./case.build ERROR: command gmake complib

[Fuhow]

Hello
When I am porting CESM 2.2.0 to a ubuntu server, I met an error during the ./case.build as shown below.

My config_machines.xml and config _compilers.xml and bld.log are attached.
Thanks for any reply.

 

[Fuhow]

These files are here.

 

[Fuhow]

I also checked similar problems posted before, some replys said that this error may be caused by the older gmake version, and my GNU make version is 4.2.1 which is the newest one.

 

[sacks]

Can you please give the output of running gfortran --version?

 

[Fuhow]

Can you please give the output of running gfortran --version?

sure, here is the gfortran version: 10.5.0, which is the newest version.

 

[slevis]

@Fuhow I don't mean to change the subject (you and @sacks can continue your troubleshooting). I just wanted to make sure that you have not made changes to the code before building, in other words you are building the default code.

 

[Fuhow]

@Fuhow I don't mean to change the subject (you and @sacks can continue your troubleshooting). I just wanted to make sure that you have not made changes to the code before building, in other words you are building the default code.

Hi, slevis, thanks for your advice, I did not change any code in the model. I checkd nearly all the similar problems on the internet, the only possibility I can think is about the configure process.

 

[sacks]

I think gcc / gfortran is up to version 13 now, but we test with version 10.1 on our main test system - so, close to what you're using. (I thought this might be due to use of an old compiler version, but that doesn't seem to be the case.)

I'm going to move this to the CAM forums, since you're getting an error in the CAM build.

 

[sacks]

CAM folks - please see the original post - any ideas what might be causing this CAM build error?

 

[peverley]

Hi @Fuhow

It looks like this might be a bug in the code. Can you try breaking up the line in question in components/cam/src/physics/cam/physics_buffer.F90.in into two lines:

i.e. change line ~764

print *,__FILE__,__LINE__,i,trim(pbufPtr%hdr%name),pbufPtr%hdr%dtype,pbufPtr%hdr%persistence,pbufPtr%hdr%dimsizes

to:

print *,__FILE__,__LINE__,i,trim(pbufPtr%hdr%name),pbufPtr%hdr%dtype, &
pbufPtr%hdr%persistence,pbufPtr%hdr%dimsizes

Once you do that, do a clean rebuild to make sure your case reflects the new code:

./case.setup --reset
./case.build --clean-all
./case.build

Let me know if that doesn't work.

Courtney

 

[Johnny]

Hello
When I am porting CESM 2.2.0 to a ubuntu server, I met an error during the ./case.build as shown below.
View attachment 4108
My config_machines.xml and config _compilers.xml and bld.log are attached.
Thanks for any reply.

When you see error on byte 132 that means the line is too long for the Fortran compiler to like.
you need -ffree-line-length-none
Please add this to your config_compilers.xml
<FFLAGS>
<append> -fallow-argument-mismatch -fallow-invalid-boz -ffree-line-length-none</append>
</FFLAGS>

 

[Fuhow]

When you see error on byte 132 that means the line is too long for the Fortran compiler to like.
you need -ffree-line-length-none
Please add this to your config_compilers.xml
<FFLAGS>
<append> -fallow-argument-mismatch -fallow-invalid-boz -ffree-line-length-none</append>
</FFLAGS>

Thank you so much! It's working now! :D

 

[Fuhow]

Hi @Fuhow

It looks like this might be a bug in the code. Can you try breaking up the line in question in components/cam/src/physics/cam/physics_buffer.F90.in into two lines:

i.e. change line ~764


to:


Once you do that, do a clean rebuild to make sure your case reflects the new code:



Let me know if that doesn't work.

Courtney

Hi, peverley.
Thanks for your reply, the problem is truly in the .F90 file. As you and Johnny said, the Fortran compiler cannot compile the long code, and my CESM is working now through Johnny's way!

 

[Johnny]

Hi, peverley.
Thanks for your reply, the problem is truly in the .F90 file. As you and Johnny said, the Fortran compiler cannot compile the long code, and my CESM is working now through Johnny's way!

Hi Fuhow,
Are you able to run a larger compset that requires multi-processor? Also when you compile hdf5, do you add in --enable-parallel? thanks.
I can run --compset QPC4 --res f45_f45_mg37 but not I2000Clm50SpGs