Re: Compiler warnings from deprecated Array<> constructor calls

[Philip Nienhuis]

Rik-4 wrote
> On 11/22/2014 01:04 AM, 

> octave-maintainers-request@

>  wrote:
>> Subject:
>> Re: Compiler warnings from deprecated Array<> constructor calls
>> From:
>> Philip Nienhuis &lt;

> address@hidden

> &gt;
>> Date:
>> 11/22/2014 12:57 AM
>>
>> To:
>> 

> octave-maintainers@

>>
>> List-Post:
>> &lt;mailto:

> octave-maintainers@

> &gt;
>> Content-Transfer-Encoding:
>> 7bit
>> Precedence:
>> list
>> MIME-Version:
>> 1.0
>> References:
>> &lt;address@hidden&gt;
>> <

> address@hidden

>>
>> <

> address@hidden

>>
>> <

> address@hidden

>>
>> In-Reply-To:
>> <

> address@hidden

>>
>> Message-ID:
>> <

> address@hidden

>>
>> Content-Type:
>> text/plain; charset=us-ascii
>> Message:
>> 9
>>
>>
>> Philip Nienhuis wrote
>>> > 
>>> > siko1056 wrote
>>>> >> 
>>>> >> Philip Nienhuis wrote
>>>>> >>> 
>>>>> >>> Rik-4 wrote
>>>>>> >>>> I just checked in Philipp Kutin's change which fixes the errors
>>>>>> when
>>>>>> >>>> '--without-qrupdate' is used during configure and the alternate
>>>>>> >>>> routines
>>>>>> >>>> are used in Octave
>>>>>> >>>> (http://hg.savannah.gnu.org/hgweb/octave/rev/d47e50953abc).
>>>>>> >>>> 
>>>>>> >>>> It might be worth updating to d47e50953abc as that will also get
>>>>>> you
>>>>>> >>>> some
>>>>>> >>>> changes that Kai made in cset 19329:1f4455ff2329 that also fix
>>>>>> problems
>>>>>> >>>> related to the deprecation of these old methods.
>>>>> >>> @Rik: was this addressing my other post 
>>>>> >>> "no matching function for call to
>>>>> 'Array&lt;std::complex&lt;double&gt;
>>>>>> >>> >::Array(octave_idx_type&, int)'"
>>>>> >>> ?
>>>>> >>> 
>>>>> >>> I just updated to "3a2a0b72b91b tip @  Swap row vector........" &
>>>>> tried
>>>>> >>> again - no luck. The exact same errors relating to complex double
>>>>> >>> appear.
>>>>> >>> 
>>>>> >>> FYI, my system is stock Mageia-4 Linux 64b with all updates. In
>>>>> fact
>>>>> >>> I've upgraded my desktop's HW and SW the last weeks and today I
>>>>> finally
>>>>> >>> got time for a first attempt to build Octave on Linux.
>>>>> >>> 
>>>>> >>> Philip
>>>> >> I replaced all constructors in this changeset
>>>> >> http://hg.savannah.gnu.org/hgweb/octave/rev/fcd9e1198847. Does it
>>>> work
>>>> >> for you?
>>> > Thanks, I'll have another try and report back soon.
>> No luck :-(
>>
>> The error messages are attached; the errors occur in
>> liboctave/array/dNDArray.cc.
>>
>> I can't rule out that something is wrong in my box; but it is intriguing
>> that up till cs 346b20b2c2ff  ("gendocprop...." it all worked fine.
>>
>> Philip
>>
>> ComplexNDArray.err
>> &lt;http://octave.1599824.n4.nabble.com/file/n4667426/ComplexNDArray.err&gt; 
>>  
>>
> 
> Philip,
> 
> It's nothing wrong with your box, in fact you're testing is exposing some
> good issues that we would have missed.  The issue in bug #43646 is that
> you
> were compiling without the qrupdate library and Octave was falling back on

In fact that was Philipp Kutin :-)   (we do not even share the exact same
name, but I think it's easy to get confused anyway; no problems!  :-) )


> routines that hardly ever get exercised.  I looked through the errors you
> posted and the issue this time is that you are not using the FFTW library,
> and again are falling back on routines that hardly ever get used.  If I

That suggests that we may have to switch off other libs too, to find out if
any other rarely used code parts got hit by Carnë's changeset.
Hmmmm... 

Anyway, your cset works:

Summary:

  PASS     12623
  FAIL         8
  XFAIL        9
  SKIPPED     48

(6 fails are due to local problems with graphicsmagick, I had to uncomment
the tests in imread.m / imwrite.m, and in hgsave.m and copyobj.m)

Thank you very much for the fix!

Philip



--
View this message in context: 
http://octave.1599824.n4.nabble.com/Re-Compiler-warnings-from-deprecated-Array-constructor-calls-tp4667413p4667439.html
Sent from the Octave - Maintainers mailing list archive at Nabble.com.