We porting an A/N solution from AIX 7.2 to redhat 8.10
big endian to little endian
We have problems with a superdescriptor based on
fields of the types A2, B8 and I4.
The superdescriptor does not work at all, a search with read with the first 2 fields filled and the i4 starts with 0 ends with 9999999 returns randomly associated records.
On AIX everything works fine, even refreshing the file and typing in new records on LINUX wont help.
Has anybody here have experienced similar behaviour?
I presume that you used ADABCK to unload the AIX database and load it into the Red Had instance, and ADABCK would have handled the endian conversion. When you refresh the file and load new records, no endian conversion is involved, so that should work, too.
But the problem must be endian-related, so perhaps it’s the superdecriptor definition. Are you using subfields of an I4 field? The documentation states “‘from’ and ‘to’ are counted from low order to high order, beginning with 1, for binary, fixed point and floating point fields”, but check your field offsets.
Have you tried an ADAINV REINVERT?