[MITgcm-devel] Problems with TAF 2.1.37

[Jean-Michel Campin]

Hi Ralf,

There are few adjoint tests that failed last night, they are
 1) compiling pkg/seaice (dynamics):  lab_sea, global_ocean.cs32x15
 2) using a f77 compiler (g77, pgf77): some on 32-bit platform
 (g77 on old aces 32 bit nodes), some on 64-bit platform (g77 on
  baudelaire, pgf77 on new aces-grid cluster).

I am getting an error message that looks like some f90 syntax
that is not understood by the compiler:

> ad_taf_output.f: In subroutine `adseaice_lsr':
> ad_taf_output.f:95852:
>                      adaushh = adau
>                      ^
> Expression at (^) has incorrect data type or rank for its context

and the declaration (line ~93841) is:
>       double precision adau(1-olx:snx+olx,1-oly:sny+oly,nsx,nsy)
>       double precision adaushh(1-olx:snx+olx,1-oly:sny+oly,nsx,nsy)

I can confirm that if I revert to TAF version 2.1.34 (which was still
used for Sunday night testing, all compiling fine) and using exactly

Here is the taf_command which is used (cat taf_command):
> tafx  -v1 -reverse -admark ad -i4 -r4 -l taf_ad.log -toplevel the_main_loop -input xx_theta_dummy,xx_salt_dummy,xx_tr1_dummy,xx_hflux_dummy,xx_sflux_dummy,xx_tauu_dummy,xx_tauv_dummy,xx_atemp_dummy,xx_aqh_dummy,xx_precip_dummy,xx_swflux_dummy,xx_swdown_dummy,xx_lwflux_dummy,xx_lwdown_dummy,xx_evap_dummy,xx_snowprecip_dummy,xx_apressure_dummy,xx_runoff_dummy,xx_uwind_dummy,xx_vwind_dummy,xx_obcsn_dummy,xx_obcss_dummy,xx_obcsw_dummy,xx_obcse_dummy,xx_sst_dummy,xx_sss_dummy,xx_depth_dummy,xx_diffkr_dummy,xx_kapgm_dummy,xx_kapredi_dummy,xx_bottomdrag_dummy,xx_efluxy_dummy,xx_efluxp_dummy,xx_edtaux_dummy,xx_edtauy_dummy,xx_uvel_dummy,xx_vvel_dummy,xx_etan_dummy,xx_gen2d_dummy,xx_gen3d_dummy,xx_siarea_dummy,xx_siheff_dummy,xx_sihsnow_dummy,xx_shifwflx_dummy,xx_relaxsst_dummy,xx_relaxsss_dummy -output fc  ad_input_code.f

Cheers,
Jean-Michel