DOS ain't dead

I have a boot sector (not MBR) written in assembler.

It works with various assemblers, and you can see it here:

https://sourceforge.net/p/pdos/gitcode/ci/master/tree/src/pbootsec.asm

I want to use the freely-available MASM 5.1. But I found it doesn't have the "tiny" memory model.

What are the implications for a bootloader if I am forced to use "small"? It didn't work at runtime when I used "small", although it did build. I was testing by booting from a floppy image using qemu.

What changes would be advised for this code so that it can work with MASM 5.1 as well as the other assemblers that it already works with?

Thanks. Paul.

> I have a boot sector (not MBR) written in assembler.
>
> It works with various assemblers, and you can see it here:
>
> https://sourceforge.net/p/pdos/gitcode/ci/master/tree/src/pbootsec.asm
>
> I want to use the freely-available MASM 5.1. But I found it doesn't have
> the "tiny" memory model.
>
> What are the implications for a bootloader if I am forced to use "small"?
> It didn't work at runtime when I used "small", although it did build. I was
> testing by booting from a floppy image using qemu.
>
> What changes would be advised for this code so that it can work with MASM
> 5.1 as well as the other assemblers that it already works with?
>
> Thanks. Paul.

Isn't it just small with the need to run "exe2bin", as far as masm is concerned?

> Isn't it just small with the need to run "exe2bin", as far as masm is
> concerned?

I don't know, but normally tiny and small are different in that
ds points to DGROUP instead of TEXT. But I don't know what
the implications of that are and how to get around that.

> > Isn't it just small with the need to run "exe2bin", as far as masm is
> > concerned?
>
> I don't know, but normally tiny and small are different in that
> ds points to DGROUP instead of TEXT. But I don't know what
> the implications of that are and how to get around that.

The second answer here stackoverflow suggests it's possible.

> The second answer here
> stackoverflow
> suggests it's possible.

Thanks - those manual segment registers fixed the problem and I have committed a fix to my source.