octave-maintainers
[Top][All Lists]
Advanced

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

Re: Octave 4.0.1 release candidate 1 available for ftp


From: Juan Pablo Carbajal
Subject: Re: Octave 4.0.1 release candidate 1 available for ftp
Date: Sun, 13 Dec 2015 22:35:58 +0100

On Sun, Dec 13, 2015 at 9:25 PM, Ben Abbott <address@hidden> wrote:
>
>> On Dec 13, 2015, at 12:45 PM, Juan Pablo Carbajal <address@hidden> wrote:
>>
>> On Sat, Dec 12, 2015 at 8:46 PM, Mike Miller <address@hidden> wrote:
>>> On Fri, Dec 11, 2015 at 16:33:17 -0500, JohnD wrote:
>>>> Using the downloaded windows installer on my win7 machine, only 1 failure
>>>> for me:
>>>>  PASS     12937
>>>>  FAIL         1
>>>>  XFAIL       16
>>>>  SKIPPED     54
>>>>
>>>> ***** testif HAVE_ARPACK, HAVE_UMFPACK
>>>> [u2,s2,v2,flag] = svds (A,k,0,opts);
>>>> s2 = diag (s2);
>>>> assert (flag, !1);
>>>> tol = 10 * eps() * norm(s2, 1);
>>>> assert (s2, s(k:-1:1), tol);
>>>>
>>>> !!!!! test failed
>>>> ASSERT errors for:  assert (s2,s (k:-1:1),tol)
>>>>
>>>>  Location  |  Observed  |  Expected  |  Reason
>>>>     .          O(6x1)       E(7x1)      Dimensions don't match
>>>
>>> I have the exact same test failure on Debian with OpenBLAS 0.2.15, and
>>> with Octave 3.8.2, 4.0.0, and default branch, so I think this change
>>> comes from an OpenBLAS update.
>>>
>>> Running with LD_LIBRARY_PATH=/usr/lib/lapack:/usr/lib/libblas makes the
>>> test pass.
>>>
>>> --
>>> mike
>>>
>>
>> In Ubuntu 15.10 as always  builds just fine and runs with proprietary drivers
>> Bur checks fails
>>
>> libinterp/dldfcn/__osmesa_print__.cc-tst ....................panic:
>> Segmentation fault -- stopping myself...
>> octave exited with signal 11
>>
>> It is not solved with
>>
>> export LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libGL.so
>>
>> that worked before, but I am reading the threads to see if there is
>> any solution. also if I preload that any plot in the gui crashes
>> octave.
>
> This is baseless speculation, but the attached patch is needed to build with 
> osmesa on MacOS X (it does not yet result in a functioning osmesa). Any 
> chance this impacts the behavior your seeing?
>
> Ben
>
Thanks Ben, after applying the patch and re-compiling it still doesn't work



reply via email to

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