Objection. Please note that when you use dmad there is a 16-bit data memory address encoded in the instruction, not "7 offset bits put into instruction opcode and segment 9 bits assumed zero." You can see this from page 1-2 of SPRU172C where 'dmad' is defined. Then you can look at an instruction definition that uses dmad such as mvmd on page 4-115 of that same document and see clearly that all 16 bits of the dmad are encoded in the 2nd word of that instruction. And you are shown this again in CPU and Periph guide 5.2: "Absolute addresses are always encoded with a length of 16 bits, so instructions that encode absolute addresses are always at least two words in length." dmad is absolute addressing. The TI tech support guy really did give you a confusing statement to go on but everything really is explained accurately in CPU & Periph Ch5. -Micah _______________________________________________________ > > Message: 1 > Date: Mon, 07 Jun 2004 08:11:36 -0000 > From: "Ilya Druker" <> > Subject: Re: 'dmad' - fundamental question to > specialists > > Now, combining the kind comments of Jeff Brower, > answer of TI's > support and my own experiments I would define dmad > addressing as follows: > dmad is the direct data-memory addressing mode > with 7 offset bits > put into instruction opcode and segment 9 bits > assumed zero. > > Any objection? > Ilya Druker > __________________________________ |
Re: 'dmad' - fundamental question to specialists
Started by ●June 7, 2004