wip

poulpy-core/README.md

@@ -52,8 +52,8 @@ fn main() {
 
     // Scratch space
     let mut scratch: ScratchOwned<FFT64> = ScratchOwned::alloc(
-        GLWECiphertext::encrypt_sk_scratch_space(&module, n, base2k, ct.k())
+        GLWECiphertext::encrypt_sk_tmp_bytes(&module, n, base2k, ct.k())
-            | GLWECiphertext::decrypt_scratch_space(&module, n, base2k, ct.k()),
+            | GLWECiphertext::decrypt_tmp_bytes(&module, n, base2k, ct.k()),
     );
 
     // Generate secret-key

poulpy-core/benches/external_product_glwe_fft64.rs

@@ -67,9 +67,9 @@ fn bench_external_product_glwe_fft64(c: &mut Criterion) {
         let pt_rgsw: ScalarZnx<Vec<u8>> = ScalarZnx::alloc(n.into(), 1);
 
         let mut scratch: ScratchOwned<FFT64Spqlios> = ScratchOwned::alloc(
-            GGSW::encrypt_sk_scratch_space(&module, &ggsw_layout)
+            GGSW::encrypt_sk_tmp_bytes(&module, &ggsw_layout)
-                | GLWE::encrypt_sk_scratch_space(&module, &glwe_in_layout)
+                | GLWE::encrypt_sk_tmp_bytes(&module, &glwe_in_layout)
-                | GLWE::external_product_scratch_space(&module, &glwe_out_layout, &glwe_in_layout, &ggsw_layout),
+                | GLWE::external_product_tmp_bytes(&module, &glwe_out_layout, &glwe_in_layout, &ggsw_layout),
         );
 
         let mut source_xs = Source::new([0u8; 32]);
@@ -167,9 +167,9 @@ fn bench_external_product_glwe_inplace_fft64(c: &mut Criterion) {
         let pt_rgsw: ScalarZnx<Vec<u8>> = ScalarZnx::alloc(n.into(), 1);
 
         let mut scratch: ScratchOwned<FFT64Spqlios> = ScratchOwned::alloc(
-            GGSW::encrypt_sk_scratch_space(&module, &ggsw_layout)
+            GGSW::encrypt_sk_tmp_bytes(&module, &ggsw_layout)
-                | GLWE::encrypt_sk_scratch_space(&module, &glwe_layout)
+                | GLWE::encrypt_sk_tmp_bytes(&module, &glwe_layout)
-                | GLWE::external_product_inplace_scratch_space(&module, &glwe_layout, &ggsw_layout),
+                | GLWE::external_product_inplace_tmp_bytes(&module, &glwe_layout, &ggsw_layout),
         );
 
         let mut source_xs: Source = Source::new([0u8; 32]);

poulpy-core/benches/keyswitch_glwe_fft64.rs

@@ -67,9 +67,9 @@ fn bench_keyswitch_glwe_fft64(c: &mut Criterion) {
         let mut ct_out: GLWE<Vec<u8>> = GLWE::alloc_from_infos(&glwe_out_layout);
 
         let mut scratch: ScratchOwned<FFT64Spqlios> = ScratchOwned::alloc(
-            GLWESwitchingKey::encrypt_sk_scratch_space(&module, &gglwe_atk_layout)
+            GLWESwitchingKey::encrypt_sk_tmp_bytes(&module, &gglwe_atk_layout)
-                | GLWE::encrypt_sk_scratch_space(&module, &glwe_in_layout)
+                | GLWE::encrypt_sk_tmp_bytes(&module, &glwe_in_layout)
-                | GLWE::keyswitch_scratch_space(
+                | GLWE::keyswitch_tmp_bytes(
                     &module,
                     &glwe_out_layout,
                     &glwe_in_layout,
@@ -178,9 +178,9 @@ fn bench_keyswitch_glwe_inplace_fft64(c: &mut Criterion) {
         let mut ct: GLWE<Vec<u8>> = GLWE::alloc_from_infos(&glwe_layout);
 
         let mut scratch: ScratchOwned<FFT64Spqlios> = ScratchOwned::alloc(
-            GLWESwitchingKey::encrypt_sk_scratch_space(&module, &gglwe_layout)
+            GLWESwitchingKey::encrypt_sk_tmp_bytes(&module, &gglwe_layout)
-                | GLWE::encrypt_sk_scratch_space(&module, &glwe_layout)
+                | GLWE::encrypt_sk_tmp_bytes(&module, &glwe_layout)
-                | GLWE::keyswitch_inplace_scratch_space(&module, &glwe_layout, &gglwe_layout),
+                | GLWE::keyswitch_inplace_tmp_bytes(&module, &glwe_layout, &gglwe_layout),
         );
 
         let mut source_xs: Source = Source::new([0u8; 32]);

poulpy-core/examples/encryption.rs

@@ -54,7 +54,7 @@ fn main() {
 
     // Scratch space
     let mut scratch: ScratchOwned<FFT64Spqlios> = ScratchOwned::alloc(
-        GLWE::encrypt_sk_scratch_space(&module, &glwe_ct_infos) | GLWE::decrypt_scratch_space(&module, &glwe_ct_infos),
+        GLWE::encrypt_sk_tmp_bytes(&module, &glwe_ct_infos) | GLWE::decrypt_tmp_bytes(&module, &glwe_ct_infos),
     );
 
     // Generate secret-key

poulpy-core/src/automorphism/gglwe_atk.rs

@@ -10,7 +10,7 @@ use poulpy_hal::{
 use crate::layouts::{AutomorphismKey, GGLWEInfos, GLWE, prepared::AutomorphismKeyPrepared};
 
 impl AutomorphismKey<Vec<u8>> {
-    pub fn automorphism_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn automorphism_tmp_bytes<B: Backend, OUT, IN, KEY>(
         module: &Module<B>,
         out_infos: &OUT,
         in_infos: &IN,
@@ -22,7 +22,7 @@ impl AutomorphismKey<Vec<u8>> {
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        GLWE::keyswitch_scratch_space(
+        GLWE::keyswitch_tmp_bytes(
             module,
             &out_infos.glwe_layout(),
             &in_infos.glwe_layout(),
@@ -30,13 +30,13 @@ impl AutomorphismKey<Vec<u8>> {
         )
     }
 
-    pub fn automorphism_inplace_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
+    pub fn automorphism_inplace_tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
     where
         OUT: GGLWEInfos,
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        AutomorphismKey::automorphism_scratch_space(module, out_infos, out_infos, key_infos)
+        AutomorphismKey::automorphism_tmp_bytes(module, out_infos, out_infos, key_infos)
     }
 }
 

poulpy-core/src/automorphism/ggsw_ct.rs

@@ -14,7 +14,7 @@ use crate::layouts::{
 };
 
 impl GGSW<Vec<u8>> {
-    pub fn automorphism_scratch_space<B: Backend, OUT, IN, KEY, TSK>(
+    pub fn automorphism_tmp_bytes<B: Backend, OUT, IN, KEY, TSK>(
         module: &Module<B>,
         out_infos: &OUT,
         in_infos: &IN,
@@ -31,17 +31,17 @@ impl GGSW<Vec<u8>> {
     {
         let out_size: usize = out_infos.size();
         let ci_dft: usize = module.bytes_of_vec_znx_dft((key_infos.rank_out() + 1).into(), out_size);
-        let ks_internal: usize = GLWE::keyswitch_scratch_space(
+        let ks_internal: usize = GLWE::keyswitch_tmp_bytes(
             module,
             &out_infos.glwe_layout(),
             &in_infos.glwe_layout(),
             key_infos,
         );
-        let expand: usize = GGSW::expand_row_scratch_space(module, out_infos, tsk_infos);
+        let expand: usize = GGSW::expand_row_tmp_bytes(module, out_infos, tsk_infos);
         ci_dft + (ks_internal | expand)
     }
 
-    pub fn automorphism_inplace_scratch_space<B: Backend, OUT, KEY, TSK>(
+    pub fn automorphism_inplace_tmp_bytes<B: Backend, OUT, KEY, TSK>(
         module: &Module<B>,
         out_infos: &OUT,
         key_infos: &KEY,
@@ -54,7 +54,7 @@ impl GGSW<Vec<u8>> {
         Module<B>:
             VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigBytesOf + VecZnxNormalizeTmpBytes + VecZnxBigNormalizeTmpBytes,
     {
-        GGSW::automorphism_scratch_space(module, out_infos, out_infos, key_infos, tsk_infos)
+        GGSW::automorphism_tmp_bytes(module, out_infos, out_infos, key_infos, tsk_infos)
     }
 }
 
@@ -115,7 +115,7 @@ impl<DataSelf: DataMut> GGSW<DataSelf> {
                 self.rank(),
                 tensor_key.rank_out()
             );
-            assert!(scratch.available() >= GGSW::automorphism_scratch_space(module, self, lhs, auto_key, tensor_key))
+            assert!(scratch.available() >= GGSW::automorphism_tmp_bytes(module, self, lhs, auto_key, tensor_key))
         };
 
         // Keyswitch the j-th row of the col 0

poulpy-core/src/automorphism/glwe_ct.rs

@@ -11,7 +11,7 @@ use poulpy_hal::{
 use crate::layouts::{GGLWEInfos, GLWE, GLWEInfos, LWEInfos, prepared::AutomorphismKeyPrepared};
 
 impl GLWE<Vec<u8>> {
-    pub fn automorphism_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn automorphism_tmp_bytes<B: Backend, OUT, IN, KEY>(
         module: &Module<B>,
         out_infos: &OUT,
         in_infos: &IN,
@@ -23,16 +23,16 @@ impl GLWE<Vec<u8>> {
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        Self::keyswitch_scratch_space(module, out_infos, in_infos, key_infos)
+        Self::keyswitch_tmp_bytes(module, out_infos, in_infos, key_infos)
     }
 
-    pub fn automorphism_inplace_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
+    pub fn automorphism_inplace_tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
     where
         OUT: GLWEInfos,
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        Self::keyswitch_inplace_scratch_space(module, out_infos, key_infos)
+        Self::keyswitch_inplace_tmp_bytes(module, out_infos, key_infos)
     }
 }
 

poulpy-core/src/conversion/gglwe_to_ggsw.rs

@@ -0,0 +1,279 @@
+use poulpy_hal::{
+    api::{
+        ModuleN, ScratchAvailable, ScratchTakeBasic, VecZnxBigBytesOf, VecZnxBigNormalize, VecZnxCopy, VecZnxDftAddInplace,
+        VecZnxDftApply, VecZnxDftBytesOf, VecZnxDftCopy, VecZnxIdftApplyTmpA, VecZnxNormalize, VecZnxNormalizeTmpBytes,
+        VmpApplyDftToDft, VmpApplyDftToDftAdd, VmpApplyDftToDftTmpBytes,
+    },
+    layouts::{Backend, DataMut, Module, Scratch, VmpPMat, ZnxInfos},
+};
+
+use crate::{
+    ScratchTakeCore,
+    layouts::{
+        GGLWE, GGLWEInfos, GGLWEToRef, GGSW, GGSWInfos, GGSWToMut, GLWEInfos, LWEInfos,
+        prepared::{TensorKeyPrepared, TensorKeyPreparedToRef},
+    },
+    operations::GLWEOperations,
+};
+
+impl GGLWE<Vec<u8>> {
+    pub fn from_gglw_tmp_bytes<R, A, M, BE: Backend>(module: &M, res_infos: &R, tsk_infos: &A) -> usize
+    where
+        M: GGSWFromGGLWE<BE>,
+        R: GGSWInfos,
+        A: GGLWEInfos,
+    {
+        module.ggsw_from_gglwe_tmp_bytes(res_infos, tsk_infos)
+    }
+}
+
+impl<D: DataMut> GGSW<D> {
+    pub fn from_gglwe<G, M, T, BE: Backend>(&mut self, module: &M, gglwe: &G, tsk: &T, scratch: &mut Scratch<BE>)
+    where
+        M: GGSWFromGGLWE<BE>,
+        G: GGLWEToRef,
+        T: TensorKeyPreparedToRef<BE>,
+        Scratch<BE>: ScratchTakeCore<BE>,
+    {
+        module.ggsw_from_gglwe(self, gglwe, tsk, scratch);
+    }
+}
+
+impl<BE: Backend> GGSWFromGGLWE<BE> for Module<BE> where Self: GGSWExpandRows<BE> + VecZnxCopy {}
+
+pub trait GGSWFromGGLWE<BE: Backend>
+where
+    Self: GGSWExpandRows<BE> + VecZnxCopy,
+{
+    fn ggsw_from_gglwe_tmp_bytes<R, A>(&self, res_infos: &R, tsk_infos: &A) -> usize
+    where
+        R: GGSWInfos,
+        A: GGLWEInfos,
+    {
+        self.ggsw_expand_rows_tmp_bytes(res_infos, tsk_infos)
+    }
+
+    fn ggsw_from_gglwe<R, A, T>(&self, res: &mut R, a: &A, tsk: &T, scratch: &mut Scratch<BE>)
+    where
+        R: GGSWToMut,
+        A: GGLWEToRef,
+        T: TensorKeyPreparedToRef<BE>,
+        Scratch<BE>: ScratchTakeCore<BE>,
+    {
+        let res: &mut GGSW<&mut [u8]> = &mut res.to_mut();
+        let a: &GGLWE<&[u8]> = &a.to_ref();
+        let tsk: &TensorKeyPrepared<&[u8], BE> = &tsk.to_ref();
+
+        assert_eq!(res.rank(), a.rank_out());
+        assert_eq!(res.dnum(), a.dnum());
+        assert_eq!(res.n(), self.n() as u32);
+        assert_eq!(a.n(), self.n() as u32);
+        assert_eq!(tsk.n(), self.n() as u32);
+
+        for row in 0..res.dnum().into() {
+            res.at_mut(row, 0).copy(self, &a.at(row, 0));
+        }
+
+        self.ggsw_expand_row(res, tsk, scratch);
+    }
+}
+
+impl<BE: Backend> GGSWExpandRows<BE> for Module<BE> where
+    Self: Sized
+        + ModuleN
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigBytesOf
+        + VecZnxNormalizeTmpBytes
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigBytesOf
+        + VecZnxNormalizeTmpBytes
+        + VecZnxDftApply<BE>
+        + VecZnxDftCopy<BE>
+        + VmpApplyDftToDft<BE>
+        + VmpApplyDftToDftAdd<BE>
+        + VecZnxDftAddInplace<BE>
+        + VecZnxBigNormalize<BE>
+        + VecZnxIdftApplyTmpA<BE>
+        + VecZnxNormalize<BE>
+{
+}
+
+pub(crate) trait GGSWExpandRows<BE: Backend>
+where
+    Self: Sized
+        + ModuleN
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigBytesOf
+        + VecZnxNormalizeTmpBytes
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigBytesOf
+        + VecZnxNormalizeTmpBytes
+        + VecZnxDftApply<BE>
+        + VecZnxDftCopy<BE>
+        + VmpApplyDftToDft<BE>
+        + VmpApplyDftToDftAdd<BE>
+        + VecZnxDftAddInplace<BE>
+        + VecZnxBigNormalize<BE>
+        + VecZnxIdftApplyTmpA<BE>
+        + VecZnxNormalize<BE>,
+{
+    fn ggsw_expand_rows_tmp_bytes<R, A>(&self, res_infos: &R, tsk_infos: &A) -> usize
+    where
+        R: GGSWInfos,
+        A: GGLWEInfos,
+    {
+        let tsk_size: usize = tsk_infos.k().div_ceil(tsk_infos.base2k()) as usize;
+        let size_in: usize = res_infos
+            .k()
+            .div_ceil(tsk_infos.base2k())
+            .div_ceil(tsk_infos.dsize().into()) as usize;
+
+        let tmp_dft_i: usize = self.bytes_of_vec_znx_dft((tsk_infos.rank_out() + 1).into(), tsk_size);
+        let tmp_a: usize = self.bytes_of_vec_znx_dft(1, size_in);
+        let vmp: usize = self.vmp_apply_dft_to_dft_tmp_bytes(
+            tsk_size,
+            size_in,
+            size_in,
+            (tsk_infos.rank_in()).into(),  // Verify if rank+1
+            (tsk_infos.rank_out()).into(), // Verify if rank+1
+            tsk_size,
+        );
+        let tmp_idft: usize = self.bytes_of_vec_znx_big(1, tsk_size);
+        let norm: usize = self.vec_znx_normalize_tmp_bytes();
+
+        tmp_dft_i + ((tmp_a + vmp) | (tmp_idft + norm))
+    }
+
+    fn ggsw_expand_row<R, T>(&self, res: &mut R, tsk: &T, scratch: &mut Scratch<BE>)
+    where
+        R: GGSWToMut,
+        T: TensorKeyPreparedToRef<BE>,
+        Scratch<BE>: ScratchTakeCore<BE>,
+    {
+        let res: &mut GGSW<&mut [u8]> = &mut res.to_mut();
+        let tsk: &TensorKeyPrepared<&[u8], BE> = &tsk.to_ref();
+
+        let basek_in: usize = res.base2k().into();
+        let basek_tsk: usize = tsk.base2k().into();
+
+        assert!(scratch.available() >= self.ggsw_expand_rows_tmp_bytes(res, tsk));
+
+        let rank: usize = res.rank().into();
+        let cols: usize = rank + 1;
+
+        let a_size: usize = (res.size() * basek_in).div_ceil(basek_tsk);
+
+        // Keyswitch the j-th row of the col 0
+        for row_i in 0..res.dnum().into() {
+            let a = &res.at(row_i, 0).data;
+
+            // Pre-compute DFT of (a0, a1, a2)
+            let (mut ci_dft, scratch_1) = scratch.take_vec_znx_dft(self, cols, a_size);
+
+            if basek_in == basek_tsk {
+                for i in 0..cols {
+                    self.vec_znx_dft_apply(1, 0, &mut ci_dft, i, a, i);
+                }
+            } else {
+                let (mut a_conv, scratch_2) = scratch_1.take_vec_znx(self, 1, a_size);
+                for i in 0..cols {
+                    self.vec_znx_normalize(basek_tsk, &mut a_conv, 0, basek_in, a, i, scratch_2);
+                    self.vec_znx_dft_apply(1, 0, &mut ci_dft, i, &a_conv, 0);
+                }
+            }
+
+            for col_j in 1..cols {
+                // Example for rank 3:
+                //
+                // Note: M is a vector (m, Bm, B^2m, B^3m, ...), so each column is
+                // actually composed of that many dnum and we focus on a specific row here
+                // implicitely given ci_dft.
+                //
+                // # Input
+                //
+                // col 0: (-(a0s0 + a1s1 + a2s2) + M[i], a0    , a1    , a2    )
+                // col 1: (0, 0, 0, 0)
+                // col 2: (0, 0, 0, 0)
+                // col 3: (0, 0, 0, 0)
+                //
+                // # Output
+                //
+                // col 0: (-(a0s0 + a1s1 + a2s2) + M[i], a0       , a1       , a2       )
+                // col 1: (-(b0s0 + b1s1 + b2s2)       , b0 + M[i], b1       , b2       )
+                // col 2: (-(c0s0 + c1s1 + c2s2)       , c0       , c1 + M[i], c2       )
+                // col 3: (-(d0s0 + d1s1 + d2s2)       , d0       , d1       , d2 + M[i])
+
+                let dsize: usize = tsk.dsize().into();
+
+                let (mut tmp_dft_i, scratch_2) = scratch_1.take_vec_znx_dft(self, cols, tsk.size());
+                let (mut tmp_a, scratch_3) = scratch_2.take_vec_znx_dft(self, 1, ci_dft.size().div_ceil(dsize));
+
+                {
+                    // Performs a key-switch for each combination of s[i]*s[j], i.e. for a0, a1, a2
+                    //
+                    // # Example for col=1
+                    //
+                    // a0 * (-(f0s0 + f1s1 + f1s2) + s0^2, f0, f1, f2) = (-(a0f0s0 + a0f1s1 + a0f1s2) + a0s0^2, a0f0, a0f1, a0f2)
+                    // +
+                    // a1 * (-(g0s0 + g1s1 + g1s2) + s0s1, g0, g1, g2) = (-(a1g0s0 + a1g1s1 + a1g1s2) + a1s0s1, a1g0, a1g1, a1g2)
+                    // +
+                    // a2 * (-(h0s0 + h1s1 + h1s2) + s0s2, h0, h1, h2) = (-(a2h0s0 + a2h1s1 + a2h1s2) + a2s0s2, a2h0, a2h1, a2h2)
+                    // =
+                    // (-(x0s0 + x1s1 + x2s2) + s0(a0s0 + a1s1 + a2s2), x0, x1, x2)
+                    for col_i in 1..cols {
+                        let pmat: &VmpPMat<&[u8], BE> = &tsk.at(col_i - 1, col_j - 1).key.data; // Selects Enc(s[i]s[j])
+
+                        // Extracts a[i] and multipies with Enc(s[i]s[j])
+                        for di in 0..dsize {
+                            tmp_a.set_size((ci_dft.size() + di) / dsize);
+
+                            // Small optimization for dsize > 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^{(dsize-1) * B}.
+                            // As such we can ignore the last dsize-2 limbs safely of the sum of vmp products.
+                            // It is possible to further ignore the last dsize-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.
+                            tmp_dft_i.set_size(tsk.size() - ((dsize - di) as isize - 2).max(0) as usize);
+
+                            self.vec_znx_dft_copy(dsize, dsize - 1 - di, &mut tmp_a, 0, &ci_dft, col_i);
+                            if di == 0 && col_i == 1 {
+                                self.vmp_apply_dft_to_dft(&mut tmp_dft_i, &tmp_a, pmat, scratch_3);
+                            } else {
+                                self.vmp_apply_dft_to_dft_add(&mut tmp_dft_i, &tmp_a, pmat, di, scratch_3);
+                            }
+                        }
+                    }
+                }
+
+                // Adds -(sum a[i] * s[i]) + m)  on the i-th column of tmp_idft_i
+                //
+                // (-(x0s0 + x1s1 + x2s2) + a0s0s0 + a1s0s1 + a2s0s2, x0, x1, x2)
+                // +
+                // (0, -(a0s0 + a1s1 + a2s2) + M[i], 0, 0)
+                // =
+                // (-(x0s0 + x1s1 + x2s2) + s0(a0s0 + a1s1 + a2s2), x0 -(a0s0 + a1s1 + a2s2) + M[i], x1, x2)
+                // =
+                // (-(x0s0 + x1s1 + x2s2), x0 + M[i], x1, x2)
+                self.vec_znx_dft_add_inplace(&mut tmp_dft_i, col_j, &ci_dft, 0);
+                let (mut tmp_idft, scratch_3) = scratch_2.take_vec_znx_big(self, 1, tsk.size());
+                for i in 0..cols {
+                    self.vec_znx_idft_apply_tmpa(&mut tmp_idft, 0, &mut tmp_dft_i, i);
+                    self.vec_znx_big_normalize(
+                        basek_in,
+                        &mut res.at_mut(row_i, col_j).data,
+                        i,
+                        basek_tsk,
+                        &tmp_idft,
+                        0,
+                        scratch_3,
+                    );
+                }
+            }
+        }
+    }
+}

poulpy-core/src/conversion/glwe_to_lwe.rs

@@ -10,7 +10,7 @@ use poulpy_hal::{
 use crate::layouts::{GGLWEInfos, GLWE, GLWEInfos, GLWELayout, LWE, LWEInfos, Rank, prepared::GLWEToLWESwitchingKeyPrepared};
 
 impl LWE<Vec<u8>> {
-    pub fn from_glwe_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn from_glwe_tmp_bytes<B: Backend, OUT, IN, KEY>(
         module: &Module<B>,
         lwe_infos: &OUT,
         glwe_infos: &IN,
@@ -34,7 +34,7 @@ impl LWE<Vec<u8>> {
             lwe_infos.base2k(),
             lwe_infos.k(),
             1u32.into(),
-        ) + GLWE::keyswitch_scratch_space(module, &glwe_layout, glwe_infos, key_infos)
+        ) + GLWE::keyswitch_tmp_bytes(module, &glwe_layout, glwe_infos, key_infos)
     }
 }
 

poulpy-core/src/conversion/lwe_to_glwe.rs

@@ -10,7 +10,7 @@ use poulpy_hal::{
 use crate::layouts::{GGLWEInfos, GLWE, GLWEInfos, GLWELayout, LWE, LWEInfos, prepared::LWEToGLWESwitchingKeyPrepared};
 
 impl GLWE<Vec<u8>> {
-    pub fn from_lwe_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn from_lwe_tmp_bytes<B: Backend, OUT, IN, KEY>(
         module: &Module<B>,
         glwe_infos: &OUT,
         lwe_infos: &IN,
@@ -28,7 +28,7 @@ impl GLWE<Vec<u8>> {
             lwe_infos.k().max(glwe_infos.k()),
             1u32.into(),
         );
-        let ks: usize = GLWE::keyswitch_inplace_scratch_space(module, glwe_infos, key_infos);
+        let ks: usize = GLWE::keyswitch_inplace_tmp_bytes(module, glwe_infos, key_infos);
         if lwe_infos.base2k() == key_infos.base2k() {
             ct + ks
         } else {

poulpy-core/src/conversion/mod.rs

@@ -1,2 +1,5 @@
+mod gglwe_to_ggsw;
 mod glwe_to_lwe;
 mod lwe_to_glwe;
+
+pub use gglwe_to_ggsw::*;

poulpy-core/src/decryption/glwe_ct.rs

@@ -9,7 +9,7 @@ use poulpy_hal::{
 use crate::layouts::{GLWE, GLWEInfos, GLWEPlaintext, LWEInfos, prepared::GLWESecretPrepared};
 
 impl GLWE<Vec<u8>> {
-    pub fn decrypt_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn decrypt_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GLWEInfos,
         Module<B>: VecZnxDftBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf,

poulpy-core/src/encryption/compressed/gglwe_atk.rs

@@ -13,13 +13,13 @@ use crate::{
 };
 
 impl AutomorphismKeyCompressed<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes + SvpPPolBytesOf,
     {
         assert_eq!(module.n() as u32, infos.n());
-        GLWESwitchingKeyCompressed::encrypt_sk_scratch_space(module, infos) + GLWESecret::bytes_of(infos.n(), infos.rank_out())
+        GLWESwitchingKeyCompressed::encrypt_sk_tmp_bytes(module, infos) + GLWESecret::bytes_of(infos.n(), infos.rank_out())
     }
 }
 
@@ -63,10 +63,10 @@ where
             assert_eq!(res.rank_out(), res.rank_in());
             assert_eq!(sk.rank(), res.rank_out());
             assert!(
-                scratch.available() >= AutomorphismKeyCompressed::encrypt_sk_scratch_space(self, res),
+                scratch.available() >= AutomorphismKeyCompressed::encrypt_sk_tmp_bytes(self, res),
-                "scratch.available(): {} < AutomorphismKey::encrypt_sk_scratch_space: {}",
+                "scratch.available(): {} < AutomorphismKey::encrypt_sk_tmp_bytes: {}",
                 scratch.available(),
-                AutomorphismKeyCompressed::encrypt_sk_scratch_space(self, res)
+                AutomorphismKeyCompressed::encrypt_sk_tmp_bytes(self, res)
             )
         }
 

poulpy-core/src/encryption/compressed/gglwe_ct.rs

@@ -34,12 +34,12 @@ impl<D: DataMut> GGLWECompressed<D> {
 }
 
 impl GGLWECompressed<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes,
     {
-        GGLWE::encrypt_sk_scratch_space(module, infos)
+        GGLWE::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 
@@ -106,10 +106,10 @@ where
             assert_eq!(res.n(), sk.n());
             assert_eq!(pt.n() as u32, sk.n());
             assert!(
-                scratch.available() >= GGLWECompressed::encrypt_sk_scratch_space(self, res),
+                scratch.available() >= GGLWECompressed::encrypt_sk_tmp_bytes(self, res),
-                "scratch.available: {} < GGLWECiphertext::encrypt_sk_scratch_space: {}",
+                "scratch.available: {} < GGLWECiphertext::encrypt_sk_tmp_bytes: {}",
                 scratch.available(),
-                GGLWECompressed::encrypt_sk_scratch_space(self, res)
+                GGLWECompressed::encrypt_sk_tmp_bytes(self, res)
             );
             assert!(
                 res.dnum().0 * res.dsize().0 * res.base2k().0 <= res.k().0,

poulpy-core/src/encryption/compressed/gglwe_ksk.rs

@@ -14,12 +14,12 @@ use crate::{
 };
 
 impl GLWESwitchingKeyCompressed<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes + SvpPPolBytesOf,
     {
-        (GGLWE::encrypt_sk_scratch_space(module, infos) | ScalarZnx::bytes_of(module.n(), 1))
+        (GGLWE::encrypt_sk_tmp_bytes(module, infos) | ScalarZnx::bytes_of(module.n(), 1))
             + ScalarZnx::bytes_of(module.n(), infos.rank_in().into())
             + GLWESecretPrepared::bytes_of(module, infos.rank_out())
     }
@@ -91,10 +91,10 @@ where
             assert!(sk_in.n().0 <= self.n() as u32);
             assert!(sk_out.n().0 <= self.n() as u32);
             assert!(
-                scratch.available() >= GLWESwitchingKey::encrypt_sk_scratch_space(self, res),
+                scratch.available() >= GLWESwitchingKey::encrypt_sk_tmp_bytes(self, res),
-                "scratch.available()={} < GLWESwitchingKey::encrypt_sk_scratch_space={}",
+                "scratch.available()={} < GLWESwitchingKey::encrypt_sk_tmp_bytes={}",
                 scratch.available(),
-                GLWESwitchingKey::encrypt_sk_scratch_space(self, res)
+                GLWESwitchingKey::encrypt_sk_tmp_bytes(self, res)
             )
         }
 

poulpy-core/src/encryption/compressed/gglwe_tsk.rs

@@ -16,12 +16,12 @@ use crate::{
 };
 
 impl TensorKeyCompressed<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: SvpPPolBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes + VecZnxBigBytesOf,
     {
-        TensorKey::encrypt_sk_scratch_space(module, infos)
+        TensorKey::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 

poulpy-core/src/encryption/compressed/ggsw_ct.rs

@@ -14,12 +14,12 @@ use crate::{
 };
 
 impl GGSWCompressed<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGSWInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf,
     {
-        GGSW::encrypt_sk_scratch_space(module, infos)
+        GGSW::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 

poulpy-core/src/encryption/compressed/glwe_ct.rs

@@ -14,12 +14,12 @@ use crate::{
 };
 
 impl GLWECompressed<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GLWEInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf,
     {
-        GLWE::encrypt_sk_scratch_space(module, infos)
+        GLWE::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 

poulpy-core/src/encryption/gglwe_atk.rs

@@ -14,7 +14,7 @@ use crate::layouts::{
 };
 
 impl AutomorphismKey<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<BE: Backend, A>(module: &Module<BE>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<BE: Backend, A>(module: &Module<BE>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<BE>: SvpPPolBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes,
@@ -24,10 +24,10 @@ impl AutomorphismKey<Vec<u8>> {
             infos.rank_out(),
             "rank_in != rank_out is not supported for GGLWEAutomorphismKey"
         );
-        GLWESwitchingKey::encrypt_sk_scratch_space(module, infos) + GLWESecret::bytes_of_from_infos(module, &infos.glwe_layout())
+        GLWESwitchingKey::encrypt_sk_tmp_bytes(module, infos) + GLWESecret::bytes_of_from_infos(module, &infos.glwe_layout())
     }
 
-    pub fn encrypt_pk_scratch_space<BE: Backend, A>(module: &Module<BE>, _infos: &A) -> usize
+    pub fn encrypt_pk_tmp_bytes<BE: Backend, A>(module: &Module<BE>, _infos: &A) -> usize
     where
         A: GGLWEInfos,
     {
@@ -36,7 +36,7 @@ impl AutomorphismKey<Vec<u8>> {
             _infos.rank_out(),
             "rank_in != rank_out is not supported for GGLWEAutomorphismKey"
         );
-        GLWESwitchingKey::encrypt_pk_scratch_space(module, _infos)
+        GLWESwitchingKey::encrypt_pk_tmp_bytes(module, _infos)
     }
 }
 
@@ -119,10 +119,10 @@ where
             assert_eq!(res.rank_out(), res.rank_in());
             assert_eq!(sk.rank(), res.rank_out());
             assert!(
-                scratch.available() >= AutomorphismKey::encrypt_sk_scratch_space(self, res),
+                scratch.available() >= AutomorphismKey::encrypt_sk_tmp_bytes(self, res),
-                "scratch.available(): {} < AutomorphismKey::encrypt_sk_scratch_space: {:?}",
+                "scratch.available(): {} < AutomorphismKey::encrypt_sk_tmp_bytes: {:?}",
                 scratch.available(),
-                AutomorphismKey::encrypt_sk_scratch_space(self, res)
+                AutomorphismKey::encrypt_sk_tmp_bytes(self, res)
             )
         }
 

poulpy-core/src/encryption/gglwe_ct.rs

@@ -13,16 +13,16 @@ use crate::{
 };
 
 impl GGLWE<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes,
     {
-        GLWE::encrypt_sk_scratch_space(module, &infos.glwe_layout())
+        GLWE::encrypt_sk_tmp_bytes(module, &infos.glwe_layout())
             + (GLWEPlaintext::bytes_of_from_infos(module, &infos.glwe_layout()) | module.vec_znx_normalize_tmp_bytes())
     }
 
-    pub fn encrypt_pk_scratch_space<B: Backend, A>(_module: &Module<B>, _infos: &A) -> usize
+    pub fn encrypt_pk_tmp_bytes<B: Backend, A>(_module: &Module<B>, _infos: &A) -> usize
     where
         A: GGLWEInfos,
     {
@@ -88,12 +88,12 @@ where
             assert_eq!(res.n(), sk.n());
             assert_eq!(pt.n() as u32, sk.n());
             assert!(
-                scratch.available() >= GGLWE::encrypt_sk_scratch_space(self, res),
+                scratch.available() >= GGLWE::encrypt_sk_tmp_bytes(self, res),
-                "scratch.available: {} < GGLWECiphertext::encrypt_sk_scratch_space(self, res.rank()={}, res.size()={}): {}",
+                "scratch.available: {} < GGLWECiphertext::encrypt_sk_tmp_bytes(self, res.rank()={}, res.size()={}): {}",
                 scratch.available(),
                 res.rank_out(),
                 res.size(),
-                GGLWE::encrypt_sk_scratch_space(self, res)
+                GGLWE::encrypt_sk_tmp_bytes(self, res)
             );
             assert!(
                 res.dnum().0 * res.dsize().0 * res.base2k().0 <= res.k().0,

poulpy-core/src/encryption/gglwe_ksk.rs

@@ -13,21 +13,21 @@ use crate::layouts::{
 };
 
 impl GLWESwitchingKey<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: SvpPPolBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes,
     {
-        (GGLWE::encrypt_sk_scratch_space(module, infos) | ScalarZnx::bytes_of(module.n(), 1))
+        (GGLWE::encrypt_sk_tmp_bytes(module, infos) | ScalarZnx::bytes_of(module.n(), 1))
             + ScalarZnx::bytes_of(module.n(), infos.rank_in().into())
             + GLWESecretPrepared::bytes_of_from_infos(module, &infos.glwe_layout())
     }
 
-    pub fn encrypt_pk_scratch_space<B: Backend, A>(module: &Module<B>, _infos: &A) -> usize
+    pub fn encrypt_pk_tmp_bytes<B: Backend, A>(module: &Module<B>, _infos: &A) -> usize
     where
         A: GGLWEInfos,
     {
-        GGLWE::encrypt_pk_scratch_space(module, _infos)
+        GGLWE::encrypt_pk_tmp_bytes(module, _infos)
     }
 }
 
@@ -66,10 +66,10 @@ impl<DataSelf: DataMut> GLWESwitchingKey<DataSelf> {
             assert!(sk_in.n().0 <= module.n() as u32);
             assert!(sk_out.n().0 <= module.n() as u32);
             assert!(
-                scratch.available() >= GLWESwitchingKey::encrypt_sk_scratch_space(module, self),
+                scratch.available() >= GLWESwitchingKey::encrypt_sk_tmp_bytes(module, self),
-                "scratch.available()={} < GLWESwitchingKey::encrypt_sk_scratch_space={}",
+                "scratch.available()={} < GLWESwitchingKey::encrypt_sk_tmp_bytes={}",
                 scratch.available(),
-                GLWESwitchingKey::encrypt_sk_scratch_space(module, self)
+                GLWESwitchingKey::encrypt_sk_tmp_bytes(module, self)
             )
         }
 

poulpy-core/src/encryption/gglwe_tsk.rs

@@ -14,7 +14,7 @@ use crate::layouts::{
 };
 
 impl TensorKey<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: SvpPPolBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes + VecZnxBigBytesOf,
@@ -24,7 +24,7 @@ impl TensorKey<Vec<u8>> {
             + module.bytes_of_vec_znx_big(1, 1)
             + module.bytes_of_vec_znx_dft(1, 1)
             + GLWESecret::bytes_of(Degree(module.n() as u32), Rank(1))
-            + GLWESwitchingKey::encrypt_sk_scratch_space(module, infos)
+            + GLWESwitchingKey::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 

poulpy-core/src/encryption/ggsw_ct.rs

@@ -14,13 +14,13 @@ use crate::{
 };
 
 impl GGSW<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGSWInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf,
     {
         let size = infos.size();
-        GLWE::encrypt_sk_scratch_space(module, &infos.glwe_layout())
+        GLWE::encrypt_sk_tmp_bytes(module, &infos.glwe_layout())
             + VecZnx::bytes_of(module.n(), (infos.rank() + 1).into(), size)
             + VecZnx::bytes_of(module.n(), 1, size)
             + module.bytes_of_vec_znx_dft((infos.rank() + 1).into(), size)

poulpy-core/src/encryption/glwe_ct.rs

@@ -19,7 +19,7 @@ use crate::{
 };
 
 impl GLWE<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GLWEInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf,
@@ -28,7 +28,7 @@ impl GLWE<Vec<u8>> {
         assert_eq!(module.n() as u32, infos.n());
         module.vec_znx_normalize_tmp_bytes() + 2 * VecZnx::bytes_of(module.n(), 1, size) + module.bytes_of_vec_znx_dft(1, size)
     }
-    pub fn encrypt_pk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_pk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GLWEInfos,
         Module<B>: VecZnxDftBytesOf + SvpPPolBytesOf + VecZnxBigBytesOf + VecZnxNormalizeTmpBytes,
@@ -147,10 +147,10 @@ where
             assert_eq!(sk.n(), self.n() as u32);
             assert_eq!(pt.n(), self.n() as u32);
             assert!(
-                scratch.available() >= GLWE::encrypt_sk_scratch_space(self, &res),
+                scratch.available() >= GLWE::encrypt_sk_tmp_bytes(self, &res),
-                "scratch.available(): {} < GLWECiphertext::encrypt_sk_scratch_space: {}",
+                "scratch.available(): {} < GLWECiphertext::encrypt_sk_tmp_bytes: {}",
                 scratch.available(),
-                GLWE::encrypt_sk_scratch_space(self, &res)
+                GLWE::encrypt_sk_tmp_bytes(self, &res)
             )
         }
 
@@ -209,10 +209,10 @@ where
             assert_eq!(res.n(), self.n() as u32);
             assert_eq!(sk.n(), self.n() as u32);
             assert!(
-                scratch.available() >= GLWE::encrypt_sk_scratch_space(self, &res),
+                scratch.available() >= GLWE::encrypt_sk_tmp_bytes(self, &res),
-                "scratch.available(): {} < GLWECiphertext::encrypt_sk_scratch_space: {}",
+                "scratch.available(): {} < GLWECiphertext::encrypt_sk_tmp_bytes: {}",
                 scratch.available(),
-                GLWE::encrypt_sk_scratch_space(self, &res)
+                GLWE::encrypt_sk_tmp_bytes(self, &res)
             )
         }
 

poulpy-core/src/encryption/glwe_pk.rs

@@ -45,7 +45,7 @@ where
         }
 
         // Its ok to allocate scratch space here since pk is usually generated only once.
-        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWE::encrypt_sk_scratch_space(self, res));
+        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWE::encrypt_sk_tmp_bytes(self, res));
 
         let mut tmp: GLWE<Vec<u8>> = GLWE::alloc_from_infos(res);
 

poulpy-core/src/encryption/glwe_to_lwe_ksk.rs

@@ -14,13 +14,13 @@ use crate::layouts::{
 };
 
 impl GLWEToLWESwitchingKey<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: SvpPPolBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes,
     {
         GLWESecretPrepared::bytes_of(module, infos.rank_in())
-            + (GLWESwitchingKey::encrypt_sk_scratch_space(module, infos) | GLWESecret::bytes_of(infos.n(), infos.rank_in()))
+            + (GLWESwitchingKey::encrypt_sk_tmp_bytes(module, infos) | GLWESecret::bytes_of(infos.n(), infos.rank_in()))
     }
 }
 

poulpy-core/src/encryption/lwe_ksk.rs

@@ -14,7 +14,7 @@ use crate::layouts::{
 };
 
 impl LWESwitchingKey<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: SvpPPolBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes,
@@ -36,7 +36,7 @@ impl LWESwitchingKey<Vec<u8>> {
         );
         GLWESecret::bytes_of(Degree(module.n() as u32), Rank(1))
             + GLWESecretPrepared::bytes_of(module, Rank(1))
-            + GLWESwitchingKey::encrypt_sk_scratch_space(module, infos)
+            + GLWESwitchingKey::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 

poulpy-core/src/encryption/lwe_to_glwe_ksk.rs

@@ -12,7 +12,7 @@ use poulpy_hal::{
 use crate::layouts::{Degree, GGLWEInfos, GLWESecret, GLWESwitchingKey, LWEInfos, LWESecret, LWEToGLWESwitchingKey, Rank};
 
 impl LWEToGLWESwitchingKey<Vec<u8>> {
-    pub fn encrypt_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn encrypt_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGLWEInfos,
         Module<B>: SvpPPolBytesOf + VecZnxNormalizeTmpBytes + VecZnxDftBytesOf + VecZnxNormalizeTmpBytes,
@@ -22,8 +22,7 @@ impl LWEToGLWESwitchingKey<Vec<u8>> {
             Rank(1),
             "rank_in != 1 is not supported for LWEToGLWESwitchingKey"
         );
-        GLWESwitchingKey::encrypt_sk_scratch_space(module, infos)
+        GLWESwitchingKey::encrypt_sk_tmp_bytes(module, infos) + GLWESecret::bytes_of(Degree(module.n() as u32), infos.rank_in())
-            + GLWESecret::bytes_of(Degree(module.n() as u32), infos.rank_in())
     }
 }
 

poulpy-core/src/external_product/gglwe_ksk.rs

@@ -18,7 +18,7 @@ where
         A: GGLWEInfos,
         B: GGSWInfos,
     {
-        self.glwe_external_product_scratch_space(res_infos, a_infos, b_infos)
+        self.glwe_external_product_tmp_bytes(res_infos, a_infos, b_infos)
     }
 
     fn gglwe_external_product<R, A, B>(&self, res: &mut R, a: &A, b: &B, scratch: &mut Scratch<BE>)

poulpy-core/src/external_product/ggsw_ct.rs

@@ -21,7 +21,7 @@ where
         A: GGSWInfos,
         B: GGSWInfos,
     {
-        self.glwe_external_product_scratch_space(res_infos, a_infos, b_infos)
+        self.glwe_external_product_tmp_bytes(res_infos, a_infos, b_infos)
     }
 
     fn ggsw_external_product<R, A, B>(&self, res: &mut R, a: &A, b: &B, scratch: &mut Scratch<BE>)

poulpy-core/src/external_product/glwe_ct.rs

@@ -15,14 +15,14 @@ use crate::{
 };
 
 impl GLWE<Vec<u8>> {
-    pub fn external_product_scratch_space<R, A, B, M, BE: Backend>(module: &M, res_infos: &R, a_infos: &A, b_infos: &B) -> usize
+    pub fn external_product_tmp_bytes<R, A, B, M, BE: Backend>(module: &M, res_infos: &R, a_infos: &A, b_infos: &B) -> usize
     where
         R: GLWEInfos,
         A: GLWEInfos,
         B: GGSWInfos,
         M: GLWEExternalProduct<BE>,
     {
-        module.glwe_external_product_scratch_space(res_infos, a_infos, b_infos)
+        module.glwe_external_product_tmp_bytes(res_infos, a_infos, b_infos)
     }
 }
 
@@ -61,7 +61,7 @@ where
         + VecZnxBigNormalize<BE>
         + VecZnxNormalize<BE>,
 {
-    fn glwe_external_product_scratch_space<R, A, B>(&self, res_infos: &R, a_infos: &A, b_infos: &B) -> usize
+    fn glwe_external_product_tmp_bytes<R, A, B>(&self, res_infos: &R, a_infos: &A, b_infos: &B) -> usize
     where
         R: GLWEInfos,
         A: GLWEInfos,
@@ -111,7 +111,7 @@ where
 
             assert_eq!(rhs.rank(), res.rank());
             assert_eq!(rhs.n(), res.n());
-            assert!(scratch.available() >= self.glwe_external_product_scratch_space(res, res, rhs));
+            assert!(scratch.available() >= self.glwe_external_product_tmp_bytes(res, res, rhs));
         }
 
         let cols: usize = (rhs.rank() + 1).into();
@@ -225,7 +225,7 @@ where
             assert_eq!(rhs.rank(), res.rank());
             assert_eq!(rhs.n(), res.n());
             assert_eq!(lhs.n(), res.n());
-            assert!(scratch.available() >= self.glwe_external_product_scratch_space(res, lhs, rhs));
+            assert!(scratch.available() >= self.glwe_external_product_tmp_bytes(res, lhs, rhs));
         }
 
         let cols: usize = (rhs.rank() + 1).into();

poulpy-core/src/glwe_packing.rs

@@ -90,13 +90,13 @@ impl GLWEPacker {
     }
 
     /// Number of scratch space bytes required to call [Self::add].
-    pub fn scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
+    pub fn tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
     where
         OUT: GLWEInfos,
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        pack_core_scratch_space(module, out_infos, key_infos)
+        pack_core_tmp_bytes(module, out_infos, key_infos)
     }
 
     pub fn galois_elements<B: Backend>(module: &Module<B>) -> Vec<i64> {
@@ -111,7 +111,7 @@ impl GLWEPacker {
     ///   of packed ciphertexts reaches N/2^log_batch is a result written.
     /// * `a`: ciphertext to pack. Can optionally give None to pack a 0 ciphertext.
     /// * `auto_keys`: a [HashMap] containing the [AutomorphismKeyExec]s.
-    /// * `scratch`: scratch space of size at least [Self::scratch_space].
+    /// * `scratch`: scratch space of size at least [Self::tmp_bytes].
     pub fn add<DataA: DataRef, DataAK: DataRef, B: Backend>(
         &mut self,
         module: &Module<B>,
@@ -177,13 +177,13 @@ impl GLWEPacker {
     }
 }
 
-fn pack_core_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
+fn pack_core_tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
 where
     OUT: GLWEInfos,
     KEY: GGLWEInfos,
     Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
 {
-    combine_scratch_space(module, out_infos, key_infos)
+    combine_tmp_bytes(module, out_infos, key_infos)
 }
 
 fn pack_core<D: DataRef, DataAK: DataRef, B: Backend>(
@@ -268,14 +268,14 @@ fn pack_core<D: DataRef, DataAK: DataRef, B: Backend>(
     }
 }
 
-fn combine_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
+fn combine_tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
 where
     OUT: GLWEInfos,
     KEY: GGLWEInfos,
     Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
 {
     GLWE::bytes_of_from_infos(module, out_infos)
-        + (GLWE::rsh_scratch_space(module.n()) | GLWE::automorphism_inplace_scratch_space(module, out_infos, key_infos))
+        + (GLWE::rsh_tmp_bytes(module.n()) | GLWE::automorphism_inplace_tmp_bytes(module, out_infos, key_infos))
 }
 
 /// [combine] merges two ciphertexts together.

poulpy-core/src/glwe_trace.rs

@@ -27,19 +27,14 @@ impl GLWE<Vec<u8>> {
         gal_els
     }
 
-    pub fn trace_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn trace_tmp_bytes<B: Backend, OUT, IN, KEY>(module: &Module<B>, out_infos: &OUT, in_infos: &IN, key_infos: &KEY) -> usize
-        module: &Module<B>,
-        out_infos: &OUT,
-        in_infos: &IN,
-        key_infos: &KEY,
-    ) -> usize
     where
         OUT: GLWEInfos,
         IN: GLWEInfos,
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        let trace: usize = Self::automorphism_inplace_scratch_space(module, out_infos, key_infos);
+        let trace: usize = Self::automorphism_inplace_tmp_bytes(module, out_infos, key_infos);
         if in_infos.base2k() != key_infos.base2k() {
             let glwe_conv: usize = VecZnx::bytes_of(
                 module.n(),
@@ -52,13 +47,13 @@ impl GLWE<Vec<u8>> {
         trace
     }
 
-    pub fn trace_inplace_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
+    pub fn trace_inplace_tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
     where
         OUT: GLWEInfos,
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        Self::trace_scratch_space(module, out_infos, out_infos, key_infos)
+        Self::trace_tmp_bytes(module, out_infos, out_infos, key_infos)
     }
 }
 

poulpy-core/src/keyswitching/gglwe_ct.rs

@@ -13,7 +13,7 @@ use crate::layouts::{
 };
 
 impl AutomorphismKey<Vec<u8>> {
-    pub fn keyswitch_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn keyswitch_tmp_bytes<B: Backend, OUT, IN, KEY>(
         module: &Module<B>,
         out_infos: &OUT,
         in_infos: &IN,
@@ -25,16 +25,16 @@ impl AutomorphismKey<Vec<u8>> {
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        GLWESwitchingKey::keyswitch_scratch_space(module, out_infos, in_infos, key_infos)
+        GLWESwitchingKey::keyswitch_tmp_bytes(module, out_infos, in_infos, key_infos)
     }
 
-    pub fn keyswitch_inplace_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
+    pub fn keyswitch_inplace_tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_infos: &KEY) -> usize
     where
         OUT: GGLWEInfos,
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        GLWESwitchingKey::keyswitch_inplace_scratch_space(module, out_infos, key_infos)
+        GLWESwitchingKey::keyswitch_inplace_tmp_bytes(module, out_infos, key_infos)
     }
 }
 
@@ -86,7 +86,7 @@ impl<DataSelf: DataMut> AutomorphismKey<DataSelf> {
 }
 
 impl GLWESwitchingKey<Vec<u8>> {
-    pub fn keyswitch_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn keyswitch_tmp_bytes<B: Backend, OUT, IN, KEY>(
         module: &Module<B>,
         out_infos: &OUT,
         in_infos: &IN,
@@ -98,16 +98,16 @@ impl GLWESwitchingKey<Vec<u8>> {
         KEY: GGLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        GLWE::keyswitch_scratch_space(module, out_infos, in_infos, key_apply)
+        GLWE::keyswitch_tmp_bytes(module, out_infos, in_infos, key_apply)
     }
 
-    pub fn keyswitch_inplace_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_apply: &KEY) -> usize
+    pub fn keyswitch_inplace_tmp_bytes<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_apply: &KEY) -> usize
     where
         OUT: GGLWEInfos + GLWEInfos,
         KEY: GGLWEInfos + GLWEInfos,
         Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
     {
-        GLWE::keyswitch_inplace_scratch_space(module, out_infos, key_apply)
+        GLWE::keyswitch_inplace_tmp_bytes(module, out_infos, key_apply)
     }
 }
 

poulpy-core/src/keyswitching/ggsw_ct.rs

@@ -1,359 +1,131 @@
-use poulpy_hal::{
+use poulpy_hal::layouts::{Backend, DataMut, Scratch, VecZnx};
-    api::{
-        ScratchAvailable, VecZnxBigAddSmallInplace, VecZnxBigBytesOf, VecZnxBigNormalize, VecZnxBigNormalizeTmpBytes, VecZnxCopy,
-        VecZnxDftAddInplace, VecZnxDftApply, VecZnxDftBytesOf, VecZnxDftCopy, VecZnxIdftApplyConsume, VecZnxIdftApplyTmpA,
-        VecZnxNormalize, VecZnxNormalizeTmpBytes, VmpApplyDftToDft, VmpApplyDftToDftAdd, VmpApplyDftToDftTmpBytes,
-    },
-    layouts::{Backend, DataMut, DataRef, Module, Scratch, VecZnx, VmpPMat, ZnxInfos},
-};
 
 use crate::{
+    GGSWExpandRows, ScratchTakeCore,
+    keyswitching::glwe_ct::GLWEKeySwitching,
     layouts::{
-        GGLWE, GGLWEInfos, GGSW, GGSWInfos, GLWE, GLWEInfos, LWEInfos,
+        GGLWEInfos, GGSW, GGSWInfos, GGSWToMut, GGSWToRef,
-        prepared::{GLWESwitchingKeyPrepared, TensorKeyPrepared},
+        prepared::{GLWESwitchingKeyPreparedToRef, TensorKeyPreparedToRef},
     },
-    operations::GLWEOperations,
 };
 
 impl GGSW<Vec<u8>> {
-    pub(crate) fn expand_row_scratch_space<B: Backend, OUT, TSK>(module: &Module<B>, out_infos: &OUT, tsk_infos: &TSK) -> usize
+    pub fn keyswitch_tmp_bytes<R, A, K, T, M, BE: Backend>(
-    where
+        module: &M,
-        OUT: GGSWInfos,
+        res_infos: &R,
-        TSK: GGLWEInfos,
+        a_infos: &A,
-        Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigBytesOf + VecZnxNormalizeTmpBytes,
+        key_infos: &K,
-    {
+        tsk_infos: &T,
-        let tsk_size: usize = tsk_infos.k().div_ceil(tsk_infos.base2k()) as usize;
-        let size_in: usize = out_infos
-            .k()
-            .div_ceil(tsk_infos.base2k())
-            .div_ceil(tsk_infos.dsize().into()) as usize;
-
-        let tmp_dft_i: usize = module.bytes_of_vec_znx_dft((tsk_infos.rank_out() + 1).into(), tsk_size);
-        let tmp_a: usize = module.bytes_of_vec_znx_dft(1, size_in);
-        let vmp: usize = module.vmp_apply_dft_to_dft_tmp_bytes(
-            tsk_size,
-            size_in,
-            size_in,
-            (tsk_infos.rank_in()).into(),  // Verify if rank+1
-            (tsk_infos.rank_out()).into(), // Verify if rank+1
-            tsk_size,
-        );
-        let tmp_idft: usize = module.bytes_of_vec_znx_big(1, tsk_size);
-        let norm: usize = module.vec_znx_normalize_tmp_bytes();
-
-        tmp_dft_i + ((tmp_a + vmp) | (tmp_idft + norm))
-    }
-
-    #[allow(clippy::too_many_arguments)]
-    pub fn keyswitch_scratch_space<B: Backend, OUT, IN, KEY, TSK>(
-        module: &Module<B>,
-        out_infos: &OUT,
-        in_infos: &IN,
-        apply_infos: &KEY,
-        tsk_infos: &TSK,
     ) -> usize
     where
-        OUT: GGSWInfos,
+        R: GGSWInfos,
-        IN: GGSWInfos,
+        A: GGSWInfos,
-        KEY: GGLWEInfos,
+        K: GGLWEInfos,
-        TSK: GGLWEInfos,
+        T: GGLWEInfos,
-        Module<B>:
+        M: GGSWKeySwitch<BE>,
-            VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigBytesOf + VecZnxNormalizeTmpBytes + VecZnxBigNormalizeTmpBytes,
     {
-        #[cfg(debug_assertions)]
+        module.ggsw_keyswitch_tmp_bytes(res_infos, a_infos, key_infos, tsk_infos)
-        {
+    }
-            assert_eq!(apply_infos.rank_in(), apply_infos.rank_out());
-            assert_eq!(tsk_infos.rank_in(), tsk_infos.rank_out());
-            assert_eq!(apply_infos.rank_in(), tsk_infos.rank_in());
 }
 
-        let rank: usize = apply_infos.rank_out().into();
+impl<D: DataMut> GGSW<D> {
+    pub fn keyswitch<M, A, K, T, BE: Backend>(&mut self, module: &M, a: &A, key: &K, tsk: &T, scratch: &mut Scratch<BE>)
+    where
+        A: GGSWToRef,
+        K: GLWESwitchingKeyPreparedToRef<BE>,
+        T: TensorKeyPreparedToRef<BE>,
+        Scratch<BE>: ScratchTakeCore<BE>,
+        M: GGSWKeySwitch<BE>,
+    {
+        module.ggsw_keyswitch(self, a, key, tsk, scratch);
+    }
 
-        let size_out: usize = out_infos.k().div_ceil(out_infos.base2k()) as usize;
+    pub fn keyswitch_inplace<M, K, T, BE: Backend>(&mut self, module: &M, key: &K, tsk: &T, scratch: &mut Scratch<BE>)
-        let res_znx: usize = VecZnx::bytes_of(module.n(), rank + 1, size_out);
+    where
-        let ci_dft: usize = module.bytes_of_vec_znx_dft(rank + 1, size_out);
+        K: GLWESwitchingKeyPreparedToRef<BE>,
-        let ks: usize = GLWE::keyswitch_scratch_space(module, out_infos, in_infos, apply_infos);
+        T: TensorKeyPreparedToRef<BE>,
-        let expand_rows: usize = GGSW::expand_row_scratch_space(module, out_infos, tsk_infos);
+        Scratch<BE>: ScratchTakeCore<BE>,
-        let res_dft: usize = module.bytes_of_vec_znx_dft(rank + 1, size_out);
+        M: GGSWKeySwitch<BE>,
+    {
+        module.ggsw_keyswitch_inplace(self, key, tsk, scratch);
+    }
+}
 
-        if in_infos.base2k() == tsk_infos.base2k() {
+pub trait GGSWKeySwitch<BE: Backend>
+where
+    Self: GLWEKeySwitching<BE> + GGSWExpandRows<BE>,
+{
+    fn ggsw_keyswitch_tmp_bytes<R, A, K, T>(&self, res_infos: &R, a_infos: &A, key_infos: &K, tsk_infos: &T) -> usize
+    where
+        R: GGSWInfos,
+        A: GGSWInfos,
+        K: GGLWEInfos,
+        T: GGLWEInfos,
+    {
+        assert_eq!(key_infos.rank_in(), key_infos.rank_out());
+        assert_eq!(tsk_infos.rank_in(), tsk_infos.rank_out());
+        assert_eq!(key_infos.rank_in(), tsk_infos.rank_in());
+
+        let rank: usize = key_infos.rank_out().into();
+
+        let size_out: usize = res_infos.k().div_ceil(res_infos.base2k()) as usize;
+        let res_znx: usize = VecZnx::bytes_of(self.n(), rank + 1, size_out);
+        let ci_dft: usize = self.bytes_of_vec_znx_dft(rank + 1, size_out);
+        let ks: usize = self.glwe_keyswitch_tmp_bytes(res_infos, a_infos, key_infos);
+        let expand_rows: usize = self.ggsw_expand_rows_tmp_bytes(res_infos, tsk_infos);
+        let res_dft: usize = self.bytes_of_vec_znx_dft(rank + 1, size_out);
+
+        if a_infos.base2k() == tsk_infos.base2k() {
             res_znx + ci_dft + (ks | expand_rows | res_dft)
         } else {
             let a_conv: usize = VecZnx::bytes_of(
-                module.n(),
+                self.n(),
                 1,
-                out_infos.k().div_ceil(tsk_infos.base2k()) as usize,
+                res_infos.k().div_ceil(tsk_infos.base2k()) as usize,
-            ) + module.vec_znx_normalize_tmp_bytes();
+            ) + self.vec_znx_normalize_tmp_bytes();
             res_znx + ci_dft + (a_conv | ks | expand_rows | res_dft)
         }
     }
 
-    #[allow(clippy::too_many_arguments)]
+    fn ggsw_keyswitch<R, A, K, T>(&self, res: &mut R, a: &A, key: &K, tsk: &T, scratch: &mut Scratch<BE>)
-    pub fn keyswitch_inplace_scratch_space<B: Backend, OUT, KEY, TSK>(
-        module: &Module<B>,
-        out_infos: &OUT,
-        apply_infos: &KEY,
-        tsk_infos: &TSK,
-    ) -> usize
     where
-        OUT: GGSWInfos,
+        R: GGSWToMut,
-        KEY: GGLWEInfos,
+        A: GGSWToRef,
-        TSK: GGLWEInfos,
+        K: GLWESwitchingKeyPreparedToRef<BE>,
-        Module<B>:
+        T: TensorKeyPreparedToRef<BE>,
-            VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigBytesOf + VecZnxNormalizeTmpBytes + VecZnxBigNormalizeTmpBytes,
+        Scratch<BE>: ScratchTakeCore<BE>,
     {
-        GGSW::keyswitch_scratch_space(module, out_infos, out_infos, apply_infos, tsk_infos)
+        let res: &mut GGSW<&mut [u8]> = &mut res.to_mut();
-    }
+        let a: &GGSW<&[u8]> = &a.to_ref();
-}
 
-impl<DataSelf: DataMut> GGSW<DataSelf> {
+        assert_eq!(res.ggsw_layout(), a.ggsw_layout());
-    pub fn from_gglwe<DataA, DataTsk, B: Backend>(
-        &mut self,
-        module: &Module<B>,
-        a: &GGLWE<DataA>,
-        tsk: &TensorKeyPrepared<DataTsk, B>,
-        scratch: &mut Scratch<B>,
-    ) where
-        DataA: DataRef,
-        DataTsk: DataRef,
-        Module<B>: VecZnxCopy
-            + VecZnxDftBytesOf
-            + VmpApplyDftToDftTmpBytes
-            + VecZnxBigBytesOf
-            + VecZnxNormalizeTmpBytes
-            + VecZnxDftApply<B>
-            + VecZnxDftCopy<B>
-            + VmpApplyDftToDft<B>
-            + VmpApplyDftToDftAdd<B>
-            + VecZnxDftAddInplace<B>
-            + VecZnxBigNormalize<B>
-            + VecZnxIdftApplyTmpA<B>
-            + VecZnxNormalize<B>,
-        Scratch<B>: ScratchAvailable,
-    {
-        #[cfg(debug_assertions)]
-        {
-            use crate::layouts::{GLWEInfos, LWEInfos};
 
-            assert_eq!(self.rank(), a.rank_out());
+        for row in 0..a.dnum().into() {
-            assert_eq!(self.dnum(), a.dnum());
-            assert_eq!(self.n(), module.n() as u32);
-            assert_eq!(a.n(), module.n() as u32);
-            assert_eq!(tsk.n(), module.n() as u32);
-        }
-        (0..self.dnum().into()).for_each(|row_i| {
-            self.at_mut(row_i, 0).copy(module, &a.at(row_i, 0));
-        });
-        self.expand_row(module, tsk, scratch);
-    }
-
-    pub fn keyswitch<DataLhs: DataRef, DataKsk: DataRef, DataTsk: DataRef, B: Backend>(
-        &mut self,
-        module: &Module<B>,
-        lhs: &GGSW<DataLhs>,
-        ksk: &GLWESwitchingKeyPrepared<DataKsk, B>,
-        tsk: &TensorKeyPrepared<DataTsk, B>,
-        scratch: &mut Scratch<B>,
-    ) where
-        Module<B>: VecZnxDftBytesOf
-            + VmpApplyDftToDftTmpBytes
-            + VecZnxBigNormalizeTmpBytes
-            + VmpApplyDftToDft<B>
-            + VmpApplyDftToDftAdd<B>
-            + VecZnxDftApply<B>
-            + VecZnxIdftApplyConsume<B>
-            + VecZnxBigAddSmallInplace<B>
-            + VecZnxBigNormalize<B>
-            + VecZnxDftBytesOf
-            + VecZnxBigBytesOf
-            + VecZnxNormalizeTmpBytes
-            + VecZnxDftCopy<B>
-            + VecZnxDftAddInplace<B>
-            + VecZnxIdftApplyTmpA<B>
-            + VecZnxNormalize<B>,
-        Scratch<B>: ScratchAvailable,
-    {
-        (0..lhs.dnum().into()).for_each(|row_i| {
             // Key-switch column 0, i.e.
             // col 0: (-(a0s0 + a1s1 + a2s2) + M[i], a0, a1, a2) -> (-(a0s0' + a1s1' + a2s2') + M[i], a0, a1, a2)
-            self.at_mut(row_i, 0)
+            self.glwe_keyswitch(&mut res.at_mut(row, 0), &a.at(row, 0), key, scratch);
-                .keyswitch(module, &lhs.at(row_i, 0), ksk, scratch);
-        });
-        self.expand_row(module, tsk, scratch);
         }
 
-    pub fn keyswitch_inplace<DataKsk: DataRef, DataTsk: DataRef, B: Backend>(
+        self.ggsw_expand_row(res, tsk, scratch);
-        &mut self,
+    }
-        module: &Module<B>,
+
-        ksk: &GLWESwitchingKeyPrepared<DataKsk, B>,
+    fn ggsw_keyswitch_inplace<R, K, T>(&self, res: &mut R, key: &K, tsk: &T, scratch: &mut Scratch<BE>)
-        tsk: &TensorKeyPrepared<DataTsk, B>,
+    where
-        scratch: &mut Scratch<B>,
+        R: GGSWToMut,
-    ) where
+        K: GLWESwitchingKeyPreparedToRef<BE>,
-        Module<B>: VecZnxDftBytesOf
+        T: TensorKeyPreparedToRef<BE>,
-            + VmpApplyDftToDftTmpBytes
+        Scratch<BE>: ScratchTakeCore<BE>,
-            + VecZnxBigNormalizeTmpBytes
-            + VmpApplyDftToDft<B>
-            + VmpApplyDftToDftAdd<B>
-            + VecZnxDftApply<B>
-            + VecZnxIdftApplyConsume<B>
-            + VecZnxBigAddSmallInplace<B>
-            + VecZnxBigNormalize<B>
-            + VecZnxDftBytesOf
-            + VecZnxBigBytesOf
-            + VecZnxNormalizeTmpBytes
-            + VecZnxDftCopy<B>
-            + VecZnxDftAddInplace<B>
-            + VecZnxIdftApplyTmpA<B>
-            + VecZnxNormalize<B>,
-        Scratch<B>: ScratchAvailable,
     {
-        (0..self.dnum().into()).for_each(|row_i| {
+        let res: &mut GGSW<&mut [u8]> = &mut res.to_mut();
+
+        for row in 0..res.dnum().into() {
             // Key-switch column 0, i.e.
             // col 0: (-(a0s0 + a1s1 + a2s2) + M[i], a0, a1, a2) -> (-(a0s0' + a1s1' + a2s2') + M[i], a0, a1, a2)
-            self.at_mut(row_i, 0)
+            self.glwe_keyswitch_inplace(&mut res.at_mut(row, 0), key, scratch);
-                .keyswitch_inplace(module, ksk, scratch);
-        });
-        self.expand_row(module, tsk, scratch);
         }
 
-    pub fn expand_row<DataTsk: DataRef, B: Backend>(
+        self.ggsw_expand_row(res, tsk, scratch);
-        &mut self,
-        module: &Module<B>,
-        tsk: &TensorKeyPrepared<DataTsk, B>,
-        scratch: &mut Scratch<B>,
-    ) where
-        Module<B>: VecZnxDftBytesOf
-            + VmpApplyDftToDftTmpBytes
-            + VecZnxBigBytesOf
-            + VecZnxNormalizeTmpBytes
-            + VecZnxDftApply<B>
-            + VecZnxDftCopy<B>
-            + VmpApplyDftToDft<B>
-            + VmpApplyDftToDftAdd<B>
-            + VecZnxDftAddInplace<B>
-            + VecZnxBigNormalize<B>
-            + VecZnxIdftApplyTmpA<B>
-            + VecZnxNormalize<B>,
-        Scratch<B>: ScratchAvailable,
-    {
-        let basek_in: usize = self.base2k().into();
-        let basek_tsk: usize = tsk.base2k().into();
-
-        assert!(scratch.available() >= GGSW::expand_row_scratch_space(module, self, tsk));
-
-        let n: usize = self.n().into();
-        let rank: usize = self.rank().into();
-        let cols: usize = rank + 1;
-
-        let a_size: usize = (self.size() * basek_in).div_ceil(basek_tsk);
-
-        // Keyswitch the j-th row of the col 0
-        for row_i in 0..self.dnum().into() {
-            let a = &self.at(row_i, 0).data;
-
-            // Pre-compute DFT of (a0, a1, a2)
-            let (mut ci_dft, scratch_1) = scratch.take_vec_znx_dft(n, cols, a_size);
-
-            if basek_in == basek_tsk {
-                for i in 0..cols {
-                    module.vec_znx_dft_apply(1, 0, &mut ci_dft, i, a, i);
-                }
-            } else {
-                let (mut a_conv, scratch_2) = scratch_1.take_vec_znx(n, 1, a_size);
-                for i in 0..cols {
-                    module.vec_znx_normalize(basek_tsk, &mut a_conv, 0, basek_in, a, i, scratch_2);
-                    module.vec_znx_dft_apply(1, 0, &mut ci_dft, i, &a_conv, 0);
     }
 }
 
-            for col_j in 1..cols {
+impl<DataSelf: DataMut> GGSW<DataSelf> {}
-                // Example for rank 3:
-                //
-                // Note: M is a vector (m, Bm, B^2m, B^3m, ...), so each column is
-                // actually composed of that many dnum and we focus on a specific row here
-                // implicitely given ci_dft.
-                //
-                // # Input
-                //
-                // col 0: (-(a0s0 + a1s1 + a2s2) + M[i], a0    , a1    , a2    )
-                // col 1: (0, 0, 0, 0)
-                // col 2: (0, 0, 0, 0)
-                // col 3: (0, 0, 0, 0)
-                //
-                // # Output
-                //
-                // col 0: (-(a0s0 + a1s1 + a2s2) + M[i], a0       , a1       , a2       )
-                // col 1: (-(b0s0 + b1s1 + b2s2)       , b0 + M[i], b1       , b2       )
-                // col 2: (-(c0s0 + c1s1 + c2s2)       , c0       , c1 + M[i], c2       )
-                // col 3: (-(d0s0 + d1s1 + d2s2)       , d0       , d1       , d2 + M[i])
-
-                let dsize: usize = tsk.dsize().into();
-
-                let (mut tmp_dft_i, scratch_2) = scratch_1.take_vec_znx_dft(n, cols, tsk.size());
-                let (mut tmp_a, scratch_3) = scratch_2.take_vec_znx_dft(n, 1, ci_dft.size().div_ceil(dsize));
-
-                {
-                    // Performs a key-switch for each combination of s[i]*s[j], i.e. for a0, a1, a2
-                    //
-                    // # Example for col=1
-                    //
-                    // a0 * (-(f0s0 + f1s1 + f1s2) + s0^2, f0, f1, f2) = (-(a0f0s0 + a0f1s1 + a0f1s2) + a0s0^2, a0f0, a0f1, a0f2)
-                    // +
-                    // a1 * (-(g0s0 + g1s1 + g1s2) + s0s1, g0, g1, g2) = (-(a1g0s0 + a1g1s1 + a1g1s2) + a1s0s1, a1g0, a1g1, a1g2)
-                    // +
-                    // a2 * (-(h0s0 + h1s1 + h1s2) + s0s2, h0, h1, h2) = (-(a2h0s0 + a2h1s1 + a2h1s2) + a2s0s2, a2h0, a2h1, a2h2)
-                    // =
-                    // (-(x0s0 + x1s1 + x2s2) + s0(a0s0 + a1s1 + a2s2), x0, x1, x2)
-                    for col_i in 1..cols {
-                        let pmat: &VmpPMat<DataTsk, B> = &tsk.at(col_i - 1, col_j - 1).key.data; // Selects Enc(s[i]s[j])
-
-                        // Extracts a[i] and multipies with Enc(s[i]s[j])
-                        for di in 0..dsize {
-                            tmp_a.set_size((ci_dft.size() + di) / dsize);
-
-                            // Small optimization for dsize > 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^{(dsize-1) * B}.
-                            // As such we can ignore the last dsize-2 limbs safely of the sum of vmp products.
-                            // It is possible to further ignore the last dsize-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.
-                            tmp_dft_i.set_size(tsk.size() - ((dsize - di) as isize - 2).max(0) as usize);
-
-                            module.vec_znx_dft_copy(dsize, dsize - 1 - di, &mut tmp_a, 0, &ci_dft, col_i);
-                            if di == 0 && col_i == 1 {
-                                module.vmp_apply_dft_to_dft(&mut tmp_dft_i, &tmp_a, pmat, scratch_3);
-                            } else {
-                                module.vmp_apply_dft_to_dft_add(&mut tmp_dft_i, &tmp_a, pmat, di, scratch_3);
-                            }
-                        }
-                    }
-                }
-
-                // Adds -(sum a[i] * s[i]) + m)  on the i-th column of tmp_idft_i
-                //
-                // (-(x0s0 + x1s1 + x2s2) + a0s0s0 + a1s0s1 + a2s0s2, x0, x1, x2)
-                // +
-                // (0, -(a0s0 + a1s1 + a2s2) + M[i], 0, 0)
-                // =
-                // (-(x0s0 + x1s1 + x2s2) + s0(a0s0 + a1s1 + a2s2), x0 -(a0s0 + a1s1 + a2s2) + M[i], x1, x2)
-                // =
-                // (-(x0s0 + x1s1 + x2s2), x0 + M[i], x1, x2)
-                module.vec_znx_dft_add_inplace(&mut tmp_dft_i, col_j, &ci_dft, 0);
-                let (mut tmp_idft, scratch_3) = scratch_2.take_vec_znx_big(n, 1, tsk.size());
-                for i in 0..cols {
-                    module.vec_znx_idft_apply_tmpa(&mut tmp_idft, 0, &mut tmp_dft_i, i);
-                    module.vec_znx_big_normalize(
-                        basek_in,
-                        &mut self.at_mut(row_i, col_j).data,
-                        i,
-                        basek_tsk,
-                        &tmp_idft,
-                        0,
-                        scratch_3,
-                    );
-                }
-            }
-        }
-    }
-}

poulpy-core/src/keyswitching/glwe_ct.rs

@@ -1,186 +1,179 @@
 use poulpy_hal::{
     api::{
-        ScratchAvailable, VecZnxBigAddSmallInplace, VecZnxBigNormalize, VecZnxBigNormalizeTmpBytes, VecZnxDftApply,
+        ModuleN, ScratchAvailable, ScratchTakeBasic, VecZnxBigAddSmallInplace, VecZnxBigNormalize, VecZnxBigNormalizeTmpBytes,
-        VecZnxDftBytesOf, VecZnxIdftApplyConsume, VecZnxNormalize, VecZnxNormalizeTmpBytes, VmpApplyDftToDft,
+        VecZnxDftApply, VecZnxDftBytesOf, VecZnxIdftApplyConsume, VecZnxNormalize, VecZnxNormalizeTmpBytes, VmpApplyDftToDft,
         VmpApplyDftToDftAdd, VmpApplyDftToDftTmpBytes,
     },
     layouts::{Backend, DataMut, DataRef, DataViewMut, Module, Scratch, VecZnx, VecZnxBig, VecZnxDft, VmpPMat, ZnxInfos},
 };
 
-use crate::layouts::{GGLWEInfos, GLWE, GLWEInfos, LWEInfos, prepared::GLWESwitchingKeyPrepared};
+use crate::{
+    ScratchTakeCore,
+    layouts::{
+        GGLWEInfos, GLWE, GLWEInfos, GLWEToMut, GLWEToRef, LWEInfos,
+        prepared::{GLWESwitchingKeyPrepared, GLWESwitchingKeyPreparedToRef},
+    },
+};
 
 impl GLWE<Vec<u8>> {
-    pub fn keyswitch_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn keyswitch_tmp_bytes<M, R, A, B, BE: Backend>(module: &M, res_infos: &R, a_infos: &A, b_infos: &B) -> usize
-        module: &Module<B>,
-        out_infos: &OUT,
-        in_infos: &IN,
-        key_apply: &KEY,
-    ) -> usize
     where
-        OUT: GLWEInfos,
+        R: GLWEInfos,
-        IN: GLWEInfos,
+        A: GLWEInfos,
-        KEY: GGLWEInfos,
+        B: GGLWEInfos,
-        Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
+        M: GLWEKeySwitching<BE>,
     {
-        let in_size: usize = in_infos
+        module.glwe_keyswitch_tmp_bytes(res_infos, a_infos, b_infos)
+    }
+}
+
+impl<D: DataMut> GLWE<D> {
+    pub fn keyswitch<A, B, M, BE: Backend>(&mut self, module: &M, a: &A, b: &B, scratch: &mut Scratch<BE>)
+    where
+        A: GLWEToRef,
+        B: GLWESwitchingKeyPreparedToRef<BE>,
+        M: GLWEKeySwitching<BE>,
+        Scratch<BE>: ScratchTakeCore<BE>,
+    {
+        module.glwe_keyswitch(self, a, b, scratch);
+    }
+
+    pub fn keyswitch_inplace<A, M, BE: Backend>(&mut self, module: &M, a: &A, scratch: &mut Scratch<BE>)
+    where
+        A: GLWESwitchingKeyPreparedToRef<BE>,
+        M: GLWEKeySwitching<BE>,
+        Scratch<BE>: ScratchTakeCore<BE>,
+    {
+        module.glwe_keyswitch_inplace(self, a, scratch);
+    }
+}
+
+impl<BE: Backend> GLWEKeySwitching<BE> for Module<BE> where
+    Self: Sized
+        + ModuleN
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigNormalizeTmpBytes
+        + VecZnxNormalizeTmpBytes
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigNormalizeTmpBytes
+        + VmpApplyDftToDft<BE>
+        + VmpApplyDftToDftAdd<BE>
+        + VecZnxDftApply<BE>
+        + VecZnxIdftApplyConsume<BE>
+        + VecZnxBigAddSmallInplace<BE>
+        + VecZnxBigNormalize<BE>
+        + VecZnxNormalize<BE>
+        + VecZnxNormalizeTmpBytes
+{
+}
+
+pub trait GLWEKeySwitching<BE: Backend>
+where
+    Self: Sized
+        + ModuleN
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigNormalizeTmpBytes
+        + VecZnxNormalizeTmpBytes
+        + VecZnxDftBytesOf
+        + VmpApplyDftToDftTmpBytes
+        + VecZnxBigNormalizeTmpBytes
+        + VmpApplyDftToDft<BE>
+        + VmpApplyDftToDftAdd<BE>
+        + VecZnxDftApply<BE>
+        + VecZnxIdftApplyConsume<BE>
+        + VecZnxBigAddSmallInplace<BE>
+        + VecZnxBigNormalize<BE>
+        + VecZnxNormalize<BE>
+        + VecZnxNormalizeTmpBytes,
+{
+    fn glwe_keyswitch_tmp_bytes<R, A, B>(&self, res_infos: &R, a_infos: &A, b_infos: &B) -> usize
+    where
+        R: GLWEInfos,
+        A: GLWEInfos,
+        B: GGLWEInfos,
+    {
+        let in_size: usize = a_infos
             .k()
-            .div_ceil(key_apply.base2k())
+            .div_ceil(b_infos.base2k())
-            .div_ceil(key_apply.dsize().into()) as usize;
+            .div_ceil(b_infos.dsize().into()) as usize;
-        let out_size: usize = out_infos.size();
+        let out_size: usize = res_infos.size();
-        let ksk_size: usize = key_apply.size();
+        let ksk_size: usize = b_infos.size();
-        let res_dft: usize = module.bytes_of_vec_znx_dft((key_apply.rank_out() + 1).into(), ksk_size); // TODO OPTIMIZE
+        let res_dft: usize = self.bytes_of_vec_znx_dft((b_infos.rank_out() + 1).into(), ksk_size); // TODO OPTIMIZE
-        let ai_dft: usize = module.bytes_of_vec_znx_dft((key_apply.rank_in()).into(), in_size);
+        let ai_dft: usize = self.bytes_of_vec_znx_dft((b_infos.rank_in()).into(), in_size);
-        let vmp: usize = module.vmp_apply_dft_to_dft_tmp_bytes(
+        let vmp: usize = self.vmp_apply_dft_to_dft_tmp_bytes(
             out_size,
             in_size,
             in_size,
-            (key_apply.rank_in()).into(),
+            (b_infos.rank_in()).into(),
-            (key_apply.rank_out() + 1).into(),
+            (b_infos.rank_out() + 1).into(),
             ksk_size,
-        ) + module.bytes_of_vec_znx_dft((key_apply.rank_in()).into(), in_size);
+        ) + self.bytes_of_vec_znx_dft((b_infos.rank_in()).into(), in_size);
-        let normalize_big: usize = module.vec_znx_big_normalize_tmp_bytes();
+        let normalize_big: usize = self.vec_znx_big_normalize_tmp_bytes();
-        if in_infos.base2k() == key_apply.base2k() {
+        if a_infos.base2k() == b_infos.base2k() {
             res_dft + ((ai_dft + vmp) | normalize_big)
-        } else if key_apply.dsize() == 1 {
+        } else if b_infos.dsize() == 1 {
             // In this case, we only need one column, temporary, that we can drop once a_dft is computed.
-            let normalize_conv: usize = VecZnx::bytes_of(module.n(), 1, in_size) + module.vec_znx_normalize_tmp_bytes();
+            let normalize_conv: usize = VecZnx::bytes_of(self.n(), 1, in_size) + self.vec_znx_normalize_tmp_bytes();
             res_dft + (((ai_dft + normalize_conv) | vmp) | normalize_big)
         } else {
             // Since we stride over a to get a_dft when dsize > 1, we need to store the full columns of a with in the base conversion.
-            let normalize_conv: usize = VecZnx::bytes_of(module.n(), (key_apply.rank_in()).into(), in_size);
+            let normalize_conv: usize = VecZnx::bytes_of(self.n(), (b_infos.rank_in()).into(), in_size);
-            res_dft + ((ai_dft + normalize_conv + (module.vec_znx_normalize_tmp_bytes() | vmp)) | normalize_big)
+            res_dft + ((ai_dft + normalize_conv + (self.vec_znx_normalize_tmp_bytes() | vmp)) | normalize_big)
         }
     }
 
-    pub fn keyswitch_inplace_scratch_space<B: Backend, OUT, KEY>(module: &Module<B>, out_infos: &OUT, key_apply: &KEY) -> usize
+    fn glwe_keyswitch<R, A, B>(&self, res: &mut R, a: &A, b: &B, scratch: &mut Scratch<BE>)
     where
-        OUT: GLWEInfos,
+        R: GLWEToMut,
-        KEY: GGLWEInfos,
+        A: GLWEToRef,
-        Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
+        B: GLWESwitchingKeyPreparedToRef<BE>,
+        Scratch<BE>: ScratchTakeCore<BE>,
     {
-        Self::keyswitch_scratch_space(module, out_infos, out_infos, key_apply)
+        let res: &mut GLWE<&mut [u8]> = &mut res.to_mut();
-    }
+        let a: &GLWE<&[u8]> = &a.to_ref();
-}
+        let b: &GLWESwitchingKeyPrepared<&[u8], BE> = &b.to_ref();
 
-impl<DataSelf: DataRef> GLWE<DataSelf> {
-    #[allow(dead_code)]
-    pub(crate) fn assert_keyswitch<B: Backend, DataLhs, DataRhs>(
-        &self,
-        module: &Module<B>,
-        lhs: &GLWE<DataLhs>,
-        rhs: &GLWESwitchingKeyPrepared<DataRhs, B>,
-        scratch: &Scratch<B>,
-    ) where
-        DataLhs: DataRef,
-        DataRhs: DataRef,
-        Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
-        Scratch<B>: ScratchAvailable,
-    {
         assert_eq!(
-            lhs.rank(),
+            a.rank(),
-            rhs.rank_in(),
+            b.rank_in(),
-            "lhs.rank(): {} != rhs.rank_in(): {}",
+            "a.rank(): {} != b.rank_in(): {}",
-            lhs.rank(),
+            a.rank(),
-            rhs.rank_in()
+            b.rank_in()
         );
         assert_eq!(
-            self.rank(),
+            res.rank(),
-            rhs.rank_out(),
+            b.rank_out(),
-            "self.rank(): {} != rhs.rank_out(): {}",
+            "res.rank(): {} != b.rank_out(): {}",
-            self.rank(),
+            res.rank(),
-            rhs.rank_out()
+            b.rank_out()
         );
-        assert_eq!(rhs.n(), self.n());
-        assert_eq!(lhs.n(), self.n());
 
-        let scrach_needed: usize = GLWE::keyswitch_scratch_space(module, self, lhs, rhs);
+        assert_eq!(res.n(), self.n() as u32);
+        assert_eq!(a.n(), self.n() as u32);
+        assert_eq!(b.n(), self.n() as u32);
+
+        let scrach_needed: usize = self.glwe_keyswitch_tmp_bytes(res, a, b);
 
         assert!(
             scratch.available() >= scrach_needed,
-            "scratch.available()={} < GLWECiphertext::keyswitch_scratch_space(
+            "scratch.available()={} < glwe_keyswitch_tmp_bytes={scrach_needed}",
-                    module,
-                    self.base2k(),
-                    self.k(),
-                    lhs.base2k(),
-                    lhs.k(),
-                    rhs.base2k(),
-                    rhs.k(),
-                    rhs.dsize(),
-                    rhs.rank_in(),
-                    rhs.rank_out(),
-                )={scrach_needed}",
             scratch.available(),
         );
-    }
 
-    #[allow(dead_code)]
+        let basek_out: usize = res.base2k().into();
-    pub(crate) fn assert_keyswitch_inplace<B: Backend, DataRhs>(
+        let base2k_out: usize = b.base2k().into();
-        &self,
-        module: &Module<B>,
-        rhs: &GLWESwitchingKeyPrepared<DataRhs, B>,
-        scratch: &Scratch<B>,
-    ) where
-        DataRhs: DataRef,
-        Module<B>: VecZnxDftBytesOf + VmpApplyDftToDftTmpBytes + VecZnxBigNormalizeTmpBytes + VecZnxNormalizeTmpBytes,
-        Scratch<B>: ScratchAvailable,
-    {
-        assert_eq!(
-            self.rank(),
-            rhs.rank_out(),
-            "self.rank(): {} != rhs.rank_out(): {}",
-            self.rank(),
-            rhs.rank_out()
-        );
 
-        assert_eq!(rhs.n(), self.n());
+        let (res_dft, scratch_1) = scratch.take_vec_znx_dft(self, (res.rank() + 1).into(), b.size()); // Todo optimise
-
+        let res_big: VecZnxBig<&mut [u8], BE> = keyswitch_internal(self, res_dft, a, b, scratch_1);
-        let scrach_needed: usize = GLWE::keyswitch_inplace_scratch_space(module, self, rhs);
+        (0..(res.rank() + 1).into()).for_each(|i| {
-
+            self.vec_znx_big_normalize(
-        assert!(
-            scratch.available() >= scrach_needed,
-            "scratch.available()={} < GLWECiphertext::keyswitch_scratch_space()={scrach_needed}",
-            scratch.available(),
-        );
-    }
-}
-
-impl<DataSelf: DataMut> GLWE<DataSelf> {
-    pub fn keyswitch<DataLhs: DataRef, DataRhs: DataRef, B: Backend>(
-        &mut self,
-        module: &Module<B>,
-        glwe_in: &GLWE<DataLhs>,
-        rhs: &GLWESwitchingKeyPrepared<DataRhs, B>,
-        scratch: &mut Scratch<B>,
-    ) where
-        Module<B>: VecZnxDftBytesOf
-            + VmpApplyDftToDftTmpBytes
-            + VecZnxBigNormalizeTmpBytes
-            + VmpApplyDftToDft<B>
-            + VmpApplyDftToDftAdd<B>
-            + VecZnxDftApply<B>
-            + VecZnxIdftApplyConsume<B>
-            + VecZnxBigAddSmallInplace<B>
-            + VecZnxBigNormalize<B>
-            + VecZnxNormalize<B>
-            + VecZnxNormalizeTmpBytes,
-        Scratch<B>: ScratchAvailable,
-    {
-        #[cfg(debug_assertions)]
-        {
-            self.assert_keyswitch(module, glwe_in, rhs, scratch);
-        }
-
-        let basek_out: usize = self.base2k().into();
-        let basek_ksk: usize = rhs.base2k().into();
-
-        let (res_dft, scratch_1) = scratch.take_vec_znx_dft(self.n().into(), (self.rank() + 1).into(), rhs.size()); // Todo optimise
-        let res_big: VecZnxBig<_, B> = glwe_in.keyswitch_internal(module, res_dft, rhs, scratch_1);
-        (0..(self.rank() + 1).into()).for_each(|i| {
-            module.vec_znx_big_normalize(
                 basek_out,
-                &mut self.data,
+                &mut res.data,
                 i,
-                basek_ksk,
+                base2k_out,
                 &res_big,
                 i,
                 scratch_1,
@@ -188,42 +181,52 @@ impl<DataSelf: DataMut> GLWE<DataSelf> {
         })
     }
 
-    pub fn keyswitch_inplace<DataRhs: DataRef, B: Backend>(
+    fn glwe_keyswitch_inplace<R, A>(&self, res: &mut R, a: &A, scratch: &mut Scratch<BE>)
-        &mut self,
+    where
-        module: &Module<B>,
+        R: GLWEToMut,
-        rhs: &GLWESwitchingKeyPrepared<DataRhs, B>,
+        A: GLWESwitchingKeyPreparedToRef<BE>,
-        scratch: &mut Scratch<B>,
+        Scratch<BE>: ScratchTakeCore<BE>,
-    ) where
-        Module<B>: VecZnxDftBytesOf
-            + VmpApplyDftToDftTmpBytes
-            + VecZnxBigNormalizeTmpBytes
-            + VmpApplyDftToDftTmpBytes
-            + VmpApplyDftToDft<B>
-            + VmpApplyDftToDftAdd<B>
-            + VecZnxDftApply<B>
-            + VecZnxIdftApplyConsume<B>
-            + VecZnxBigAddSmallInplace<B>
-            + VecZnxBigNormalize<B>
-            + VecZnxNormalize<B>
-            + VecZnxNormalizeTmpBytes,
-        Scratch<B>: ScratchAvailable,
     {
-        #[cfg(debug_assertions)]
+        let res: &mut GLWE<&mut [u8]> = &mut res.to_mut();
-        {
+        let a: &GLWESwitchingKeyPrepared<&[u8], BE> = &a.to_ref();
-            self.assert_keyswitch_inplace(module, rhs, scratch);
-        }
 
-        let basek_in: usize = self.base2k().into();
+        assert_eq!(
-        let basek_ksk: usize = rhs.base2k().into();
+            res.rank(),
+            a.rank_in(),
+            "res.rank(): {} != a.rank_in(): {}",
+            res.rank(),
+            a.rank_in()
+        );
+        assert_eq!(
+            res.rank(),
+            a.rank_out(),
+            "res.rank(): {} != b.rank_out(): {}",
+            res.rank(),
+            a.rank_out()
+        );
 
-        let (res_dft, scratch_1) = scratch.take_vec_znx_dft(self.n().into(), (self.rank() + 1).into(), rhs.size()); // Todo optimise
+        assert_eq!(res.n(), self.n() as u32);
-        let res_big: VecZnxBig<_, B> = self.keyswitch_internal(module, res_dft, rhs, scratch_1);
+        assert_eq!(a.n(), self.n() as u32);
-        (0..(self.rank() + 1).into()).for_each(|i| {
+
-            module.vec_znx_big_normalize(
+        let scrach_needed: usize = self.glwe_keyswitch_tmp_bytes(res, res, a);
-                basek_in,
+
-                &mut self.data,
+        assert!(
+            scratch.available() >= scrach_needed,
+            "scratch.available()={} < glwe_keyswitch_tmp_bytes={scrach_needed}",
+            scratch.available(),
+        );
+
+        let base2k_in: usize = res.base2k().into();
+        let base2k_out: usize = a.base2k().into();
+
+        let (res_dft, scratch_1) = scratch.take_vec_znx_dft(self, (res.rank() + 1).into(), a.size()); // Todo optimise
+        let res_big: VecZnxBig<&mut [u8], BE> = keyswitch_internal(self, res_dft, res, a, scratch_1);
+        (0..(res.rank() + 1).into()).for_each(|i| {
+            self.vec_znx_big_normalize(
+                base2k_in,
+                &mut res.data,
                 i,
-                basek_ksk,
+                base2k_out,
                 &res_big,
                 i,
                 scratch_1,
@@ -232,129 +235,72 @@ impl<DataSelf: DataMut> GLWE<DataSelf> {
     }
 }
 
-impl<D: DataRef> GLWE<D> {
+impl GLWE<Vec<u8>> {}
-    pub(crate) fn keyswitch_internal<B: Backend, DataRes, DataKey>(
+
-        &self,
+impl<DataSelf: DataMut> GLWE<DataSelf> {}
-        module: &Module<B>,
+
-        res_dft: VecZnxDft<DataRes, B>,
+fn keyswitch_internal<BE: Backend, M, DR, DA, DB>(
-        rhs: &GLWESwitchingKeyPrepared<DataKey, B>,
+    module: &M,
-        scratch: &mut Scratch<B>,
+    mut res: VecZnxDft<DR, BE>,
-    ) -> VecZnxBig<DataRes, B>
+    a: &GLWE<DA>,
+    b: &GLWESwitchingKeyPrepared<DB, BE>,
+    scratch: &mut Scratch<BE>,
+) -> VecZnxBig<DR, BE>
 where
-        DataRes: DataMut,
+    DR: DataMut,
-        DataKey: DataRef,
+    DA: DataRef,
-        Module<B>: VecZnxDftBytesOf
+    DB: DataRef,
+    M: ModuleN
+        + VecZnxDftBytesOf
         + VmpApplyDftToDftTmpBytes
         + VecZnxBigNormalizeTmpBytes
         + VmpApplyDftToDftTmpBytes
-            + VmpApplyDftToDft<B>
+        + VmpApplyDftToDft<BE>
-            + VmpApplyDftToDftAdd<B>
+        + VmpApplyDftToDftAdd<BE>
-            + VecZnxDftApply<B>
+        + VecZnxDftApply<BE>
-            + VecZnxIdftApplyConsume<B>
+        + VecZnxIdftApplyConsume<BE>
-            + VecZnxBigAddSmallInplace<B>
+        + VecZnxBigAddSmallInplace<BE>
-            + VecZnxBigNormalize<B>
+        + VecZnxBigNormalize<BE>
-            + VecZnxNormalize<B>,
+        + VecZnxNormalize<BE>,
-        Scratch<B>:,
+    Scratch<BE>: ScratchTakeCore<BE>,
 {
-        if rhs.dsize() == 1 {
+    let base2k_in: usize = a.base2k().into();
-            return keyswitch_vmp_one_digit(
+    let base2k_out: usize = b.base2k().into();
-                module,
+    let cols: usize = (a.rank() + 1).into();
-                self.base2k().into(),
+    let a_size: usize = (a.size() * base2k_in).div_ceil(base2k_out);
-                rhs.base2k().into(),
+    let pmat: &VmpPMat<DB, BE> = &b.key.data;
-                res_dft,
-                &self.data,
-                &rhs.key.data,
-                scratch,
-            );
-        }
 
-        keyswitch_vmp_multiple_digits(
+    if b.dsize() == 1 {
-            module,
+        let (mut ai_dft, scratch_1) = scratch.take_vec_znx_dft(module, cols - 1, a.size());
-            self.base2k().into(),
-            rhs.base2k().into(),
-            res_dft,
-            &self.data,
-            &rhs.key.data,
-            rhs.dsize().into(),
-            scratch,
-        )
-    }
-}
 
-fn keyswitch_vmp_one_digit<B: Backend, DataRes, DataIn, DataVmp>(
+        if base2k_in == base2k_out {
-    module: &Module<B>,
-    basek_in: usize,
-    basek_ksk: usize,
-    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>: VecZnxDftBytesOf
-        + VecZnxDftApply<B>
-        + VmpApplyDftToDft<B>
-        + VecZnxIdftApplyConsume<B>
-        + VecZnxBigAddSmallInplace<B>
-        + VecZnxNormalize<B>,
-    Scratch<B>:,
-{
-    let cols: usize = a.cols();
-
-    let a_size: usize = (a.size() * basek_in).div_ceil(basek_ksk);
-    let (mut ai_dft, scratch_1) = scratch.take_vec_znx_dft(a.n(), cols - 1, a.size());
-
-    if basek_in == basek_ksk {
             (0..cols - 1).for_each(|col_i| {
-            module.vec_znx_dft_apply(1, 0, &mut ai_dft, col_i, a, col_i + 1);
+                module.vec_znx_dft_apply(1, 0, &mut ai_dft, col_i, a.data(), col_i + 1);
             });
         } else {
-        let (mut a_conv, scratch_2) = scratch_1.take_vec_znx(a.n(), 1, a_size);
+            let (mut a_conv, scratch_2) = scratch_1.take_vec_znx(module, 1, a_size);
             (0..cols - 1).for_each(|col_i| {
-            module.vec_znx_normalize(basek_ksk, &mut a_conv, 0, basek_in, a, col_i + 1, scratch_2);
+                module.vec_znx_normalize(
+                    base2k_out,
+                    &mut a_conv,
+                    0,
+                    base2k_in,
+                    a.data(),
+                    col_i + 1,
+                    scratch_2,
+                );
                 module.vec_znx_dft_apply(1, 0, &mut ai_dft, col_i, &a_conv, 0);
             });
         }
 
-    module.vmp_apply_dft_to_dft(&mut res_dft, &ai_dft, mat, scratch_1);
+        module.vmp_apply_dft_to_dft(&mut res, &ai_dft, pmat, scratch_1);
-    let mut res_big: VecZnxBig<DataRes, B> = module.vec_znx_idft_apply_consume(res_dft);
+    } else {
-    module.vec_znx_big_add_small_inplace(&mut res_big, 0, a, 0);
+        let dsize: usize = b.dsize().into();
-    res_big
-}
 
-#[allow(clippy::too_many_arguments)]
+        let (mut ai_dft, scratch_1) = scratch.take_vec_znx_dft(module, cols - 1, a_size.div_ceil(dsize));
-fn keyswitch_vmp_multiple_digits<B: Backend, DataRes, DataIn, DataVmp>(
-    module: &Module<B>,
-    basek_in: usize,
-    basek_ksk: usize,
-    mut res_dft: VecZnxDft<DataRes, B>,
-    a: &VecZnx<DataIn>,
-    mat: &VmpPMat<DataVmp, B>,
-    dsize: usize,
-    scratch: &mut Scratch<B>,
-) -> VecZnxBig<DataRes, B>
-where
-    DataRes: DataMut,
-    DataIn: DataRef,
-    DataVmp: DataRef,
-    Module<B>: VecZnxDftBytesOf
-        + VecZnxDftApply<B>
-        + VmpApplyDftToDft<B>
-        + VmpApplyDftToDftAdd<B>
-        + VecZnxIdftApplyConsume<B>
-        + VecZnxBigAddSmallInplace<B>
-        + VecZnxNormalize<B>,
-    Scratch<B>:,
-{
-    let cols: usize = a.cols();
-    let a_size: usize = (a.size() * basek_in).div_ceil(basek_ksk);
-    let (mut ai_dft, scratch_1) = scratch.take_vec_znx_dft(a.n(), cols - 1, a_size.div_ceil(dsize));
         ai_dft.data_mut().fill(0);
 
-    if basek_in == basek_ksk {
+        if base2k_in == base2k_out {
             for di in 0..dsize {
                 ai_dft.set_size((a_size + di) / dsize);
 
@@ -365,22 +311,30 @@ where
                 // It is possible to further ignore the last dsize-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() - ((dsize - di) as isize - 2).max(0) as usize);
+                res.set_size(pmat.size() - ((dsize - di) as isize - 2).max(0) as usize);
 
                 for j in 0..cols - 1 {
-                module.vec_znx_dft_apply(dsize, dsize - di - 1, &mut ai_dft, j, a, j + 1);
+                    module.vec_znx_dft_apply(dsize, dsize - di - 1, &mut ai_dft, j, a.data(), j + 1);
                 }
 
                 if di == 0 {
-                module.vmp_apply_dft_to_dft(&mut res_dft, &ai_dft, mat, scratch_1);
+                    module.vmp_apply_dft_to_dft(&mut res, &ai_dft, pmat, scratch_1);
                 } else {
-                module.vmp_apply_dft_to_dft_add(&mut res_dft, &ai_dft, mat, di, scratch_1);
+                    module.vmp_apply_dft_to_dft_add(&mut res, &ai_dft, pmat, di, scratch_1);
                 }
             }
         } else {
-        let (mut a_conv, scratch_2) = scratch_1.take_vec_znx(a.n(), cols - 1, a_size);
+            let (mut a_conv, scratch_2) = scratch_1.take_vec_znx(module, cols - 1, a_size);
             for j in 0..cols - 1 {
-            module.vec_znx_normalize(basek_ksk, &mut a_conv, j, basek_in, a, j + 1, scratch_2);
+                module.vec_znx_normalize(
+                    base2k_out,
+                    &mut a_conv,
+                    j,
+                    base2k_in,
+                    a.data(),
+                    j + 1,
+                    scratch_2,
+                );
             }
 
             for di in 0..dsize {
@@ -393,22 +347,24 @@ where
                 // It is possible to further ignore the last dsize-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() - ((dsize - di) as isize - 2).max(0) as usize);
+                res.set_size(pmat.size() - ((dsize - di) as isize - 2).max(0) as usize);
 
                 for j in 0..cols - 1 {
                     module.vec_znx_dft_apply(dsize, dsize - di - 1, &mut ai_dft, j, &a_conv, j);
                 }
 
                 if di == 0 {
-                module.vmp_apply_dft_to_dft(&mut res_dft, &ai_dft, mat, scratch_2);
+                    module.vmp_apply_dft_to_dft(&mut res, &ai_dft, pmat, scratch_2);
                 } else {
-                module.vmp_apply_dft_to_dft_add(&mut res_dft, &ai_dft, mat, di, scratch_2);
+                    module.vmp_apply_dft_to_dft_add(&mut res, &ai_dft, pmat, di, scratch_2);
                 }
             }
         }
 
-    res_dft.set_size(res_dft.max_size());
+        res.set_size(res.max_size());
-    let mut res_big: VecZnxBig<DataRes, B> = module.vec_znx_idft_apply_consume(res_dft);
+    }
-    module.vec_znx_big_add_small_inplace(&mut res_big, 0, a, 0);
+
+    let mut res_big: VecZnxBig<DR, BE> = module.vec_znx_idft_apply_consume(res);
+    module.vec_znx_big_add_small_inplace(&mut res_big, 0, a.data(), 0);
     res_big
 }

poulpy-core/src/keyswitching/lwe_ct.rs

@@ -10,7 +10,7 @@ use poulpy_hal::{
 use crate::layouts::{GGLWEInfos, GLWE, GLWELayout, LWE, LWEInfos, Rank, TorusPrecision, prepared::LWESwitchingKeyPrepared};
 
 impl LWE<Vec<u8>> {
-    pub fn keyswitch_scratch_space<B: Backend, OUT, IN, KEY>(
+    pub fn keyswitch_tmp_bytes<B: Backend, OUT, IN, KEY>(
         module: &Module<B>,
         out_infos: &OUT,
         in_infos: &IN,
@@ -50,7 +50,7 @@ impl LWE<Vec<u8>> {
 
         let glwe_in: usize = GLWE::bytes_of_from_infos(module, &glwe_in_infos);
         let glwe_out: usize = GLWE::bytes_of_from_infos(module, &glwe_out_infos);
-        let ks: usize = GLWE::keyswitch_scratch_space(module, &glwe_out_infos, &glwe_in_infos, key_infos);
+        let ks: usize = GLWE::keyswitch_tmp_bytes(module, &glwe_out_infos, &glwe_in_infos, key_infos);
 
         glwe_in + glwe_out + ks
     }
@@ -84,7 +84,7 @@ impl<DLwe: DataMut> LWE<DLwe> {
         {
             assert!(self.n() <= module.n() as u32);
             assert!(a.n() <= module.n() as u32);
-            assert!(scratch.available() >= LWE::keyswitch_scratch_space(module, self, a, ksk));
+            assert!(scratch.available() >= LWE::keyswitch_tmp_bytes(module, self, a, ksk));
         }
 
         let max_k: TorusPrecision = self.k().max(a.k());

poulpy-core/src/lib.rs

@@ -14,6 +14,7 @@ mod utils;
 
 pub use operations::*;
 pub mod layouts;
+pub use conversion::*;
 pub use dist::*;
 pub use external_product::*;
 pub use glwe_packing::*;

poulpy-core/src/noise/gglwe_ct.rs

@@ -35,7 +35,7 @@ impl<D: DataRef> GGLWE<D> {
         let dsize: usize = self.dsize().into();
         let base2k: usize = self.base2k().into();
 
-        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWE::decrypt_scratch_space(module, self));
+        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWE::decrypt_tmp_bytes(module, self));
         let mut pt: GLWEPlaintext<Vec<u8>> = GLWEPlaintext::alloc_from_infos(module, self);
 
         (0..self.rank_in().into()).for_each(|col_i| {

poulpy-core/src/noise/ggsw_ct.rs

@@ -48,7 +48,7 @@ impl<D: DataRef> GGSW<D> {
         let mut pt_big: VecZnxBig<Vec<u8>, B> = module.vec_znx_big_alloc(1, self.size());
 
         let mut scratch: ScratchOwned<B> =
-            ScratchOwned::alloc(GLWE::decrypt_scratch_space(module, self) | module.vec_znx_normalize_tmp_bytes());
+            ScratchOwned::alloc(GLWE::decrypt_tmp_bytes(module, self) | module.vec_znx_normalize_tmp_bytes());
 
         (0..(self.rank() + 1).into()).for_each(|col_j| {
             (0..self.dnum().into()).for_each(|row_i| {
@@ -120,7 +120,7 @@ impl<D: DataRef> GGSW<D> {
         let mut pt_big: VecZnxBig<Vec<u8>, B> = module.vec_znx_big_alloc(1, self.size());
 
         let mut scratch: ScratchOwned<B> =
-            ScratchOwned::alloc(GLWE::decrypt_scratch_space(module, self) | module.vec_znx_normalize_tmp_bytes());
+            ScratchOwned::alloc(GLWE::decrypt_tmp_bytes(module, self) | module.vec_znx_normalize_tmp_bytes());
 
         (0..(self.rank() + 1).into()).for_each(|col_j| {
             (0..self.dnum().into()).for_each(|row_i| {

poulpy-core/src/noise/glwe_ct.rs

@@ -61,7 +61,7 @@ impl<D: DataRef> GLWE<D> {
             + VecZnxNormalizeInplace<B>,
         B: Backend + ScratchOwnedAllocImpl<B> + ScratchOwnedBorrowImpl<B>,
     {
-        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWE::decrypt_scratch_space(module, self));
+        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWE::decrypt_tmp_bytes(module, self));
         let noise_have: f64 = self.noise(module, sk_prepared, pt_want, scratch.borrow());
         assert!(noise_have <= max_noise, "{noise_have} {max_noise}");
     }

poulpy-core/src/operations/glwe.rs

@@ -246,10 +246,10 @@ pub trait GLWEOperations: GLWEToMut + GLWEInfos + SetGLWEInfos + Sized {
         });
     }
 
-    fn copy<A, B: Backend>(&mut self, module: &Module<B>, a: &A)
+    fn copy<A, M>(&mut self, module: &M, a: &A)
     where
         A: GLWEToRef + GLWEInfos,
-        Module<B>: VecZnxCopy,
+        M: VecZnxCopy,
     {
         #[cfg(debug_assertions)]
         {
@@ -319,8 +319,8 @@ pub trait GLWEOperations: GLWEToMut + GLWEInfos + SetGLWEInfos + Sized {
 }
 
 impl GLWE<Vec<u8>> {
-    pub fn rsh_scratch_space(n: usize) -> usize {
+    pub fn rsh_tmp_bytes(n: usize) -> usize {
-        VecZnx::rsh_scratch_space(n)
+        VecZnx::rsh_tmp_bytes(n)
     }
 }
 

poulpy-core/src/tests/test_suite/automorphism/gglwe_atk.rs

@@ -120,9 +120,9 @@ where
             let mut source_xa: Source = Source::new([0u8; 32]);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                AutomorphismKey::encrypt_sk_scratch_space(module, &auto_key_in_infos)
+                AutomorphismKey::encrypt_sk_tmp_bytes(module, &auto_key_in_infos)
-                    | AutomorphismKey::encrypt_sk_scratch_space(module, &auto_key_apply_infos)
+                    | AutomorphismKey::encrypt_sk_tmp_bytes(module, &auto_key_apply_infos)
-                    | AutomorphismKey::automorphism_scratch_space(
+                    | AutomorphismKey::automorphism_tmp_bytes(
                         module,
                         &auto_key_out_infos,
                         &auto_key_in_infos,
@@ -319,9 +319,9 @@ where
             let mut source_xa: Source = Source::new([0u8; 32]);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                AutomorphismKey::encrypt_sk_scratch_space(module, &auto_key)
+                AutomorphismKey::encrypt_sk_tmp_bytes(module, &auto_key)
-                    | AutomorphismKey::encrypt_sk_scratch_space(module, &auto_key_apply)
+                    | AutomorphismKey::encrypt_sk_tmp_bytes(module, &auto_key_apply)
-                    | AutomorphismKey::automorphism_inplace_scratch_space(module, &auto_key, &auto_key_apply),
+                    | AutomorphismKey::automorphism_inplace_tmp_bytes(module, &auto_key, &auto_key_apply),
             );
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&auto_key);

poulpy-core/src/tests/test_suite/automorphism/ggsw_ct.rs

@@ -139,10 +139,10 @@ where
             let mut source_xa: Source = Source::new([0u8; 32]);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ct_in)
+                GGSW::encrypt_sk_tmp_bytes(module, &ct_in)
-                    | AutomorphismKey::encrypt_sk_scratch_space(module, &auto_key)
+                    | AutomorphismKey::encrypt_sk_tmp_bytes(module, &auto_key)
-                    | TensorKey::encrypt_sk_scratch_space(module, &tensor_key)
+                    | TensorKey::encrypt_sk_tmp_bytes(module, &tensor_key)
-                    | GGSW::automorphism_scratch_space(module, &ct_out, &ct_in, &auto_key, &tensor_key),
+                    | GGSW::automorphism_tmp_bytes(module, &ct_out, &ct_in, &auto_key, &tensor_key),
             );
 
             let var_xs: f64 = 0.5;
@@ -319,10 +319,10 @@ where
             let mut source_xa: Source = Source::new([0u8; 32]);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ct)
+                GGSW::encrypt_sk_tmp_bytes(module, &ct)
-                    | AutomorphismKey::encrypt_sk_scratch_space(module, &auto_key)
+                    | AutomorphismKey::encrypt_sk_tmp_bytes(module, &auto_key)
-                    | TensorKey::encrypt_sk_scratch_space(module, &tensor_key)
+                    | TensorKey::encrypt_sk_tmp_bytes(module, &tensor_key)
-                    | GGSW::automorphism_inplace_scratch_space(module, &ct, &auto_key, &tensor_key),
+                    | GGSW::automorphism_inplace_tmp_bytes(module, &ct, &auto_key, &tensor_key),
             );
 
             let var_xs: f64 = 0.5;

poulpy-core/src/tests/test_suite/automorphism/glwe_ct.rs

@@ -112,10 +112,10 @@ where
             module.vec_znx_fill_uniform(base2k, &mut pt_want.data, 0, &mut source_xa);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                AutomorphismKey::encrypt_sk_scratch_space(module, &autokey)
+                AutomorphismKey::encrypt_sk_tmp_bytes(module, &autokey)
-                    | GLWE::decrypt_scratch_space(module, &ct_out)
+                    | GLWE::decrypt_tmp_bytes(module, &ct_out)
-                    | GLWE::encrypt_sk_scratch_space(module, &ct_in)
+                    | GLWE::encrypt_sk_tmp_bytes(module, &ct_in)
-                    | GLWE::automorphism_scratch_space(module, &ct_out, &ct_in, &autokey),
+                    | GLWE::automorphism_tmp_bytes(module, &ct_out, &ct_in, &autokey),
             );
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&ct_out);
@@ -246,10 +246,10 @@ where
             module.vec_znx_fill_uniform(base2k, &mut pt_want.data, 0, &mut source_xa);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                AutomorphismKey::encrypt_sk_scratch_space(module, &autokey)
+                AutomorphismKey::encrypt_sk_tmp_bytes(module, &autokey)
-                    | GLWE::decrypt_scratch_space(module, &ct)
+                    | GLWE::decrypt_tmp_bytes(module, &ct)
-                    | GLWE::encrypt_sk_scratch_space(module, &ct)
+                    | GLWE::encrypt_sk_tmp_bytes(module, &ct)
-                    | GLWE::automorphism_inplace_scratch_space(module, &ct, &autokey),
+                    | GLWE::automorphism_inplace_tmp_bytes(module, &ct, &autokey),
             );
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&ct);

poulpy-core/src/tests/test_suite/conversion.rs

@@ -96,9 +96,9 @@ where
     };
 
     let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-        LWEToGLWESwitchingKey::encrypt_sk_scratch_space(module, &lwe_to_glwe_infos)
+        LWEToGLWESwitchingKey::encrypt_sk_tmp_bytes(module, &lwe_to_glwe_infos)
-            | GLWE::from_lwe_scratch_space(module, &glwe_infos, &lwe_infos, &lwe_to_glwe_infos)
+            | GLWE::from_lwe_tmp_bytes(module, &glwe_infos, &lwe_infos, &lwe_to_glwe_infos)
-            | GLWE::decrypt_scratch_space(module, &glwe_infos),
+            | GLWE::decrypt_tmp_bytes(module, &glwe_infos),
     );
 
     let mut sk_glwe: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_infos);
@@ -213,9 +213,9 @@ where
     let mut source_xe: Source = Source::new([0u8; 32]);
 
     let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-        GLWEToLWESwitchingKey::encrypt_sk_scratch_space(module, &glwe_to_lwe_infos)
+        GLWEToLWESwitchingKey::encrypt_sk_tmp_bytes(module, &glwe_to_lwe_infos)
-            | LWE::from_glwe_scratch_space(module, &lwe_infos, &glwe_infos, &glwe_to_lwe_infos)
+            | LWE::from_glwe_tmp_bytes(module, &lwe_infos, &glwe_infos, &glwe_to_lwe_infos)
-            | GLWE::decrypt_scratch_space(module, &glwe_infos),
+            | GLWE::decrypt_tmp_bytes(module, &glwe_infos),
     );
 
     let mut sk_glwe: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_infos);

poulpy-core/src/tests/test_suite/encryption/gglwe_atk.rs

@@ -90,7 +90,7 @@ where
             let mut source_xe: Source = Source::new([0u8; 32]);
             let mut source_xa: Source = Source::new([0u8; 32]);
 
-            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(AutomorphismKey::encrypt_sk_scratch_space(
+            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(AutomorphismKey::encrypt_sk_tmp_bytes(
                 module, &atk_infos,
             ));
 
@@ -192,7 +192,7 @@ where
             let mut source_xs: Source = Source::new([0u8; 32]);
             let mut source_xe: Source = Source::new([0u8; 32]);
 
-            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(AutomorphismKey::encrypt_sk_scratch_space(
+            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(AutomorphismKey::encrypt_sk_tmp_bytes(
                 module, &atk_infos,
             ));
 

poulpy-core/src/tests/test_suite/encryption/gglwe_ct.rs

@@ -87,7 +87,7 @@ where
                 let mut source_xe: Source = Source::new([0u8; 32]);
                 let mut source_xa: Source = Source::new([0u8; 32]);
 
-                let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKey::encrypt_sk_scratch_space(
+                let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKey::encrypt_sk_tmp_bytes(
                     module,
                     &gglwe_infos,
                 ));
@@ -179,7 +179,7 @@ where
                 let mut source_xs: Source = Source::new([0u8; 32]);
                 let mut source_xe: Source = Source::new([0u8; 32]);
 
-                let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKeyCompressed::encrypt_sk_scratch_space(
+                let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKeyCompressed::encrypt_sk_tmp_bytes(
                     module,
                     &gglwe_infos,
                 ));

poulpy-core/src/tests/test_suite/encryption/ggsw_ct.rs

@@ -55,7 +55,7 @@ where
 
             pt_scalar.fill_ternary_hw(0, n, &mut source_xs);
 
-            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GGSW::encrypt_sk_scratch_space(module, &ggsw_infos));
+            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GGSW::encrypt_sk_tmp_bytes(module, &ggsw_infos));
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&ggsw_infos);
             sk.fill_ternary_prob(0.5, &mut source_xs);
@@ -144,7 +144,7 @@ where
 
             pt_scalar.fill_ternary_hw(0, n, &mut source_xs);
 
-            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GGSWCompressed::encrypt_sk_scratch_space(
+            let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(GGSWCompressed::encrypt_sk_tmp_bytes(
                 module,
                 &ggsw_infos,
             ));

poulpy-core/src/tests/test_suite/encryption/glwe_ct.rs

@@ -85,7 +85,7 @@ where
         let mut source_xa: Source = Source::new([0u8; 32]);
 
         let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-            GLWE::encrypt_sk_scratch_space(module, &glwe_infos) | GLWE::decrypt_scratch_space(module, &glwe_infos),
+            GLWE::encrypt_sk_tmp_bytes(module, &glwe_infos) | GLWE::decrypt_tmp_bytes(module, &glwe_infos),
         );
 
         let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_infos);
@@ -178,7 +178,7 @@ where
         let mut source_xa: Source = Source::new([0u8; 32]);
 
         let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-            GLWECompressed::encrypt_sk_scratch_space(module, &glwe_infos) | GLWE::decrypt_scratch_space(module, &glwe_infos),
+            GLWECompressed::encrypt_sk_tmp_bytes(module, &glwe_infos) | GLWE::decrypt_tmp_bytes(module, &glwe_infos),
         );
 
         let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_infos);
@@ -269,7 +269,7 @@ where
         let mut source_xa: Source = Source::new([0u8; 32]);
 
         let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-            GLWE::decrypt_scratch_space(module, &glwe_infos) | GLWE::encrypt_sk_scratch_space(module, &glwe_infos),
+            GLWE::decrypt_tmp_bytes(module, &glwe_infos) | GLWE::encrypt_sk_tmp_bytes(module, &glwe_infos),
         );
 
         let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_infos);
@@ -349,9 +349,9 @@ where
         let mut source_xu: Source = Source::new([0u8; 32]);
 
         let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-            GLWE::encrypt_sk_scratch_space(module, &glwe_infos)
+            GLWE::encrypt_sk_tmp_bytes(module, &glwe_infos)
-                | GLWE::decrypt_scratch_space(module, &glwe_infos)
+                | GLWE::decrypt_tmp_bytes(module, &glwe_infos)
-                | GLWE::encrypt_pk_scratch_space(module, &glwe_infos),
+                | GLWE::encrypt_pk_tmp_bytes(module, &glwe_infos),
         );
 
         let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_infos);

poulpy-core/src/tests/test_suite/encryption/glwe_tsk.rs

@@ -86,7 +86,7 @@ where
         let mut source_xe: Source = Source::new([0u8; 32]);
         let mut source_xa: Source = Source::new([0u8; 32]);
 
-        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(TensorKey::encrypt_sk_scratch_space(
+        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(TensorKey::encrypt_sk_tmp_bytes(
             module,
             &tensor_key_infos,
         ));
@@ -204,7 +204,7 @@ where
         let mut source_xs: Source = Source::new([0u8; 32]);
         let mut source_xe: Source = Source::new([0u8; 32]);
 
-        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(TensorKeyCompressed::encrypt_sk_scratch_space(
+        let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(TensorKeyCompressed::encrypt_sk_tmp_bytes(
             module,
             &tensor_key_infos,
         ));

poulpy-core/src/tests/test_suite/external_product/gglwe_ksk.rs

@@ -121,14 +121,14 @@ where
                 let mut source_xa: Source = Source::new([0u8; 32]);
 
                 let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                    GLWESwitchingKey::encrypt_sk_scratch_space(module, &gglwe_in_infos)
+                    GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &gglwe_in_infos)
-                        | GLWESwitchingKey::external_product_scratch_space(
+                        | GLWESwitchingKey::external_product_tmp_bytes(
                             module,
                             &gglwe_out_infos,
                             &gglwe_in_infos,
                             &ggsw_infos,
                         )
-                        | GGSW::encrypt_sk_scratch_space(module, &ggsw_infos),
+                        | GGSW::encrypt_sk_tmp_bytes(module, &ggsw_infos),
                 );
 
                 let r: usize = 1;
@@ -292,9 +292,9 @@ where
                 let mut source_xa: Source = Source::new([0u8; 32]);
 
                 let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                    GLWESwitchingKey::encrypt_sk_scratch_space(module, &gglwe_out_infos)
+                    GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &gglwe_out_infos)
-                        | GLWESwitchingKey::external_product_inplace_scratch_space(module, &gglwe_out_infos, &ggsw_infos)
+                        | GLWESwitchingKey::external_product_inplace_tmp_bytes(module, &gglwe_out_infos, &ggsw_infos)
-                        | GGSW::encrypt_sk_scratch_space(module, &ggsw_infos),
+                        | GGSW::encrypt_sk_tmp_bytes(module, &ggsw_infos),
                 );
 
                 let r: usize = 1;

poulpy-core/src/tests/test_suite/external_product/ggsw_ct.rs

@@ -128,9 +128,9 @@ where
             pt_apply.to_mut().raw_mut()[k] = 1; //X^{k}
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ggsw_apply_infos)
+                GGSW::encrypt_sk_tmp_bytes(module, &ggsw_apply_infos)
-                    | GGSW::encrypt_sk_scratch_space(module, &ggsw_in_infos)
+                    | GGSW::encrypt_sk_tmp_bytes(module, &ggsw_in_infos)
-                    | GGSW::external_product_scratch_space(module, &ggsw_out_infos, &ggsw_in_infos, &ggsw_apply_infos),
+                    | GGSW::external_product_tmp_bytes(module, &ggsw_out_infos, &ggsw_in_infos, &ggsw_apply_infos),
             );
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc(n.into(), rank.into());
@@ -282,9 +282,9 @@ where
             pt_apply.to_mut().raw_mut()[k] = 1; //X^{k}
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ggsw_apply_infos)
+                GGSW::encrypt_sk_tmp_bytes(module, &ggsw_apply_infos)
-                    | GGSW::encrypt_sk_scratch_space(module, &ggsw_out_infos)
+                    | GGSW::encrypt_sk_tmp_bytes(module, &ggsw_out_infos)
-                    | GGSW::external_product_inplace_scratch_space(module, &ggsw_out_infos, &ggsw_apply_infos),
+                    | GGSW::external_product_inplace_tmp_bytes(module, &ggsw_out_infos, &ggsw_apply_infos),
             );
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc(n.into(), rank.into());

poulpy-core/src/tests/test_suite/external_product/glwe_ct.rs

@@ -116,9 +116,9 @@ where
             pt_ggsw.raw_mut()[k] = 1; // X^{k}
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ggsw_apply_infos)
+                GGSW::encrypt_sk_tmp_bytes(module, &ggsw_apply_infos)
-                    | GLWE::encrypt_sk_scratch_space(module, &glwe_in_infos)
+                    | GLWE::encrypt_sk_tmp_bytes(module, &glwe_in_infos)
-                    | GLWE::external_product_scratch_space(module, &glwe_out_infos, &glwe_in_infos, &ggsw_apply_infos),
+                    | GLWE::external_product_tmp_bytes(module, &glwe_out_infos, &glwe_in_infos, &ggsw_apply_infos),
             );
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc(n.into(), rank.into());
@@ -259,9 +259,9 @@ where
             pt_ggsw.raw_mut()[k] = 1; // X^{k}
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ggsw_apply_infos)
+                GGSW::encrypt_sk_tmp_bytes(module, &ggsw_apply_infos)
-                    | GLWE::encrypt_sk_scratch_space(module, &glwe_out_infos)
+                    | GLWE::encrypt_sk_tmp_bytes(module, &glwe_out_infos)
-                    | GLWE::external_product_inplace_scratch_space(module, &glwe_out_infos, &ggsw_apply_infos),
+                    | GLWE::external_product_inplace_tmp_bytes(module, &glwe_out_infos, &ggsw_apply_infos),
             );
 
             let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc(n.into(), rank.into());

poulpy-core/src/tests/test_suite/keyswitch/gglwe_ct.rs

@@ -119,11 +119,11 @@ where
                     let mut source_xa: Source = Source::new([0u8; 32]);
 
                     let mut scratch_enc: ScratchOwned<B> = ScratchOwned::alloc(
-                        GLWESwitchingKey::encrypt_sk_scratch_space(module, &gglwe_s0s1_infos)
+                        GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &gglwe_s0s1_infos)
-                            | GLWESwitchingKey::encrypt_sk_scratch_space(module, &gglwe_s1s2_infos)
+                            | GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &gglwe_s1s2_infos)
-                            | GLWESwitchingKey::encrypt_sk_scratch_space(module, &gglwe_s0s2_infos),
+                            | GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &gglwe_s0s2_infos),
                     );
-                    let mut scratch_apply: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKey::keyswitch_scratch_space(
+                    let mut scratch_apply: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKey::keyswitch_tmp_bytes(
                         module,
                         &gglwe_s0s1_infos,
                         &gglwe_s0s2_infos,
@@ -274,10 +274,10 @@ where
                 let mut source_xa: Source = Source::new([0u8; 32]);
 
                 let mut scratch_enc: ScratchOwned<B> = ScratchOwned::alloc(
-                    GLWESwitchingKey::encrypt_sk_scratch_space(module, &gglwe_s0s1_infos)
+                    GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &gglwe_s0s1_infos)
-                        | GLWESwitchingKey::encrypt_sk_scratch_space(module, &gglwe_s1s2_infos),
+                        | GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &gglwe_s1s2_infos),
                 );
-                let mut scratch_apply: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKey::keyswitch_inplace_scratch_space(
+                let mut scratch_apply: ScratchOwned<B> = ScratchOwned::alloc(GLWESwitchingKey::keyswitch_inplace_tmp_bytes(
                     module,
                     &gglwe_s0s1_infos,
                     &gglwe_s1s2_infos,

poulpy-core/src/tests/test_suite/keyswitch/ggsw_ct.rs

@@ -132,10 +132,10 @@ where
             let mut source_xa: Source = Source::new([0u8; 32]);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ggsw_in_infos)
+                GGSW::encrypt_sk_tmp_bytes(module, &ggsw_in_infos)
-                    | GLWESwitchingKey::encrypt_sk_scratch_space(module, &ksk_apply_infos)
+                    | GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &ksk_apply_infos)
-                    | TensorKey::encrypt_sk_scratch_space(module, &tsk_infos)
+                    | TensorKey::encrypt_sk_tmp_bytes(module, &tsk_infos)
-                    | GGSW::keyswitch_scratch_space(
+                    | GGSW::keyswitch_tmp_bytes(
                         module,
                         &ggsw_out_infos,
                         &ggsw_in_infos,
@@ -310,10 +310,10 @@ where
             let mut source_xa: Source = Source::new([0u8; 32]);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GGSW::encrypt_sk_scratch_space(module, &ggsw_out_infos)
+                GGSW::encrypt_sk_tmp_bytes(module, &ggsw_out_infos)
-                    | GLWESwitchingKey::encrypt_sk_scratch_space(module, &ksk_apply_infos)
+                    | GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &ksk_apply_infos)
-                    | TensorKey::encrypt_sk_scratch_space(module, &tsk_infos)
+                    | TensorKey::encrypt_sk_tmp_bytes(module, &tsk_infos)
-                    | GGSW::keyswitch_inplace_scratch_space(module, &ggsw_out_infos, &ksk_apply_infos, &tsk_infos),
+                    | GGSW::keyswitch_inplace_tmp_bytes(module, &ggsw_out_infos, &ksk_apply_infos, &tsk_infos),
             );
 
             let var_xs: f64 = 0.5;

poulpy-core/src/tests/test_suite/keyswitch/glwe_ct.rs

@@ -112,9 +112,9 @@ where
                 module.vec_znx_fill_uniform(base2k, &mut pt_want.data, 0, &mut source_xa);
 
                 let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                    GLWESwitchingKey::encrypt_sk_scratch_space(module, &key_apply)
+                    GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &key_apply)
-                        | GLWE::encrypt_sk_scratch_space(module, &glwe_in_infos)
+                        | GLWE::encrypt_sk_tmp_bytes(module, &glwe_in_infos)
-                        | GLWE::keyswitch_scratch_space(module, &glwe_out_infos, &glwe_in_infos, &key_apply),
+                        | GLWE::keyswitch_tmp_bytes(module, &glwe_out_infos, &glwe_in_infos, &key_apply),
                 );
 
                 let mut sk_in: GLWESecret<Vec<u8>> = GLWESecret::alloc(n.into(), rank_in.into());
@@ -244,9 +244,9 @@ where
             module.vec_znx_fill_uniform(base2k, &mut pt_want.data, 0, &mut source_xa);
 
             let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-                GLWESwitchingKey::encrypt_sk_scratch_space(module, &key_apply_infos)
+                GLWESwitchingKey::encrypt_sk_tmp_bytes(module, &key_apply_infos)
-                    | GLWE::encrypt_sk_scratch_space(module, &glwe_out_infos)
+                    | GLWE::encrypt_sk_tmp_bytes(module, &glwe_out_infos)
-                    | GLWE::keyswitch_inplace_scratch_space(module, &glwe_out_infos, &key_apply_infos),
+                    | GLWE::keyswitch_inplace_tmp_bytes(module, &glwe_out_infos, &key_apply_infos),
             );
 
             let mut sk_in: GLWESecret<Vec<u8>> = GLWESecret::alloc(n.into(), rank.into());

poulpy-core/src/tests/test_suite/keyswitch/lwe_ct.rs

@@ -99,8 +99,8 @@ where
     };
 
     let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-        LWESwitchingKey::encrypt_sk_scratch_space(module, &key_apply_infos)
+        LWESwitchingKey::encrypt_sk_tmp_bytes(module, &key_apply_infos)
-            | LWE::keyswitch_scratch_space(module, &lwe_out_infos, &lwe_in_infos, &key_apply_infos),
+            | LWE::keyswitch_tmp_bytes(module, &lwe_out_infos, &lwe_in_infos, &key_apply_infos),
     );
 
     let mut sk_lwe_in: LWESecret<Vec<u8>> = LWESecret::alloc(n_lwe_in.into());

poulpy-core/src/tests/test_suite/packing.rs

@@ -105,9 +105,9 @@ where
     };
 
     let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-        GLWE::encrypt_sk_scratch_space(module, &glwe_out_infos)
+        GLWE::encrypt_sk_tmp_bytes(module, &glwe_out_infos)
-            | AutomorphismKey::encrypt_sk_scratch_space(module, &key_infos)
+            | AutomorphismKey::encrypt_sk_tmp_bytes(module, &key_infos)
-            | GLWEPacker::scratch_space(module, &glwe_out_infos, &key_infos),
+            | GLWEPacker::tmp_bytes(module, &glwe_out_infos, &key_infos),
     );
 
     let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_out_infos);

poulpy-core/src/tests/test_suite/trace.rs

@@ -107,10 +107,10 @@ where
         let mut source_xa: Source = Source::new([0u8; 32]);
 
         let mut scratch: ScratchOwned<B> = ScratchOwned::alloc(
-            GLWE::encrypt_sk_scratch_space(module, &glwe_out_infos)
+            GLWE::encrypt_sk_tmp_bytes(module, &glwe_out_infos)
-                | GLWE::decrypt_scratch_space(module, &glwe_out_infos)
+                | GLWE::decrypt_tmp_bytes(module, &glwe_out_infos)
-                | AutomorphismKey::encrypt_sk_scratch_space(module, &key_infos)
+                | AutomorphismKey::encrypt_sk_tmp_bytes(module, &key_infos)
-                | GLWE::trace_inplace_scratch_space(module, &glwe_out_infos, &key_infos),
+                | GLWE::trace_inplace_tmp_bytes(module, &glwe_out_infos, &key_infos),
         );
 
         let mut sk: GLWESecret<Vec<u8>> = GLWESecret::alloc_from_infos(&glwe_out_infos);

poulpy-hal/src/layouts/vec_znx.rs

@@ -110,7 +110,7 @@ impl<D: DataRef> ZnxView for VecZnx<D> {
 }
 
 impl VecZnx<Vec<u8>> {
-    pub fn rsh_scratch_space(n: usize) -> usize {
+    pub fn rsh_tmp_bytes(n: usize) -> usize {
         n * std::mem::size_of::<i64>()
     }
 }

poulpy-hal/src/layouts/zn.rs

@@ -98,7 +98,7 @@ impl<D: DataRef> ZnxView for Zn<D> {
 }
 
 impl Zn<Vec<u8>> {
-    pub fn rsh_scratch_space(n: usize) -> usize {
+    pub fn rsh_tmp_bytes(n: usize) -> usize {
         n * std::mem::size_of::<i64>()
     }
 }

poulpy-schemes/src/tfhe/blind_rotation/cggi_algo.rs

@@ -21,7 +21,7 @@ use crate::tfhe::blind_rotation::{
 };
 
 #[allow(clippy::too_many_arguments)]
-pub fn cggi_blind_rotate_scratch_space<B: Backend, OUT, GGSW>(
+pub fn cggi_blind_rotate_tmp_bytes<B: Backend, OUT, GGSW>(
     module: &Module<B>,
     block_size: usize,
     extension_factor: usize,
@@ -61,7 +61,7 @@ where
             + vmp_xai
             + (vmp | (acc_big + (module.vec_znx_big_normalize_tmp_bytes() | module.vec_znx_idft_apply_tmp_bytes())))
     } else {
-        GLWE::bytes_of(glwe_infos) + GLWE::external_product_inplace_scratch_space(module, glwe_infos, brk_infos)
+        GLWE::bytes_of(glwe_infos) + GLWE::external_product_inplace_tmp_bytes(module, glwe_infos, brk_infos)
     }
 }
 

poulpy-schemes/src/tfhe/blind_rotation/cggi_key.rs

@@ -43,12 +43,12 @@ impl BlindRotationKeyAlloc for BlindRotationKey<Vec<u8>, CGGI> {
 }
 
 impl BlindRotationKey<Vec<u8>, CGGI> {
-    pub fn generate_from_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn generate_from_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGSWInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf,
     {
-        GGSW::encrypt_sk_scratch_space(module, infos)
+        GGSW::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 
@@ -145,12 +145,12 @@ impl BlindRotationKeyCompressed<Vec<u8>, CGGI> {
         }
     }
 
-    pub fn generate_from_sk_scratch_space<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
+    pub fn generate_from_sk_tmp_bytes<B: Backend, A>(module: &Module<B>, infos: &A) -> usize
     where
         A: GGSWInfos,
         Module<B>: VecZnxNormalizeTmpBytes + VecZnxDftBytesOf,
     {
-        GGSWCompressed::encrypt_sk_scratch_space(module, infos)
+        GGSWCompressed::encrypt_sk_tmp_bytes(module, infos)
     }
 }
 

poulpy-schemes/src/tfhe/blind_rotation/tests/generic_blind_rotation.rs

@@ -19,7 +19,7 @@ use poulpy_hal::{
 
 use crate::tfhe::blind_rotation::{
     BlincRotationExecute, BlindRotationKey, BlindRotationKeyAlloc, BlindRotationKeyEncryptSk, BlindRotationKeyLayout,
-    BlindRotationKeyPrepared, CGGI, LookUpTable, cggi_blind_rotate_scratch_space, mod_switch_2n,
+    BlindRotationKeyPrepared, CGGI, LookUpTable, cggi_blind_rotate_tmp_bytes, mod_switch_2n,
 };
 
 use poulpy_core::layouts::{
@@ -123,7 +123,7 @@ where
         base2k: base2k.into(),
     };
 
-    let mut scratch: ScratchOwned<B> = ScratchOwned::<B>::alloc(BlindRotationKey::generate_from_sk_scratch_space(
+    let mut scratch: ScratchOwned<B> = ScratchOwned::<B>::alloc(BlindRotationKey::generate_from_sk_tmp_bytes(
         module, &brk_infos,
     ));
 
@@ -134,7 +134,7 @@ where
     let mut sk_lwe: LWESecret<Vec<u8>> = LWESecret::alloc(n_lwe.into());
     sk_lwe.fill_binary_block(block_size, &mut source_xs);
 
-    let mut scratch_br: ScratchOwned<B> = ScratchOwned::<B>::alloc(cggi_blind_rotate_scratch_space(
+    let mut scratch_br: ScratchOwned<B> = ScratchOwned::<B>::alloc(cggi_blind_rotate_tmp_bytes(
         module,
         block_size,
         extension_factor,