Re: [Pfunit-support] gfortran 5.2 ICE

[Florian O. <flo...@gm...>]

Tom,

if that is relatively painless for you, that would be great. Really, I'm
just concerned with the longevity of the version that works for me right
now on the development branch. I just need a place to download that from
whenever I build my project (which needs to be built on several different
machines).

Thanks!
florian

On 1 October 2015 at 16:34, Tom Clune <Tho...@na...> wrote:

> Florian,
>
> No worries - glad we resolved the issue.
>
> With regard to merging - I’ve been a bit distracted.  I was hoping to
> capture PGI and Cray support in the next release, as those seemed
> imminent.  PGI probably is, but progress by Cray in fixing their issues has
> been silent for a while.
>
> If it is crucial, I’m happy to do a release to master for the latest
> development branch.  Let me know.
>
> Cheers,
>
> - Tom
>
>
> On Oct 1, 2015, at 10:31 AM, Florian Oswald <flo...@gm...>
> wrote:
>
> Hi Tom,
>
> yes you are right - a clean build resolved this issue. I apologise if I
> was a bit too quick with hitting the mailing list with that. Everything
> runs perfect now.
>
> Just asking: Do you plan to merge that development branch that I'm running
> on right now into your master at some point, or what's the state of affairs
> with respect to the gfortran 5.2. issue?
>
> cheers
> florian
>
> On 1 October 2015 at 16:23, Tom Clune <Tho...@na...> wrote:
>
>> Hi Florian,
>>
>> OK -  our regression tests now include 5.2 on the development branch and
>> did not report any errors.
>>
>> I then tried a manual build to see if I got the message you reported and
>> the scripts ignored simply because the return status was 0.
>> Unfortunately, I don’t see that message.
>>
>> Is it possible that you had some detritus from previous attempts (either
>> the other branch or another compiler)?     If so, the tests should
>> certainly
>> be improved not to do this.  (Tests that involve the file system are a
>> pain
>> in general.)   Please let me know if this persists with a fresh checkout
>> and build,
>> and if it does, provide more details about your OS.
>>
>> Cheers,
>>
>> - Tom
>>
>>
>>
>> On Sep 30, 2015, at 8:21 AM, Florian Oswald <flo...@gm...>
>> wrote:
>>
>> Hi Tom,
>>
>> brilliant, thanks a lot.
>>
>> Florian
>>
>> On 30 September 2015 at 13:53, Tom Clune <Tho...@na...> wrote:
>>
>>> Florian,
>>>
>>> Thanks for the report.
>>>
>>> I don’t have any immediate ideas, but there was a bit of activity in the
>>> XML printer layer in the last month or two.   Since we’d not added 5.2 to
>>> our regression test scripts, we had not noticed this on our end.  I’ve
>>> asked a developer to modify the regression scripts and we’ll investigate.
>>>
>>
>>> Cheers,
>>>
>>> - Tom
>>>
>>>
>>>
>>> On Sep 29, 2015, at 4:09 AM, Florian Oswald <flo...@gm...>
>>> wrote:
>>>
>>> Hi there,
>>>
>>> I just wanted to report that this bug:
>>> http://sourceforge.net/p/pfunit/bugs/47/
>>>
>>> persists. I tried with the recommended strategy of using the development
>>> branch, and the tests build and pass. However, they error out with
>>>
>>>
>>> ................................................................................
>>>
>>> ................................................................................
>>> .....................sh: tests/test_xmlprinter_output.tmp: No such file
>>> or directory
>>> At line 124 of file
>>> /Users/florian.oswald/git/pfunit/tests/Test_XmlPrinter.F90 (unit =
>>> 538976288, file = 'fort.538976288')
>>> Fortran runtime error: End of file
>>>
>>> any ideas where that is coming from?
>>>
>>> thanks!
>>> florian
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> _______________________________________________
>>> Pfunit-support mailing list
>>> Pfu...@li...
>>> https://lists.sourceforge.net/lists/listinfo/pfunit-support
>>>
>>>
>>> *Thomas Clune, Ph. D.  <Tho...@na...
>>> <Tho...@na...>>Head ASTG,Code 606NASA GSFC MS 610.8 B33-C128*
>>> *Greenbelt, MD 20771*
>>> *301-286-4635*
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>
>> *Thomas Clune, Ph. D.  <Tho...@na...
>> <Tho...@na...>>Head ASTG,Code 606NASA GSFC MS 610.8 B33-C128*
>> *Greenbelt, MD 20771*
>> *301-286-4635*
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
> *Thomas Clune, Ph. D.  <Tho...@na...
> <Tho...@na...>>Head ASTG,Code 606NASA GSFC MS 610.8 B33-C128*
> *Greenbelt, MD 20771*
> *301-286-4635*
>
>
>
>
>
>
>
>
>
>
>