🚀 go-pugleaf

Steve Siperas <hsi!hsi86!siperas@UUNET.UU.NET> writes:

>   I was wondering if anyone would know if the following code
>   particularly the Branch with indexed register
>   instruction would act any different if the code was
>   assembled with XA above the line opposed to assembling the code
>   without XA below the line.
>
>   00023          L     R3,VERS                R3 -> VERSION (0 - 3)
>   00024          PACK  DWORD,0(2,R3)          PACK VERSION IN DWORD
>   00025          CVB   R3,DWORD
>   00026          SLL   R3,2                   MULTIPLY VERSION BY 4 FOR OFFSET
>   00027          B     ROUTE(R3)
>   00028 *
>   00029 ROUTE    B     STEP0                  (0)BRANCH TO STEP0
>   00030          B     STEP1                  (4)BRANCH TO STEP1
>   00031          B     STEP2                  (8)BRANCH TO STEP2
>   00032          B     STEP3                  (12)BRANCH TO STEP3

To my knowledge it should work identically in 24 and 31-bit mode.
CVB generates a full 32-bit signed value and the SLL also operates
on 32 bits.  The potential exposure on the branch is the possibility
that the high-order byte contains garbage, and I don't see that
happening here.
           - Pete

=====================================================================
= Peter Flass                               flass@nyslbd.gov        =
= Systems Programmer                        tel (518)458-5114       =
= NYS Legislative Bill Drafting Commission                          =
= 1450 Western Ave.                                                 =
= Albany, NY  12203                                                 =
=====================================================================