That dwarf2read patch i just submitted

[Daniel Berlin]

Kevin Buettner <kevinb@cygnus.com> writes:

> On Oct 24,  3:30pm, Stan Shebs wrote:
> 
> > > > I also modified every line of the patch, literally.
> > > 
> > > It is true that you modified many of the lines, but you didn't touch
> > > *all* of them.
> > 
> > If it were me making the decision, I would not require the copyright
> > assignment for this patch.  The original contribution is small, it's
> > been mostly rewritten, it's not a significant new algorithm, and the
> > contributor seems to have disappeared.
> 
> I am in agreement regarding the signifcance of the algorithm.  Also,
> even if we were to somehow do a clean room implementation, it would
> likely look a lot like Brian Nettleton's patch as adapted by Daniel
> Berlin.
> 

Unless I rewrote the dwarf2 reader, you are correct. There actually is
no other good way to add .debug_loc support than what he did.
It makes no sense to do it any other way, really.
> Assuming that we come to agree that patch can go in, I have the
> following questions:
> 
>  - How would you structure the ChangeLog entry?  I feel very strongly
>    that the original contributor should given credit in some form.

Same here. 
> 
>  - Brian submitted a number of other patches from the same time frame
>    that have remained dormant.  Assuming that they are technically
>    sound, should we attempt to adapt and commit the rest of these
>    patches too?

Dunno. I looked at them when I looked at the patch i cleaned up, but
none seemed all that useful, mainly fixes for some things ADA uses.
.debug_loc support is something that all compilers can benefit
from. It can give you a very big size savings. I believe my thinking
was that the GNAT guys, if they ever got around to copyright
assignment and submitting patches to the mainline, would have
the same type of fixes brian submitted, and we'd get around to
integrating them then.
> 
> Kevin