noel

Same issue as above comment for `SMS_D_R_Ld5.ne4_ne4.FSCM-ARM97.pm-cpu_intel.eam-scm` with master of June 17th

No surprise I'm also seeing this with the highres runs using 1950 compset. `SMS_Ld2_D_PT.ne120np4_oRRS18to6v3_ICG.A_WCYCL1950S_CMIP6_HR.cori-knl_gnu.cam-cosplite` ``` 169: At line 938 of file /global/cscratch1/sd/ndk/wacmy/m67-mar1/components/cam/src/physics/clubb/advance_clubb_core_module.F90 169: Fortran runtime error: Index '1' of dimension...

I then added some prints, like so: ``` write(*,*) " ndk size(wp4_zt), size(wprtp2), size(wp2rtp)=", size(wp4_zt), size(wprtp2), size(wp2rtp) ``` Most of them are 73,73,73, but several print 73,1,1 So that would...

I also ran with Intel debug and it looks like the value is always 73. I will see if value is 73 for GNU when NOT using DEBUG. etc... Note...

I meant to say thanks @worleyph for the link. That does look very relevant. I just haven't had time to look at this closer, but this does seem like something...

I tried the high res case (that I'm most interested in) with DEBUG=TRUE and with only 1 thread and I got a different error. ``` 659: Program received signal SIGFPE:...

Looking back at the error with ne16/ne30 and the issue Pat mentioned, I'm now trying the following: ``` if (gr%nz > 0) then Skw_zt(1:gr%nz) = Skw_func( wp2_zt(1:gr%nz), wp3(1:gr%nz) ) !...

Yea I tried that and it failed in same location with same error. I'm now trying to compile this file differently in hopes I can let the run continue.

``` Skw(1:size(wp2)) ! Result Skw [-] 1 Error: 'array' argument of 'size' intrinsic at (1) must be an array ```

Hmm, just got: ``` 337: 337: Program received signal SIGSEGV: Segmentation fault - invalid memory reference. ```