[coyotos-dev] Preemption again

[Jonathan S. Shapiro]

On Tue, 2008-02-05 at 07:52 -0600, Trey Boudreau wrote:
> This implies that a debugger would have to manage both the primary debug
> target and the proxy receiver, right?

Probably. Had not considered it.

> Can the proxy receiver have a
> proxy receiver?

Yes, but probably not usefully. In that design, the proxy receiver would
only get the message if the special bit in the entry cap was set.

> What manages the proxy receiver slot: the language runtime or the
> program itself?

>From the kernel perspective, the two cannot be distinguished.

> I assume the proxy receiver slot can change over the life of the
> program.  How will a debugger track this change?

Not considered yet.

As I wrote that note, I came to the provisional conclusion that the
whole thing was too complicated and the simpler (second) design was
probably better.

shap