From: notsaying@nospicedham.invalid.org
"wolfgang kern" wrote in
news:okdbps$1hi6$1@gioia.aioe.org:
> Kerr Mudd-John wrote:
> ...
>
>> An alternative (I don't know if it's better or worse for your fixups)
>> 2C 06 sub al,06
>> 98 cbw
>> 41 inc cx
>> 41 inc cx
>> D3 E0 shl ax,cl
>> 08 25 or [di],ah
>> 47 inc di
>> 83 E1 06 and cx,0006
>> 88 05 mov [di],al
>> 75 01 jnz 0137
>> 4F dec di
>> AC lodsb
>> EB d5 jmp top+2
>>
>> 9806 e0d3 08 e183 8806 05 01 AC D5EB
>> 9 maybe not.
>
> easy to check: 9806+7862=1068 so not pairable by sub. Only values
> between a9xx..05xx and xxbf..xx1c will be pairable with my current
> DX=7862 subs. So I use the coincidental given AH=A1 to cover 80..a8
> singles.
Another alternative: (needs cl=7 at start and 7 on the stack).
123 98 cbw
124 49 dec cx
125 D3 C8 ror ax,cl
127 08 05 or [di],al
129 47 inc di
12A 88 25 mov [di],ah
12C 49 dec cx
12D 79 03 jns 0132
12F 4F dec di
130 59 pop cx ; OR push dx
131 51 push cx ; pop cx
132 AC lodsb
133 3C 3D cmp al,3D
135 75 D4 jnz 010B
137 59 pop cx ; tidy stack on exit
or could set dl=7 and "push dx/pop cx" to save the extra 2 stack ops?
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
