Re: [PATCH v6 07/13] arm64: add basic pointer authentication support

[Mark Rutland]

On Fri, Jan 04, 2019 at 10:33:40AM +0100, Pavel Machek wrote:
> On Fri 2019-01-04 09:21:30, Marc Zyngier wrote:
> > On 03/01/2019 20:29, Pavel Machek wrote:
> > > On Fri 2018-12-07 18:39:25, Kristina Martsenko wrote:
> > >> From: Mark Rutland <mark.rutland@xxxxxxx>
> > >> 
> > >> This patch adds basic support for pointer authentication,
> > >> allowing userspace to make use of APIAKey, APIBKey, APDAKey,
> > >> APDBKey, and APGAKey. The kernel maintains key values for each
> > >> process (shared by all threads within), which are initialised to
> > >> random values at exec()
> > > time.
> > > 
> > > ...
> > > 
> > >> +/* + * We give each process its own keys, which are shared by
> > >> all threads. The keys + * are inherited upon fork(), and
> > >> reinitialised upon exec*(). + */ +struct ptrauth_keys { +	struct
> > >> ptrauth_key apia; +	struct ptrauth_key apib; +	struct ptrauth_key
> > >> apda; +	struct ptrauth_key apdb; +	struct ptrauth_key apga; +};
> > > 
> > > intstruction_a, data_a, generic_a? Should be easier to understand
> > > than "apdb" ...
> > 
> > ... until you realize that these names do match the documentation,
> > which makes it even easier to understand how the code uses the
> > architecture.
> 
> See how not even the commit log matches the documentation then?

The commit message exactly matches the documentation, as it refers to:

	APIAKey, APIBKey, APDAKey, APDBKey, and APGAKey

... which are the architected names for those registers, in all the
documentation.

Searching "apga" in the ARM ARM finds all of the relevant information on
APGAKey_EL1. Searching "generic_a" finds precisely nothing, as it's a
term which you invented, that no-one else has previously used.

Likewise for the other key names.

> Naming something "apdb" is just bad... Just because the documentation
> is evil does not mean it should be followed...

It is in no way evil to use the documented names for things.

It is unhelpful to make up terminology that no-one else uses.

Mark.