From a6a991b384380ecc5206b4832f9278650ab73620 Mon Sep 17 00:00:00 2001 From: Andrew Milson Date: Tue, 30 Apr 2024 23:42:59 -0400 Subject: [PATCH] Add bit_reverse for SIMD backend --- .../src/core/backend/simd/bit_reverse.rs | 88 +++++++++---------- crates/prover/src/core/backend/simd/mod.rs | 50 ++++++++++- 2 files changed, 90 insertions(+), 48 deletions(-) diff --git a/crates/prover/src/core/backend/simd/bit_reverse.rs b/crates/prover/src/core/backend/simd/bit_reverse.rs index 94158818a..4f311b4c2 100644 --- a/crates/prover/src/core/backend/simd/bit_reverse.rs +++ b/crates/prover/src/core/backend/simd/bit_reverse.rs @@ -1,14 +1,19 @@ -use std::simd::{u32x16, Swizzle}; +use std::array; +use std::mem::transmute; +use std::simd::Swizzle; +use super::utils::{LoHiInterleaveHiHi, LoLoInterleaveHiLo}; use super::PackedBaseField; -use crate::core::backend::simd::utils::{LoHiInterleaveHiHi, LoLoInterleaveHiLo}; use crate::core::utils::bit_reverse_index; const VEC_BITS: u32 = 4; + const W_BITS: u32 = 3; + pub const MIN_LOG_SIZE: u32 = 2 * W_BITS + VEC_BITS; -/// Bit reverses packed M31 values. +/// Bit reverses M31 values. +/// /// Given an array `A[0..2^n)`, computes `B[i] = A[bit_reverse(i)]`. pub fn bit_reverse_m31(data: &mut [PackedBaseField]) { assert!(data.len().is_power_of_two()); @@ -18,16 +23,16 @@ pub fn bit_reverse_m31(data: &mut [PackedBaseField]) { // |v_h| = |v_l| = VEC_BITS, |w_h| = |w_l| = W_BITS, |a| = n - 2*W_BITS - VEC_BITS. // The loops go over a, w_l, w_h, and then swaps the 16 by 16 values at: // * w_h a w_l * <-> * rev(w_h a w_l) *. - // These are 1 or 2 chunks of 2^W_BITS contiguous AVX512 vectors. + // These are 1 or 2 chunks of 2^W_BITS contiguous `u32x16` vectors. let log_size = data.len().ilog2(); let a_bits = log_size - 2 * W_BITS - VEC_BITS; // TODO(spapini): when doing multithreading, do it over a. - for a in 0u32..(1 << a_bits) { - for w_l in 0u32..(1 << W_BITS) { - let w_l_rev = w_l.reverse_bits() >> (32 - W_BITS); - for w_h in 0u32..(w_l_rev + 1) { + for a in 0u32..1 << a_bits { + for w_l in 0u32..1 << W_BITS { + let w_l_rev = w_l.reverse_bits() >> (u32::BITS - W_BITS); + for w_h in 0..w_l_rev + 1 { let idx = ((((w_h << a_bits) | a) << W_BITS) | w_l) as usize; let idx_rev = bit_reverse_index(idx, log_size - VEC_BITS); @@ -38,7 +43,7 @@ pub fn bit_reverse_m31(data: &mut [PackedBaseField]) { // Read first chunk. // TODO(spapini): Think about optimizing a_bits. - let chunk0 = std::array::from_fn(|i| unsafe { + let chunk0 = array::from_fn(|i| unsafe { *data.get_unchecked(idx + (i << (2 * W_BITS + a_bits))) }); let values0 = bit_reverse16(chunk0); @@ -56,7 +61,7 @@ pub fn bit_reverse_m31(data: &mut [PackedBaseField]) { } // Read bit reversed chunk. - let chunk1 = std::array::from_fn(|i| unsafe { + let chunk1 = array::from_fn(|i| unsafe { *data.get_unchecked(idx_rev + (i << (2 * W_BITS + a_bits))) }); let values1 = bit_reverse16(chunk1); @@ -75,23 +80,7 @@ pub fn bit_reverse_m31(data: &mut [PackedBaseField]) { /// Bit reverses 256 M31 values, packed in 16 words of 16 elements each. fn bit_reverse16(data: [PackedBaseField; 16]) -> [PackedBaseField; 16] { - let mut data: [u32x16; 16] = unsafe { std::mem::transmute(data) }; - // L is an input to _mm512_permutex2var_epi32, and it is used to - // interleave the first half of a with the first half of b. - const _L: u32x16 = unsafe { - core::mem::transmute([ - 0b00000, 0b10000, 0b00001, 0b10001, 0b00010, 0b10010, 0b00011, 0b10011, 0b00100, - 0b10100, 0b00101, 0b10101, 0b00110, 0b10110, 0b00111, 0b10111, - ]) - }; - // H is an input to _mm512_permutex2var_epi32, and it is used to interleave the second half - // interleave the second half of a with the second half of b. - const _H: u32x16 = unsafe { - core::mem::transmute([ - 0b01000, 0b11000, 0b01001, 0b11001, 0b01010, 0b11010, 0b01011, 0b11011, 0b01100, - 0b11100, 0b01101, 0b11101, 0b01110, 0b11110, 0b01111, 0b11111, - ]) - }; + let mut data = data.map(PackedBaseField::into_simd); // Denote the index of each element in the 16 packed M31 words as abcd:0123, // where abcd is the index of the packed word and 0123 is the index of the element in the word. @@ -105,7 +94,7 @@ fn bit_reverse16(data: [PackedBaseField; 16]) -> [PackedBaseField; 16] { // 3210:dcba for _ in 0..4 { // Apply the abcd:0123 => 0abc:123d permutation. - // _mm512_permutex2var_epi32() with L allows us to interleave the first half of 2 words. + // `LoLoInterleaveHiLo` allows us to interleave the first half of 2 words. // For example, the second call interleaves 0010:0xyz (low half of register 2) with // 0011:0xyz (low half of register 3), and stores the result in register 1 (0001). // This results in @@ -131,38 +120,45 @@ fn bit_reverse16(data: [PackedBaseField; 16]) -> [PackedBaseField; 16] { LoHiInterleaveHiHi::concat_swizzle(data[14], data[15]), ]; } - unsafe { std::mem::transmute(data) } + + unsafe { transmute(data) } } #[cfg(test)] mod tests { - use super::{bit_reverse16, bit_reverse_m31}; - use crate::core::backend::simd::BaseFieldVec; + use std::array; + use std::mem::transmute; + + use aligned::{Aligned, A64}; + + use super::bit_reverse16; + use crate::core::backend::simd::bit_reverse::bit_reverse_m31; + use crate::core::backend::simd::column::BaseFieldVec; use crate::core::backend::Column; use crate::core::fields::m31::BaseField; - use crate::core::utils::bit_reverse; + use crate::core::utils::bit_reverse as ground_truth_bit_reverse; #[test] fn test_bit_reverse16() { - let data: [u32; 256] = std::array::from_fn(|i| i as u32); - let expected: [u32; 256] = std::array::from_fn(|i| (i as u32).reverse_bits() >> 24); - unsafe { - let data = bit_reverse16(std::mem::transmute(data)); - assert_eq!(std::mem::transmute::<_, [u32; 256]>(data), expected); - } + let data: Aligned = Aligned(array::from_fn(|i| i as u32)); + let mut expected: Aligned = data; + ground_truth_bit_reverse(&mut *expected); + + let res = unsafe { transmute::<_, [u32; 256]>(bit_reverse16(transmute(data))) }; + + assert_eq!(res, *expected); } #[test] - fn test_bit_reverse() { + fn bit_reverse_m31_works() { const SIZE: usize = 1 << 15; - let data: Vec<_> = (0..SIZE as u32) - .map(BaseField::from_u32_unchecked) - .collect(); + let data: Vec<_> = (0..SIZE).map(BaseField::from).collect(); let mut expected = data.clone(); - bit_reverse(&mut expected); - let mut data: BaseFieldVec = data.into_iter().collect(); + ground_truth_bit_reverse(&mut expected); + + let mut res: BaseFieldVec = data.into_iter().collect(); + bit_reverse_m31(&mut res.data[..]); - bit_reverse_m31(&mut data.data[..]); - assert_eq!(data.to_cpu(), expected); + assert_eq!(res.to_cpu(), expected); } } diff --git a/crates/prover/src/core/backend/simd/mod.rs b/crates/prover/src/core/backend/simd/mod.rs index 15b80900f..ae6abb386 100644 --- a/crates/prover/src/core/backend/simd/mod.rs +++ b/crates/prover/src/core/backend/simd/mod.rs @@ -1,3 +1,4 @@ +use self::bit_reverse::bit_reverse_m31; use self::column::{BaseFieldVec, SecureFieldVec}; use self::m31::PackedBaseField; use self::qm31::PackedSecureField; @@ -5,6 +6,7 @@ use super::ColumnOps; use crate::core::fields::m31::BaseField; use crate::core::fields::qm31::SecureField; use crate::core::fields::{FieldExpOps, FieldOps}; +use crate::core::utils::bit_reverse as cpu_bit_reverse; pub mod bit_reverse; pub mod cm31; @@ -19,8 +21,14 @@ pub struct SimdBackend; impl ColumnOps for SimdBackend { type Column = BaseFieldVec; - fn bit_reverse_column(_column: &mut Self::Column) { - todo!() + fn bit_reverse_column(column: &mut Self::Column) { + // Fallback to cpu bit_reverse. + if column.data.len().ilog2() < bit_reverse::MIN_LOG_SIZE { + cpu_bit_reverse(column.as_mut()); + return; + } + + bit_reverse_m31(&mut column.data); } } @@ -43,3 +51,41 @@ impl FieldOps for SimdBackend { PackedSecureField::batch_inverse(&column.data, &mut dst.data); } } + +#[cfg(test)] +mod tests { + use itertools::Itertools; + + use super::bit_reverse; + use crate::core::backend::simd::column::BaseFieldVec; + use crate::core::backend::simd::SimdBackend; + use crate::core::backend::{Column, ColumnOps}; + use crate::core::fields::m31::BaseField; + use crate::core::utils::bit_reverse as ground_truth_bit_reverse; + + #[test] + fn bit_reverse_small_column_works() { + const LOG_SIZE: u32 = bit_reverse::MIN_LOG_SIZE - 1; + let column = (0..1 << LOG_SIZE).map(BaseField::from).collect_vec(); + let mut expected = column.clone(); + ground_truth_bit_reverse(&mut expected); + + let mut res = column.iter().copied().collect::(); + >::bit_reverse_column(&mut res); + + assert_eq!(res.to_cpu(), expected); + } + + #[test] + fn bit_reverse_large_column_works() { + const LOG_SIZE: u32 = bit_reverse::MIN_LOG_SIZE; + let column = (0..1 << LOG_SIZE).map(BaseField::from).collect_vec(); + let mut expected = column.clone(); + ground_truth_bit_reverse(&mut expected); + + let mut res = column.iter().copied().collect::(); + >::bit_reverse_column(&mut res); + + assert_eq!(res.to_cpu(), expected); + } +}