On Thu, May 11, 2017 at 12:32 PM, Charles Mills <[email protected]> wrote:
> And one of these days the architecture starts allowing EX of an EX and it > fails the bite test. > I really doubt that IBM would make such a change to the existing EX instruction. IBM well knows that people (idiots? like me) use EX 0,* as a one instruction guaranteed abend. If they did do such a thing, I would expect them to deliberately check for that specific case and force a S0C3 (actually PIC 3 - not COBOL). I don't actually like the EX instructions. I remember the implementation of such in the Xerox Sigma series. Every instruction on it was 4 bytes long and so would fit into one of the general purpose registers. That would allow you to "execute" any instruction, which was created inside any register. (another weird aspect is that "absolute" address 0x00 .. 0x0F actually mapped to general registers 0x00..0x0F, so you couldn't really use the 16 bytes of memory as actual memory. And for those 16 addresses, the "memory" accessed 32 bits instead of 8 bits.) Seeing as how the longest instruction on the z is 48 bits (6 bytes), such an implementation might even possible in today's IBM z by using a 64 bit register. > Charles > -- Advertising is a valuable economic factor because it is the cheapest way of selling goods, particularly if the goods are worthless. -- Sinclair Lewis Maranatha! <>< John McKown
