working rlwe encryption example with interleaved polynomial

base2k/examples/rlwe_encrypt.rs

@@ -13,7 +13,7 @@ fn main() {
     let log_scale: usize = msg_size * log_base2k - 5;
     let module: Module<FFT64> = Module::<FFT64>::new(n);
 
-    let mut carry: Vec<u8> = alloc_aligned(module.vec_znx_big_normalize_tmp_bytes(2));
+    let mut carry: Vec<u8> = alloc_aligned(module.vec_znx_big_normalize_tmp_bytes());
 
     let seed: [u8; 32] = [0; 32];
     let mut source: Source = Source::new(seed);
@@ -28,69 +28,95 @@ fn main() {
     // s_ppol <- DFT(s)
     module.svp_prepare(&mut s_ppol, &s);
 
-    // ct = (c0, c1)
-    let mut ct: VecZnx = module.new_vec_znx(2, ct_size);
+    // Allocates a VecZnx with two columns: ct=(0, 0)
+    let mut ct: VecZnx = module.new_vec_znx(
+        2,       // Number of columns
+        ct_size, // Number of small poly per column
+    );
 
-    // Fill c1 with random values
+    // Fill the second column with random values: ct = (0, a)
     module.fill_uniform(log_base2k, &mut ct, 1, ct_size, &mut source);
 
     // Scratch space for DFT values
-    let mut buf_dft: VecZnxDft<FFT64> = module.new_vec_znx_dft(1, ct.size());
-
-    // Applies buf_dft <- s * c1
-    module.svp_apply_dft(
-        &mut buf_dft, // DFT(c1 * s)
-        &s_ppol,
-        &ct,
-        1, // c1
+    let mut buf_dft: VecZnxDft<FFT64> = module.new_vec_znx_dft(
+        1,         // Number of columns
+        ct.size(), // Number of polynomials per column
     );
 
-    // Alias scratch space (VecZnxDftis always at least as big as VecZnxBig)
+    // Applies DFT(ct[1]) * DFT(s)
+    module.svp_apply_dft(
+        &mut buf_dft, // DFT(ct[1] * s)
+        &s_ppol,      // DFT(s)
+        &ct,
+        1, // Selects the second column of ct
+    );
+
+    // Alias scratch space (VecZnxDft<B> is always at least as big as VecZnxBig<B>)
     let mut buf_big: VecZnxBig<FFT64> = buf_dft.as_vec_znx_big();
 
-    // BIG(c1 * s) <- IDFT(DFT(c1 * s)) (not normalized)
+    // BIG(ct[1] * s) <- IDFT(DFT(ct[1] * s)) (not normalized)
     module.vec_znx_idft_tmp_a(&mut buf_big, &mut buf_dft);
 
-    // m <- (0)
-    let mut m: VecZnx = module.new_vec_znx(1, msg_size);
+    // Creates a plaintext: VecZnx with 1 column
+    let mut m: VecZnx = module.new_vec_znx(
+        1,        // Number of columns
+        msg_size, // Number of small polynomials
+    );
     let mut want: Vec<i64> = vec![0; n];
     want.iter_mut()
         .for_each(|x| *x = source.next_u64n(16, 15) as i64);
     m.encode_vec_i64(0, log_base2k, log_scale, &want, 4);
     m.normalize(log_base2k, &mut carry);
 
-    // m - BIG(c1 * s)
-    module.vec_znx_big_sub_small_ab_inplace(&mut buf_big, &m);
+    // m - BIG(ct[1] * s)
+    module.vec_znx_big_sub_small_a_inplace(
+        &mut buf_big,
+        0, // Selects the first column of the receiver
+        &m,
+        0, // Selects the first column of the message
+    );
 
-    // c0 <- m - BIG(c1 * s)
-    module.vec_znx_big_normalize(log_base2k, &mut ct, &buf_big, &mut carry);
+    // Normalizes back to VecZnx
+    // ct[0] <- m - BIG(c1 * s)
+    module.vec_znx_big_normalize(
+        log_base2k, &mut ct, 0, // Selects the first column of ct (ct[0])
+        &buf_big, 0, // Selects the first column of buf_big
+        &mut carry,
+    );
 
-    ct.print(ct.sl());
-
-    // (c0 + e, c1)
+    // Add noise to ct[0]
+    // ct[0] <- ct[0] + e
     module.add_normal(
         log_base2k,
         &mut ct,
-        0, // c0
-        log_base2k * ct_size,
+        0,                    // Selects the first column of ct (ct[0])
+        log_base2k * ct_size, // Scaling of the noise: 2^{-log_base2k * limbs}
         &mut source,
-        3.2,
-        19.0,
+        3.2,  // Standard deviation
+        19.0, // Truncatation bound
     );
 
-    // Decrypt
+    // Final ciphertext: ct = (-a * s + m + e, a)
+
+    // Decryption
+
+    // DFT(ct[1] * s)
+    module.svp_apply_dft(
+        &mut buf_dft,
+        &s_ppol,
+        &ct,
+        1, // Selects the second column of ct (ct[1])
+    );
 
-    // DFT(c1 * s)
-    module.svp_apply_dft(&mut buf_dft, &s_ppol, &ct, 1);
     // BIG(c1 * s) = IDFT(DFT(c1 * s))
     module.vec_znx_idft_tmp_a(&mut buf_big, &mut buf_dft);
 
-    // BIG(c1 * s) + c0
-    module.vec_znx_big_add_small_inplace(&mut buf_big, &ct);
+    // BIG(c1 * s) + ct[0]
+    module.vec_znx_big_add_small_inplace(&mut buf_big, 0, &ct, 0);
 
-    // m + e <- BIG(c1 * s + c0)
+    // m + e <- BIG(ct[1] * s + ct[0])
     let mut res: VecZnx = module.new_vec_znx(1, ct_size);
-    module.vec_znx_big_normalize(log_base2k, &mut res, &buf_big, &mut carry);
+    module.vec_znx_big_normalize(log_base2k, &mut res, 0, &buf_big, 0, &mut carry);
 
     // have = m * 2^{log_scale} + e
     let mut have: Vec<i64> = vec![i64::default(); n];

base2k/examples/vector_matrix_product.rs

@@ -46,7 +46,7 @@ fn main() {
     module.vec_znx_idft_tmp_a(&mut c_big, &mut c_dft);
 
     let mut res: VecZnx = module.new_vec_znx(1, limbs_vec);
-    module.vec_znx_big_normalize(log_base2k, &mut res, &c_big, &mut buf);
+    module.vec_znx_big_normalize(log_base2k, &mut res, 0, &c_big, 0, &mut buf);
 
     let mut values_res: Vec<i64> = vec![i64::default(); n];
     res.decode_vec_i64(0, log_base2k, log_k, &mut values_res);

base2k/src/vec_znx_big_ops.rs

@@ -1,8 +1,5 @@
-use std::cmp::min;
-
 use crate::ffi::vec_znx;
-use crate::internals::{apply_binary_op, apply_unary_op, ffi_binary_op_factory_type_1, ffi_ternary_op_factory};
-use crate::{Backend, FFT64, Module, VecZnx, VecZnxBig, VecZnxOps, ZnxBase, ZnxBasics, ZnxInfos, ZnxLayout, assert_alignement};
+use crate::{Backend, FFT64, Module, VecZnx, VecZnxBig, VecZnxOps, ZnxBase, ZnxInfos, ZnxLayout, assert_alignement};
 
 pub trait VecZnxBigOps<B: Backend> {
     /// Allocates a vector Z[X]/(X^N+1) that stores not normalized values.
@@ -41,40 +38,80 @@ pub trait VecZnxBigOps<B: Backend> {
     fn bytes_of_vec_znx_big(&self, cols: usize, size: usize) -> usize;
 
     /// Adds `a` to `b` and stores the result on `c`.
-    fn vec_znx_big_add(&self, c: &mut VecZnxBig<B>, a: &VecZnxBig<B>, b: &VecZnxBig<B>);
+    fn vec_znx_big_add(
+        &self,
+        res: &mut VecZnxBig<B>,
+        col_res: usize,
+        a: &VecZnxBig<B>,
+        col_a: usize,
+        b: &VecZnxBig<B>,
+        col_b: usize,
+    );
 
     /// Adds `a` to `b` and stores the result on `b`.
-    fn vec_znx_big_add_inplace(&self, b: &mut VecZnxBig<B>, a: &VecZnxBig<B>);
+    fn vec_znx_big_add_inplace(&self, res: &mut VecZnxBig<B>, col_res: usize, a: &VecZnxBig<B>, col_a: usize);
 
     /// Adds `a` to `b` and stores the result on `c`.
-    fn vec_znx_big_add_small(&self, c: &mut VecZnxBig<B>, a: &VecZnx, b: &VecZnxBig<B>);
+    fn vec_znx_big_add_small(
+        &self,
+        res: &mut VecZnxBig<B>,
+        col_res: usize,
+        a: &VecZnx,
+        col_a: usize,
+        b: &VecZnxBig<B>,
+        col_b: usize,
+    );
 
     /// Adds `a` to `b` and stores the result on `b`.
-    fn vec_znx_big_add_small_inplace(&self, b: &mut VecZnxBig<B>, a: &VecZnx);
+    fn vec_znx_big_add_small_inplace(&self, res: &mut VecZnxBig<B>, col_res: usize, a: &VecZnx, col_a: usize);
 
     /// Subtracts `a` to `b` and stores the result on `c`.
-    fn vec_znx_big_sub(&self, c: &mut VecZnxBig<B>, a: &VecZnxBig<B>, b: &VecZnxBig<B>);
+    fn vec_znx_big_sub(
+        &self,
+        res: &mut VecZnxBig<B>,
+        col_res: usize,
+        a: &VecZnxBig<B>,
+        col_a: usize,
+        b: &VecZnxBig<B>,
+        col_b: usize,
+    );
 
     /// Subtracts `a` to `b` and stores the result on `b`.
-    fn vec_znx_big_sub_ab_inplace(&self, b: &mut VecZnxBig<B>, a: &VecZnxBig<B>);
+    fn vec_znx_big_sub_ab_inplace(&self, res: &mut VecZnxBig<B>, col_res: usize, a: &VecZnxBig<B>, col_a: usize);
 
     /// Subtracts `b` to `a` and stores the result on `b`.
-    fn vec_znx_big_sub_ba_inplace(&self, b: &mut VecZnxBig<B>, a: &VecZnxBig<B>);
+    fn vec_znx_big_sub_ba_inplace(&self, res: &mut VecZnxBig<B>, col_res: usize, a: &VecZnxBig<B>, col_a: usize);
 
     /// Subtracts `b` to `a` and stores the result on `c`.
-    fn vec_znx_big_sub_small_ab(&self, c: &mut VecZnxBig<B>, a: &VecZnx, b: &VecZnxBig<B>);
+    fn vec_znx_big_sub_small_a(
+        &self,
+        res: &mut VecZnxBig<B>,
+        col_res: usize,
+        a: &VecZnx,
+        col_a: usize,
+        b: &VecZnxBig<B>,
+        col_b: usize,
+    );
 
     /// Subtracts `a` to `b` and stores the result on `b`.
-    fn vec_znx_big_sub_small_ab_inplace(&self, b: &mut VecZnxBig<B>, a: &VecZnx);
+    fn vec_znx_big_sub_small_a_inplace(&self, res: &mut VecZnxBig<B>, col_res: usize, a: &VecZnx, col_a: usize);
 
     /// Subtracts `b` to `a` and stores the result on `c`.
-    fn vec_znx_big_sub_small_ba(&self, c: &mut VecZnxBig<B>, a: &VecZnxBig<B>, b: &VecZnx);
+    fn vec_znx_big_sub_small_b(
+        &self,
+        res: &mut VecZnxBig<B>,
+        col_res: usize,
+        a: &VecZnxBig<B>,
+        col_a: usize,
+        b: &VecZnx,
+        col_b: usize,
+    );
 
     /// Subtracts `b` to `a` and stores the result on `b`.
-    fn vec_znx_big_sub_small_ba_inplace(&self, b: &mut VecZnxBig<B>, a: &VecZnx);
+    fn vec_znx_big_sub_small_b_inplace(&self, res: &mut VecZnxBig<B>, col_res: usize, a: &VecZnx, col_a: usize);
 
     /// Returns the minimum number of bytes to apply [VecZnxBigOps::vec_znx_big_normalize].
-    fn vec_znx_big_normalize_tmp_bytes(&self, cols: usize) -> usize;
+    fn vec_znx_big_normalize_tmp_bytes(&self) -> usize;
 
     /// Normalizes `a` and stores the result on `b`.
     ///
@@ -82,13 +119,21 @@ pub trait VecZnxBigOps<B: Backend> {
     ///
     /// * `log_base2k`: normalization basis.
     /// * `tmp_bytes`: scratch space of size at least [VecZnxBigOps::vec_znx_big_normalize].
-    fn vec_znx_big_normalize(&self, log_base2k: usize, b: &mut VecZnx, a: &VecZnxBig<B>, tmp_bytes: &mut [u8]);
+    fn vec_znx_big_normalize(
+        &self,
+        log_base2k: usize,
+        res: &mut VecZnx,
+        col_res: usize,
+        a: &VecZnxBig<B>,
+        col_a: usize,
+        tmp_bytes: &mut [u8],
+    );
 
     /// Applies the automorphism X^i -> X^ik on `a` and stores the result on `b`.
-    fn vec_znx_big_automorphism(&self, k: i64, b: &mut VecZnxBig<B>, a: &VecZnxBig<B>);
+    fn vec_znx_big_automorphism(&self, k: i64, res: &mut VecZnxBig<B>, col_res: usize, a: &VecZnxBig<B>, col_a: usize);
 
     /// Applies the automorphism X^i -> X^ik on `a` and stores the result on `a`.
-    fn vec_znx_big_automorphism_inplace(&self, k: i64, a: &mut VecZnxBig<B>);
+    fn vec_znx_big_automorphism_inplace(&self, k: i64, a: &mut VecZnxBig<B>, col_a: usize);
 }
 
 impl VecZnxBigOps<FFT64> for Module<FFT64> {
@@ -108,170 +153,267 @@ impl VecZnxBigOps<FFT64> for Module<FFT64> {
         VecZnxBig::bytes_of(self, cols, size)
     }
 
-    fn vec_znx_big_add(&self, c: &mut VecZnxBig<FFT64>, a: &VecZnxBig<FFT64>, b: &VecZnxBig<FFT64>) {
-        let op = ffi_ternary_op_factory(
-            self.ptr,
-            c.size(),
-            c.sl(),
-            a.size(),
-            a.sl(),
-            b.size(),
-            b.sl(),
-            vec_znx::vec_znx_add,
-        );
-        apply_binary_op::<FFT64, VecZnxBig<FFT64>, VecZnxBig<FFT64>, VecZnxBig<FFT64>, false>(self, c, a, b, op);
-    }
-
-    fn vec_znx_big_add_inplace(&self, b: &mut VecZnxBig<FFT64>, a: &VecZnxBig<FFT64>) {
-        unsafe {
-            let b_ptr: *mut VecZnxBig<FFT64> = b as *mut VecZnxBig<FFT64>;
-            Self::vec_znx_big_add(self, &mut *b_ptr, a, &*b_ptr);
-        }
-    }
-
-    fn vec_znx_big_sub(&self, c: &mut VecZnxBig<FFT64>, a: &VecZnxBig<FFT64>, b: &VecZnxBig<FFT64>) {
-        let op = ffi_ternary_op_factory(
-            self.ptr,
-            c.size(),
-            c.sl(),
-            a.size(),
-            a.sl(),
-            b.size(),
-            b.sl(),
-            vec_znx::vec_znx_sub,
-        );
-        apply_binary_op::<FFT64, VecZnxBig<FFT64>, VecZnxBig<FFT64>, VecZnxBig<FFT64>, true>(self, c, a, b, op);
-    }
-
-    fn vec_znx_big_sub_ab_inplace(&self, b: &mut VecZnxBig<FFT64>, a: &VecZnxBig<FFT64>) {
-        unsafe {
-            let b_ptr: *mut VecZnxBig<FFT64> = b as *mut VecZnxBig<FFT64>;
-            Self::vec_znx_big_sub(self, &mut *b_ptr, a, &*b_ptr);
-        }
-    }
-
-    fn vec_znx_big_sub_ba_inplace(&self, b: &mut VecZnxBig<FFT64>, a: &VecZnxBig<FFT64>) {
-        unsafe {
-            let b_ptr: *mut VecZnxBig<FFT64> = b as *mut VecZnxBig<FFT64>;
-            Self::vec_znx_big_sub(self, &mut *b_ptr, &*b_ptr, a);
-        }
-    }
-
-    fn vec_znx_big_sub_small_ba(&self, c: &mut VecZnxBig<FFT64>, a: &VecZnxBig<FFT64>, b: &VecZnx) {
-        let op = ffi_ternary_op_factory(
-            self.ptr,
-            c.size(),
-            c.sl(),
-            a.size(),
-            a.sl(),
-            b.size(),
-            b.sl(),
-            vec_znx::vec_znx_sub,
-        );
-        apply_binary_op::<FFT64, VecZnxBig<FFT64>, VecZnxBig<FFT64>, VecZnx, true>(self, c, a, b, op);
-    }
-
-    fn vec_znx_big_sub_small_ba_inplace(&self, b: &mut VecZnxBig<FFT64>, a: &VecZnx) {
-        unsafe {
-            let b_ptr: *mut VecZnxBig<FFT64> = b as *mut VecZnxBig<FFT64>;
-            Self::vec_znx_big_sub_small_ba(self, &mut *b_ptr, &*b_ptr, a);
-        }
-    }
-
-    fn vec_znx_big_sub_small_ab(&self, c: &mut VecZnxBig<FFT64>, a: &VecZnx, b: &VecZnxBig<FFT64>) {
-        let op = ffi_ternary_op_factory(
-            self.ptr,
-            c.size(),
-            c.sl(),
-            a.size(),
-            a.sl(),
-            b.size(),
-            b.sl(),
-            vec_znx::vec_znx_sub,
-        );
-        apply_binary_op::<FFT64, VecZnxBig<FFT64>, VecZnx, VecZnxBig<FFT64>, true>(self, c, a, b, op);
-    }
-
-    fn vec_znx_big_sub_small_ab_inplace(&self, b: &mut VecZnxBig<FFT64>, a: &VecZnx) {
-        unsafe {
-            let b_ptr: *mut VecZnxBig<FFT64> = b as *mut VecZnxBig<FFT64>;
-            Self::vec_znx_big_sub_small_ab(self, &mut *b_ptr, a, &*b_ptr);
-        }
-    }
-
-    fn vec_znx_big_add_small(&self, c: &mut VecZnxBig<FFT64>, a: &VecZnx, b: &VecZnxBig<FFT64>) {
-        let op = ffi_ternary_op_factory(
-            self.ptr,
-            c.size(),
-            c.sl(),
-            a.size(),
-            a.sl(),
-            b.size(),
-            b.sl(),
-            vec_znx::vec_znx_add,
-        );
-        apply_binary_op::<FFT64, VecZnxBig<FFT64>, VecZnx, VecZnxBig<FFT64>, false>(self, c, a, b, op);
-    }
-
-    fn vec_znx_big_add_small_inplace(&self, b: &mut VecZnxBig<FFT64>, a: &VecZnx) {
-        unsafe {
-            let b_ptr: *mut VecZnxBig<FFT64> = b as *mut VecZnxBig<FFT64>;
-            Self::vec_znx_big_add_small(self, &mut *b_ptr, a, &*b_ptr);
-        }
-    }
-
-    fn vec_znx_big_normalize_tmp_bytes(&self, cols: usize) -> usize {
-        Self::vec_znx_normalize_tmp_bytes(self, cols)
-    }
-
-    fn vec_znx_big_normalize(&self, log_base2k: usize, b: &mut VecZnx, a: &VecZnxBig<FFT64>, tmp_bytes: &mut [u8]) {
+    fn vec_znx_big_add(
+        &self,
+        res: &mut VecZnxBig<FFT64>,
+        col_res: usize,
+        a: &VecZnxBig<FFT64>,
+        col_a: usize,
+        b: &VecZnxBig<FFT64>,
+        col_b: usize,
+    ) {
         #[cfg(debug_assertions)]
         {
-            assert!(tmp_bytes.len() >= Self::vec_znx_big_normalize_tmp_bytes(&self, a.cols()));
+            assert_eq!(a.n(), self.n());
+            assert_eq!(b.n(), self.n());
+            assert_eq!(res.n(), self.n());
+            assert_ne!(a.as_ptr(), b.as_ptr());
+        }
+        unsafe {
+            vec_znx::vec_znx_add(
+                self.ptr,
+                res.at_mut_ptr(col_res, 0),
+                res.size() as u64,
+                res.sl() as u64,
+                a.at_ptr(col_a, 0),
+                a.size() as u64,
+                a.sl() as u64,
+                b.at_ptr(col_b, 0),
+                b.size() as u64,
+                b.sl() as u64,
+            )
+        }
+    }
+
+    fn vec_znx_big_add_inplace(&self, res: &mut VecZnxBig<FFT64>, col_res: usize, a: &VecZnxBig<FFT64>, col_a: usize) {
+        unsafe {
+            let res_ptr: *mut VecZnxBig<FFT64> = res as *mut VecZnxBig<FFT64>;
+            Self::vec_znx_big_add(self, &mut *res_ptr, col_res, a, col_a, &*res_ptr, col_res);
+        }
+    }
+
+    fn vec_znx_big_sub(
+        &self,
+        res: &mut VecZnxBig<FFT64>,
+        col_res: usize,
+        a: &VecZnxBig<FFT64>,
+        col_a: usize,
+        b: &VecZnxBig<FFT64>,
+        col_b: usize,
+    ) {
+        #[cfg(debug_assertions)]
+        {
+            assert_eq!(a.n(), self.n());
+            assert_eq!(b.n(), self.n());
+            assert_eq!(res.n(), self.n());
+            assert_ne!(a.as_ptr(), b.as_ptr());
+        }
+        unsafe {
+            vec_znx::vec_znx_sub(
+                self.ptr,
+                res.at_mut_ptr(col_res, 0),
+                res.size() as u64,
+                res.sl() as u64,
+                a.at_ptr(col_a, 0),
+                a.size() as u64,
+                a.sl() as u64,
+                b.at_ptr(col_b, 0),
+                b.size() as u64,
+                b.sl() as u64,
+            )
+        }
+    }
+
+    fn vec_znx_big_sub_ab_inplace(&self, res: &mut VecZnxBig<FFT64>, col_res: usize, a: &VecZnxBig<FFT64>, col_a: usize) {
+        unsafe {
+            let res_ptr: *mut VecZnxBig<FFT64> = res as *mut VecZnxBig<FFT64>;
+            Self::vec_znx_big_sub(self, &mut *res_ptr, col_res, a, col_a, &*res_ptr, col_res);
+        }
+    }
+
+    fn vec_znx_big_sub_ba_inplace(&self, res: &mut VecZnxBig<FFT64>, col_res: usize, a: &VecZnxBig<FFT64>, col_a: usize) {
+        unsafe {
+            let res_ptr: *mut VecZnxBig<FFT64> = res as *mut VecZnxBig<FFT64>;
+            Self::vec_znx_big_sub(self, &mut *res_ptr, col_res, &*res_ptr, col_res, a, col_a);
+        }
+    }
+
+    fn vec_znx_big_sub_small_b(
+        &self,
+        res: &mut VecZnxBig<FFT64>,
+        col_res: usize,
+        a: &VecZnxBig<FFT64>,
+        col_a: usize,
+        b: &VecZnx,
+        col_b: usize,
+    ) {
+        #[cfg(debug_assertions)]
+        {
+            assert_eq!(a.n(), self.n());
+            assert_eq!(b.n(), self.n());
+            assert_eq!(res.n(), self.n());
+            assert_ne!(a.as_ptr(), b.as_ptr());
+        }
+        unsafe {
+            vec_znx::vec_znx_sub(
+                self.ptr,
+                res.at_mut_ptr(col_res, 0),
+                res.size() as u64,
+                res.sl() as u64,
+                a.at_ptr(col_a, 0),
+                a.size() as u64,
+                a.sl() as u64,
+                b.at_ptr(col_b, 0),
+                b.size() as u64,
+                b.sl() as u64,
+            )
+        }
+    }
+
+    fn vec_znx_big_sub_small_b_inplace(&self, res: &mut VecZnxBig<FFT64>, col_res: usize, a: &VecZnx, col_a: usize) {
+        unsafe {
+            let res_ptr: *mut VecZnxBig<FFT64> = res as *mut VecZnxBig<FFT64>;
+            Self::vec_znx_big_sub_small_b(self, &mut *res_ptr, col_res, &*res_ptr, col_res, a, col_a);
+        }
+    }
+
+    fn vec_znx_big_sub_small_a(
+        &self,
+        res: &mut VecZnxBig<FFT64>,
+        col_res: usize,
+        a: &VecZnx,
+        col_a: usize,
+        b: &VecZnxBig<FFT64>,
+        col_b: usize,
+    ) {
+        #[cfg(debug_assertions)]
+        {
+            assert_eq!(a.n(), self.n());
+            assert_eq!(b.n(), self.n());
+            assert_eq!(res.n(), self.n());
+            assert_ne!(a.as_ptr(), b.as_ptr());
+        }
+        unsafe {
+            vec_znx::vec_znx_sub(
+                self.ptr,
+                res.at_mut_ptr(col_res, 0),
+                res.size() as u64,
+                res.sl() as u64,
+                a.at_ptr(col_a, 0),
+                a.size() as u64,
+                a.sl() as u64,
+                b.at_ptr(col_b, 0),
+                b.size() as u64,
+                b.sl() as u64,
+            )
+        }
+    }
+
+    fn vec_znx_big_sub_small_a_inplace(&self, res: &mut VecZnxBig<FFT64>, col_res: usize, a: &VecZnx, col_a: usize) {
+        unsafe {
+            let res_ptr: *mut VecZnxBig<FFT64> = res as *mut VecZnxBig<FFT64>;
+            Self::vec_znx_big_sub_small_a(self, &mut *res_ptr, col_res, a, col_a, &*res_ptr, col_res);
+        }
+    }
+
+    fn vec_znx_big_add_small(
+        &self,
+        res: &mut VecZnxBig<FFT64>,
+        col_res: usize,
+        a: &VecZnx,
+        col_a: usize,
+        b: &VecZnxBig<FFT64>,
+        col_b: usize,
+    ) {
+        #[cfg(debug_assertions)]
+        {
+            assert_eq!(a.n(), self.n());
+            assert_eq!(b.n(), self.n());
+            assert_eq!(res.n(), self.n());
+            assert_ne!(a.as_ptr(), b.as_ptr());
+        }
+        unsafe {
+            vec_znx::vec_znx_add(
+                self.ptr,
+                res.at_mut_ptr(col_res, 0),
+                res.size() as u64,
+                res.sl() as u64,
+                a.at_ptr(col_a, 0),
+                a.size() as u64,
+                a.sl() as u64,
+                b.at_ptr(col_b, 0),
+                b.size() as u64,
+                b.sl() as u64,
+            )
+        }
+    }
+
+    fn vec_znx_big_add_small_inplace(&self, res: &mut VecZnxBig<FFT64>, col_res: usize, a: &VecZnx, a_col: usize) {
+        unsafe {
+            let res_ptr: *mut VecZnxBig<FFT64> = res as *mut VecZnxBig<FFT64>;
+            Self::vec_znx_big_add_small(self, &mut *res_ptr, col_res, a, a_col, &*res_ptr, col_res);
+        }
+    }
+
+    fn vec_znx_big_normalize_tmp_bytes(&self) -> usize {
+        Self::vec_znx_normalize_tmp_bytes(self)
+    }
+
+    fn vec_znx_big_normalize(
+        &self,
+        log_base2k: usize,
+        res: &mut VecZnx,
+        col_res: usize,
+        a: &VecZnxBig<FFT64>,
+        col_a: usize,
+        tmp_bytes: &mut [u8],
+    ) {
+        #[cfg(debug_assertions)]
+        {
+            assert_eq!(a.n(), self.n());
+            assert_eq!(res.n(), self.n());
+            assert!(tmp_bytes.len() >= Self::vec_znx_normalize_tmp_bytes(&self));
             assert_alignement(tmp_bytes.as_ptr());
         }
-
-        let a_size: usize = a.size();
-        let b_size: usize = b.size();
-        let a_sl: usize = a.sl();
-        let b_sl: usize = b.sl();
-        let a_cols: usize = a.cols();
-        let b_cols: usize = b.cols();
-        let min_cols: usize = min(a_cols, b_cols);
-        (0..min_cols).for_each(|i| unsafe {
+        unsafe {
             vec_znx::vec_znx_normalize_base2k(
                 self.ptr,
                 log_base2k as u64,
-                b.at_mut_ptr(i, 0),
-                b_size as u64,
-                b_sl as u64,
-                a.at_ptr(i, 0),
-                a_size as u64,
-                a_sl as u64,
+                res.at_mut_ptr(col_res, 0),
+                res.size() as u64,
+                res.sl() as u64,
+                a.at_ptr(col_a, 0),
+                a.size() as u64,
+                a.sl() as u64,
                 tmp_bytes.as_mut_ptr(),
             );
-        });
-
-        (min_cols..b_cols).for_each(|i| (0..b_size).for_each(|j| b.zero_at(i, j)));
+        }
     }
 
-    fn vec_znx_big_automorphism(&self, k: i64, b: &mut VecZnxBig<FFT64>, a: &VecZnxBig<FFT64>) {
-        let op = ffi_binary_op_factory_type_1(
-            self.ptr,
-            k,
-            b.size(),
-            b.sl(),
-            a.size(),
-            a.sl(),
-            vec_znx::vec_znx_automorphism,
-        );
-        apply_unary_op::<FFT64, VecZnxBig<FFT64>>(self, b, a, op);
+    fn vec_znx_big_automorphism(&self, k: i64, res: &mut VecZnxBig<FFT64>, col_res: usize, a: &VecZnxBig<FFT64>, col_a: usize) {
+        #[cfg(debug_assertions)]
+        {
+            assert_eq!(a.n(), self.n());
+            assert_eq!(res.n(), self.n());
+        }
+        unsafe {
+            vec_znx::vec_znx_automorphism(
+                self.ptr,
+                k,
+                res.at_mut_ptr(col_res, 0),
+                res.size() as u64,
+                res.sl() as u64,
+                a.at_ptr(col_a, 0),
+                a.size() as u64,
+                a.sl() as u64,
+            )
+        }
     }
 
-    fn vec_znx_big_automorphism_inplace(&self, k: i64, a: &mut VecZnxBig<FFT64>) {
+    fn vec_znx_big_automorphism_inplace(&self, k: i64, a: &mut VecZnxBig<FFT64>, col_a: usize) {
         unsafe {
             let a_ptr: *mut VecZnxBig<FFT64> = a as *mut VecZnxBig<FFT64>;
-            Self::vec_znx_big_automorphism(self, k, &mut *a_ptr, &*a_ptr);
+            Self::vec_znx_big_automorphism(self, k, &mut *a_ptr, col_a, &*a_ptr, col_a);
         }
     }
 }