Auto merge of #127809 - Urgau:blake2-stable-hash, r=<try>

[perf] Change stable hasher to Blake2s Based on rust-lang/rustc-stable-hash@main...Urgau:rustc-stable-hash:blake2 With inputs from https://jszym.com/blog/short_input_hash/ and #127754. cc `@michaelwoerister` `@Kobzol` r? ghost
rust-lang · Jul 16, 2024 · 4172c7c · 4172c7c
2 parents 5572759 + 9b14ce5
commit 4172c7c
Show file tree

Hide file tree

Showing 7 changed files with 62 additions and 11 deletions.
diff --git a/Cargo.lock b/Cargo.lock
@@ -362,6 +362,15 @@ version = "2.5.0"
 source = "registry+https:/rust-lang/crates.io-index"
 checksum = "cf4b9d6a944f767f8e5e0db018570623c85f3d925ac718db4e06d0187adb21c1"
 
+[[package]]
+name = "blake2"
+version = "0.10.6"
+source = "registry+https:/rust-lang/crates.io-index"
+checksum = "46502ad458c9a52b69d4d4d32775c788b7a1b85e8bc9d482d92250fc0e3f8efe"
+dependencies = [
+ "digest",
+]
+
 [[package]]
 name = "block-buffer"
 version = "0.10.4"
@@ -1131,6 +1140,7 @@ checksum = "9ed9a281f7bc9b7576e61468ba615a66a5c8cfdff42420a70aa82701a3b1e292"
 dependencies = [
  "block-buffer",
  "crypto-common",
+ "subtle",
 ]
 
 [[package]]
@@ -3518,8 +3528,10 @@ checksum = "5be1bdc7edf596692617627bbfeaba522131b18e06ca4df2b6b689e3c5d5ce84"
 [[package]]
 name = "rustc-stable-hash"
 version = "0.1.0"
-source = "registry+https:/rust-lang/crates.io-index"
-checksum = "e5c9f15eec8235d7cb775ee6f81891db79b98fd54ba1ad8fae565b88ef1ae4e2"
+source = "git+https:/Urgau/rustc-stable-hash.git?rev=5dbe93b#5dbe93b5959372a64f6f7ea60382fe35be55dcb3"
+dependencies = [
+ "blake2",
+]
 
 [[package]]
 name = "rustc-std-workspace-alloc"
@@ -5408,6 +5420,12 @@ dependencies = [
  "syn 1.0.109",
 ]
 
+[[package]]
+name = "subtle"
+version = "2.6.1"
+source = "registry+https:/rust-lang/crates.io-index"
+checksum = "13c2bddecc57b384dee18652358fb23172facb8a2c51ccc10d74c157bdea3292"
+
 [[package]]
 name = "suggest-tests"
 version = "0.1.0"

diff --git a/compiler/rustc_data_structures/Cargo.toml b/compiler/rustc_data_structures/Cargo.toml
@@ -15,7 +15,7 @@ jobserver_crate = { version = "0.1.28", package = "jobserver" }
 measureme = "11"
 rustc-hash = "1.1.0"
 rustc-rayon = { version = "0.5.0", optional = true }
-rustc-stable-hash = { version = "0.1.0", features = ["nightly"] }
+rustc-stable-hash = { git = "https:/Urgau/rustc-stable-hash.git", rev = "5dbe93b", features = ["blake2", "nightly"] }
 rustc_arena = { path = "../rustc_arena" }
 rustc_graphviz = { path = "../rustc_graphviz" }
 rustc_index = { path = "../rustc_index", package = "rustc_index" }

diff --git a/compiler/rustc_data_structures/src/fingerprint.rs b/compiler/rustc_data_structures/src/fingerprint.rs
@@ -158,8 +158,15 @@ impl FromStableHash for Fingerprint {
  type Hash = StableHasherHash;
 
  #[inline]
- fn from(StableHasherHash([_0, _1]): Self::Hash) -> Self {
- Fingerprint(_0, _1)
+ fn from(StableHasherHash(bytes): Self::Hash) -> Self {
+ let p0 = u64::from_le_bytes(bytes[0..8].try_into().unwrap());
+ let p1 = u64::from_le_bytes(bytes[8..16].try_into().unwrap());
+ let p2 = u64::from_le_bytes(bytes[16..24].try_into().unwrap());
+ let p3 = u64::from_le_bytes(bytes[24..32].try_into().unwrap());
+
+ // See https://stackoverflow.com/a/27952689 on why this function is
+ // implemented this way.
+ Fingerprint(p0.wrapping_mul(3).wrapping_add(p1), p2.wrapping_mul(3).wrapping_add(p3))
  }
 }
 

diff --git a/compiler/rustc_data_structures/src/hashes.rs b/compiler/rustc_data_structures/src/hashes.rs
@@ -60,8 +60,19 @@ impl FromStableHash for Hash64 {
  type Hash = StableHasherHash;
 
  #[inline]
- fn from(StableHasherHash([_0, __1]): Self::Hash) -> Self {
- Self { inner: _0 }
+ fn from(StableHasherHash(bytes): Self::Hash) -> Self {
+ let p0 = u64::from_le_bytes(bytes[0..8].try_into().unwrap());
+ let p1 = u64::from_le_bytes(bytes[8..16].try_into().unwrap());
+ let p2 = u64::from_le_bytes(bytes[16..24].try_into().unwrap());
+ let p3 = u64::from_le_bytes(bytes[24..32].try_into().unwrap());
+
+ // See https://stackoverflow.com/a/27952689 on why this function is
+ // implemented this way.
+ let m0 = p0.wrapping_mul(3).wrapping_add(p1);
+ let m1 = p2.wrapping_mul(3).wrapping_add(p3);
+ let h = m0.wrapping_mul(3).wrapping_add(m1);
+
+ Self { inner: h }
  }
 }
 
@@ -127,8 +138,18 @@ impl FromStableHash for Hash128 {
  type Hash = StableHasherHash;
 
  #[inline]
- fn from(StableHasherHash([_0, _1]): Self::Hash) -> Self {
- Self { inner: u128::from(_0) | (u128::from(_1) << 64) }
+ fn from(StableHasherHash(bytes): Self::Hash) -> Self {
+ let p0 = u64::from_le_bytes(bytes[0..8].try_into().unwrap());
+ let p1 = u64::from_le_bytes(bytes[8..16].try_into().unwrap());
+ let p2 = u64::from_le_bytes(bytes[16..24].try_into().unwrap());
+ let p3 = u64::from_le_bytes(bytes[24..32].try_into().unwrap());
+
+ // See https://stackoverflow.com/a/27952689 on why this function is
+ // implemented this way.
+ let upper = p0.wrapping_mul(3).wrapping_add(p1);
+ let lower = p2.wrapping_mul(3).wrapping_add(p3);
+
+ Self { inner: u128::from(lower) | (u128::from(upper) << 64) }
  }
 }
 

diff --git a/compiler/rustc_data_structures/src/stable_hasher.rs b/compiler/rustc_data_structures/src/stable_hasher.rs
@@ -11,9 +11,9 @@ mod tests;
 
 pub use crate::hashes::{Hash128, Hash64};
 
+pub use rustc_stable_hash::hashers::Blake2s256Hash as StableHasherHash;
+pub use rustc_stable_hash::hashers::StableBlake2sHasher256 as StableHasher;
 pub use rustc_stable_hash::FromStableHash;
-pub use rustc_stable_hash::SipHasher128Hash as StableHasherHash;
-pub use rustc_stable_hash::StableSipHasher128 as StableHasher;
 
 /// Something that implements `HashStable<CTX>` can be hashed in a way that is
 /// stable across multiple compilation sessions.

diff --git a/src/tools/tidy/src/deps.rs b/src/tools/tidy/src/deps.rs
@@ -94,6 +94,7 @@ const EXCEPTIONS: ExceptionList = &[
  ("ryu", "Apache-2.0 OR BSL-1.0"), // BSL is not acceptble, but we use it under Apache-2.0 // cargo/... (because of serde)
  ("self_cell", "Apache-2.0"), // rustc (fluent translations)
  ("snap", "BSD-3-Clause"), // rustc
+ ("subtle", "BSD-3-Clause"), // rustc blake2 deps
  ("wasm-encoder", "Apache-2.0 WITH LLVM-exception"), // rustc
  ("wasmparser", "Apache-2.0 WITH LLVM-exception"), // rustc
  // tidy-alphabetical-end
@@ -235,6 +236,7 @@ const PERMITTED_RUSTC_DEPENDENCIES: &[&str] = &[
  "arrayvec",
  "autocfg",
  "bitflags",
+ "blake2",
  "block-buffer",
  "byteorder", // via ruzstd in object in thorin-dwp
  "cc",
@@ -376,6 +378,7 @@ const PERMITTED_RUSTC_DEPENDENCIES: &[&str] = &[
  "stacker",
  "static_assertions",
  "strsim",
+ "subtle",
  "syn",
  "synstructure",
  "tempfile",

diff --git a/src/tools/tidy/src/extdeps.rs b/src/tools/tidy/src/extdeps.rs
@@ -9,6 +9,8 @@ const ALLOWED_SOURCES: &[&str] = &[
  r#""registry+https:/rust-lang/crates.io-index""#,
  // This is `rust_team_data` used by `site` in src/tools/rustc-perf,
  r#""git+https:/rust-lang/team#a5260e76d3aa894c64c56e6ddc8545b9a98043ec""#,
+ // WIP Blake2 in rustc-stable-hash
+ r#""git+https:/Urgau/rustc-stable-hash.git?rev=5dbe93b#5dbe93b5959372a64f6f7ea60382fe35be55dcb3""#,
 ];
 
 /// Checks for external package sources. `root` is the path to the directory that contains the