1 |
EL Wed, 1 Jun 2011 06:57:03 +1200 |
2 |
Francois Bissey <fbissey@××××××××××××.nz> escrigué: |
3 |
|
4 |
> > Hi, |
5 |
> > |
6 |
> > While updating sage and python to 4.7 and 2.7, I hit a variation of |
7 |
> > an old bug (see http://bugs.gentoo.org/195619): scipy compiles with |
8 |
> > lapack-reference but is unusable. Running in python |
9 |
> > |
10 |
> > import scipy.optimize |
11 |
> > |
12 |
> > fails with |
13 |
> > |
14 |
> > [...] |
15 |
> > from scipy.linalg import clapack |
16 |
> > ImportError: /usr/lib64/python2.7/site-packages/scipy/linalg/clapack.so: |
17 |
> > undefined symbol: clapack_sgesv |
18 |
> > |
19 |
> > I tried rebuilding numpy and scipy with lapack provided (eselect) by |
20 |
> > lapack-reference, blas provided by blas-reference, cblas provided by |
21 |
> > gsl and with/without clapack with the same result. |
22 |
> > |
23 |
> > Then I tried rebuilding numpy and scipy with lapack, blas and cblas |
24 |
> > provided (eselect) by atlas and scipy worked flawlessly. |
25 |
> > |
26 |
> > Should scipy depend on a specific lapack provider instead of a |
27 |
> > virtual? Enable it at compile time? Is it an unfortunate |
28 |
> > combination what fails? |
29 |
> > |
30 |
> > I can provide build logs if you need them. |
31 |
> > |
32 |
> Hi Marc, |
33 |
> |
34 |
> Just one thing that I can think of on the top of my head. We recently |
35 |
> made scipy depend on arpack and superlu. They also have dependencies |
36 |
> on blas/lapack if you didn't rebuild them as well they may cause |
37 |
> trouble. |
38 |
> |
39 |
> I had a similar problem recently but eselecting everything properly |
40 |
> worked for me. A last piece of info you could provide: are you using |
41 |
> the main tree implementation or bicatali's overlay implementation? |
42 |
> |
43 |
> Francois |
44 |
> |
45 |
|
46 |
Hi Francois, |
47 |
|
48 |
I tried rebuilding arpack and superlu without luck. Since you mentioned |
49 |
that eselect fixed your problems, I also tried installing |
50 |
cblas-reference and using it instead of gsl, also without luck. |
51 |
|
52 |
I'm using the main tree implementation. |
53 |
|
54 |
Best, |
55 |
Marc. |