Re: VAX addressing modes

["Maciej W. Rozycki" <macro%linux-mips.org@localhost>]

On Thu, 10 Aug 2017, Johnny Billquist wrote:

> You need to read the VARM better. You have three fixed arguments, 
> followed by a variable number of arguments/entries in the instruction 
> flow. They are all a part of the instruction.

 Do you assert that the whole jump table is fetched by the IFU before 
execution of the instruction starts?

 Iff it was so, then it would be considered a part of the instruction 
itself.  Otherwise it's just data, whose address happens to be implied by 
the instruction rather than specified with an operand; not uncommon a case 
across processor architectures (e.g. the x86 XLATB instruction).  Even the 
architecture document previously referred (on p. 3-63) is clear about it:

"1. After operand evaluation, PC is pointing at displ[0], not at
    the next instruction.  The branch displacements are relative
    to the address of displ[0]."

-- so no, `displ[0]', etc. are not operands, they are data.

 That's my point of view anyway.  You have the right to disagree of 
course.

  Maciej