Re: Explicit pointer provenance APIs for CHERI

--------
Jessica Clarke writes:

> On 24 Jun 2026, at 17:34, Poul-Henning Kamp <[email protected]> wrote:

> > As long as there is a cheri(7) manualpage to get people started...
>
> Why?

Because the way we talk about this is "CHERI this" and "CHERI that" so
like it or not, "CHERI" is the "industry standard" term for this stuff
and hiding the consequences away in N different manpages makes no sense,
unless you somewhere tell people where to look.

Poul-Henning

PS:

I find your "mmu" strawman argument entirely unconvincing, because
we have plenty of "abstract concept" manual pages already.

See for instance intro(7) which points to:

           arch(7)          supported CPU architectures and platforms

           c(7)             the C programming language

           clocks(7)        system timekeeping clocks available in FreeBSD

           crypto(7)        cryptographic algorithms provided by OpenCrypto in
                            FreeBSD

           development(7)   development introduction to FreeBSD

           firewall(7)      simple firewalls under FreeBSD

           networking(7)    network connection quickstart guide

           release(7)       layout of FreeBSD releases and snapshots

           ports(7)         introduction to the ports infrastructure of
                            FreeBSD

           security(7)      security features available in FreeBSD

           stats(7)         statistics utilities available in FreeBSD

           tests(7)         introduction to the FreeBSD Test Suite

           tracing(7)       introduction to tracing and performance monitoring
                            facilities

           tuning(7)        general advice on tuning FreeBSD


And given that we have:

	pmap(9) - machine-dependent portion of virtual memory subsystem

and

	vm_map(9) - virtual address space portion of virtual memory subsystem

I think it would be fair to argue that vm(7) and vm(9) are both missing.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
[email protected]         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.