home bbs files messages ]

Forums before death by AOL, social media and spammers... "We can't have nice things"

   alt.os.development      Operating system development chatter      4,255 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 3,910 of 4,255   
   Robert Pengelly to Robert Pengelly   
   Re: COM1 interrupt for 16-bit OS   
   12 Nov 23 11:13:14   
   
   From: robertapengelly@gmail.com   
      
   On Sunday, 12 November 2023 at 19:04:55 UTC, Robert Pengelly wrote:   
   > On Sunday, 12 November 2023 at 15:18:45 UTC, wolfgang kern wrote:    
   > > On 12/11/2023 15:34, Robert Pengelly wrote:    
   > > ...    
   > > >>>> the mouse sends a sequence of bytes where every byte causes in IRQ.    
   > > >>>> so you first need to fill a buffer with the gotten bytes before you   
   can    
   > > >>>> calculate/scale/interpret the values. mouse IRQs can easy come out of    
   > > >>>> sync, so checking always for the first byte mark is a good idea.    
   > > >>> Oh I thought you had to get them all at the same time when the   
   interrupt was fired. As for filling the buffer how would I tell the interrupts   
   apart? I can test for 0x40 and 0x01 (which I'm already doing) I'm just a   
   little confused how I would    
   know whether it's the 1st, 2nd and 3rd bytes.    
   > > >> there is a mark for the first byte (reread the RBIL I posted)    
   > > >> I use one variable for count down (easier than count up) and use this    
   > > >> with BX as an index into my buffer.    
   > > >> and when the buffer is filled I adjust all my mouse-variables which I    
   > > >> later use on my mSec timer-base schedule to update cursor and act on    
   > > >> buttons.    
   > > > Yeah, there's bit 6 for the initial bit, I weren't sure if there's   
   anything identifying the others. So basically have two variables one being a   
   counter and just write to the buffer until I get three (or if it's counting   
   down then it would be until    
   I hit zero)?    
   > > yes.    
   > > As for:    
   > > >> I use one variable for count down (easier than count up) and use this    
   > > >> with BX as an index into my buffer.    
   > >    
   > > > How do you use it for the buffer if your counting down? If you were   
   counting up you could shl by 8 and add that to the buffer offset.    
   > > no shift required here at all:    
   > >    
   > > push ...ds,ax,bx,cx,dx    
   > > ... ds become ptr to my data    
   > > IN AL,port    
   > > TEST AL,0x40    
   > > jz skip_init    
   > > xor bx,bx    
   > > MOV CX,3 ;(or eight in my case)    
   > > MOV [count],cl    
   > > MOV [index].bx    
   > > skip_init:    
   > > MOV BX,[index]    
   > > MOV [BX+buffer],AL    
   > > INC word[index] ;or: INC BX |MOV [index],BX    
   > > DEC byte[count]    
   > > pop ... dx,cx,bx    
   > > MOV AL,0x20    
   > > out 0xA0,AL    
   > > jnz done_it    
   > > got_all:    
   > > ;store/modify, update only variables, actions are done in idle queue.    
   > > done_it:    
   > > pop ax    
   > > pop ds    
   > > iret    
   > > __    
   > > wolfgang   
   > Ahh you have 3 variables instead of 2, right I think I understand. What   
   should I do about the buffer though? Like should I get them in say a 1Ch   
   handler to move the cursor in there?   
   That should have been "What should I do about the values though?"  Like should   
   I parse the buffer in say 1Ch and move the cursor in there?   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca