[Octave-bug-tracker] [bug #51630] BIST test for log2 fails on Windows

[Markus Mützel]

URL:
  <http://savannah.gnu.org/bugs/?51630>

                 Summary: BIST test for log2 fails on Windows
                 Project: GNU Octave
            Submitted by: mmuetzel
            Submitted on: Mon 31 Jul 2017 05:56:07 PM UTC
                Category: Octave Function
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: Regression
                  Status: None
             Assigned to: None
         Originator Name: 
        Originator Email: 
             Open/Closed: Open
         Discussion Lock: Any
                 Release: dev
        Operating System: Microsoft Windows

    _______________________________________________________

Details:

The BIST test for log2 is failing on Windows since a few weeks (tested with hg
id e56d84a16a2e):

>>>>> processing
D:\Repositories\Octave\test\octave-2017-07-30-17-21_gcc5\octave-2017-07-30-17-21\share\octave\4.3.0+\etc\tests\libinterp\corefcn\data.cc-tst
***** test
 [f, e] = log2 ([0,-1; 2,-4; Inf,-Inf]);
 assert (f, [0,-0.5; 0.5,-0.5; Inf,-Inf]);
 assert (e(1:2,:), [0,1;2,3]);
!!!!! test failed
ASSERT errors for:  assert (f,[0, -0.5; 0.5, -0.5; Inf, -Inf])

  Location  |  Observed  |  Expected  |  Reason
   (3,1)         NaN          Inf        'NaN' mismatch
   (3,2)         NaN          -Inf       'NaN' mismatch
shared variables   scalar structure containing the fields:

    xs =

Compressed Column Sparse (rows = 1, cols = 4, nnz = 3 [75%])

  (1, 2) ->  1
  (1, 3) ->  2
  (1, 4) ->  3

***** test
 [f, e] = log2 (complex (zeros (3, 2), [0,-1; 2,-4; Inf,-Inf]));
 assert (f, complex (zeros (3, 2), [0,-0.5; 0.5,-0.5; Inf,-Inf]));
 assert (e(1:2,:), [0,1; 2,3]);
!!!!! test failed
ASSERT errors for:  assert (f,complex (zeros (3, 2), [0, -0.5; 0.5, -0.5; Inf,
-Inf]))

  Location  |  Observed  |  Expected  |  Reason
   (3,1)       NaN+NaNi      0+Infi      'NaN' mismatch
   (3,2)       NaN+NaNi      0-Infi      'NaN' mismatch
shared variables   scalar structure containing the fields:

    xs =

Compressed Column Sparse (rows = 1, cols = 4, nnz = 3 [75%])

  (1, 2) ->  1
  (1, 3) ->  2
  (1, 4) ->  3


CC'ing Rik because this might be related to the recent change to use standard
library functions:
http://hg.savannah.gnu.org/hgweb/octave/rev/4c048a2792bc

Calling "log2" with one or two output argument gives these results on
Windows:

>> log2 (Inf)
ans = Inf
>> log2 (-Inf)
ans = Inf + 4.532i
>> [a,b]=log2 (Inf)
a = NaN
b = -1
>> [a,b]=log2 (-Inf)
a = NaN
b = -1


The same on Linux (correct):

>> log2 (Inf)
ans = Inf
>> log2 (-Inf)
ans = Inf + 4.532i
>> [a,b]=log2 (Inf)
a = Inf
b = 0
>> [a,b]=log2 (-Inf)
a = -Inf
b = 0


The results when calling with one output argument are correct.
The results with two output arguments (effectively calling frexp) fail on
Windows.
With 4.2.1 (and a dev build from March I found on my disk, hg id
a470f7fee1cf), the results were still correct on Windows.
Should we return to use gnulib for frexp?

According to [1], behaviour for +/-Inf is implementation specific. Is
mingw/gcc/glibc/Windows not IEC 60559 compliant? 

[1] http://en.cppreference.com/w/cpp/numeric/math/frexp




    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?51630>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/