[MITgcm-support] compiling problems

[Andreas Klocker]

Hi Oliver,
Yes, this helped, and the model compiled! Thanks a lot!!
I use a different opt file which has been put together by some people specifically for the machine we're using, but it didn't have the -mcmodel flag.
Andreas

________________________________
From: MITgcm-support <mitgcm-support-bounces at mitgcm.org> on behalf of Oliver Jahn <jahn at mit.edu>
Sent: Tuesday, 20 October 2020 12:54 AM
To: mitgcm-support at mitgcm.org <mitgcm-support at mitgcm.org>
Subject: Re: [MITgcm-support] compiling problems

Hi Andreas,

I agree with Martin that this points to an executable size issue. With
ifort, this can usually be resolved with "-mcmodel=medium". This seems
to have helped others with the same error:

https://bb.cgd.ucar.edu/cesm/threads/failed-to-convert-gotpcrel-relocation-relink-with-no-relax.4494/<https://bb.cgd.ucar.edu/cesm/threads/failed-to-convert-gotpcrel-relocation-relink-with-no-relax.4494>

The linux_amd64_ifort+impi optfile has this already, but maybe you are
using a different one?

Cheers,
Oliver


On 2020-10-19 02:18, Andreas Klocker wrote:
> Hi Martin,
>
> That's what happens if I add this flag:
>
> mpif90 -WB -convert big_endian -assume byterecl -g -traceback -fPIC
> --no-relax -O2 -align -xHost -ip  -c write_state.f
> ifort: command line warning #10006: ignoring unknown option '-fno-relax'
> ifort: command line warning #10006: ignoring unknown option '-fno-relax'
> ifort: command line warning #10006: ignoring unknown option '-fno-relax'
> ifort: command line warning #10006: ignoring unknown option '-fno-relax'
> ifort: command line warning #10006: ignoring unknown option '-fno-relax'
> ifort: command line warning #10006: ignoring unknown option '-fno-relax'
>
> So no luck here..
> A>
>
> ------------------------------------------------------------------------
> *From:* MITgcm-support <mitgcm-support-bounces at mitgcm.org> on behalf of
> Martin Losch <Martin.Losch at awi.de>
> *Sent:* Monday, 19 October 2020 5:05 PM
> *To:* MITgcm Support <mitgcm-support at mitgcm.org>
> *Subject:* Re: [MITgcm-support] compiling problems
>
> Hi Andreas,
>
> I have no idea, what compiler you are using, but “relocation” would make
> me think that the compiler/loader cannot allocate the size of the
> executable/fields within the executable (that’s an error that I often
> get with gfortran/ifort). Since reducing the tiles makes the problem go
> away, I would guess that size is the problem and you may need additional
> compiler flags (‘—no-relax’?).
>
> Martin
>
>> On 19. Oct 2020, at 07:09, Andreas Klocker
> <andreas.klocker at utas.edu.au> wrote:
>>
>> Hi everyone,
>>
>> During compiling MITgcm I currently get this error which makes no
> sense to me:
>>
>> /bin/ld: failed to convert GOTPCREL relocation; relink with --no-relax
>> make[1]: *** [Makefile:1226: mitgcmuv] Error 1
>> make[1]: Leaving directory
> '/scratch/ds0/axk157/mitgcm/codebase/MITgcm/exp/channel_jet_global/build'
>> make: *** [Makefile:1222: fwd_exe_target] Error 2
>>
>> Nevertheless, if I make tiles smaller in SIZE.h compilation works
> fine. So, this problem seems to depend on the exact grid configuration.
>>
>> Does anyone know what's going on here?
>>
>> cheers,
>> Andreas
>>
>>
>> University of Tasmania Electronic Communications Policy (December, 2014).
>> This email is confidential, and is for the intended recipient only.
> Access, disclosure, copying, distribution, or reliance on any of it by
> anyone outside the intended recipient organisation is prohibited and may
> be a criminal offence. Please delete if obtained in error and email
> confirmation to the sender. The views expressed in this email are not
> necessarily the views of the University of Tasmania, unless clearly
> intended otherwise.
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support<http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support<http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support<http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support>
>
_______________________________________________
MITgcm-support mailing list
MITgcm-support at mitgcm.org
http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support<http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20201019/23b827ee/attachment.html>