run riverplume case firstly

Bug reports, work arounds and fixes

Moderators: arango, robertson

Post Reply
Message
Author
okami_zxh

run riverplume case firstly

#1 Unread post by okami_zxh »

Hi
I used build.bash to compile ROMS , I got oceanM, then ran riverplume case. But I got these errors, I son't know what cased this kind of errors.

forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
oceanM 00000000004E0FE5 Unknown Unknown Unknown
oceanM 00000000004D5A03 Unknown Unknown Unknown
oceanM 00000000004B96EB Unknown Unknown Unknown
oceanM 00000000004241EC Unknown Unknown Unknown
oceanM 000000000042400B Unknown Unknown Unknown
oceanM 0000000000423E5C Unknown Unknown Unknown
libc.so.6 0000003083A1D994 Unknown Unknown Unknown
oceanM 0000000000423D69 Unknown Unknown Unknown
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
oceanM 00000000004E0FE5 Unknown Unknown Unknown
oceanM 00000000004D5A03 Unknown Unknown Unknown
oceanM 00000000004B96EB Unknown Unknown Unknown
oceanM 00000000004241EC Unknown Unknown Unknown
oceanM 000000000042400B Unknown Unknown Unknown
oceanM 0000000000423E5C Unknown Unknown Unknown
libc.so.6 0000003083A1D994 Unknown Unknown Unknown
oceanM 0000000000423D69 Unknown Unknown Unknown
--------------------------------------------------------------------------
mpirun has exited due to process rank 1 with PID 25228 on
node compute-0-3.local exiting without calling "finalize". This may
have caused other processes in the application to be
terminated by signals sent by mpirun (as reported here).

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: run riverplume case firstly

#2 Unread post by kate »

You had better recompile with USE_DEBUG and see if it gives you some useful line numbers.

okami_zxh

Re: run riverplume case firstly

#3 Unread post by okami_zxh »

Hi Kate,
I used oceanG then it works. What is the reason?
There is another question. I used makefile to generate excutable file. But when I used build.bash to generate oceanG, it failed. The error is as below:
cd /home/zhou/roms/MyTest/upwelling/Build; /usr/mpi/intel/openmpi-1.3.2/bin/mpif90 -c -heap-arrays -fp-model precise -g -check bounds -traceback -check uninit -warn interfaces,nouncalled -gen-interfaces -free biology.f90
biology.f90(864): error #6404: This name does not have a type, and must have an explicit type. [R_DATE]
CALL caldate (r_date, tdays(ng), year, yday, month, iday, &
--------------------------^
compilation aborted for biology.f90 (code 1)
make: *** [/home/zhou/roms/MyTest/upwelling/Build/biology.o] Error 1

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: run riverplume case firstly

#4 Unread post by kate »

Your compile-time error shouldn't depend on using the makefile vs. build.bash. Remove the old build directory and start fresh.

I have occasionally seen code which fails when optimized, runs in debug mode. It could be that you have a compiler bug, but to be sure, make sure everything is with a fresh build.

Hernan changed the clock code recently, so now we need to know exactly which version you are using, also which ecosystem model.

okami_zxh

Re: run riverplume case firstly

#5 Unread post by okami_zxh »

Hi Kate,
I refreshed the Build directory and compiled in a debug mode, the error of compiling still occurred, The version of ROMS is 3.7, and the error is as blow:

cd /home/zhou/roms/MyTest/upwelling/Build; /usr/mpi/intel/openmpi-1.3.2/bin/mpif90 -c -heap-arrays -fp-model precise -g -check bounds -traceback -check uninit -warn interfaces,nouncalled -gen-interfaces -free biology.f90
biology.f90(864): error #6404: This name does not have a type, and must have an explicit type. [R_DATE]
CALL caldate (r_date, tdays(ng), year, yday, month, iday, &
--------------------------^
compilation aborted for biology.f90 (code 1)
make: *** [/home/zhou/roms/MyTest/upwelling/Build/biology.o] Error 1

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: run riverplume case firstly

#6 Unread post by kate »

Again I need to ask - exactly what version are we talking about and which NPZ model? In the latest trunk code, a grep for r_date brings up nothing. Can you update your code?

In my Biology directory, there are matches to r_date from bestnpz.h and goanpz.h. I haven't gotten updates on those models in many a moon.

okami_zxh

Re: run riverplume case firstly

#7 Unread post by okami_zxh »

Hi Kate
I update the latest version, at version 849. Where is NPZ model? And I didn't find the biology directory under roms.

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: run riverplume case firstly

#8 Unread post by kate »

You are failing to compile biology.f90. This means you have turned on one of the NPZ-type biology models in ROMS/Nonlinear/Biology. Did your problem go away when you updated?

okami_zxh

Re: run riverplume case firstly

#9 Unread post by okami_zxh »

Hi Kate,
After I updated, the problem is still here.

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: run riverplume case firstly

#10 Unread post by kate »

You need to learn to take a little more control - what test problem, what ecosystem model, what the heck are you planning to do? Is this a problem out of the box or did you change something?

okami_zxh

Re: run riverplume case firstly

#11 Unread post by okami_zxh »

Hi Kate,
I am just testing the riverplume/upwelling case. Actually I didn't change anything. In the build.bash, I just changed the routine of directory. For the riverplume case, it works. But for the upwelling case, this error still occured. I also cheched the directory ROMS/Nonlinear/Biology, there are no bestnpz.h and goanpz.h files.

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: run riverplume case firstly

#12 Unread post by kate »

Looking at Hernan's upwelling.h online, there is no ecosystem model defined although there are tests:

Code: Select all

54	#if defined BIO_FENNEL  || defined ECOSIM || \
55	    defined NPZD_POWELL || defined NEMURO
I can compile it just fine and biology.f90 contains the start and end of the biology module, nothing else.

Post Reply