Re: [PATCH v2 14/28] arm64/sve: Backend logic for setting the vector length

[Alan Hayward <Alan.Hayward@arm.com>]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="utf-8", Size: 1314 bytes --]


> On 21 Sep 2017, at 12:19, Dave Martin <Dave.Martin@arm.com> wrote:
> 
> On Wed, Sep 20, 2017 at 06:08:21PM +0000, Alan Hayward wrote:
>> 
>>> On 20 Sep 2017, at 12:09, Dave Martin <dave.martin@foss.arm.com> wrote:
> 
> [...]
> 
>>>> Given, sve_set_vector_length is called when setting the vector length in
>>>> PTRACE_SETREGSET, it looks to me like if you set VL to a value that’s not
>>>> supported by the hardware, then it’s going to round down to the previous value.
>>>> Is that correct? I’m not sure if that’s explained in the docs?
>>> 
>>> Does this cover it?
>>> 
>>> "On success, the calling thread's vector length is changed to the
>>> largest value supported by the system that is less than or equal to vl."
>>> 
>>> (For ptrace, I just cross-reference the PR_SVE_SET_VL behaviour, above.)
>> 
>> For ptrace is it worth mentioning user should do a GET after a SET to confirm
>> what VL value was actually set?
> 
> This seems worth a clarification -- I'd thought this was already
> mentioned, but it isn't.
> 
> How about:
> 
>  The caller must make a further GETREGSET call if it needs to know what VL is
>  actually set by SETREGSET, unless is it known in advance that the requested
>  VL is supported.
> 

Looks good to me.


Alan.\x16º&ÖëzÛ«ŸŽ´Ó‰b²Ö«r\x18\x1d