incomplete:
      size of /home/palmieri/mark/sage-5.3.beta1-ATLAS/local/lib/liblapack.so is 360k (expected >4000k)

    Follow the instructions in the KNOWN PROBLEMS section of the file
    numpy/INSTALL.txt.
*********************************************************************

If you have an account on skynet, feel free to look around /home/palmieri/mark/sage-5.3.beta1-ATLAS/.

[fbissey]

Yes totally ATLAS related. May you could also post ATLAS's log or at least the end bit of it where it builds lapack.

[vbraun]

I've added archdefs for Itanium2, now build time is a lot faster:

vbraun@iras:~/opt/iras/sage-5.3.beta0> SAGE_ATLAS_ARCH=fast ./sage -f http://www.stp.dias.ie/~vbraun/Sage/spkg/atlas-3.10.0.spkg
[...]
real    44m44.170s
user    41m47.156s
sys     2m42.456s

Though if you don't specify SAGE_ATLAS_ARCH it will still run, by default, a full tuning. If that fails for any reason it will try SAGE_ATLAS_ARCH=fast next.

[jhpalmieri]

Replying to fbissey:

Yes totally ATLAS related. May you could also post ATLAS's log or at least the end bit of it where it builds lapack.

It's ​here.

[fbissey]

I am not sure what happens at all. I am not convinced that the original lapack archive is getting relinked along with the C interface from ATLAS but that's not easy to establish from the log because of the size. I would need to spend some time looking at it carefully.

[kcrisman]

The OSX 10.4 issue is a cvxopt bug. Where is BLAS in the ancient vecLib and why didn't Apple call it libblas? In any case I don't have access to a museum Mac, so somebody else will have to figure that out.

Reporting for duty! Currently building MPIR, with with changes in this ticket (except the doctest/doc patch, which is no biggie).

Would this issue only come up if someone did set SAGE_ATLAS_ARCH=base or at any rate set it? I though that using the system ATLAS was still the default on Mac after this ticket - was that a misunderstanding? I might find out before I leave work today... but if it doesn't break without resetting that variable, I don't think that is as bad, since hopefully no one using this old of a machine is depending on it for fast linear algebra.

[fbissey]

Replying to kcrisman:

The OSX 10.4 issue is a cvxopt bug. Where is BLAS in the ancient vecLib and why didn't Apple call it libblas? In any case I don't have access to a museum Mac, so somebody else will have to figure that out.

Reporting for duty! Currently building MPIR, with with changes in this ticket (except the doctest/doc patch, which is no biggie).

Would this issue only come up if someone did set SAGE_ATLAS_ARCH=base or at any rate set it? I though that using the system ATLAS was still the default on Mac after this ticket - was that a misunderstanding? I might find out before I leave work today... but if it doesn't break without resetting that variable, I don't think that is as bad, since hopefully no one using this old of a machine is depending on it for fast linear algebra.

That would be a start but most of the spkg using blas have specific rules for OS X and we would need to change them before they start using ATLAS instead of veclib. From the top of my head:

• iml (actually that would solve some problems)
• linbox
• numpy
• scipy
• cvxopt
• R?

I believe sage itself can cope with the change as is but it should be checked.

[kcrisman]

My point was not that we should switch to ATLAS, but rather that it might be possible to have a different ticket for making ATLAS work optionally on Mac this way, if Sage builds and tests fine with this new ATLAS in the case where Mac does not actually build it. I don't know if that's true, though, since jdemeyer's report could perhaps have been with that env var set.

[kcrisman]

Well, it turns out that it must be removing the other spkgs that does it.

building 'gsl' extension
creating build/temp.macosx-10.4-ppc-2.7
creating build/temp.macosx-10.4-ppc-2.7/C
gcc -fno-strict-aliasing -fwrapv -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -I/Users/student/Desktop/sage-5.3.rc0/local/include -I/Users/student/Desktop/sage-5.3.rc0/local/include/python2.7 -c C/gsl.c -o build/temp.macosx-10.4-ppc-2.7/C/gsl.o
C/gsl.c: In function 'initgsl':
C/gsl.c:198:13: warning: variable 'm' set but not used [-Wunused-but-set-variable]
gcc -bundle -undefined dynamic_lookup -L/Users/student/Desktop/sage-5.3.rc0/local/lib build/temp.macosx-10.4-ppc-2.7/C/gsl.o -L/Users/student/Desktop/sage-5.3.rc0/local/lib -L/Users/student/Desktop/sage-5.3.rc0/local/lib -lm -lgsl -lblas -lgfortran -lm -o build/lib.macosx-10.4-ppc-2.7/cvxopt/gsl.so
/usr/bin/ld: can't locate file for: -lblas
collect2: ld returned 1 exit status
error: command 'gcc' failed with exit status 1
Error building/installing cvxopt

real    0m3.866s
user    0m1.281s
sys     0m1.359s
************************************************************************
Error installing package cvxopt-1.1.5
************************************************************************

Now I understand Volker's comment. Maybe we can just have a different configuration when using old Macs, as in François' comment. But I wouldn't know how to do that.

[kcrisman]

With SAGE_ATLAS_ARCH=base on the same machine, I get

Base configuration on PPC.
Running configure with arch = POWER4, isa extensions (), archdef dir None
Running ../src/configure --prefix=/Users/student/Desktop/sage-5.3.rc0/local --with-netlib-lapack-tarfile=/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/lapack-3.4.1.tgz --cc="gcc" -Si latune 0 -Fa alg -fPIC -C if "/Users/student/Desktop/sage-5.3.rc0/local/bin/sage_fortran" -C acg "/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc" -b 32 -O 12 -A 2 -V 0
gcc -I/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/include  -g -w -c /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/src/atlconf_misc.c
gcc -I/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/include  -g -w -o xconfig /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/src/config.c atlconf_misc.o 
./xconfig -d s /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src/ -d b /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build  -Si lapackref 1 -Si latune 0 -Fa alg -fPIC -C if /Users/student/Desktop/sage-5.3.rc0/local/bin/sage_fortran -C acg /Users/student/Desktop/sage-5.3.rc0/local/bin/gcc -b 32 -O 12 -A 2 -V 0
gcc -I/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/include  -g -w -o xisgcc /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/src/IsGcc.c atlconf_misc.o 
gcc -I/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/include  -g -w -c /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/src/probe_comp.c
gcc -I/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build/../src//CONFIG/include  -g -w -o xprobe_comp probe_comp.o atlconf_misc.o 
rm -f config1.out
make -j1 atlas_run atldir=/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build exe=xprobe_comp redir=config1.out \
                args="-v 0 -o atlconf.txt -O 12 -A 2 -Si nof77 0 -V 0  -C ic '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa ic '-fPIC' -C sm '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa sm '-fPIC' -C dm '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa dm '-fPIC' -C sk '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa sk '-fPIC' -C dk '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa dk '-fPIC' -C xc '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa xc '-fPIC' -C gc '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa gc '-fPIC' -C if '/Users/student/Desktop/sage-5.3.rc0/local/bin/sage_fortran' -Fa if '-fPIC' -b 32 -d b /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build"
cd /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build ; ./xprobe_comp -v 0 -o atlconf.txt -O 12 -A 2 -Si nof77 0 -V 0  -C ic '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa ic '-fPIC' -C sm '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa sm '-fPIC' -C dm '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa dm '-fPIC' -C sk '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa sk '-fPIC' -C dk '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa dk '-fPIC' -C xc '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa xc '-fPIC' -C gc '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa gc '-fPIC' -C if '/Users/student/Desktop/sage-5.3.rc0/local/bin/sage_fortran' -Fa if '-fPIC' -b 32 -d b /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/ATLAS-build > config1.out
/bin/sh: line 1: ./xctest: Bad CPU type in executable
make[5]: *** [atlas_run] Error 126
make[4]: *** [IRunCComp] Error 2


Unable to find usable compiler for ICC; abortingMake sure compilers are in your path, and specify good compilers to configure
(see INSTALL.txt or 'configure --help' for details)make[3]: *** [atlas_run] Error 1
make[2]: *** [IRun_comp] Error 2
ERROR 512 IN SYSCMND: 'make IRun_comp args="-v 0 -o atlconf.txt -O 12 -A 2 -Si nof77 0 -V 0  -C ic '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa ic '-fPIC' -C sm '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa sm '-fPIC' -C dm '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa dm '-fPIC' -C sk '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa sk '-fPIC' -C dk '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa dk '-fPIC' -C xc '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa xc '-fPIC' -C gc '/Users/student/Desktop/sage-5.3.rc0/local/bin/gcc' -Fa gc '-fPIC' -C if '/Users/student/Desktop/sage-5.3.rc0/local/bin/sage_fortran' -Fa if '-fPIC' -b 32"'

Assembly configured as GAS_PPC (4)

Bad VECFLAG value=0, res='VECFLAG=0
'

Vector ISA Extension configured as   (0,0)

Clock rate configured as 700Mhz

Maximum number of threads configured as  1
Cannot detect CPU throttling.
mkdir src bin tune interfaces
cd src ; mkdir testing auxil blas lapack pthreads threads
cd src/blas ; \
           mkdir f77reference reference gemv ger gemm kbmm \
                 level1 level2 level3 pklevel3
cd src/blas/reference ; mkdir level1 level2 level3
cd src/blas/level2 ; mkdir kernel
cd src/blas/pklevel3 ; mkdir gpmm sprk
cd src/blas/level3 ; mkdir rblas kernel
cd src/pthreads ; mkdir blas misc
cd src/pthreads/blas ; mkdir level1 level2 level3
cd src/threads ; mkdir blas lapack
cd src/threads/blas ; mkdir level3 level2
cd tune ; mkdir blas sysinfo lapack threads
cd tune/blas ; mkdir gemm gemv ger level1 level3
cd interfaces ; mkdir blas lapack
cd interfaces/lapack ; mkdir C F77
cd interfaces/lapack/C ; mkdir src testing
cd interfaces/lapack/F77 ; mkdir src testing
cd interfaces/blas ; mkdir C F77
cd interfaces/blas/C ; mkdir src testing
cd interfaces/blas/F77 ; mkdir src testing
cd interfaces/lapack ; mkdir C2F
cd interfaces/lapack/C2F ; mkdir src
mkdir ARCHS
make -j1 -f Make.top startup
Make.top:1: Make.inc: No such file or directory
Make.top:325: warning: overriding commands for target `/AtlasTest'
Make.top:76: warning: ignoring old commands for target `/AtlasTest'
make[3]: *** No rule to make target `Make.inc'.  Stop.
make[2]: *** [startup] Error 2
mv: rename lib/Makefile to lib/Make.tmp: No such file or directory
../src/configure: line 450: lib/Makefile: No such file or directory
../src/configure: line 451: lib/Makefile: No such file or directory
../src/configure: line 452: lib/Makefile: No such file or directory
../src/configure: line 453: lib/Makefile: No such file or directory
../src/configure: line 509: lib/Makefile: No such file or directory
DONE configure
Finished configuring ATLAS.
Running make -j1
make -j1 -f Make.top build
Make.top:1: Make.inc: No such file or directory
Make.top:325: warning: overriding commands for target `/AtlasTest'
Make.top:76: warning: ignoring old commands for target `/AtlasTest'
make[3]: *** No rule to make target `Make.inc'.  Stop.
make[2]: *** [build] Error 2
------------------------------------------------------------
  File "./spkg-install", line 476, in <module>
    assert_success(rc, bad='Failed to build ATLAS.', good='Finished building ATLAS core.')
  File "./spkg-install", line 76, in assert_success
    traceback.print_stack(file=sys.stdout)
------------------------------------------------------------
Error:  Failed to build ATLAS.

real    1m2.106s
user    0m15.388s
sys     0m25.736s
************************************************************************
Error installing package atlas-3.10.0
************************************************************************

[vbraun]

Oh I take it your museum piece doesn't have a Power4 CPU. Can you try

SAGE_ATLAS_ARCH=PPCG4 sage -f http://www.stp.dias.ie/~vbraun/Sage/spkg/atlas-3.10.0.spkg

[kcrisman]

Correct, PowerPC G4, which I think is slightly different. It's currently compiling fine, I'll let you know the results.

---

In other news, #13408 seems to be a dup of this ticket. I don't know if we support "rhel 5" systems currently, nor whether this particular spkg fixes that person's problem, though.

[kcrisman]

Finished installing shared ATLAS library.
Copying /Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0/patches/atlas-config to /Users/student/Desktop/sage-5.3.rc0/local/bin

real    1558m23.512s
user    1148m7.019s
sys     243m20.134s
Successfully installed atlas-3.10.0
Deleting temporary build directory
/Users/student/Desktop/sage-5.3.rc0/spkg/build/atlas-3.10.0
Making Python scripts relocatable...
Finished installing atlas-3.10.0.spkg

Good as far as it goes! Is this the record for longest ATLAS build ever?

Typing make after that still leads to the cvxopt can't find file for -lblas error; can someone remind me what the command would be to force rebuilding of dependencies (if that would even make a difference)?

[vbraun]

So all we have to do is build ATLAS on OSX 10.4, then we don't have to figure out how to drive the ancient accelerate framework. Dima, can you change cvxopt to link against atlas on 10.4?

[kcrisman]

I see, so that's why the problem is the same. Sounds like either we would do this, or fbissey's idea of a different configuration now.

I've touched cvxopt and am trying to build the rest of Sage now, so don't hold your breath yet that this was a permanent fix (in addition to roughly doubling the build time of Sage, not that this would matter on this platform).

[vbraun]

I've changed the spkg to build by default on PPC OSX machines (Darwin + uname -p equals powerpc).

[kcrisman]

Like I said, don't count those chickens. I think this is pretty clearly related, given the libatlas reference.

building package 'graphics'
mkdir ../../../library/graphics
mkdir ../../../library/graphics/R
mkdir ../../../library/graphics/po
byte-compiling package 'graphics'
Error in dyn.load(file, DLLpath = DLLpath, ...) : 
  unable to load shared object '/Users/student/Desktop/sage-5.3.rc0/spkg/build/r-2.14.0.p3/src/library/grDevices/libs/grDevices.so':
  dlopen(/Users/student/Desktop/sage-5.3.rc0/spkg/build/r-2.14.0.p3/src/library/grDevices/libs/grDevices.so, 6): Symbol not found: _ATL_DecAtomicCount
  Referenced from: /Users/student/Desktop/sage-5.3.rc0/local/lib/libatlas.2.dylib
  Expected in: dynamic lookup
Calls: <Anonymous> ... namespaceImport -> loadNamespace -> library.dynam -> dyn.load
Execution halted
make[6]: *** [../../../library/graphics/R/graphics.rdb] Error 1
make[5]: *** [all] Error 2
make[4]: *** [R] Error 1
make[3]: *** [R] Error 1
make[2]: *** [R] Error 1
Error building R.

real    44m11.283s
user    28m7.249s
sys     6m59.266s
************************************************************************
Error installing package r-2.14.0.p3
************************************************************************

I think that's the last spkg other than rubiks and sagetex, actually, so cvxopt and r might be the only problems with this approach.

[kcrisman]

Haha, I was wrong. Same error in Scipy:

23, in <module>
    from numpy.linalg import lapack_lite
ImportError: dlopen(/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so, 2): Symbol not found: _ATL_DecAtomicCount
  Referenced from: /Users/student/Desktop/sage-5.3.rc0/local/lib/libatlas.2.dylib
  Expected in: dynamic lookup

Error building scipy.

real    0m3.779s
user    0m0.437s
sys     0m0.840s
************************************************************************
Error installing package scipy-0.9.p1
************************************************************************

[kcrisman]

And finally:

spkg/pipestatus "./sage --docbuild --no-pdf-links all html  2>&1" "tee -a dochtml.log"
Warning: Could not import sage.calculus.riemann dlopen(/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so, 2): Symbol not found: _ATL_DecAtomicCount
  Referenced from: /Users/student/Desktop/sage-5.3.rc0/local/lib/libatlas.2.dylib
  Expected in: dynamic lookup

Traceback (most recent call last):
  File "/Users/student/Desktop/sage-5.3.rc0/devel/sage/doc/common/builder.py", line 1098, in <module>
    getattr(get_builder(name), type)()
  File "/Users/student/Desktop/sage-5.3.rc0/devel/sage/doc/common/builder.py", line 243, in _wrapper
    getattr(get_builder(document), name)(*args, **kwds)
  File "/Users/student/Desktop/sage-5.3.rc0/devel/sage/doc/common/builder.py", line 363, in _wrapper
    self.write_auto_rest_file(module_name)
  File "/Users/student/Desktop/sage-5.3.rc0/devel/sage/doc/common/builder.py", line 622, in write_auto_rest_file
    title = self.get_module_docstring_title(module_name)
  File "/Users/student/Desktop/sage-5.3.rc0/devel/sage/doc/common/builder.py", line 580, in get_module_docstring_title
    __import__(module_name)
  File "numpy.pxd", line 154, in init sage.calculus.interpolators (sage/calculus/interpolators.c:5716)
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__init__.py", line 136, in <module>
    import add_newdocs
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/add_newdocs.py", line 9, in <module>
    from numpy.lib import add_newdoc
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/__init__.py", line 13, in <module>
    from polynomial import *
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/polynomial.py", line 11, in <module>
    import numpy.core.numeric as NX
AttributeError: 'module' object has no attribute 'core'
make: *** [doc-html] Error 1

so it won't even build doc, which is weird, since it did say that Sage started properly.

[fbissey]

The last two errors are really a misconfiguration/miscompilation of numpy. If you look at /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so with otool -L (I think) you may see something. In fact the deeper problem seem to be that the dynamic lookup doesn't find libatlas.dylib. I am not sure if explicit linking would solve the problem. We should look at the numpy spkg and also atyour numpy log to see how lapack_lite.so was linked.

[kcrisman]

:~/Desktop/sage-5.3.rc0 student$ otool -L local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so 
local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so:
        /System/Library/Frameworks/Accelerate.framework/Versions/A/Accelerate (compatibility version 1.0.0, current version 4.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /Users/student/Desktop/sage-5.3.rc0/local/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.1.12)

I've uploaded the log to ​http://sage.math.washington.edu/home/kcrisman/numpy-1.5.1.p1.log as well. What do you think the R problem is?

[fbissey]

For R it looks like a similar problem. R uses lapack and compile its own subset if not present. It also looks for ATLAS specific bits. Did you run otool from the sage shell? In this case I don't think it matters but you should look at those library problem from a sage shell to get the right LD_LIBRARY_PATH settings.

[kcrisman]

$ otool -L local/lib/python2.7/site-pack
ages/numpy/linalg/lapack_lite.so 
local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so:
        /System/Library/Frameworks/Accelerate.framework/Versions/A/Accelerate (compatibility version 1.0.0, current version 4.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /Users/student/Desktop/sage-5.3.rc0/local/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.1.12)

[fbissey]

OK that's weird. According to your log of building numpy, lapack_lite.so was compiled against veclib rather than atlas and this is in accord with your otool output. So the problem may steem from a relinking or hiding of the veclib library by atlas. It is probably finding the symbols in the ATLAS install first. The problem in turn seems to be that libatlas.2.dylib is missing a symbol or that it should be provided by an extra library not provided. otool -L -r on libatlas.2.dylib? (I hope the -r does what I except: same thing as -r in ldd in linux), actually on lapack_lite.so while you are at it.

[kcrisman]

I'm not sure that was what you wanted.

$ otool -L -r local/lib/libatlas.dylib local/lib/libatlas.dylib:
        /Users/student/Desktop/sage-5.3.rc0/local/lib/libatlas.2.dylib (compatibility version 4.0.0, current version 4.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /Users/student/Desktop/sage-5.3.rc0/local/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.1.12)
External relocation information 0 entries
address  pcrel length extern type    scattered symbolnum/value
Local relocation information 1107 entries
address  pcrel length extern type    scattered symbolnum/value
005a7000 0     2      0      0       0         1
005a7c4c 0     2      0      0       0         1
005a7c48 0     2      0      0       0         1
005a7c44 0     2      0      0       0         1

This continues for a long, long while.

[fbissey]

It's not. I am not sure there is an otool equivalent for "ldd -r". Another point: from the error you get I am expecting that if you start python from the sage shell and then "import numpy" if will fail with a similar error. Starting python with "-v" may show us what libraries get loaded exactly when you import numpy.

[kcrisman]

According to the man page for ldd, that option only works on ELF binaries (whatever that means), and ​this blog post confirms that Mac doesn't use them. ​This other blog post suggests that dyld might be useful...

Numpy results coming up.

[kcrisman]

Same symbol not found, all the way at the end. If this is too long let me know and I'll edit (!) this comment.

>>> import numpy
import numpy # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__init__.py
import numpy # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__init__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__config__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__config__.py
import numpy.__config__ # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__config__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/version.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/version.py
import numpy.version # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/version.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/_import_tools.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/_import_tools.py
import numpy._import_tools # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/_import_tools.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/add_newdocs.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/add_newdocs.py
import numpy.add_newdocs # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/add_newdocs.pyc
import numpy.lib # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/__init__.py
import numpy.lib # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/__init__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/info.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/info.py
import numpy.lib.info # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/info.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/type_check.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/type_check.py
import numpy.lib.type_check # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/type_check.pyc
import numpy.core # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/__init__.py
import numpy.core # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/__init__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/info.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/info.py
import numpy.core.info # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/info.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/multiarray.so", 2);
import numpy.core.multiarray # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/multiarray.so
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/umath.so", 2);
import numpy.core.umath # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/umath.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/_internal.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/_internal.py
import numpy.core._internal # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/_internal.pyc
import numpy.compat # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/__init__.py
import numpy.compat # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/__init__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/_inspect.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/_inspect.py
import numpy.compat._inspect # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/_inspect.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/py3k.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/py3k.py
import numpy.compat.py3k # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/compat/py3k.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/numerictypes.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/numerictypes.py
import numpy.core.numerictypes # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/numerictypes.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/_sort.so", 2);
import numpy.core._sort # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/_sort.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/numeric.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/numeric.py
import numpy.core.numeric # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/numeric.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/_dotblas.so", 2);
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/arrayprint.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/arrayprint.py
import numpy.core.arrayprint # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/arrayprint.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/fromnumeric.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/fromnumeric.py
import numpy.core.fromnumeric # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/fromnumeric.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/cPickle.so", 2);
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/cStringIO.so", 2);
import cStringIO # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/cStringIO.so
import cPickle # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/cPickle.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/defchararray.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/defchararray.py
import numpy.core.defchararray # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/defchararray.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/records.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/records.py
import numpy.core.records # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/records.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/memmap.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/memmap.py
import numpy.core.memmap # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/memmap.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/scalarmath.so", 2);
import numpy.core.scalarmath # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/scalarmath.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/function_base.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/function_base.py
import numpy.core.function_base # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/function_base.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/machar.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/machar.py
import numpy.core.machar # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/machar.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/getlimits.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/getlimits.py
import numpy.core.getlimits # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/getlimits.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/shape_base.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/shape_base.py
import numpy.core.shape_base # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/core/shape_base.pyc
import numpy.testing # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/__init__.py
import numpy.testing # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/__init__.pyc
import unittest # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/__init__.py
import unittest # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/__init__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/result.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/result.py
import unittest.result # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/result.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/StringIO.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/StringIO.py
import StringIO # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/StringIO.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/util.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/util.py
import unittest.util # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/util.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/collections.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/collections.py
import collections # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/collections.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_collections.so", 2);
import _collections # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_collections.so
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/operator.so", 2);
import operator # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/operator.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/keyword.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/keyword.py
import keyword # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/keyword.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/heapq.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/heapq.py
import heapq # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/heapq.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/itertools.so", 2);
import itertools # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/itertools.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/bisect.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/bisect.py
import bisect # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/bisect.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_bisect.so", 2);
import _bisect # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_bisect.so
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_heapq.so", 2);
import _heapq # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_heapq.so
import thread # builtin
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/functools.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/functools.py
import functools # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/functools.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_functools.so", 2);
import _functools # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/_functools.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/case.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/case.py
import unittest.case # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/case.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/difflib.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/difflib.py
import difflib # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/difflib.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/pprint.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/pprint.py
import pprint # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/pprint.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/suite.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/suite.py
import unittest.suite # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/suite.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/loader.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/loader.py
import unittest.loader # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/loader.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/fnmatch.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/fnmatch.py
import fnmatch # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/fnmatch.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/main.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/main.py
import unittest.main # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/main.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/runner.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/runner.py
import unittest.runner # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/runner.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/time.so", 2);
import time # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/time.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/signals.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/signals.py
import unittest.signals # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/unittest/signals.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python/weakref.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python/weakref.py
import weakref # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python/weakref.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/decorators.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/decorators.py
import numpy.testing.decorators # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/decorators.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/utils.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/utils.py
import numpy.testing.utils # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/utils.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/nosetester.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/nosetester.py
import numpy.testing.nosetester # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/nosetester.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/numpytest.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/numpytest.py
import numpy.testing.numpytest # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/testing/numpytest.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/ufunclike.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/ufunclike.py
import numpy.lib.ufunclike # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/ufunclike.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/index_tricks.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/index_tricks.py
import numpy.lib.index_tricks # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/index_tricks.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/math.so", 2);
import math # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/lib-dynload/math.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/function_base.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/function_base.py
import numpy.lib.function_base # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/function_base.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/twodim_base.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/twodim_base.py
import numpy.lib.twodim_base # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/twodim_base.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/_compiled_base.so", 2);
import numpy.lib._compiled_base # dynamically loaded from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/_compiled_base.so
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/arraysetops.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/arraysetops.py
import numpy.lib.arraysetops # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/arraysetops.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/utils.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/utils.py
import numpy.lib.utils # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/utils.pyc
import numpy.matrixlib # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/matrixlib
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/matrixlib/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/matrixlib/__init__.py
import numpy.matrixlib # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/matrixlib/__init__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/matrixlib/defmatrix.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/matrixlib/defmatrix.py
import numpy.matrixlib.defmatrix # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/matrixlib/defmatrix.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/shape_base.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/shape_base.py
import numpy.lib.shape_base # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/shape_base.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py
import numpy.lib.stride_tricks # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/scimath.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/scimath.py
import numpy.lib.scimath # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/scimath.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/polynomial.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/polynomial.py
import numpy.lib.polynomial # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/polynomial.pyc
import numpy.linalg # directory /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/__init__.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/__init__.py
import numpy.linalg # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/__init__.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/info.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/info.py
import numpy.linalg.info # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/info.pyc
# /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/linalg.pyc matches /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/linalg.py
import numpy.linalg.linalg # precompiled from /Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/linalg.pyc
dlopen("/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so", 2);
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/__init__.py", line 136, in <module>
    import add_newdocs
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/add_newdocs.py", line 9, in <module>
    from numpy.lib import add_newdoc
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/__init__.py", line 13, in <module>
    from polynomial import *
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/lib/polynomial.py", line 17, in <module>
    from numpy.linalg import eigvals, lstsq
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/__init__.py", line 48, in <module>
    from linalg import *
  File "/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/linalg.py", line 23, in <module>
    from numpy.linalg import lapack_lite
ImportError: dlopen(/Users/student/Desktop/sage-5.3.rc0/local/lib/python2.7/site-packages/numpy/linalg/lapack_lite.so, 2): Symbol not found: _ATL_DecAtomicCount
  Referenced from: /Users/student/Desktop/sage-5.3.rc0/local/lib/libatlas.2.dylib
  Expected in: dynamic lookup

[GeorgSWeber]

Hi,

now that I re-read the comments here again, I seem to faintly remember this problem circle from my adventure with lmonade (Gentoo Prefix derivative) on OS X 10.4 last year. See my travel account about this, points "22." and "26." and especially the rant following that in:

​https://groups.google.com/d/msg/lmnd-devel/1ZqHx1JBY2Q/syAOZYtZN4IJ

1. If my memory is correct, the symbol "_ATL_DecAtomicCount" comes from/is resolved in libblas(.dylib), this could be checked independently from the underlying OS. Where is libblas.dylib (libblas.so under Linux) located, is it something like .../lib/blas/reference/?

2. Karl-Dieter, I try to catch up with what you installed and tried, but till then could you please try

"sage -R"

anyway in the meantime (from the shell prompt) and print the result, probably an error message from the OS X 10.4 less-than-optimal dyld dynamic loader?

3. If this really is what the problem is about, some symlinks might suffice, but I do not remember exactly how and for which libraries (both R and maxima were affected in a weird way, and Volker's compilerwrapper intervenes in the lmonade setting, too).

Best regards,

Georg

P.S.: Later last year, I ultimately managed somehow to get some 64-bit Sage (lmonade flavour) working on the 32-bit OS X 10.4, so in the course, I definitely had to successfully build and link/reference some new "self-made" 64-bit atlas/blas, since OS X 10.4 only ships with a 32-bit "system copy" of atlas/blas, so this is definitely doable. But even Python itself is not officially supported as 64-bit version on OS X 10.4, so quite some more hackery was needed. (This only might be interesting, if one day the Sage project wants to get rid of the dependency of certain Apple system library headers, which come only with "XCode", or more precisely the "command line tools", on newer OS X versions --- currently these are a prerequisite for building Sage on OS X.)

[fbissey]

Yeah I remember those bits vaguely now. But lmnd never used libvec unless you hacked something (I have an ebuild to allow that but no-one else has seen it). The location you mention in (1) is specific to gentoo and correspond to netlib blas (otherwise called blas reference) current lmnd uses a different scheme. A lmnd user could pick openblas or gotoblas these days.

The link idea is interesting, having $SAGE_LOCAL/lib/libblas.dylib pointing to libvec could help.

[kcrisman]

(sage-sh) student@Dasher-03:sage-5.3.rc0$ ./sage -R
/Users/student/Desktop/sage-5.3.rc0/spkg/bin/sage: line 496: /Users/student/Desktop/sage-5.3.rc0/local/bin/R: No such file or directory

But that's expected, since I didn't succeed in building R with this.

[GeorgSWeber]

With OS X 10.4 on my MacIntel?, I ran with sage-5.3.rc1 into the same problem as earlier reported here (for Sage-5.2 and Sage-5.1, I can confirm the problem exists already for Sage-5.1):

​https://groups.google.com/forum/?fromgroups=#!searchin/sage-devel/5.1$205.2/sage-devel/0T4djTzK0BE/4d7dnBJq9YQJ

The fix is trivial (once one knows what the problem is), in module_list.py there is a "-lpari" option missing for the cremona.mat module (for the other cremona.xxx modules, the "-lpari" option is listed with a comment about "Cygiwn might need this" --- obviously the dynamic linker/loader on OS X 10.4 MacIntel? is comparably dumb). Once I sorted out creating a corresponding ticket, uploaded the fix there, and answered to the thread above, I'll continue with the atlas story here.

[GeorgSWeber]

Hi all, it is really puzzling, why with current Sage-5.3 on OS X 10.4 PPC (yesterday released, it's up on the servers), the cvxopt-1.1.5 spkg finds during its build the blas.dylib --- but with the changes of this ticket here, cvxopt-1.1.5 does not find it anymore (see comments #142 and #170 here).

Some facts:

• On OS X 10.4 PPC, under the directory "/usr/lib/", there are no "atlas.dylib", "blas.dylib", "cblas.dylib", "lapack.dylib" files or links (on OS X 10.4 Intel, these four entries *do* exist there, as links).
• On OS X 10.4 (both PPC and Intel), Sage historically does not build atlas nor blas (nor cblas) nor lapack, so there are not corresponding dylibs either under "$SAGE_ROOT/local/lib/".
• In $SAGE_ROOT/devel/sage-main/module_list.py, there is some magic for the configuration of the "BLAS" and "BLAS2" macros. Given the above fact(s), on OS X 10.4 PPC, I read these as resulting to use "gslblas" (!!) provided by Sage, on OS X 10.4 Intel, I read these as resulting to use the system "cblas" and "atlas".
• On OS X 10.4 Intel, these "system" atlas, blas, and cblas libraries actually all link to one and the same "blas.dylib" located under /System/Libraries/Frameworks/Accelerate?.framework/version/A/vecLIB.framwork/version/A/ (or some permutation thereof, I can't look up the exact path right now at the moment). These libraries are also located there under OS X 10.4 PPC, but there do not seem to be any links.
• The old cvxopt-1.1.4(.p1) lists "gslblas" as library to be linked to, but actually, it is unclear at least to me, which kind of "blas" library really got used in the end on OS X 10.4 --- maybe on PPC some different one than on Intel.

Is it possible that currently, the build of cvxopt-1.1.5 on OS X 10.4 PPC succeeds only during a "whole" build of Sagte-5.3, the linker somehow caching the libraries it finds in the course of the build (especially "blas.dylib" somewhen before cvxopt is built) --- and with the changes of this ticket, the build sequence changes somehow so suddenly, "blas.dylib" is not anymore in this cache (or whatever kind of mechansim with the same result)?

[fbissey]

Before this ticket netlib's blas (blas spkg) would be build on OS X if I am not mistaken. So an unoptimized blas would be present to be picked in SAGE_LOCAL/lib. cblas would indeed be provided by gslblas if nothing else.

[dimpase]

Replying to fbissey:

Before this ticket netlib's blas (blas spkg) would be build on OS X if I am not mistaken. So an unoptimized blas would be present to be picked in SAGE_LOCAL/lib. cblas would indeed be provided by gslblas if nothing else.

hmm, no, I think on OSX there is native blas/lapack (veclib) which was used (and still is used, given the right settings used to build Sage).

[vbraun]

I agree with Francois, we probably never used the ancient accelerate (OSX 10.4) framework correctly. In fact, this ticket makes it pretty clear that we did not.

[dimpase]

Replying to vbraun:

I agree with Francois, we probably never used the ancient accelerate (OSX 10.4) framework correctly. In fact, this ticket makes it pretty clear that we did not.

as far as CVXOPT is concerned, it does use MacOSX veclib (there was a OSX 10.7-related ticket which dealt with the fact that veclib changed some undefined behaviour when 10.7 was released, and CVXOPT got hit by this issue).

[vbraun]

Replying to dimpase:

as far as CVXOPT is concerned, it does use MacOSX veclib

Are you sure? I think OSX 10.4 is rather different from modern OSX accelerate. Post the output of otool -L on OSX 10.4 PPC or it didn't happen ;)

[kcrisman]

Two hundred comments!

as far as CVXOPT is concerned, it does use MacOSX veclib

Are you sure? I think OSX 10.4 is rather different from modern OSX accelerate. Post the output of otool -L on OSX 10.4 PPC or it didn't happen ;)

Tell me the exact command (otool -L what_file?) and I'll do it as soon as I get to work.

[dimpase]

Replying to kcrisman:

Two hundred comments!

as far as CVXOPT is concerned, it does use MacOSX veclib

Are you sure? I think OSX 10.4 is rather different from modern OSX accelerate. Post the output of otool -L on OSX 10.4 PPC or it didn't happen ;)

Tell me the exact command (otool -L what_file?) and I'll do it as soon as I get to work.

$ cd $SAGE_LOCAL/lib/python/site-packages/cvxopt
$ otool -L lapack.so

[kcrisman]

So who was right?

Dasher-03:~/Desktop/sage-4.8/local/lib/python/site-packages/cvxopt student$ otool -L lapack.so 
lapack.so:
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.1.12)
        /Users/student/Desktop/sage-4.8/local/lib/libgsl.0.dylib (compatibility version 17.0.0, current version 17.0.0)
        /Users/student/Desktop/sage-4.8/local/lib/libgslcblas.0.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)

This is with Sage 4.8, the last working binary I have on this box. I foolishly deleted all newer ones in the meantime. Sage 5.3 works on it, of course - in fact, I just used the app bundle yesterday - but it would take some time to start a fresh 5.3 or 5.4.beta0.

[dimpase]

Replying to kcrisman:

So who was right?

Dasher-03:~/Desktop/sage-4.8/local/lib/python/site-packages/cvxopt student$ otool -L lapack.so 
lapack.so:
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.1.12)
        /Users/student/Desktop/sage-4.8/local/lib/libgsl.0.dylib (compatibility version 17.0.0, current version 17.0.0)
        /Users/student/Desktop/sage-4.8/local/lib/libgslcblas.0.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)

looks like this particular system does not use the veclib framework. Is it Intel, or PPC?

[kcrisman]

Dasher-03:~/Desktop/sage-4.8/local/lib/python/site-packages/cvxopt student$ uname -a
Darwin Dasher-03.local 8.11.0 Darwin Kernel Version 8.11.0: Wed Oct 10 18:26:00 PDT 2007; root:xnu-792.24.17~1/RELEASE_PPC Power Macintosh powerpc

At 700 MHz and half a gig of ram, probably one of the lowest-capability machines running Sage today... but it works fine, if slow. But we're just spoiled - I'm sure Gauss or Newton would have been happy to have this :)

[GeorgSWeber]

Replying to kcrisman:

So who was right?

Dasher-03:~/Desktop/sage-4.8/local/lib/python/site-packages/cvxopt student$ otool -L lapack.so 
lapack.so:
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.1.12)
        /Users/student/Desktop/sage-4.8/local/lib/libgsl.0.dylib (compatibility version 17.0.0, current version 17.0.0)
        /Users/student/Desktop/sage-4.8/local/lib/libgslcblas.0.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)

This is with Sage 4.8, the last working binary I have on this box. I foolishly deleted all newer ones in the meantime. Sage 5.3 works on it, of course - in fact, I just used the app bundle yesterday - but it would take some time to start a fresh 5.3 or 5.4.beta0.

It is not necessary to "start a fresh 5.3 or 5.4.beta0". Instead, one can go to some Sage OS X PPC download page, ​http://boxen.math.washington.edu/home/sagemath/sage-mirror/osx/powerpc/index.html , download from there e.g. "sage-5.3-OSX-32bit-10.4-PowerMacintosh?-Darwin.dmg", and look inside --- on any OS X system/version. The result is ... surprising, to say the least:

sagebuilder$ otool -L lapack.so
lapack.so:
        /Users/buildbot/build/sage/moufang-1/moufang_binary/build/sage-5.3/local/lib/libgfortran.3.dylib (compatibility version 4.0.0, current version 4.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /Users/buildbot/build/sage/moufang-1/moufang_binary/build/sage-5.3/local/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.1.12)

The two occurences of "libgcc_s.1.dylib" are somewhat fishy, but apparently nothing new, and I'd vote for that topic being a complete different story, and not to be treated here any further. The library "libSystem.B.dylib" is the C runtime on OS X, AFAIK. But then only libfortran.dylib --- and no trace whatsoever of any *las.dylib?!?!? For comparison, on OS X 10.4 Intel, for Sage-5.3.rc1 the analoguous result is:

sagebuilder$ otool -L lapack.so
lapack.so:
        /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libLAPACK.dylib (compatibility version 1.0.0, current version 192.15.0)
        /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib (compatibility version 1.0.0, current version 192.15.0)
        /Users/Shared/sage/sage-5.3.rc1/local/lib/libgfortran.3.dylib (compatibility version 4.0.0, current version 4.0.0)
        /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /Users/Shared/sage/sage-5.3.rc1/local/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)
        /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 88.3.11)

I would conclude that it is problematic, that the OS X 10.4 PPC linker/loader builds at all the current cvxopt-1.1.5 spkg contained in Sage-5.3, producing results with such incomplete dependency information. Looking closely on the command line, it is told to use "-undefined dynamic_lookup" (which seems to save us at runtime later, however only by chance, I guess), so essentially it does what it should --- but then, why does that linker/loader under some very comparable circumstances error out with "/usr/bin/ld: can't locate file for: -lblas"???

Of the different alternatives to come to an end with this ticket here (which is about the new atlas-3.10 spkg), I like the "link idea" (see the comment 191 from Francois), i.e. the atlas spkg, generally on OS X (!), doing no more and no less than creating links in $SAGE_LOCAL/lib/ --- the same as exist in /usr/lib/, at least on OS X 10.4 Intel. More precisely, three links pointing to the "framework" blas library:

libatlas.dylib -> /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
libblas.dylib -> /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
libcblas.dylib -> /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib

and one more link for the "framework" lapack library (all link targets being absolute, not relative):

liblapack.dylib -> /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libLAPACK.dylib

A quick check shows that at least on OS X Intel, the absolute path to these framework libraries is the same for OS X 10.4, 10.5, 10.6, and 10.7 (I can't check for OS X 10.8), and even on OS X 10.7.3, there still is only one version "A", which is the "Current" version of that framework.

This change, for all OS X variants, would be a bit intrusive. But to me, it seems to bring the OS X world a bit closer to the situation on the other OSes. I offer to test a respectively updated atlas spkg on the OS X systems/versions I have access to. (In the "patches" subdir, one also might wish to delete the "edit" duplicates of two files, those with a tilde at the end of their file name.)

Thoughts?

[kcrisman]

I can't pretend to understand exactly what all the linking is doing here - it seems very much a black box. But those two absolute paths certainly exist on my box, and seem like a good alternative to trying to get ATLAS to build from scratch on OS X for now. I will naturally also volunteer to test on anything I have access to.

[dimpase]

Replying to vbraun:

Replying to dimpase:

as far as CVXOPT is concerned, it does use MacOSX veclib

Are you sure? I think OSX 10.4 is rather different from modern OSX accelerate. Post the output of otool -L on OSX 10.4 PPC or it didn't happen ;)

On 10.5 PPC it is veclib all right...

$ otool -L lapack.so 
lapack.so:
        /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libLAPACK.dylib (compatibility version 1.0.0, current version 218.0.0)

[kcrisman]

So you mean that after

if conf['Darwin?'] \
        and not os.environ.has_key('SAGE_ATLAS_ARCH'):
    print 'Skipping build of ATLAS on intel OS X'
    sys.exit(0)

in the spkg-install (well, before the exit) we should add the linking described comment:205? That would seem easy enough, though I don't know what the syntax would be (not just ln, I assume).

[vbraun]

I don't think we should be symlinking system libraries into local/lib if we can avoid it. Really cvxopt is broken, it should be linking against the accelerate blas but doesn't. In particular, we should correctly set BLAS_LIB_DIR in cvxopt on Darwin. The update spkg fixes this. I've also changed the atlas spkg back to not build by default on Darwin. Please test the combination of both spkgs on your OSX machines.