You might want to place Zero Initialized (ZI) data in load region LR1, and use a relative base address for the next load region LR2, for example:

LR1 0x8000
{
    er_progbits +0
    {
        *(+RO,+RW) ; Takes space in the Load Region
    }
    er_zi +0
    {
        *(+ZI) ; Takes no space in the Load Region
    }
}
LR2 +0 ; Load Region follows immediately from LR1
{
    er_moreprogbits +0
    {
        file1.o(+RO) ; Takes space in the Load Region
    }
}

Because the linker does not adjust the base address of LR2 to account for ZI data, the execution region er_zi overlaps the execution region er_moreprogbits. This generates an error when linking.

To correct this, use the ImageLimit() function with the name of the ZI execution region to calculate the base address of LR2. For example:

LR1 0x8000
{
    er_progbits +0
    {
        *(+RO,+RW) ; Takes space in the Load Region
    }
    er_zi +0
    {
        *(+ZI) ; Takes no space in the Load Region
    }
}
LR2 ImageLimit(er_zi) ; Set the address of LR2 to limit of er_zi
{
    er_moreprogbits +0
    {
        file1.o(+RO) ; Takes space in the Load Region
    }
}