Segmentation fault error in int2lm – in #9: CCLM

in #9: CCLM

<p> Dear all, </p> <p> recently, I want to set up simulation with <span class="caps"> COSMO </span> - <span class="caps"> CLM </span> in <span class="caps"> CORDEX </span> -EU domain. The driven data is <span class="caps"> ERAI </span> nterim. I use chain to submit jobs. Then, the std out is correct with no error. But the std err always gets error as following. I don’t how to fix this problem. Is there special parameter should to switch on? I attach all the files. </p> <p> —- <br/> forrtl: severe (174): <span class="caps"> SIGSEGV </span> , segmentation fault occurred <br/> Image PC Routine Line Source <br/> int2lm.exe 0000000000F918F5 Unknown Unknown Unknown <br/> int2lm.exe 0000000000F8F6B7 Unknown Unknown Unknown <br/> int2lm.exe 0000000000F43D14 Unknown Unknown Unknown </p> <p> ….. </p> <p> <span class="caps"> MPT </span> : #11 0×0000000000d1b0e4 in src_read_coarse_grid::scatter_data (procarrays=…, <br/> MPT: index=12, iloc=28, irank=3, ilev=58, leof=.FALSE., lrequired=4294967295, <br/> MPT: istat=0) <br/> MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/src_read_coarse_grid.f90:4195 <br/> MPT: #12 0×0000000000cc5771 in src_read_coarse_grid::org_read_coarse_grid (nnow=0, <br/> MPT: lfirst=4294967295, ydate=…, .tmp.YDATE.len_V$f9f=14) <br/> MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/src_read_coarse_grid.f90:1773 <br/> MPT: #13 0×000000000075dab4 in int2lm_org () <br/> MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/int2lm_org.f90:265 <br/> MPT: #14 0×000000000040cc1e in main () <br/> MPT: #15 0×00002aaab09a8c36 in __libc_start_main () from /lib64/libc.so.6 <br/> MPT: #16 0×000000000040cb29 in _start () at ../sysdeps/x86_64/elf/start.S:113 <br/> MPT: (gdb) A debugging session is active. </p> <p> —— </p> <p> Greetings, <br/> Bo </p>

  @bohuang in #d06d3e6

<p> Dear all, </p> <p> recently, I want to set up simulation with <span class="caps"> COSMO </span> - <span class="caps"> CLM </span> in <span class="caps"> CORDEX </span> -EU domain. The driven data is <span class="caps"> ERAI </span> nterim. I use chain to submit jobs. Then, the std out is correct with no error. But the std err always gets error as following. I don’t how to fix this problem. Is there special parameter should to switch on? I attach all the files. </p> <p> —- <br/> forrtl: severe (174): <span class="caps"> SIGSEGV </span> , segmentation fault occurred <br/> Image PC Routine Line Source <br/> int2lm.exe 0000000000F918F5 Unknown Unknown Unknown <br/> int2lm.exe 0000000000F8F6B7 Unknown Unknown Unknown <br/> int2lm.exe 0000000000F43D14 Unknown Unknown Unknown </p> <p> ….. </p> <p> <span class="caps"> MPT </span> : #11 0×0000000000d1b0e4 in src_read_coarse_grid::scatter_data (procarrays=…, <br/> MPT: index=12, iloc=28, irank=3, ilev=58, leof=.FALSE., lrequired=4294967295, <br/> MPT: istat=0) <br/> MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/src_read_coarse_grid.f90:4195 <br/> MPT: #12 0×0000000000cc5771 in src_read_coarse_grid::org_read_coarse_grid (nnow=0, <br/> MPT: lfirst=4294967295, ydate=…, .tmp.YDATE.len_V$f9f=14) <br/> MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/src_read_coarse_grid.f90:1773 <br/> MPT: #13 0×000000000075dab4 in int2lm_org () <br/> MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/int2lm_org.f90:265 <br/> MPT: #14 0×000000000040cc1e in main () <br/> MPT: #15 0×00002aaab09a8c36 in __libc_start_main () from /lib64/libc.so.6 <br/> MPT: #16 0×000000000040cb29 in _start () at ../sysdeps/x86_64/elf/start.S:113 <br/> MPT: (gdb) A debugging session is active. </p> <p> —— </p> <p> Greetings, <br/> Bo </p>

Segmentation fault error in int2lm

BH wrote

Dear all,

recently, I want to set up simulation with COSMO - CLM in CORDEX -EU domain. The driven data is ERAI nterim. I use chain to submit jobs. Then, the std out is correct with no error. But the std err always gets error as following. I don’t how to fix this problem. Is there special parameter should to switch on? I attach all the files.

—-
forrtl: severe (174): SIGSEGV , segmentation fault occurred
Image PC Routine Line Source
int2lm.exe 0000000000F918F5 Unknown Unknown Unknown
int2lm.exe 0000000000F8F6B7 Unknown Unknown Unknown
int2lm.exe 0000000000F43D14 Unknown Unknown Unknown

…..

MPT : #11 0×0000000000d1b0e4 in src_read_coarse_grid::scatter_data (procarrays=…,
MPT: index=12, iloc=28, irank=3, ilev=58, leof=.FALSE., lrequired=4294967295,
MPT: istat=0)
MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/src_read_coarse_grid.f90:4195
MPT: #12 0×0000000000cc5771 in src_read_coarse_grid::org_read_coarse_grid (nnow=0,
MPT: lfirst=4294967295, ydate=…, .tmp.YDATE.len_V$f9f=14)
MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/src_read_coarse_grid.f90:1773
MPT: #13 0×000000000075dab4 in int2lm_org ()
MPT: at /work/bhuang/cclm-sp_2.4/src/int2lm/src/int2lm_org.f90:265
MPT: #14 0×000000000040cc1e in main ()
MPT: #15 0×00002aaab09a8c36 in __libc_start_main () from /lib64/libc.so.6
MPT: #16 0×000000000040cb29 in _start () at ../sysdeps/x86_64/elf/start.S:113
MPT: (gdb) A debugging session is active.

——

Greetings,
Bo

View in channel
<p> Can you provide the related joblog file, too, please. </p>

  @burkhardtrockel in #9fc33b3

<p> Can you provide the related joblog file, too, please. </p>
BR replied

Can you provide the related joblog file, too, please.

View in channel
<p> Dear Burkhardt, </p> <p> actually, I used the cclm-sp_2.4 version from <span class="caps"> CLM </span> Training 2017. Firstly, I tested the ext data (europe_0440.nc) and gcm data (ncep). It works. The simulation finished and the post-process also has done. It does not work when I change the gcm data ncep to erainterim. I only change the <span class="caps"> GRID </span> _IN in int2lm.job.tmpl. Attached the joblog file for both simulations. How to fix this problem? Thank you very much. </p> <p> <strong> ncep: </strong> <br/> &amp;GRID_IN pcontrol_fi =30000., ie_in_tot = 192, je_in_tot = 94, ke_in_tot = 28, startlat_in_tot = -88.542, startlon_in_tot = -180., endlat_in_tot = 88.542, endlon_in_tot = 178.125, pollat_in = 90.0, pollon_in = 180.0, lushift_in=.FALSE.,.FALSE., lvshift_in=.FALSE.,.FALSE., ke_soil_in=1, czml_soil_in=0.05, 1.05, /END </p> <p> <strong> erainterim </strong> <br/> &amp;GRID_IN pcontrol_fi =30000., ie_in_tot = 512, je_in_tot = 256, ke_in_tot = 60, startlat_in_tot = -89.46282, startlon_in_tot = -180.0, endlat_in_tot = 89.46282, endlon_in_tot = 179.2969, pollat_in = 90.0, pollon_in = 180.0, lushift_in=.FALSE.,.FALSE., lvshift_in=.FALSE.,.FALSE., ke_soil_in=3, czml_soil_in=0.035, 0.175, 0.64, 1.775, /END </p> <p> Greetings, <br/> Bo </p>

  @bohuang in #3e899e3

<p> Dear Burkhardt, </p> <p> actually, I used the cclm-sp_2.4 version from <span class="caps"> CLM </span> Training 2017. Firstly, I tested the ext data (europe_0440.nc) and gcm data (ncep). It works. The simulation finished and the post-process also has done. It does not work when I change the gcm data ncep to erainterim. I only change the <span class="caps"> GRID </span> _IN in int2lm.job.tmpl. Attached the joblog file for both simulations. How to fix this problem? Thank you very much. </p> <p> <strong> ncep: </strong> <br/> &amp;GRID_IN pcontrol_fi =30000., ie_in_tot = 192, je_in_tot = 94, ke_in_tot = 28, startlat_in_tot = -88.542, startlon_in_tot = -180., endlat_in_tot = 88.542, endlon_in_tot = 178.125, pollat_in = 90.0, pollon_in = 180.0, lushift_in=.FALSE.,.FALSE., lvshift_in=.FALSE.,.FALSE., ke_soil_in=1, czml_soil_in=0.05, 1.05, /END </p> <p> <strong> erainterim </strong> <br/> &amp;GRID_IN pcontrol_fi =30000., ie_in_tot = 512, je_in_tot = 256, ke_in_tot = 60, startlat_in_tot = -89.46282, startlon_in_tot = -180.0, endlat_in_tot = 89.46282, endlon_in_tot = 179.2969, pollat_in = 90.0, pollon_in = 180.0, lushift_in=.FALSE.,.FALSE., lvshift_in=.FALSE.,.FALSE., ke_soil_in=3, czml_soil_in=0.035, 0.175, 0.64, 1.775, /END </p> <p> Greetings, <br/> Bo </p>
BH replied

Dear Burkhardt,

actually, I used the cclm-sp_2.4 version from CLM Training 2017. Firstly, I tested the ext data (europe_0440.nc) and gcm data (ncep). It works. The simulation finished and the post-process also has done. It does not work when I change the gcm data ncep to erainterim. I only change the GRID _IN in int2lm.job.tmpl. Attached the joblog file for both simulations. How to fix this problem? Thank you very much.

ncep:
&GRID_IN pcontrol_fi =30000., ie_in_tot = 192, je_in_tot = 94, ke_in_tot = 28, startlat_in_tot = -88.542, startlon_in_tot = -180., endlat_in_tot = 88.542, endlon_in_tot = 178.125, pollat_in = 90.0, pollon_in = 180.0, lushift_in=.FALSE.,.FALSE., lvshift_in=.FALSE.,.FALSE., ke_soil_in=1, czml_soil_in=0.05, 1.05, /END

erainterim
&GRID_IN pcontrol_fi =30000., ie_in_tot = 512, je_in_tot = 256, ke_in_tot = 60, startlat_in_tot = -89.46282, startlon_in_tot = -180.0, endlat_in_tot = 89.46282, endlon_in_tot = 179.2969, pollat_in = 90.0, pollon_in = 180.0, lushift_in=.FALSE.,.FALSE., lvshift_in=.FALSE.,.FALSE., ke_soil_in=3, czml_soil_in=0.035, 0.175, 0.64, 1.775, /END

Greetings,
Bo

View in channel
<p> Can you send me the <code> preperai044.job </code> from your <code> jobs </code> directory too? </p>

  @burkhardtrockel in #7876d44

<p> Can you send me the <code> preperai044.job </code> from your <code> jobs </code> directory too? </p>
BR replied

Can you send me the preperai044.job from your jobs directory too?

View in channel
<p> Hi, </p> <p> can you try <br/> lprog_qi=.TRUE. <br/> in the &amp;CONTRL section for the europe_0440.nc case with <span class="caps"> ERAI </span> nterim data? </p> <p> I’ve just tried it and it seems to work for me. </p> <p> Probably related to this <br/> [[http://redc.clm-community.eu/boards/12/topics/417]] </p> <p> <span class="caps"> NCEP </span> does not have QI, while <span class="caps"> ERAI </span> nterim has. Somehow our setup in the environment/compiler/library/settings at vilje (our <span class="caps"> HPC </span> ) seems to have problems with ignoring forcing data that’s available. </p>

  @andreasdobler in #9be0cf2

<p> Hi, </p> <p> can you try <br/> lprog_qi=.TRUE. <br/> in the &amp;CONTRL section for the europe_0440.nc case with <span class="caps"> ERAI </span> nterim data? </p> <p> I’ve just tried it and it seems to work for me. </p> <p> Probably related to this <br/> [[http://redc.clm-community.eu/boards/12/topics/417]] </p> <p> <span class="caps"> NCEP </span> does not have QI, while <span class="caps"> ERAI </span> nterim has. Somehow our setup in the environment/compiler/library/settings at vilje (our <span class="caps"> HPC </span> ) seems to have problems with ignoring forcing data that’s available. </p>
AD replied

Hi,

can you try
lprog_qi=.TRUE.
in the &CONTRL section for the europe_0440.nc case with ERAI nterim data?

I’ve just tried it and it seems to work for me.

Probably related to this
[[http://redc.clm-community.eu/boards/12/topics/417]]

NCEP does not have QI, while ERAI nterim has. Somehow our setup in the environment/compiler/library/settings at vilje (our HPC ) seems to have problems with ignoring forcing data that’s available.

View in channel
<p> Hi Burkhardt, </p> <p> enclosed the prep and int2lm job for both erainterim and ncep. Please find them. </p> <p> Greetings, <br/> Bo </p>

  @bohuang in #552ece4

<p> Hi Burkhardt, </p> <p> enclosed the prep and int2lm job for both erainterim and ncep. Please find them. </p> <p> Greetings, <br/> Bo </p>
BH replied

Hi Burkhardt,

enclosed the prep and int2lm job for both erainterim and ncep. Please find them.

Greetings,
Bo

View in channel
<p> 1. Andreas is right, you have to use <code> lprog_qi=.TRUE. </code> for <span class="caps"> ERAI </span> nterim, otherwise the simulation will fail. </p> <p> 2. Please follow the instruction in <code> cclm-sp_2.4/chain/extensions/boundary_data/erainterim/README </code> <br/> You have not made the correct settings in prep and int2lm templates as provided in the <code> cclm-sp_2.4/chain/extensions/boundary_data/erainterim/templates </code> directory. </p>

  @burkhardtrockel in #64845e3

<p> 1. Andreas is right, you have to use <code> lprog_qi=.TRUE. </code> for <span class="caps"> ERAI </span> nterim, otherwise the simulation will fail. </p> <p> 2. Please follow the instruction in <code> cclm-sp_2.4/chain/extensions/boundary_data/erainterim/README </code> <br/> You have not made the correct settings in prep and int2lm templates as provided in the <code> cclm-sp_2.4/chain/extensions/boundary_data/erainterim/templates </code> directory. </p>
BR replied

1. Andreas is right, you have to use lprog_qi=.TRUE. for ERAI nterim, otherwise the simulation will fail.

2. Please follow the instruction in cclm-sp_2.4/chain/extensions/boundary_data/erainterim/README
You have not made the correct settings in prep and int2lm templates as provided in the cclm-sp_2.4/chain/extensions/boundary_data/erainterim/templates directory.

View in channel
<p> Hi Andreas, </p> <p> it works now. That is the reason for the error. Thank you so much. </p> <p> Greetings, <br/> Bo </p>

  @bohuang in #96af69a

<p> Hi Andreas, </p> <p> it works now. That is the reason for the error. Thank you so much. </p> <p> Greetings, <br/> Bo </p>
BH replied

Hi Andreas,

it works now. That is the reason for the error. Thank you so much.

Greetings,
Bo

View in channel
<p> By the way, we had this problem already in the int2lm forum: <br/> http://redc.clm-community.eu/boards/12/topics/444 </p>

  @burkhardtrockel in #cc76b7c

<p> By the way, we had this problem already in the int2lm forum: <br/> http://redc.clm-community.eu/boards/12/topics/444 </p>
BR replied

By the way, we had this problem already in the int2lm forum:
http://redc.clm-community.eu/boards/12/topics/444

View in channel
<p> I’ll try to remember this next time ;-) </p>

  @andreasdobler in #0ccc071

<p> I’ll try to remember this next time ;-) </p>
AD replied

I’ll try to remember this next time ;-)

View in channel
<p> Hi Burkhardt, </p> <p> thank you very much for your suggestions and tips. I will follow the instruction to set the model. </p> <p> Greetings, <br/> Bo </p>

  @bohuang in #a79f877

<p> Hi Burkhardt, </p> <p> thank you very much for your suggestions and tips. I will follow the instruction to set the model. </p> <p> Greetings, <br/> Bo </p>
BH replied

Hi Burkhardt,

thank you very much for your suggestions and tips. I will follow the instruction to set the model.

Greetings,
Bo

View in channel