Re: [PATCH v2 03/14] x86/shstk: Introduce Supervisor Shadow Stack support

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 29/05/2020 12:59, Jan Beulich wrote:
> On 28.05.2020 20:10, Andrew Cooper wrote:
>> On 28/05/2020 11:25, Jan Beulich wrote:
>>> On 27.05.2020 21:18, Andrew Cooper wrote:
>>>> --- a/xen/arch/x86/Kconfig
>>>> +++ b/xen/arch/x86/Kconfig
>>>> @@ -34,6 +34,10 @@ config ARCH_DEFCONFIG
>>>>  config INDIRECT_THUNK
>>>>    def_bool $(cc-option,-mindirect-branch-register)
>>>>  
>>>> +config HAS_AS_CET
>>>> +  # binutils >= 2.29 and LLVM >= 7
>>>> +  def_bool $(as-instr,wrssq %rax$(comma)0;setssbsy;endbr64)
>>> So you put me in a really awkward position: I'd really like to see
>>> this series go in for 4.14, yet I've previously indicated I want the
>>> underlying concept to first be agreed upon, before any uses get
>>> introduced.
>> There are already users.  One of them is even in context.
> Hmm, indeed. I clearly didn't notice this aspect when reviewing
> Anthony's series.
>
>> I don't see that there is anything open for dispute in the first place. 
>> Being able to do exactly this was a one key driving factor to a newer
>> Kconfig, because it is superior mechanism to the ad-hoc mess we had
>> previously (not to mention, a vast detriment to build time).
> This "key driving factor" was presumably from your perspective.
> Could you point me to a discussion (and resulting decision) that
> this is an explicit goal of that work? I don't recall any, and
> hence I also don't recall having been given a chance in influence
> the direction, decision, and overall outcome.

It took up a large chunk of the build system design session in Chicago.

>
> In the interest of getting this series in for 4.14, and on the
> assumption that you're willing to have a discussion on the
> direction wrt storing tool chain capabilities in .config before
> any further uses get added (and with the potential need to undo
> the ones we have / gain here)
> Reviewed-by: Jan Beulich <jbeulich@xxxxxxxx>

Thanks.