avx merkle

benches/merkle.rs

@@ -7,11 +7,13 @@ pub fn cpu_merkle(c: &mut criterion::Criterion) {
     use itertools::Itertools;
     use num_traits::Zero;
     use stwo::commitment_scheme::ops::MerkleOps;
-    use stwo::core::backend::CPUBackend;
+    use stwo::core::backend::avx512::AVX512Backend;
+    use stwo::core::backend::{CPUBackend, Col};
     use stwo::core::fields::m31::BaseField;
+    use stwo::platform;
 
     const N_COLS: usize = 1 << 8;
-    const LOG_SIZE: u32 = 20;
+    const LOG_SIZE: u32 = 16;
     let cols = (0..N_COLS)
         .map(|_| {
             (0..(1 << LOG_SIZE))
@@ -30,6 +32,23 @@ pub fn cpu_merkle(c: &mut criterion::Criterion) {
             CPUBackend::commit_on_layer(LOG_SIZE, None, &cols.iter().collect_vec());
         })
     });
+
+    if !platform::avx512_detected() {
+        return;
+    }
+    let cols = (0..N_COLS)
+        .map(|_| {
+            (0..(1 << LOG_SIZE))
+                .map(|_| BaseField::zero())
+                .collect::<Col<AVX512Backend, BaseField>>()
+        })
+        .collect::<Vec<_>>();
+
+    group.bench_function("avx merkle", |b| {
+        b.iter(|| {
+            AVX512Backend::commit_on_layer(LOG_SIZE, None, &cols.iter().collect_vec());
+        })
+    });
 }
 
 #[cfg(target_arch = "x86_64")]

src/commitment_scheme/blake2_merkle.rs

@@ -1,10 +1,8 @@
-use itertools::Itertools;
 use num_traits::Zero;
 
 use super::blake2_hash::Blake2sHash;
 use super::blake2s_ref::compress;
-use super::ops::{MerkleHasher, MerkleOps};
-use crate::core::backend::CPUBackend;
+use super::ops::MerkleHasher;
 use crate::core::fields::m31::BaseField;
 
 #[derive(Copy, Clone, Debug, PartialEq, Eq, Default)]
@@ -39,23 +37,6 @@ impl MerkleHasher for Blake2sMerkleHasher {
     }
 }
 
-impl MerkleOps<Blake2sMerkleHasher> for CPUBackend {
-    fn commit_on_layer(
-        log_size: u32,
-        prev_layer: Option<&Vec<Blake2sHash>>,
-        columns: &[&Vec<BaseField>],
-    ) -> Vec<Blake2sHash> {
-        (0..(1 << log_size))
-            .map(|i| {
-                Blake2sMerkleHasher::hash_node(
-                    prev_layer.map(|prev_layer| (prev_layer[2 * i], prev_layer[2 * i + 1])),
-                    &columns.iter().map(|column| column[i]).collect_vec(),
-                )
-            })
-            .collect()
-    }
-}
-
 #[cfg(test)]
 mod tests {
     use std::collections::BTreeMap;

src/core/backend/avx512/blake2s.rs

@@ -0,0 +1,88 @@
+use std::arch::x86_64::__m512i;
+
+use itertools::Itertools;
+
+use super::blake2s_avx::{compress16, set1, transpose_msgs, untranspose_states};
+use super::{AVX512Backend, VECS_LOG_SIZE};
+use crate::commitment_scheme::blake2_hash::Blake2sHash;
+use crate::commitment_scheme::blake2_merkle::Blake2sMerkleHasher;
+use crate::commitment_scheme::ops::MerkleOps;
+use crate::core::backend::{Col, ColumnOps};
+use crate::core::fields::m31::BaseField;
+
+impl ColumnOps<Blake2sHash> for AVX512Backend {
+    type Column = Vec<Blake2sHash>;
+
+    fn bit_reverse_column(_column: &mut Self::Column) {
+        unimplemented!()
+    }
+}
+
+impl MerkleOps<Blake2sMerkleHasher> for AVX512Backend {
+    fn commit_on_layer(
+        log_size: u32,
+        prev_layer: Option<&Vec<Blake2sHash>>,
+        columns: &[&Col<AVX512Backend, BaseField>],
+    ) -> Vec<Blake2sHash> {
+        // Pad prev_layer if too small.
+        let mut padded_buffer = vec![];
+        let prev_layer = if log_size < 4 {
+            prev_layer.map(|prev_layer| {
+                padded_buffer = prev_layer
+                    .iter()
+                    .copied()
+                    .chain(std::iter::repeat(Blake2sHash::default()))
+                    .collect_vec();
+                &padded_buffer
+            })
+        } else {
+            prev_layer
+        };
+
+        // Commit to columns.
+        let mut res = Vec::with_capacity(1 << log_size);
+        for i in 0..(1 << (log_size - VECS_LOG_SIZE as u32)) {
+            let mut state: [__m512i; 8] = unsafe { std::mem::zeroed() };
+            // Hash prev_layer.
+            if let Some(prev_layer) = prev_layer {
+                let ptr = prev_layer[(i << 5)..(i << 5) + 32].as_ptr() as *const __m512i;
+                let msgs: [__m512i; 16] = std::array::from_fn(|j| unsafe { *ptr.add(j) });
+                state = unsafe {
+                    compress16(
+                        state,
+                        transpose_msgs(msgs),
+                        set1(0),
+                        set1(0),
+                        set1(0),
+                        set1(0),
+                    )
+                };
+            }
+
+            // Hash columns in chunks of 16.
+            let mut col_chunk_iter = columns.array_chunks();
+            for col_chunk in &mut col_chunk_iter {
+                let msgs = col_chunk.map(|column| column.data[i].0);
+                state = unsafe { compress16(state, msgs, set1(0), set1(0), set1(0), set1(0)) };
+            }
+
+            // Hash remaining columns.
+            let remainder = col_chunk_iter.remainder();
+            if !remainder.is_empty() {
+                let msgs = remainder
+                    .iter()
+                    .map(|column| column.data[i].0)
+                    .chain(std::iter::repeat(unsafe { set1(0) }))
+                    .take(16)
+                    .collect_vec()
+                    .try_into()
+                    .unwrap();
+                state = unsafe { compress16(state, msgs, set1(0), set1(0), set1(0), set1(0)) };
+            }
+            let state: [Blake2sHash; 16] =
+                unsafe { std::mem::transmute(untranspose_states(state)) };
+            res.extend_from_slice(&state);
+        }
+        res
+    }
+}

src/core/backend/avx512/circle.rs

@@ -132,7 +132,10 @@ impl PolyOps for AVX512Backend {
     ) -> CircleEvaluation<Self, BaseField, BitReversedOrder> {
         // TODO(spapini): Optimize.
         let eval = CPUBackend::new_canonical_ordered(coset, as_cpu_vec(values));
-        CircleEvaluation::new(eval.domain, Col::<AVX512Backend, _>::from_iter(eval.values))
+        CircleEvaluation::new(
+            eval.domain,
+            Col::<AVX512Backend, BaseField>::from_iter(eval.values),
+        )
     }
 
     fn interpolate(

src/core/backend/avx512/mod.rs

@@ -1,4 +1,5 @@
 pub mod bit_reverse;
+mod blake2s;
 pub mod blake2s_avx;
 pub mod circle;
 pub mod cm31;
@@ -202,7 +203,7 @@ mod tests {
         for i in 1..16 {
             let len = 1 << i;
             let mut col = Col::<B, BaseField>::from_iter((0..len).map(BaseField::from));
-            B::bit_reverse_column(&mut col);
+            <B as ColumnOps<BaseField>>::bit_reverse_column(&mut col);
             assert_eq!(
                 col.to_vec(),
                 (0..len)

src/core/backend/cpu/blake2s.rs

@@ -0,0 +1,24 @@
+use itertools::Itertools;
+
+use crate::commitment_scheme::blake2_hash::Blake2sHash;
+use crate::commitment_scheme::blake2_merkle::Blake2sMerkleHasher;
+use crate::commitment_scheme::ops::{MerkleHasher, MerkleOps};
+use crate::core::backend::CPUBackend;
+use crate::core::fields::m31::BaseField;
+
+impl MerkleOps<Blake2sMerkleHasher> for CPUBackend {
+    fn commit_on_layer(
+        log_size: u32,
+        prev_layer: Option<&Vec<Blake2sHash>>,
+        columns: &[&Vec<BaseField>],
+    ) -> Vec<Blake2sHash> {
+        (0..(1 << log_size))
+            .map(|i| {
+                Blake2sMerkleHasher::hash_node(
+                    prev_layer.map(|prev_layer| (prev_layer[2 * i], prev_layer[2 * i + 1])),
+                    &columns.iter().map(|column| column[i]).collect_vec(),
+                )
+            })
+            .collect()
+    }
+}

src/core/backend/cpu/mod.rs

@@ -1,3 +1,4 @@
+mod blake2s;
 mod circle;
 mod fri;
 pub mod quotients;