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

FOOTBALL-related fixes

Browse source
This commit is contained in:
Jeff Parsons 2016-02-08 15:36:23 -08:00
commit ac51075c39
7 changed files with 956 additions and 885 deletions
Download patch file Download diff file Expand all files Collapse all files

30
modules/pcjs/lib/video.js
View file

@ -4598,6 +4598,7 @@ Video.prototype.getCardAccess = function()
*
* @this {Video}
* @param {number|undefined} nAccess (one of the Card.ACCESS.* constants)
* @return {boolean} true if access may have changed, false if not
*/
Video.prototype.setCardAccess = function(nAccess)
{
@ -4618,7 +4619,9 @@ Video.prototype.setCardAccess = function(nAccess)
* before choking.
*/
this.bus.setMemoryAccess(card.addrBuffer, card.sizeBuffer, card.getMemoryAccess(), true);
return true;
}
return false;
};
/**
@ -6308,7 +6311,32 @@ Video.prototype.outSEQData = function(port, bOut, addrFrom)
this.cardEGA.nSeqMapMask = Video.aEGAByteToDW[bOut & Card.SEQ.MAPMASK.MAPS];
break;
case Card.SEQ.MEMMODE.INDX:
this.setCardAccess(this.getCardAccess());
if (this.setCardAccess(this.getCardAccess())) {
/*
* When switching screens (via SysReq) on early revisions of OS/2 (eg, FOOTBALL), the screen would go
* blank; this appeared to be because when the card is reprogrammed, we first think the card is going into
* graphics mode, then we reverse course when it becomes clear that the card is going back into text mode,
* but unfortunately, at that precise moment, the Sequencer hasn't been fully reprogrammed, so when we're
* reading screen memory, we're getting back ZEROS for every odd byte (which are the text attribute bytes),
* so the screen is redrawn as black-on-black.
*
* My solution was to change setCardAccess() to indicate whether it actually altered the video buffer
* address and/or format, and if so, then force another screen update.
*
* TODO: This scenario does not seem unique; it suggests that we should generally force a screen update
* whenever the video buffer has undergone a significant change.
*
* UPDATE: This change was NOT sufficient to resolve the OS/2 screen-switching bug described above; in fact,
* it's apparently not even necessary, because the REAL problem was caused by PAGED blocks with stale
* physical video memory blocks; the solution was for the Bus addMemory() and removeMemory() functions to
* call the the CPU flushPageBlocks() function. With that change in place, the window now stays in sync
* with the buffer.
*
* However, calling updateScreen() here still seems like a good idea, and it shouldn't hurt performance,
* since we're doing it only when setCardAccess() indicates a change, so I'm leaving this addition in place.
*/
this.updateScreen(true);
}
break;
default:
break;