feat: Make boojum compile with modern compiler

.github/workflows/ci.yaml

@@ -17,25 +17,6 @@ jobs:
       - uses: actions-rust-lang/setup-rust-toolchain@v1
       - run: cargo build --verbose
       - run: cargo test --verbose --all
-
-  build_old:
-    name: cargo build and test (packed_simd)
-    strategy:
-      matrix:
-        # Needs big runners to run tests
-        # Only macos-13-xlarge is Apple Silicon, as per:
-        # https://docs.github.com/en/actions/using-github-hosted-runners/about-larger-runners/about-larger-runners#about-macos-larger-runners
-        os: [ubuntu-22.04-github-hosted-16core, macos-13-xlarge]
-    runs-on: ${{ matrix.os }}
-    steps:
-      - uses: actions/checkout@v4
-      - uses: actions-rust-lang/setup-rust-toolchain@v1
-        with:
-          toolchain: nightly-2023-05-31 
-
-      # Still compile the old rust nightly with packed simd - until we have a good replacement in poseidon.
-      - run: RUSTFLAGS=-Awarnings cargo +nightly-2023-05-31 build --features include_packed_simd
-      - run: RUSTFLAGS=-Awarnings cargo +nightly-2023-05-31 test --features include_packed_simd
 
   formatting:
     name: cargo fmt

Cargo.toml

@@ -30,7 +30,6 @@ itertools = "0.10"
 blake2 = "0.10"
 sha2 = "0.10"
 num-modular = "0.5.1"
-packed_simd = { version = "0.3.9" , optional = true}
 pairing = { package = "pairing_ce", version = "=0.28.6" }
 crypto-bigint = "0.5"
 convert_case = "0.6"
@@ -61,9 +60,5 @@ opt-level = 3
 [features]
 # If enabled, logs will be using trace, if disabled, they will be printed to stdout.
 log_tracing = ["tracing"]
-# Currently packed_simd is no longer working with the newest nightly.
-# But we still keep it as a feature, as we didn't migrate all the code, and 
-# some people might want to use older rust nightly, to be able to gain some performance.
-include_packed_simd = ["packed_simd"]
 cr_paranoia_mode = []
 debug_track = []

rust-toolchain.toml

@@ -1,2 +1,2 @@
 [toolchain]
-channel = "nightly-2024-05-07"
+channel = "nightly-2024-08-01"

src/cs/implementations/fast_serialization.rs

@@ -272,7 +272,7 @@ where
 {
     fn write_into_buffer<W: Write>(&self, mut dst: W) -> Result<(), Box<dyn Error>> {
         // we avoid transmute here
-        let flattened_self = self[..].flatten();
+        let flattened_self = self[..].as_flattened();
 
         let len_as_base = flattened_self.len();
         let len_le_bytes = (len_as_base as u64).to_le_bytes();
@@ -346,7 +346,7 @@ where
 {
     fn write_into_buffer<W: Write>(&self, mut dst: W) -> Result<(), Box<dyn Error>> {
         // we avoid transmute here
-        let flattened_self = self[..].flatten();
+        let flattened_self = self[..].as_flattened();
 
         let len_as_base = flattened_self.len();
         let len_le_bytes = (len_as_base as u64).to_le_bytes();