[RFC] sketch of an unwinder cache interface for elfutils libdwfl

[Frank Ch. Eigler]

Hi -

> [...]
> If there's something we can do here on the glibc side, please let us
> know.  I think by default, dlclose needs to unmap the address space, so
> options might be somewhat limited.  But for debugging purposes, I've
> long since wanted a mode where dlclose keeps the address space reserved,
> so that it becomes easier to diagnose dlclose misuse.

Silly question, but would remapping that area with a /dev/null
PROT_NONE be compliant with tradition?

- FChE