help-octave
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: SLATEC error: Chebyshev series too short for specified accuracy


From: Ben Abbott
Subject: Re: SLATEC error: Chebyshev series too short for specified accuracy
Date: Wed, 23 Feb 2011 18:48:44 -0500

On Feb 23, 2011, at 6:19 PM, Camillo wrote:

> On 23/feb/2011, at 18.09, bpabbott wrote:
> 
>> On Feb 23, 2011, at 11:13 AM, Camillo <address@hidden> wrote:
>> 
>>> On 23/feb/2011, at 02.11, Ben Abbott wrote:
>>> 
>>>> On Feb 22, 2011, at 9:32 AM, Camillo wrote:
>>>> 
>>>>> On 22/feb/2011, at 14.12, Ben Abbott wrote:
>>>>> 
>>>>>> On Feb 22, 2011, at 6:53 AM, Camillo wrote:
>>>>>> 
>>>>>>> Hello,
>>>>>>> 
>>>>>>> I am working on a Mac build of Octave 3.4.0. It's almost complete: it 
>>>>>>> passes 6908 tests, has 2 known failures, and 5 unexpected ones. All of 
>>>>>>> these five are due to the same issue, an exception inside SLATEC: 
>>>>>>> "Chebyshev series too short for specified accuracy". I don't know 
>>>>>>> Fortran and I'm not sure how to best debug the issue. The compilers I 
>>>>>>> am using are gcc and gfortran 4.2.1.
>>>>>>> 
>>>>>>> Any ideas?
>>>>>>> 
>>>>>>> Best regards,
>>>>>>> Camillo Lugaresi
>>>>>> 
>>>>>> I've not seen this error.
>>>>>> 
>>>>>> Where did you get your gfortran?
>>>>> 
>>>>> http://r.research.att.com/gfortran-42-5664.pkg
>>>>> 
>>>>>> What dependencies did you install?
>>>>> 
>>>>> All, except for gnuplot and ghostscript.
>>>>> 
>>>>>> What was the result of running /configure? (it gives a summary at the 
>>>>>> end)
>>>>>> 
>>>>>> Finally, please copy-n-paste the actual error.
>>>>> 
>>>>> I'll attach the configure summary and the output of make check. I'm 
>>>>> throwing in fntests.log as well.
>>>>> 
>>>>> Thanks,
>>>>> Camillo
>>>>> 
>>>>> <configure summary><tests><fntests.log>
>>>> 
>>>> There may be another problem, but since you're building a 64bit app you'll 
>>>> need to take care of a bug in Apple's blas.
>>>> 
>>>> The problem is discussed in the developer's copy of README.MacOS
>>>> 
>>>>    http://hg.savannah.gnu.org/hgweb/octave/file/64ae43e0e1c0/README.MacOS
>>>> 
>>>> There is a patch intended for the developer's sources at the link below 
>>>> that will allow the configure script to detect the problem and apply the 
>>>> wrapper.
>>>> 
>>>>    https://savannah.gnu.org/patch/?7468
>>> 
>>> I was aware of the problem with BLAS, and I am using a small wrapper 
>>> library for the affected routines (as suggested in 
>>> <http://developer.apple.com/hardwaredrivers/ve/errata.html#fortran_conventions>);
>>>  this is the "dotwrp" library you can see in the configure summary.
>>> 
>>> I didn't know that there were issues with LAPACK routines as well. Do you 
>>> have a test case?
>>> 
>>> CL
>>  
>> No I don't. I've only been reading along on the developer's mail-list.
>> 
>> I'll forward the link to those working on the wrapper which has been 
>> submitted to the Octave patch tracker.
>> 
>> Ben
> 
> Thank you. Should I subscribe to that mailing list as well?
> Apart from those 5 test failures, I think I put together a pretty nice Mac 
> build of Octave. Perhaps it could be useful to others.
> 
> CL

Yes, please subscribe to the maintainer's list and hopefully contribute.

        https://mailman.cae.wisc.edu/listinfo/octave-maintainers

If you're interested in documenting what you've done, the README.MacOS could be 
expanded to describe how to manually build a fully functional Octave (we'd like 
to keep the minimal dependency description that it there now).

The build instructions on the wiki could be updated to Mac OS 10.6 as well.

        http://wiki.octave.org/wiki.pl?OctaveForMac

Some low hanging fruit that would helpful, would be to test the blas-wrapper 
and confirm it works for you.

        https://savannah.gnu.org/patch/?7468

If you're not up for any of that, no problem. You're welcome to follow along or 
contribute as much as you like.

Ben




reply via email to

[Prev in Thread] Current Thread [Next in Thread]