* Linus Torvalds <torva...@linux-foundation.org> wrote:
> On Mon, Aug 17, 2015 at 3:59 AM, Denys Vlasenko <dvlas...@redhat.com> wrote:
> >
> > I have it running linux 2.6.20 and busybox here:
> >
> > http://busybox.net/live_bbox/live_bbox.html
> >
> > (or rather, *you* will have it running linux 2.6.20 inside your browser,
> > after you click on that link)
>
> Heh. I'm not sure that's a very useful thing, but if somebody can make a kvm
> image or something with an old distribution that is known to work with FPU
> emulation, maybe we should verify that the current code at least works.
>
> Because even if we decide that just deleting it is the right thing for
> long-term
> maintainability, it would be better if we delete it in a state where it is
> known
> to work about as well as it ever did. So that *if* we have to resurrect it,
> we
> know that it at least was working at the point where it was deleted.
>
> I hate deleting code because it got broken. In contrast, I don't mind
> deleting
> code that no longer makes sense to maintain. The two are supposed to be very
> different things.
Absolutely agreed - it was my other motivator to try to fix math-emu in the FPU
series.
Another example is the low level assembly code which we are converting to C
code,
we (tried to) fix and deobfuscate and fix it before moving to C for similar
reasons.
Thanks,
Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
