Re: Boot code rewritten for GAS

Zack Weinberg (zack@bitmover.com)
Mon, 02 Aug 1999 09:41:08 -0700

Jeff Garzik wrote:
>
> So we have
> as86 - works, but poor docs and poor interface (raw binary code?? :))
> nasm - portable, maintained, has docs, but an additional dependency
> gas - requires bleeding edge version, strange 16-bit syntax
>
> IMHO nasm seems the best choice. Its 16-bit support will be familiar to
> many, and also has been stable for longer than gas's 16-bit support has
> existed.

The existing boot code works. It practically never needs to be
changed. Therefore, in my opinion, the poor docs and missing features
of as86 do not matter. Replacing a tool everyone already has (as86)
with a tool most people do not have (NASM) would only make sense if it
offered substantial benefits to maintainability - which it doesn't.

Changing to gas *does* make sense because it reduces the number of
different programs needed to compile the kernel. The problem with
that is the bleeding edge version. It might be reasonable to ship two
versions of the boot code and use the GAS one if a sufficiently
up-to-date assembler is found.

zw

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/