Re: [FYI] Inlining support, rough patch

[Tom Tromey <tromey@redhat.com>]

Resurrecting an old thread...

Mark> It's the ID construction that I'm worried about.  It is the very core
Mark> of the unwinding code.  I really think your diff violates the most
Mark> fundamental principle of this bit of code and in that way, makes it
Mark> much harder to understand it.

Daniel> I don't understand what you mean when you say this makes the generic
Daniel> code any harder to understand.  Can you point to lines for me?

[...]

Daniel> If you can think of a way to do this that doesn't involve complicating
Daniel> the generic unwind machinery - exactly what we're both trying to avoid
Daniel> - I'll give it another shot.

Mark, could you answer Daniel's questions?  This patch has been in
limbo since last July.  I'd like to at least know what needs to be
done to move forward on this.

FWIW, we're shipping this in Archer.  I think other organizations are
shipping it as well.  Debugging inlined functions nicely is a
frequently requested feature; I answer questions about it on irc at
least once a week.

Tom