From: king621@nospicedham.comcast.net   
      
   On 04/24/2018 07:08 PM, Rod Pemberton wrote:   
   > On Tue, 24 Apr 2018 16:20:47 -0700   
   > bilsch  wrote:   
   >   
   >> On 04/24/2018 04:04 AM, Bartc wrote:   
   >>> On 24/04/2018 10:48, bilsch wrote:   
   >   
   >>>> The boot sector jumps to this code.   
   >>>> It should print junk6789 but prints only j   
   >>>> Why it not print junk6789 ?   TIA.   Bill S.   
   >>>   
   >>>>   
   >>>> pbuf    times 80 db 0   
   >>>> string    db 'junk'   
   >>>   
   >>>>             mov si,pbuf   
   >>>>             mov di,string   
   >>>> lup8:     mov al,[di]   
   >>>>             mov byte[si],al   
   >>>>             inc si   
   >>>>             inc di   
   >>>>             cmp di,4   
   >>>>             jbe lup8   
   >>>   
   >>> di starts at label 'string'. Is that address 0? If not then   
   >>> comparing it with 4 looks odd.   
   >>>   
   >>> If string has an address of at least 80 as seems likely, then this   
   >>> will compare false and only one character is copied to the buffer.   
   >>>   
   >>   
   >> Indeed.  Here's the fix to the code above.  It prints 'junk' but not   
   >> 'junk6789'.  My real question is why doesn't it print the 6789 part?   
   >>   
   >> string    db 'junk',0   
   >>   
   >>             mov si,pbuf   
   >>             mov di,string   
   >> lup8:     mov al,[di]   
   >>             cmp al,0   
   >>             jz  ovr1   
   >>             move byte[si],al   
   >>             inc si   
   >>             inc di   
   >>             jmp lup8   
   >> ovr1:   
   >>   
   >   
   > Your 'junk' loop at 'lup8' already increments 'si' _after_ storing al,   
   > and then you increment 'si' again at 'xo5', but /prior/ to storing 'al'.   
   > So, you've skipped copying a character into one location.  Since pbuf   
   > buffer is zero'd initially, you've "inserted" a nul character between   
   > 'junk' and '6789', due to this skip or off-by-one error.  Your prstr   
   > routine terminates at nul characters, which then skips everything after   
   > 'junk' since 'junk' is followed by a nul character.  I.e., the 'inc si'   
   > at 'xo5' should be below the next instruction.   
   >   
   >   
   > Rod Pemberton   
   >   
   I added 'dec si' at ovr1 and it works now.   
   Good news: this is what was causing my original problem in my real   
   program.  I fixed it also by adding a dec si line.   
   Thanks for the help.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)