[Libre-soc-dev] [RFC] SVP64 on branch instructions

[lkcl]

On August 3, 2021 6:55:21 PM UTC, Luke Kenneth Casson Leighton 
<lkcl at lkcl.net> wrote:

>... you get the general idea, though: that with sz and SNZ there's a
>way for predicate masks to interact with the CR Vector, to create
>Vec-AND and VEC-OR behaviour that, at the same time, still allows CTR
>the option of counting masked-in elements or all elements.

sigh.

we've run out of bits, and i have a feeling that it is more useful to have the option of updating the CR field being tested, taking predicate masks into account, than it is say to keep the Absolute Address functionality of branch.

AA is something that is only used in Hypervisor mode, for interrupt tables or OS source, and is otherwise very much wasted in userspace.

normally it is a Hard Rule that under no circumstances should SVP64 alter the base operation.  this so that when talking about it, and advocating it, we may state, plainly "base. loop. simple".

the moment the word "except" has to go into that sentence, it will make people nervous when it cones to adoption.

in this particular case however the entire Branch has to be *replaced*.

thoughts?

l.