[Carpet] portland compiler and ADMConstraints runtime error

[I.Hawke]

This is a problem with strings continuing over lines; certain versions
of the portland compilers split the strings incorrectly. There are two
solutions:

1) Explicitly modify the code to include continuation lines of less than
72 characters. You'll need to modify InitSymBound.F of ADMConstraints
lines 51, 60, 67 and 74, plus make similar modifications to (at least)
BSSN_MoL and Whisky, I believe.
2) Modify your setup options to use a better compiler / preprocessor
set; this may not always be possible.

As -Mextend should allow for up to 132 characters my guess is that the
preprocessor is causing the problems by including whitespace.

Ian

On Thu, 2007-07-12 at 11:39 +0200, sebastiano bernuzzi wrote:
> dear all
> 
> I'm trying to compile a standard cactus-carpet-whisky (dev)
> configuration cheked out recently,
> with these options:
> 
> F77=pgf77
> F90=pgf90
> F90FLAGS=-Mx,125,0x200 -Mextend
> F77FLAGS=-Mx,125,0x200 -Mextend 
> 
> it compiles but I obtain this runtime error:
> 
> s', 'WHISKY::rho', 'ADMCONSTRAINTS::momz', 'ADMCONSTRAINTS::momy',
> 'ADMCONSTRAINTS::momx', 'ADMCONSTRAINTS::ham', 'ADMBASE::betaz',
> 'ADMBASE::betay', 'ADMBASE::betax', 'ADMBASE::alp', 'ADMBASE::gzz',
> 'ADMBASE::gyz', 'ADMBASE::gyy', 'ADMBASE::gxz', 'ADMBASE::gxy',
> 'ADMBASE::gxx' 
> INFO (CarpetIOASCII): I/O Method 'IOASCII_2D' registered: 2D AMR
> output of grid variables to ASCII files
> INFO (CarpetIOASCII): I/O Method 'IOASCII_3D' registered: 3D AMR
> output of grid variables to ASCII files 
> INFO (WaveExtractGeneral): setup GH
> INFO (WaveExtractGeneral): Output directory is './StarDAT/'
> INFO (WaveExtractGeneral): Setting out_every to 1
> WARNING[L1,P0] (CartGrid3D): Unknown variable
> 'admconstraints::hamnormalize
> d' in SetCartSymVN 
> WARNING level 0 in thorn ADMConstraints processor 0 host
> master.albert2n.pr.infn.it
>   (line 53 of InitSymBound.F):
>   -> Error set
> ting symmetries for ADMConstraints::hamnormalized 
> WARNING level 0 in thorn ADMConstraints processor 0 host
> master.albert2n.pr.infn.it
>   (line 53 of InitSymBound.F):
>   -> Error set
> ting symmetries for ADMConstraints::hamnormalized 
> 
> does someone know why?
> 
> thank you
> 
> 
> -- 
> Sebastiano Bernuzzi
> SKYPE: sebastiano.bernuzzi
> ------------------------------------------------------
> www.fis.unipr.it/~bernuzzi
> www.mojoshouse.net
> ------------------------------------------------------ 
> _______________________________________________
> developers mailing list
> developers at lists.carpetcode.org
> http://lists.carpetcode.org/listinfo/developers