Re: [PATCH 2/3] Introduce <pointer_guard.h>, extracted from <sysdep.h>

[Florian Weimer <fweimer@redhat.com>]

* Adhemerval Zanella Netto:

> On 13/10/22 09:20, Florian Weimer via Libc-alpha wrote:
>> This allows us to define a generic no-op version of PTR_MANGLE and
>> PTR_DEMANGLE.  In the future, we can use PTR_MANGLE and PTR_DEMANGLE
>> unconditionally in C sources, avoiding an unintended loss of hardening
>> due to missing include files or unlucky header inclusion ordering.
>
> Could we also improve the generic implementation to always XOR with the
> pointer guard and move it to be a proper static inline function?  I think
> we can then remove a lot of boilerplate code each architecture need to
> have.

I think the assembler implementation in part aims to avoid leaving the
plain guard value behind in a register, at least on some targets.

>> In i386 and x86_64, we can avoid a <tls.h> dependency in the C
>> code by using the computed constant from <tcb-offsets.h>.  <sysdep.h>
>> no longer includes these definitions, so there is no cyclic dependency
>> anymore when computing the <tcb-offsets.h> constants.
>
> LGTM, thanks.
>
> Reviewed-by: Adhemerval Zanella  <adhemerval.zanella@linaro.org>

Thanks,
Florian