[MITgcm-devel] compiling error : for_diags_intel.c:(.text+0xa4c): relocation truncated to fit

[Matthew Mazloff]

Hi An

Welcome to the support board! In the future I encourage you to use the search feature as this error has been discussed before  :o)


Yes Chen, it just means your compiler thinks your executable is getting too big

add
-mcmodel=medium

Matt



> On Jun 20, 2016, at 9:27 PM, An Nguyen <antnguyen13 at gmail.com> wrote:
> 
> Hello,
> 
> In the past couple of days I and a grad student here at UT-Austin encountered on two different systems (I on NAS machine and Chen on the UT-austin machine) the following error related to "relocation truncated to fit" .  Has anyone encountered the same issue?  And could you help us figuring out where this error might come from?
> 
> Attached below are mine and Chen's error message, both encountered during "make adall" or "make"
> 
> Thanks,
> An
> 
>> 1. issue related to truncation: R_X86_64_PC32
>> This occurs during make adall on pleiades:
>> 
>>> make adall
>>> ...
>>> ad_taf_output.o: In function `cal_intdays_':
>>> ad_taf_output.f:(.text+0x1a): relocation truncated to fit: R_X86_64_PC32 against symbol `calendar_i_' defined in COMMON section in ad_taf_output.o
>>> ad_taf_output.o: In function `cal_subdates_':
>>> ad_taf_output.o: In function `cal_intmonths_':
>>> ad_taf_output.o: In function `cal_intyears_':
>>> ad_taf_output.o: In function `cal_printdate_':
>>> ad_taf_output.f:(.text+0x6a5): relocation truncated to fit: R_X86_64_PC32 against symbol `eeparams_i_' defined in COMMON section in ad_taf_output.o
>>> ad_taf_output.o: In function `cal_timestamp_':
>>> ad_taf_output.o: In function `calc_grad_phi_fv_.R':
>>> ad_taf_output.f:(.text+0xe00): relocation truncated to fit: R_X86_64_PC32 against symbol `surf_fixed_' defined in COMMON section in ad_taf_output.o
>>> ad_taf_output.f:(.text+0x1bd5): additional relocation overflows omitted from the output
>>> make[1]: *** [mitgcmuv_ad] Error 1
>>> make[1]: Leaving directory `/nobackupp2/atnguye4/MITgcm_c65q/mysetups/aste_270x450x180/build_c65q_bp'
>>> make: *** [ad_exe_target] Error 2
> 
>> Begin forwarded message:
>> 
>> From: Chen Chen <chen636489 at utexas.edu>
>> Subject: compiling error : for_diags_intel.c:(.text+0xa4c): relocation truncated to fit
>> Date: June 20, 2016 at 5:13:41 PM EDT
>> To: An T Nguyen <atnguyen at ices.utexas.edu>
>> 
>> An,
>> 
>> I ran into an compiling error.
>> 
>> /opt/apps/intel/15/composer_xe_2015.2.164/compiler/lib/intel64/libifcore.a(for_diags_intel.o): In function `for__io_return':
>> for_diags_intel.c:(.text+0xa4c): relocation truncated to fit: R_X86_64_PC32 against symbol `message_catalog' defined in COMMON section in /opt/apps/intel/15/composer_xe_2015.2.164/compiler/lib/intel64/libifcore.a(for_diags_intel.o)
>> 
>> (1) gfd+exch2(without blank tiles) run successfully on stampede, with SIZE.h_Nr=50.
>> (2) gfd+exch2(with blank tiles) run successfully on stampede, with SIZE.h_Nr=50.
>> (3) gfd+exch2(with blank tiles) failed, with SIZE.h_Nr=106.
>>      ==> error: relocation truncated to fit
>> (4) gfd+exch2(with blank tiles) + diagnostics + ggl90 + cal 
>> 
>>      Is it because of 2GB static limitations during compiling? Should I add -mcmodel=large -shared_intel during linking, and how? 
>> 
>> PS: I can drop by your office any time today if you feel inconvenient to reply.
>> 
>> 
>> Bests,
>> ==============
>> Chen CHEN
>> Graduate Student in EM&ICES
>> POB 3.402J
>> The University of Texas at Austin
>> 
>> Phone: +1 512-968-2126
>> Email : chen636489 at utexas.edu
>>            chen6366489 at gmail.com
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel