NPTL libthread_db changes for static debugging

[Roland McGrath]

> It's actually not that general.  It delays thread debugging
> initialization until it sees a non-zero ti_tid.

Ok.  That is nonetheless pretty safe and stable, and compatible with
anything libthread_db is likely ever to do.

> > If the gdb change is a nonregressing robustness improvement with respect to
> > libthread_db behavior, might as well put it in without delay regardless, no?
> 
> Yes.  Thanks for your help; I'll check it in today.

Thanks.


Roland