I was just working in Assembler and contemplating how powerful D is for it. In fact, if I could write Assembler using the D language without GC or heap/stack I would use it for everything including boot code.

Right now it uses much the same system for addressing structures as NASM does, which is a kludge at best.  I want this:

lea eax, MyStruct.myFavoriteMember;

But if we do this:

lea eax, MyStruct.mySubStruct.myMember;

I know it will fail.  Why?  Because we're accessing a struct, and then getting a member from that struct to get another struct.  We need to first get mySubStruct, and THEN we can get myMember.  It needs two lea's.

lea eax, MyStruct.mySubStruct;
lea eax, eax.myMember;

That would make programming assembler in D unbelievably cool.  Unfortunately, I know about zilch about the D implementation code.

My only other gripe is the lack of assembler coloring in any D editors I've encountered.

If you're using TextPad, it's pretty easy to add keywords to the D.syn file; it's just text.


"Dan" <Dan_member@pathlink.com> wrote in message news:djmffm$1io7$1@digitaldaemon.com...
>
> I was just working in Assembler and contemplating how
> powerful D is for it.
> In fact, if I could write Assembler using the D language
> without GC or
> heap/stack I would use it for everything including boot
> code.
>
> Right now it uses much the same system for addressing
> structures as NASM does,
> which is a kludge at best.  I want this:
>
> lea eax, MyStruct.myFavoriteMember;
>
> But if we do this:
>
> lea eax, MyStruct.mySubStruct.myMember;
>
> I know it will fail.  Why?  Because we're accessing a
> struct, and then getting a
> member from that struct to get another struct.  We need to
> first get
> mySubStruct, and THEN we can get myMember.  It needs two
> lea's.
>
> lea eax, MyStruct.mySubStruct;
> lea eax, eax.myMember;
>
> That would make programming assembler in D unbelievably
> cool.  Unfortunately, I
> know about zilch about the D implementation code.
>
> My only other gripe is the lack of assembler coloring in
> any D editors I've
> encountered.
>
>