Retro/pcjs
Watch
1
0
Fork
You've already forked pcjs
 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

PDP-10 multiply diagnostic testing finally underway; fixed one bug related to the KA10 (not that anyone really cares anymore)

Browse source
This commit is contained in:
Jeff 2017-03-30 23:46:01 -07:00 • committed by Jeff Parsons
commit 307517fd76
9 changed files with 20151 additions and 45 deletions
Download patch file Download diff file Expand all files Collapse all files

33
modules/pdp10/lib/cpuops.js
View file

@ -6323,6 +6323,33 @@ PDP10.doDIV = function(dst, ext, src)
* number (base 2^18). Each individual multiplication of these 18-bit "digits" will produce
* a result within 2^36, well within JavaScript integer accuracy.
*
* PDP-10 "DAKAK" Diagnostic Notes
* -------------------------------
*
* 036174: 200240 043643 MOVE 5,43643 ; [43643] = 400000000000
* 036175: 200300 043603 MOVE 6,43603 ; [43603] = 777777777777
* 036176: 200140 043604 MOVE 3,43604 ; [43604] = 000000000001
* 036177: 224240 000003 MUL 5,3 ; Multiply 400000000000 by 000000000001
* 036200: 312240 043604 CAME 5,43604 ; high order result in AC should be: 000000000001
* 036201: 003240 033721 UUO 5,33721 ;
* 036202: 312300 043602 CAME 6,43602 ; low order result in AC+1 should be: 000000000000
*
* The "natural" result is:
*
* 05=777777777777 06=400000000000
*
* And SIMH seems to agree. So why does the DEC diagnostic expect:
*
* 05=000000000001 06=000000000000
*
* The answer can be found in the June 1982 "DECSYSTEM-10 and DECSYSTEM-20 Processor Reference Manual",
* in the description of the MUL instruction:
*
* CAUTION: In the KA10, an AC operand of 2^35 is treated as though it were +2^35, producing the
* incorrect sign in the product.
*
* This behavior is now simulated below for MODEL_KA10, at least to the extent that the diagnostic is happy.
*
* @this {CPUStatePDP10}
* @param {number} dst (36-bit value)
* @param {number} src (36-bit value)
@ -6339,8 +6366,10 @@ PDP10.doMUL = function(dst, src, fTruncate)
* we'll negate the result afterward if necessary.
*/
if (n1 > PDP10.INT_MASK) {
n1 = PDP10.WORD_LIMIT - n1;
fNeg = !fNeg;
if (this.model != PDP10.MODEL_KA10 || n1 != PDP10.INT_LIMIT) {
n1 = PDP10.WORD_LIMIT - n1;
fNeg = !fNeg;
}
}
if (n2 > PDP10.INT_MASK) {