octave-bug-tracker
[Top][All Lists]
Advanced

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

[Octave-bug-tracker] [bug #35596] Some 'make check' tests fail


From: Lluís Batlle
Subject: [Octave-bug-tracker] [bug #35596] Some 'make check' tests fail
Date: Thu, 23 Feb 2012 11:33:17 +0000
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:8.0.1) Gecko/20111124 Firefox/8.0.1

URL:
  <http://savannah.gnu.org/bugs/?35596>

                 Summary: Some 'make check' tests fail
                 Project: GNU Octave
            Submitted by: viric
            Submitted on: dj 23 feb 2012 11:33:15 GMT
                Category: None
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: Incorrect Result
                  Status: None
             Assigned to: None
         Originator Name: 
        Originator Email: 
             Open/Closed: Open
         Discussion Lock: Any
                 Release: 3.6.0
        Operating System: GNU/Linux

    _______________________________________________________

Details:

I'm trying octave built with liblapack 3.4.0 on GNU/Linux x86_64 and qrupdate
1.1.2. qrupdate, lapack, and atlas tests work all fine. I get the same outcome
with octave 3.4.3 and 3.6.0.

Some tests of the octave "make check" fail for me, related to lapack. I don't
list them all here.

I pick one of the tests as an example, as at least someone got it working fine
in his system, but not me. I run this test:

A = [       0.091364   0.613038   0.999083
       0.594638   0.425302   0.603537
       0.383594   0.291238   0.085574
       0.265712   0.268003   0.238409
       0.669966   0.743851   0.445057
];
 AA = A.';
 i = 2; j = 4; p = [1:i-1, shift(i:j,-1), j+1:5];
 [Q,R] = qr(AA);
 [Q,R] = qrshift(Q,R,i,j);
 assert(norm(vec(Q'*Q - eye(3)),Inf) < 1e1*eps)
 assert(norm(vec(triu(R)-R),Inf) == 0)
 assert(norm(vec(Q*R - AA(:,p)),Inf) < norm(AA)*1e1*eps)


This fails for me this way:

 ** On entry to DORGLQ parameter number  5 had an illegal value
error: norm: exception encountered in Fortran subroutine dgesvd_


Building lapack and octave with debug symobls, I could check with gdb where
all this happens. Backtrace:

#0 dorglq (m=5, n=5, k=3, a=..., lda=1, tau=..., work=..., lwork=-1, info=0)
at lapack-3.4.0/SRC/dorglq.f:128
#1 dgesvd (jobu=..., jobvt=..., m=3, n=5, a=..., lda=3, s=..., u=..., ldu=3,
vt=..., ldvt=1, work=..., lwork=-1, info=0, _jobu=1, _jobvt=1) at
lapack-3.4.0/SRC/dgesvd.f:480
#2 SVD::init (this=0x7fffffffbf00, a=..., svd_type=SVD::sigma_only,
svd_driver=SVD::GESVD) at dbleSVD.cc:156
#3 SVD::SVD (this=0x7fffffffbf00, a=..., svd_type=SVD::sigma_only,
svd_driver=SVD::GESVD) at dbleSVD.h:56


jwe at #octave claimed that the test works for his octave.

Can someone reproduce this?




    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?35596>

_______________________________________________
  Missatge enviat a través de/per Savannah
  http://savannah.gnu.org/




reply via email to

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