feat(vcr-ra): ship immediate-shift folding default-on — v0.15.0 (#390, #242) (#467)

avrabe · claude · web-flow · commit cc203eca52e4 · 2026-06-24T11:56:11.000-05:00
Flips the immediate-shift folding peephole default-on (PR #463 landed it flag-off). A constant shift amount the stack selector materialized into a scratch register (`movw rM,#C; lsl rD,rN,rM`) now folds to the immediate form (`lsl rD,rN,#C`), removing the dead `movw` — −1 instruction, −1 live register per folded shift. Flip: arm_backend.rs default-ON with opt-out SYNTH_NO_IMM_SHIFT_FOLD=1. Re-froze the ARM goldens (control_step 316→304, flight_seam 866→774, flight_seam_flat 1006→910 = −200 B; signed_div_const unchanged — no register-shift folds). RV32 gate untouched (ARM-only peephole). Results preserved across the byte change: control_step 0x00210A55 (differential 13/13), flat+inlined flight_algo 0x07FDF307 (MATCH); opt-out restores the v0.14.0 bytes; full workspace suite green. Validated bit-identical + a net cycle win on the dissolved hot path (−2 cyc/call, .text 100→90 B on gust_mix). Cumulative dissolved hot-path: 64.0 → 58.0 (cmp→select) → 50.0 (local promotion) → 48.0 cyc/call. Pin-swept 0.14.0→0.15.0; CHANGELOG added. Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -7,6 +7,29 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.15.0] - 2026-06-24
+
+**Immediate-shift folding is now DEFAULT-ON (VCR-RA, #390, epic #242).** The stack
+selector lowers `i32.shl/shr_s/shr_u` to register-shift forms, materializing a
+constant shift amount into a scratch register (`movw rM,#C; lsl rD,rN,rM`); the
+peephole now folds that to the immediate form (`lsl rD,rN,#C`) and removes the dead
+`movw` — −1 instruction and −1 live register per folded shift. This is a
+**byte-changing** release on top of v0.14.0: `.text` shrinks on shift-heavy
+functions — control_step 316→304 B, flight_seam 866→774, flight_seam_flat 1006→910
+(−200 B across the frozen fixtures); signed_div_const (no register-shift folds)
+unchanged.
+
+Execution results are unchanged: control_step `0x00210A55` (differential 13/13) and
+flat+inlined flight_algo `0x07FDF307` are preserved. Validated bit-identical and a
+net cycle win on the dissolved hot path (−2 cyc/call, `.text` 100→90 B on gust_mix).
+Cumulative dissolved hot-path improvement across the codegen levers:
+64.0 → 58.0 (cmp→select) → 50.0 (local promotion) → 48.0 cyc/call.
+
+Soundness: only folds shift amounts in [1,31] (where register- and immediate-shift
+forms provably agree), the shift-amount register must be dead after the shift, and
+the rewrite is removal-only (offset-neutral before branch resolution). Escape hatch:
+`SYNTH_NO_IMM_SHIFT_FOLD=1` restores the register-shift form.
+
 ## [0.14.0] - 2026-06-24
 
 **i32 local promotion is now DEFAULT-ON (VCR-RA-001, #390, epic #242).** The ARM
diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
@@ -27,7 +27,7 @@ resolver = "2"
 # semver to publish, so the convention now catches up: workspace
 # version follows the release tag, bumped pre-tag in the release
 # checklist. See docs/release-process.md.
-version = "0.14.0"
+version = "0.15.0"
 edition = "2024"
 rust-version = "1.88"
 authors = ["PulseEngine Team"]
diff --git a/MODULE.bazel b/MODULE.bazel
@@ -7,7 +7,7 @@ module(
     name = "synth",
     # Kept in lockstep with [workspace.package] version in Cargo.toml.
     # Both are bumped pre-tag — see docs/release-process.md.
-    version = "0.14.0",
+    version = "0.15.0",
 )
 
 # Bazel dependencies
diff --git a/crates/synth-backend-awsm/Cargo.toml b/crates/synth-backend-awsm/Cargo.toml
@@ -11,6 +11,6 @@ categories.workspace = true
 description = "aWsm backend integration for the Synth compiler"
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.14.0" }
+synth-core = { path = "../synth-core", version = "0.15.0" }
 anyhow.workspace = true
 thiserror.workspace = true
diff --git a/crates/synth-backend-riscv/Cargo.toml b/crates/synth-backend-riscv/Cargo.toml
@@ -11,8 +11,8 @@ categories.workspace = true
 description = "RISC-V encoder, ELF builder, PMP allocator, and bare-metal startup for synth"
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.14.0" }
-synth-synthesis = { path = "../synth-synthesis", version = "0.14.0" }
+synth-core = { path = "../synth-core", version = "0.15.0" }
+synth-synthesis = { path = "../synth-synthesis", version = "0.15.0" }
 anyhow.workspace = true
 thiserror.workspace = true
 tracing.workspace = true
diff --git a/crates/synth-backend-wasker/Cargo.toml b/crates/synth-backend-wasker/Cargo.toml
@@ -11,6 +11,6 @@ categories.workspace = true
 description = "Wasker backend integration for the Synth compiler"
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.14.0" }
+synth-core = { path = "../synth-core", version = "0.15.0" }
 anyhow.workspace = true
 thiserror.workspace = true
diff --git a/crates/synth-backend/Cargo.toml b/crates/synth-backend/Cargo.toml
@@ -15,7 +15,7 @@ default = ["arm-cortex-m"]
 arm-cortex-m = ["synth-synthesis"]
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.14.0" }
-synth-synthesis = { path = "../synth-synthesis", version = "0.14.0", optional = true }
+synth-core = { path = "../synth-core", version = "0.15.0" }
+synth-synthesis = { path = "../synth-synthesis", version = "0.15.0", optional = true }
 anyhow.workspace = true
 thiserror.workspace = true
diff --git a/crates/synth-backend/src/arm_backend.rs b/crates/synth-backend/src/arm_backend.rs
@@ -492,11 +492,10 @@ fn compile_wasm_to_arm(
     // stack selector materialized into a scratch register (`movw rM,#C; lsl rD,rN,rM`)
     // folds to the immediate form (`lsl rD,rN,#C`), removing the dead `movw` — −1
     // instruction, −1 live register. Removal-only (offset-neutral before branch
-    // resolution, like the dead-store pass). BEHIND `SYNTH_IMM_SHIFT_FOLD=1`
-    // (opt-in, off by default ⇒ bit-identical) while it earns the execution
-    // differential + gale's G474RE DWT gate — the same gated path local promotion
-    // and cmp→select took before shipping default-on. gale-named lever toward ≤1.3×.
-    let arm_instrs = if std::env::var("SYNTH_IMM_SHIFT_FOLD").is_ok() {
+    // resolution, like the dead-store pass). DEFAULT-ON as of v0.15.0: validated
+    // bit-identical results + a net cycle win on the dissolved hot path (−2
+    // cyc/call, .text 100→90 B on gust_mix). Escape hatch: `SYNTH_NO_IMM_SHIFT_FOLD=1`.
+    let arm_instrs = if std::env::var("SYNTH_NO_IMM_SHIFT_FOLD").is_err() {
         let (out, folds) = synth_synthesis::liveness::fold_immediate_shifts(&arm_instrs);
         if std::env::var("SYNTH_FUSE_STATS").is_ok() {
             eprintln!(
diff --git a/crates/synth-cli/Cargo.toml b/crates/synth-cli/Cargo.toml
@@ -27,18 +27,18 @@ verify = ["synth-verify"]
 # Path deps carry `version` so `cargo publish` rewrites them to the
 # crates.io coordinate. Bumping the workspace version requires
 # updating these in lockstep — see docs/release-process.md.
-synth-core = { path = "../synth-core", version = "0.14.0" }
-synth-frontend = { path = "../synth-frontend", version = "0.14.0" }
-synth-synthesis = { path = "../synth-synthesis", version = "0.14.0" }
-synth-backend = { path = "../synth-backend", version = "0.14.0" }
+synth-core = { path = "../synth-core", version = "0.15.0" }
+synth-frontend = { path = "../synth-frontend", version = "0.15.0" }
+synth-synthesis = { path = "../synth-synthesis", version = "0.15.0" }
+synth-backend = { path = "../synth-backend", version = "0.15.0" }
 
 # Optional external backends
-synth-backend-awsm = { path = "../synth-backend-awsm", version = "0.14.0", optional = true }
-synth-backend-wasker = { path = "../synth-backend-wasker", version = "0.14.0", optional = true }
-synth-backend-riscv = { path = "../synth-backend-riscv", version = "0.14.0", optional = true }
+synth-backend-awsm = { path = "../synth-backend-awsm", version = "0.15.0", optional = true }
+synth-backend-wasker = { path = "../synth-backend-wasker", version = "0.15.0", optional = true }
+synth-backend-riscv = { path = "../synth-backend-riscv", version = "0.15.0", optional = true }
 
 # Optional verification (requires z3)
-synth-verify = { path = "../synth-verify", version = "0.14.0", optional = true, features = ["z3-solver", "arm"] }
+synth-verify = { path = "../synth-verify", version = "0.15.0", optional = true, features = ["z3-solver", "arm"] }
 
 # Optional PulseEngine WASM optimizer
 # Uncomment when loom crate is available:
diff --git a/crates/synth-cli/tests/frozen_codegen_bytes.rs b/crates/synth-cli/tests/frozen_codegen_bytes.rs
@@ -68,6 +68,7 @@ fn text_sha256(wasm: &str, backend: &str, target: &str) -> (String, usize) {
     let out = Command::new(synth())
         .env_remove("SYNTH_NO_CMP_SELECT_FUSE")
         .env_remove("SYNTH_NO_LOCAL_PROMOTE")
+        .env_remove("SYNTH_NO_IMM_SHIFT_FOLD")
         .env_remove("SYNTH_CONST_CSE")
         .args([
             "compile",
@@ -129,33 +130,33 @@ fn assert_frozen(cases: &[(&str, &str, usize)], backend: &str, target: &str) {
 /// the `.py` differentials cover): control_step ↔ 0x00210A55, flight_seam_flat ↔
 /// flat+inlined flight_algo 0x07FDF307, plus flight_seam and the div seam.
 ///
-/// Goldens RE-FROZEN for v0.14.0 (#390): local promotion is now default-on (on top
-/// of v0.13.0 cmp→select), so these lock the PROMOTED+FUSED .text. The execution
-/// RESULTS are preserved — re-verified on this commit with both default-on:
-/// control_step still 0x00210A55 (control_step_differential.py 13/13), flat+inlined
-/// flight_algo still 0x07FDF307 (flight_seam_differential.py MATCH). .text shrank
-/// again (stack spill/reloads eliminated): control_step 324→316, flight_seam
-/// 902→866, flight_seam_flat 1122→1006 (−154 B total); signed_div_const (no
-/// promotable i32 locals) unchanged. gale G474RE DWT: gust_mix 58→50 cyc/call
-/// (−14%), 5→0 [sp] traffic. Prior cmp→select-only goldens were on main @ 377b93e
-/// (v0.13.0), 2026-06-24.
+/// Goldens RE-FROZEN for v0.15.0 (#390): immediate-shift folding is now default-on
+/// (on top of v0.13.0 cmp→select + v0.14.0 local promotion), so these lock the
+/// folded+promoted+fused .text. The execution RESULTS are preserved — re-verified
+/// on this commit: control_step still 0x00210A55 (control_step_differential.py
+/// 13/13), flat+inlined flight_algo still 0x07FDF307 (flight_seam_differential.py
+/// MATCH). .text shrank again (constant shift-amount `movw`s removed): control_step
+/// 316→304, flight_seam 866→774, flight_seam_flat 1006→910 (−200 B total);
+/// signed_div_const (no register-shift folds) unchanged. Measured −2 cyc/call on the
+/// dissolved hot path (.text 100→90 B on gust_mix). Prior promotion goldens were on
+/// main @ 6b46f09 (v0.14.0), 2026-06-24.
 #[test]
 fn frozen_fixtures_text_is_bit_identical_oracle_001() {
     let cases = [
         (
             "control_step.wasm",
-            "cd929e7d91a8f7aad93f0e1cf0c93ecf3ccc6584ee94fb32e68d591134ed1410",
-            316usize,
+            "1a97711cfb4754794a8577814388f08b81eff444edcba3de7d3e3d18ff435183",
+            304usize,
         ),
         (
             "flight_seam.wasm",
-            "52b19365e32bcd9d5a4be74565d0fa467517eb3a07648a3e1ccd0a67556c1948",
-            866,
+            "9e73eea3867ba085820329951e84a7d650c38a7fc78d9d03a6a83d02963f9670",
+            774,
         ),
         (
             "flight_seam_flat.wasm",
-            "fa019f18cbc93869fff51630c6ab9cff6c4e052e22d783fe741b663ece49fa1e",
-            1006,
+            "887ea546429a4569112147fdc94b0ba90f02a6ccd2b511aa2ca48dab017dbc2c",
+            910,
         ),
         (
             "signed_div_const.wasm",
diff --git a/crates/synth-frontend/Cargo.toml b/crates/synth-frontend/Cargo.toml
@@ -14,7 +14,7 @@ description = "WASM/WAT parser and module decoder frontend for the Synth compile
 # Internal path deps carry an explicit version so `cargo publish`
 # can rewrite to the crates.io coordinate. `path` is used for
 # in-workspace builds; `version` is what crates.io sees.
-synth-core = { path = "../synth-core", version = "0.14.0" }
+synth-core = { path = "../synth-core", version = "0.15.0" }
 
 wasmparser.workspace = true
 wasm-encoder.workspace = true
diff --git a/crates/synth-opt/Cargo.toml b/crates/synth-opt/Cargo.toml
@@ -11,7 +11,7 @@ categories.workspace = true
 description = "Peephole optimization passes for the Synth compiler"
 
 [dependencies]
-synth-cfg = { path = "../synth-cfg", version = "0.14.0" }
+synth-cfg = { path = "../synth-cfg", version = "0.15.0" }
 
 [dev-dependencies]
 criterion = { version = "0.8", features = ["html_reports"] }
diff --git a/crates/synth-synthesis/Cargo.toml b/crates/synth-synthesis/Cargo.toml
@@ -11,9 +11,9 @@ categories.workspace = true
 description = "WASM-to-ARM instruction selection and peephole optimizer"
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.14.0" }
-synth-cfg = { path = "../synth-cfg", version = "0.14.0" }
-synth-opt = { path = "../synth-opt", version = "0.14.0" }
+synth-core = { path = "../synth-core", version = "0.15.0" }
+synth-cfg = { path = "../synth-cfg", version = "0.15.0" }
+synth-opt = { path = "../synth-opt", version = "0.15.0" }
 serde.workspace = true
 anyhow.workspace = true
 thiserror.workspace = true
diff --git a/crates/synth-verify/Cargo.toml b/crates/synth-verify/Cargo.toml
@@ -17,12 +17,12 @@ arm = ["synth-synthesis"]
 
 [dependencies]
 # Core dependencies (always required)
-synth-core = { path = "../synth-core", version = "0.14.0" }
-synth-cfg = { path = "../synth-cfg", version = "0.14.0" }
-synth-opt = { path = "../synth-opt", version = "0.14.0" }
+synth-core = { path = "../synth-core", version = "0.15.0" }
+synth-cfg = { path = "../synth-cfg", version = "0.15.0" }
+synth-opt = { path = "../synth-opt", version = "0.15.0" }
 
 # ARM synthesis (optional, behind 'arm' feature)
-synth-synthesis = { path = "../synth-synthesis", version = "0.14.0", optional = true }
+synth-synthesis = { path = "../synth-synthesis", version = "0.15.0", optional = true }
 
 # SMT solver for formal verification
 z3 = { version = "0.19", features = ["static-link-z3"], optional = true }

Original file line number	Diff line number	Diff line change
`@@ -7,7 +7,7 @@ module(`
`7`	`7`	`name = "synth",`
`8`	`8`	`# Kept in lockstep with [workspace.package] version in Cargo.toml.`
`9`	`9`	`# Both are bumped pre-tag — see docs/release-process.md.`
`10`		`- version = "0.14.0",`
	`10`	`+ version = "0.15.0",`
`11`	`11`	`)`
`12`	`12`
`13`	`13`	`# Bazel dependencies`