updated repo for publishing (#74)

This commit is contained in:
Jean-Philippe Bossuat
2025-08-17 14:57:39 +02:00
committed by GitHub
parent 0be569eca0
commit 62eb87cc07
244 changed files with 374 additions and 539 deletions

View File

@@ -0,0 +1,306 @@
use poulpy_backend::hal::{
api::{
DataViewMut, ScratchAvailable, TakeVecZnxDft, VecZnxBigAddSmallInplace, VecZnxBigNormalize, VecZnxBigNormalizeTmpBytes,
VecZnxDftAllocBytes, VecZnxDftFromVecZnx, VecZnxDftToVecZnxBigConsume, VmpApply, VmpApplyAdd, VmpApplyTmpBytes, ZnxInfos,
},
layouts::{Backend, DataMut, DataRef, Module, Scratch, VecZnx, VecZnxBig, VecZnxDft, VmpPMat},
};
use crate::layouts::{GLWECiphertext, Infos, prepared::GGLWESwitchingKeyPrepared};
impl GLWECiphertext<Vec<u8>> {
#[allow(clippy::too_many_arguments)]
pub fn keyswitch_scratch_space<B: Backend>(
module: &Module<B>,
n: usize,
basek: usize,
k_out: usize,
k_in: usize,
k_ksk: usize,
digits: usize,
rank_in: usize,
rank_out: usize,
) -> usize
where
Module<B>: VecZnxDftAllocBytes + VmpApplyTmpBytes + VecZnxBigNormalizeTmpBytes,
{
let in_size: usize = k_in.div_ceil(basek).div_ceil(digits);
let out_size: usize = k_out.div_ceil(basek);
let ksk_size: usize = k_ksk.div_ceil(basek);
let res_dft: usize = module.vec_znx_dft_alloc_bytes(n, rank_out + 1, ksk_size); // TODO OPTIMIZE
let ai_dft: usize = module.vec_znx_dft_alloc_bytes(n, rank_in, in_size);
let vmp: usize = module.vmp_apply_tmp_bytes(
n,
out_size,
in_size,
in_size,
rank_in,
rank_out + 1,
ksk_size,
) + module.vec_znx_dft_alloc_bytes(n, rank_in, in_size);
let normalize: usize = module.vec_znx_big_normalize_tmp_bytes(n);
res_dft + ((ai_dft + vmp) | normalize)
}
pub fn keyswitch_inplace_scratch_space<B: Backend>(
module: &Module<B>,
n: usize,
basek: usize,
k_out: usize,
k_ksk: usize,
digits: usize,
rank: usize,
) -> usize
where
Module<B>: VecZnxDftAllocBytes + VmpApplyTmpBytes + VecZnxBigNormalizeTmpBytes,
{
Self::keyswitch_scratch_space(module, n, basek, k_out, k_out, k_ksk, digits, rank, rank)
}
}
impl<DataSelf: DataRef> GLWECiphertext<DataSelf> {
#[allow(dead_code)]
pub(crate) fn assert_keyswitch<B: Backend, DataLhs, DataRhs>(
&self,
module: &Module<B>,
lhs: &GLWECiphertext<DataLhs>,
rhs: &GGLWESwitchingKeyPrepared<DataRhs, B>,
scratch: &Scratch<B>,
) where
DataLhs: DataRef,
DataRhs: DataRef,
Module<B>: VecZnxDftAllocBytes + VmpApplyTmpBytes + VecZnxBigNormalizeTmpBytes,
Scratch<B>: ScratchAvailable,
{
let basek: usize = self.basek();
assert_eq!(
lhs.rank(),
rhs.rank_in(),
"lhs.rank(): {} != rhs.rank_in(): {}",
lhs.rank(),
rhs.rank_in()
);
assert_eq!(
self.rank(),
rhs.rank_out(),
"self.rank(): {} != rhs.rank_out(): {}",
self.rank(),
rhs.rank_out()
);
assert_eq!(self.basek(), basek);
assert_eq!(lhs.basek(), basek);
assert_eq!(rhs.n(), self.n());
assert_eq!(lhs.n(), self.n());
assert!(
scratch.available()
>= GLWECiphertext::keyswitch_scratch_space(
module,
self.n(),
self.basek(),
self.k(),
lhs.k(),
rhs.k(),
rhs.digits(),
rhs.rank_in(),
rhs.rank_out(),
),
"scratch.available()={} < GLWECiphertext::keyswitch_scratch_space(
module,
self.basek(),
self.k(),
lhs.k(),
rhs.k(),
rhs.digits(),
rhs.rank_in(),
rhs.rank_out(),
)={}",
scratch.available(),
GLWECiphertext::keyswitch_scratch_space(
module,
self.n(),
self.basek(),
self.k(),
lhs.k(),
rhs.k(),
rhs.digits(),
rhs.rank_in(),
rhs.rank_out(),
)
);
}
}
impl<DataSelf: DataMut> GLWECiphertext<DataSelf> {
pub fn keyswitch<DataLhs: DataRef, DataRhs: DataRef, B: Backend>(
&mut self,
module: &Module<B>,
lhs: &GLWECiphertext<DataLhs>,
rhs: &GGLWESwitchingKeyPrepared<DataRhs, B>,
scratch: &mut Scratch<B>,
) where
Module<B>: VecZnxDftAllocBytes
+ VmpApplyTmpBytes
+ VecZnxBigNormalizeTmpBytes
+ VmpApplyTmpBytes
+ VmpApply<B>
+ VmpApplyAdd<B>
+ VecZnxDftFromVecZnx<B>
+ VecZnxDftToVecZnxBigConsume<B>
+ VecZnxBigAddSmallInplace<B>
+ VecZnxBigNormalize<B>,
Scratch<B>: ScratchAvailable + TakeVecZnxDft<B>,
{
#[cfg(debug_assertions)]
{
self.assert_keyswitch(module, lhs, rhs, scratch);
}
let (res_dft, scratch1) = scratch.take_vec_znx_dft(self.n(), self.cols(), rhs.size()); // Todo optimise
let res_big: VecZnxBig<_, B> = lhs.keyswitch_internal(module, res_dft, rhs, scratch1);
(0..self.cols()).for_each(|i| {
module.vec_znx_big_normalize(self.basek(), &mut self.data, i, &res_big, i, scratch1);
})
}
pub fn keyswitch_inplace<DataRhs: DataRef, B: Backend>(
&mut self,
module: &Module<B>,
rhs: &GGLWESwitchingKeyPrepared<DataRhs, B>,
scratch: &mut Scratch<B>,
) where
Module<B>: VecZnxDftAllocBytes
+ VmpApplyTmpBytes
+ VecZnxBigNormalizeTmpBytes
+ VmpApplyTmpBytes
+ VmpApply<B>
+ VmpApplyAdd<B>
+ VecZnxDftFromVecZnx<B>
+ VecZnxDftToVecZnxBigConsume<B>
+ VecZnxBigAddSmallInplace<B>
+ VecZnxBigNormalize<B>,
Scratch<B>: ScratchAvailable + TakeVecZnxDft<B>,
{
unsafe {
let self_ptr: *mut GLWECiphertext<DataSelf> = self as *mut GLWECiphertext<DataSelf>;
self.keyswitch(module, &*self_ptr, rhs, scratch);
}
}
}
impl<D: DataRef> GLWECiphertext<D> {
pub(crate) fn keyswitch_internal<B: Backend, DataRes, DataKey>(
&self,
module: &Module<B>,
res_dft: VecZnxDft<DataRes, B>,
rhs: &GGLWESwitchingKeyPrepared<DataKey, B>,
scratch: &mut Scratch<B>,
) -> VecZnxBig<DataRes, B>
where
DataRes: DataMut,
DataKey: DataRef,
Module<B>: VecZnxDftAllocBytes
+ VmpApplyTmpBytes
+ VecZnxBigNormalizeTmpBytes
+ VmpApplyTmpBytes
+ VmpApply<B>
+ VmpApplyAdd<B>
+ VecZnxDftFromVecZnx<B>
+ VecZnxDftToVecZnxBigConsume<B>
+ VecZnxBigAddSmallInplace<B>
+ VecZnxBigNormalize<B>,
Scratch<B>: TakeVecZnxDft<B>,
{
if rhs.digits() == 1 {
return keyswitch_vmp_one_digit(module, res_dft, &self.data, &rhs.key.data, scratch);
}
keyswitch_vmp_multiple_digits(
module,
res_dft,
&self.data,
&rhs.key.data,
rhs.digits(),
scratch,
)
}
}
fn keyswitch_vmp_one_digit<B: Backend, DataRes, DataIn, DataVmp>(
module: &Module<B>,
mut res_dft: VecZnxDft<DataRes, B>,
a: &VecZnx<DataIn>,
mat: &VmpPMat<DataVmp, B>,
scratch: &mut Scratch<B>,
) -> VecZnxBig<DataRes, B>
where
DataRes: DataMut,
DataIn: DataRef,
DataVmp: DataRef,
Module<B>:
VecZnxDftAllocBytes + VecZnxDftFromVecZnx<B> + VmpApply<B> + VecZnxDftToVecZnxBigConsume<B> + VecZnxBigAddSmallInplace<B>,
Scratch<B>: TakeVecZnxDft<B>,
{
let cols: usize = a.cols();
let (mut ai_dft, scratch1) = scratch.take_vec_znx_dft(a.n(), cols - 1, a.size());
(0..cols - 1).for_each(|col_i| {
module.vec_znx_dft_from_vec_znx(1, 0, &mut ai_dft, col_i, a, col_i + 1);
});
module.vmp_apply(&mut res_dft, &ai_dft, mat, scratch1);
let mut res_big: VecZnxBig<DataRes, B> = module.vec_znx_dft_to_vec_znx_big_consume(res_dft);
module.vec_znx_big_add_small_inplace(&mut res_big, 0, a, 0);
res_big
}
fn keyswitch_vmp_multiple_digits<B: Backend, DataRes, DataIn, DataVmp>(
module: &Module<B>,
mut res_dft: VecZnxDft<DataRes, B>,
a: &VecZnx<DataIn>,
mat: &VmpPMat<DataVmp, B>,
digits: usize,
scratch: &mut Scratch<B>,
) -> VecZnxBig<DataRes, B>
where
DataRes: DataMut,
DataIn: DataRef,
DataVmp: DataRef,
Module<B>: VecZnxDftAllocBytes
+ VecZnxDftFromVecZnx<B>
+ VmpApply<B>
+ VmpApplyAdd<B>
+ VecZnxDftToVecZnxBigConsume<B>
+ VecZnxBigAddSmallInplace<B>,
Scratch<B>: TakeVecZnxDft<B>,
{
let cols: usize = a.cols();
let size: usize = a.size();
let (mut ai_dft, scratch1) = scratch.take_vec_znx_dft(a.n(), cols - 1, size.div_ceil(digits));
ai_dft.data_mut().fill(0);
(0..digits).for_each(|di| {
ai_dft.set_size((size + di) / digits);
// Small optimization for digits > 2
// VMP produce some error e, and since we aggregate vmp * 2^{di * B}, then
// we also aggregate ei * 2^{di * B}, with the largest error being ei * 2^{(digits-1) * B}.
// As such we can ignore the last digits-2 limbs safely of the sum of vmp products.
// It is possible to further ignore the last digits-1 limbs, but this introduce
// ~0.5 to 1 bit of additional noise, and thus not chosen here to ensure that the same
// noise is kept with respect to the ideal functionality.
res_dft.set_size(mat.size() - ((digits - di) as isize - 2).max(0) as usize);
(0..cols - 1).for_each(|col_i| {
module.vec_znx_dft_from_vec_znx(digits, digits - di - 1, &mut ai_dft, col_i, a, col_i + 1);
});
if di == 0 {
module.vmp_apply(&mut res_dft, &ai_dft, mat, scratch1);
} else {
module.vmp_apply_add(&mut res_dft, &ai_dft, mat, di, scratch1);
}
});
res_dft.set_size(res_dft.max_size());
let mut res_big: VecZnxBig<DataRes, B> = module.vec_znx_dft_to_vec_znx_big_consume(res_dft);
module.vec_znx_big_add_small_inplace(&mut res_big, 0, a, 0);
res_big
}