chore(release): v0.41.0 — f32 op-completeness (#719) + f32 soundness CI gate (#712) (#722)

#719: f32.store/abs/neg/copysign + local.set/tee + mixed f32/int AAPCS-VFP params
(156/156 vs wasmtime, ~18 falcon f32 skips cleared; f32-across-call declined
loudly, f64 = phase 2). #712: all six f32 compares gated (84/84) + the f32
soundness differentials CI-wired + NaN-aware compare (§4.3.3 non-determinism).
Pin sweep 0.40.0→0.41.0 incl. npm; claim gate 17/17.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

Author: avrabe

CHANGELOG.md

@@ -7,6 +7,37 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.41.0] - 2026-07-11
+
+**f32 op-completeness for falcon + a soundness-gate hardening.**
+
+### Added
+
+- **f32 op-completeness (#719, #369, GI-FPU-002).** After v0.40.0 landed f32
+  load/reinterpret, falcon's float functions skipped on the next unsupported op.
+  Now lowered: `f32.store` (`VMOV Rn,Sn` + the proven i32.store address path),
+  `f32.abs`/`f32.neg`/`f32.copysign` (sign-bit ops, verified against ±0.0, ±inf,
+  NaN-sign), `f32.local.set`/`tee` (VFP home write-back), and **mixed f32/int
+  parameter lists** (AAPCS-VFP independent register pools — `(param i32 f32)` →
+  i32 in R0, f32 in S0). 156/156 bit-exact vs wasmtime on cortex-m4f. This clears
+  ~18 of falcon's remaining f32 skips. `f32` live across a call (S0–S15
+  caller-saved spill/rehome) is declined LOUDLY — the honest-skip, a documented
+  follow-on; f64 is phase 2.
+
+### Changed
+
+- **The f32 soundness oracles are now CI-gated (#712).** #712 reported all six
+  f32 comparisons returning value-independent wrong results — the `MOVS`-after-
+  `VMRS` flag clobber, already fixed in v0.40.0 (#716) but only *gated* on `flt`/
+  `fgt`, leaving `eq`/`ne`/`le`/`ge` fixed-but-ungated (the vacuous-harness
+  pattern that let the bug ship). Now: the compare oracle covers all six
+  (84/84), and the f32 soundness differentials (compares + trunc-trap +
+  load/reinterpret + the #719 op set) run in CI on every PR — the f32 path can no
+  longer silently regress. Also made the differential NaN-aware: WASM leaves a
+  NaN result's sign/payload non-deterministic (§4.3.3), so `f32.div(-0.0,0.0)` is
+  compared as NaN==NaN, not bit-exact (which diverged by wasmtime version), while
+  signed-zero and ±inf stay bit-exact.
+
 ## [0.40.0] - 2026-07-11
 
 **A five-lane North-Star feature hub: the f32 hard-float path completes and

Cargo.lock

@@ -28,7 +28,7 @@ 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 = "0.40.0"
+version = "0.41.0"
 edition = "2024"
 rust-version = "1.88"
 authors = ["PulseEngine Team"]

Cargo.toml

@@ -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 = "0.40.0",
+    version = "0.41.0",
 )
 
 # Bazel dependencies

MODULE.bazel

@@ -11,6 +11,6 @@ categories.workspace = true
 description = "AArch64 (A64) host-native backend for synth — integer subset (milestone 1, #538)"
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.40.0" }
+synth-core = { path = "../synth-core", version = "0.41.0" }
 thiserror.workspace = true
 tracing.workspace = true

crates/synth-backend-aarch64/Cargo.toml

@@ -11,6 +11,6 @@ categories.workspace = true
 description = "aWsm backend integration for the Synth compiler"
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.40.0" }
+synth-core = { path = "../synth-core", version = "0.41.0" }
 anyhow.workspace = true
 thiserror.workspace = true

crates/synth-backend-awsm/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 = "0.40.0" }
-synth-synthesis = { path = "../synth-synthesis", version = "0.40.0" }
+synth-core = { path = "../synth-core", version = "0.41.0" }
+synth-synthesis = { path = "../synth-synthesis", version = "0.41.0" }
 anyhow.workspace = true
 thiserror.workspace = true
 tracing.workspace = true

crates/synth-backend-riscv/Cargo.toml

@@ -11,6 +11,6 @@ categories.workspace = true
 description = "Wasker backend integration for the Synth compiler"
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.40.0" }
+synth-core = { path = "../synth-core", version = "0.41.0" }
 anyhow.workspace = true
 thiserror.workspace = true

crates/synth-backend-wasker/Cargo.toml

@@ -15,13 +15,13 @@ default = ["arm-cortex-m"]
 arm-cortex-m = ["synth-synthesis"]
 
 [dependencies]
-synth-core = { path = "../synth-core", version = "0.40.0" }
-synth-synthesis = { path = "../synth-synthesis", version = "0.40.0", optional = true }
+synth-core = { path = "../synth-core", version = "0.41.0" }
+synth-synthesis = { path = "../synth-synthesis", version = "0.41.0", optional = true }
 anyhow.workspace = true
 thiserror.workspace = true
 
 [dev-dependencies]
 # #667 move 2: the i64 pseudo-op expansion certification oracle
 # (tests/i64_expansion_certification.rs) feeds THIS crate's emitted encoder
 # bytes to the synth-verify expansion validator. Dev-only — no prod-dep edge.
-synth-verify = { path = "../synth-verify", version = "0.40.0", features = ["arm"] }
+synth-verify = { path = "../synth-verify", version = "0.41.0", features = ["arm"] }

crates/synth-backend/Cargo.toml

@@ -44,23 +44,23 @@ 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 = "0.40.0" }
-synth-frontend = { path = "../synth-frontend", version = "0.40.0" }
-synth-synthesis = { path = "../synth-synthesis", version = "0.40.0" }
-synth-backend = { path = "../synth-backend", version = "0.40.0" }
+synth-core = { path = "../synth-core", version = "0.41.0" }
+synth-frontend = { path = "../synth-frontend", version = "0.41.0" }
+synth-synthesis = { path = "../synth-synthesis", version = "0.41.0" }
+synth-backend = { path = "../synth-backend", version = "0.41.0" }
 
 # AArch64 host-native backend (#538) — small pure-Rust crate, always on.
-synth-backend-aarch64 = { path = "../synth-backend-aarch64", version = "0.40.0" }
+synth-backend-aarch64 = { path = "../synth-backend-aarch64", version = "0.41.0" }
 
 # Optional external backends
-synth-backend-awsm = { path = "../synth-backend-awsm", version = "0.40.0", optional = true }
-synth-backend-wasker = { path = "../synth-backend-wasker", version = "0.40.0", optional = true }
-synth-backend-riscv = { path = "../synth-backend-riscv", version = "0.40.0", optional = true }
+synth-backend-awsm = { path = "../synth-backend-awsm", version = "0.41.0", optional = true }
+synth-backend-wasker = { path = "../synth-backend-wasker", version = "0.41.0", optional = true }
+synth-backend-riscv = { path = "../synth-backend-riscv", version = "0.41.0", optional = true }
 
 # Optional translation validation — pure-Rust ordeal engine by default (#553),
 # no C++ toolchain needed. For the Z3 differential oracle build with
 # `--features verify,synth-verify/z3-solver` (+ SYNTH_SOLVER_DIFF=1 at runtime).
-synth-verify = { path = "../synth-verify", version = "0.40.0", optional = true, features = ["arm"] }
+synth-verify = { path = "../synth-verify", version = "0.41.0", optional = true, features = ["arm"] }
 
 # Optional PulseEngine WASM optimizer
 # Uncomment when loom crate is available: