Integrate PoL into consensus #721
Conversation
|
Did you try |
zeegomo
force-pushed
the
add-cl-crate
branch
from
49c16cb
to
2192c8b
Compare
| @@ -29,6 +29,8 @@ jobs: | |||
| profile: minimal | |||
There was a problem hiding this comment.
Not sure if its worth it, but we may have a separated workflow for risc0 if needed. Wdyt @bacv ?
| let res = tokio::task::spawn_blocking(move || { | ||
| Risc0LeaderProof::build( | ||
| public_inputs, | ||
| LeaderPrivate { | ||
| input, | ||
| input_cm_path, | ||
| }, | ||
| ) | ||
| }) | ||
| .await; |
There was a problem hiding this comment.
Is this a heavy operation? I don't get the spawn_blocking to immediately await.
is this the proving part?
There was a problem hiding this comment.
yes, this is heavy and cpu intenstive (i.e. does not yield).spawn_blocking uses a dedicated thread for this to avoid blocking other tasks
Verifying is cheap right? |
yes, verifying takes ~13ms on my laptop |
|
Apparently risc0 does not work on Windows. How would you be with dropping that for now? |
It doesnt work here or risc0 itself does not support windows? The later would be a big problem imo |
|
I'm afraid it's the latter but:
|
There should be no problem in moving away from windows temporary. But whatever we choose should be available widely across OSs. I'll bubble up this in our channels. |
zeegomo
force-pushed
the
add-cl-crate
branch
from
8f8348a
to
6661b67
Compare
zeegomo
force-pushed
the
add-cl-crate
branch
from
b9b9fc7
to
d75c0e8
Compare
Unfortunately, testing now takes a lot longer since proving PoL on M2 takes ~10s, not sure about CI