The general idea is to multiply ah
by 10 then add to al
. Not that the implementation in x86 assembly is somewhat tedious, because the multiply instructions are limited in what they can operate on. As an alternative you could break down the multiplication into shifts and adds, like:
add ah, ah ; ah*2
add al, ah ; al + 2*ah
shl ah, 2 ; 8*ah
add al, ah ; al + 10*ah
Note that the comments refer to the original values, and that this is a 8-bit version.