feat(integer): add is_even/is_odd functions #1488
Conversation
| ); | ||
| } | ||
| CudaBootstrappingKey::MultiBit(d_multibit_bsk) => { | ||
| apply_univariate_lut_kb_async( |
There was a problem hiding this comment.
is there also bootstrap then keyswitch ? I don't know the cuda code
There was a problem hiding this comment.
I think they only do ks_pbs not pbs_ks
There was a problem hiding this comment.
that's weird it seems there is a PBS order in the struct
There was a problem hiding this comment.
Yep its in the struct, but its not really used, a quick grep showed me that the inly place where its use is in the gpu's create_trivial_radix function
There was a problem hiding this comment.
Well then we likely need a panic in the GPU key decompression to avoid running stuff if we have the wrong pbs order
There was a problem hiding this comment.
Can be a different pr to check the logic with the gpu team
| if !ct.block_carries_are_empty() { | ||
| if ct | ||
| .blocks() | ||
| .iter() | ||
| .any(|block| self.key.max_degree.validate(block.degree).is_err()) |
There was a problem hiding this comment.
any reason for this change ?
There was a problem hiding this comment.
its related to this comment #1488 (comment)
this function does not care about carries, so a smart variant does no make sense unless there is some carry cleaning (or we accept to put a #[allow] of some sort
There was a problem hiding this comment.
ah right, I think we have some unused &mut on smart variants where cleaning is not done we end up with an allow IIRC
tmontaigu
force-pushed
the
tm/is-even-odd
branch
from
0607889 to
cf3c28b
Compare
|
good to squash, I'll approve the squashed commits |
These ones are pretty simple and so are also directly done for GPU
tmontaigu
force-pushed
the
tm/is-even-odd
branch
from
cf3c28b to
ab461d5
Compare
Closes https://github.com/zama-ai/tfhe-rs-internal/issues/639