[Libre-soc-dev] gigabit router design decisions
programmerjake at gmail.com
Tue Nov 2 04:09:37 GMT 2021
On Mon, Nov 1, 2021, 18:09 Jacob Lifshay <programmerjake at gmail.com> wrote:
> On Mon, Nov 1, 2021, 17:52 <bugzilla-daemon at libre-soc.org> wrote:
>> --- Comment #11 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
>> (In reply to Jacob Lifshay from comment #10)
>> > we could have the fetch pipe fetch ahead...execution could still always
>> > stall rather than speculating -- all that would happen is it will have
>> > fetched the correct target instead of always obliviously fetching the
>> > instructions immediately following the branch so a branch doesn't always
>> > require flushing the entire fetch pipe... that will speed the whole cpu
>> > by probably 50%.
>> this is an optimistion, and it is scope-creep.
Since iirc you were complaining about words with no code:
If I went ahead and implemented a whole fetch/decode pipe (would just need
the mmu/cache-fetch pipe to fit into it), would you be ok with it having a
branch predictor then?
I have one partially completed for reg_rename_demo_cpu and it can be
adapted to work with the no-speculation in-order execution design you
More information about the Libre-soc-dev