[Libre-soc-dev] memcpy optimization
Luke Kenneth Casson Leighton
lkcl at lkcl.net
Sun Dec 13 19:02:41 GMT 2020
On 12/13/20, Richard Wilbur <richard.wilbur at gmail.com> wrote:
>> On Dec 13, 2020, at 11:08, Luke Kenneth Casson Leighton <lkcl at lkcl.net>
>> On 12/13/20, Richard Wilbur <richard.wilbur at gmail.com> wrote:
>>> Why is memcpy still doing the vector flag 0 bytes (vseq.vi)?
>> because i forgot to delete it?
> No problem. I was trying to figure out if it was there to fill a time slot
> á la a RISC processor of yore. Thankfully not.
> I think I understand a little bit about this architecture—and from what I
> can see, I like it!
cool. it.. takes getting used to. byte-level write mask granularity
is pretty normal on memory but certainly not on a register file.
hardware for-loops pausing the PC and running a Sub-PC? simple to
conceptualise, you wonder why the hell nobody's thought in those terms
before, and then you see the implementation details needed for a high
performance design and start screaming silently in shock :)
> Some very useful things done in one opcode that not so
> long ago required loops!
that's the idea. and to keep it paralleliseable at the hardware end.
> Brava and bravo to the team!
>> i need to find some puppies and fluffy bunnies to play with and de-stress.
> Great idea!
> Libre-soc-dev mailing list
> Libre-soc-dev at lists.libre-soc.org
More information about the Libre-soc-dev