I'm wondering if the error gcc is throwing is really a valid error. These mismatches are coming because an INGEGER*8 is being passed via %VAL() so this is actually passing by value (instead of the usual fortran pass-by-reference).
For the first error at line 248, - PIXMAP is initialised by a GRGMEM call at line 238 which is a C function (sys/grgmem.c). - the GRGIO3 function declares the relevant argument type BYTE. It's a similar situation for the other errors. I think there's a resonable case for handling this FTBFS by adding -fallow-argument-mismatch to the compiler arguments for the make rules that build this module. Vince
