[Scipy-tickets] [SciPy] #1737: OS X Mountain Lion test failures

SciPy Trac scipy-tickets@scipy....
Thu Oct 4 11:53:03 CDT 2012


#1737: OS X Mountain Lion test failures
--------------------------+-------------------------------------------------
 Reporter:  aleach        |       Owner:  cdavid     
     Type:  defect        |      Status:  new        
 Priority:  normal        |   Milestone:  Unscheduled
Component:  Build issues  |     Version:  0.10.0     
 Keywords:  OSX, ATLAS    |  
--------------------------+-------------------------------------------------

Comment(by aleach):

 No worries! Seen the PR; ta for the acknowledgement! Glad to help :)

 Whilst my Atlas Lapack is down, I thought I'd try again with the
 Accelerate framework. I tried a couple times, but get pretty much all the
 errors as reported in [http://projects.scipy.org/scipy/ticket/1729 ticket
 1729]'s report. I built scipy once using numpy's default flags, and once
 with FFLAGS="-ff2c -fno-second-underscore -fPIC". Under both conditions,
 the Accelerate framework causes loads of arpack errors, and also some of
 those ValueErrors and IndexErrrors in my first log, which you said came
 from numpy.

 Test summary with default fcompiler=gnu95 flags:-

 {{{
 Ran 5535 tests in 66.337s
 FAILED (KNOWNFAIL=13, SKIP=42, errors=12, failures=82)

 }}}

 Test summary when FFLAGS="-ff2c -fno second-underscore -fPIC" and
 fcompiler=gnu95:-

 {{{
 Ran 5535 tests in 66.018s
 FAILED (KNOWNFAIL=13, SKIP=42, errors=11, failures=73)

 }}}

 If you're interested in seeing the differences between the two builds test
 results, I'll upload a diff of the exception messages.

 I'd prefer to work with and report on build with the full ATLAS though.
 I'm still getting grief with the tuned lapack archive, which is hanging in
 slartg, but a manually built lapack seems to work fine. I'll see how that
 goes and will report back shortly if I get any other problems wrong with
 scipy.

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/1737#comment:6>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list