Improve Support for WASM (#60)

* feat: use `XofReader` instead of `io::Read` for no_std compatibility * feat: remove `thiserror` dependency thiserror is not no_std safe * feat: move deps std features behind new `std` feature - add new `std` feature - add `simd_backend` - move `colored` dep behind profile feature * feat: use old rand(v0.7) for cubic example. since curve25519-dalek (v3) uses old rand(v0.7) we need this. should upgrade curve25519-dalek to v4 once it out of pre release * feat: only build bench & profile if std in enabled * feat: remove rand_core as dependency * feat(ci): add job to test wasm build * fix: rollback rand to v7 and update debug test * fix(ci): Cargo.toml patching * feat: make clippy happy * feat: add wasm doc in readme * feat: readme formatting * feat: derive `Default` for `ProofVerifyError`
microsoft · Jan 17, 2023 · 2be72b2 · 2be72b2
1 parent 2dee78c
commit 2be72b2
Show file tree

Hide file tree

Showing 9 changed files with 187 additions and 87 deletions.
diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml
@@ -28,3 +28,39 @@ jobs:
     - name: Check clippy warnings
       run: cargo clippy --all-targets --all-features -- -D warnings
 
+  build_nightly_wasm:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v2
+
+    - name: Install
+      run: rustup default nightly
+
+    - name: Build without std
+      run: cargo build --no-default-features --verbose
+
+    - name: Run tests without std
+      run: cargo test --no-default-features --verbose
+
+    - name: Build examples without std
+      run: cargo build --examples --no-default-features --verbose
+
+    - name: Install wasm32-wasi target
+      run: rustup target add wasm32-wasi
+
+    - name: Install wasm32-unknown-unknown target
+      run: rustup target add wasm32-unknown-unknown
+
+    - name: Build for target wasm-wasi
+      run: RUSTFLAGS="" cargo build --target=wasm32-wasi --no-default-features --verbose
+
+    - name: Patch Cargo.toml for wasm-bindgen
+      run: |
+        echo "[dependencies.getrandom]" >> Cargo.toml
+        echo "version = \"0.1\"" >> Cargo.toml
+        echo "default-features = false" >> Cargo.toml
+        echo "features = [\"wasm-bindgen\"]" >> Cargo.toml
+
+    - name: Build for target wasm32-unknown-unknown
+      run: RUSTFLAGS="" cargo build --target=wasm32-unknown-unknown --no-default-features --verbose
+
diff --git a/Cargo.toml b/Cargo.toml
@@ -11,22 +11,24 @@ license-file = "LICENSE"
 keywords = ["zkSNARKs", "cryptography", "proofs"]
 
 [dependencies]
-curve25519-dalek = {version = "3.2.0", features = ["serde"]}
-merlin = "3.0.0"
-rand = "0.7.3"
-digest = "0.8.1"
-sha3 = "0.8.2"
-byteorder = "1.3.4"
+curve25519-dalek = { version = "3.2.0", features = [
+    "serde",
+    "u64_backend",
+    "alloc",
+], default-features = false }
+merlin = { version = "3.0.0", default-features = false }
+rand = { version = "0.7.3", features = ["getrandom"], default-features = false }
+digest = { version = "0.8.1", default-features = false }
+sha3 = { version = "0.8.2", default-features = false }
+byteorder = { version = "1.3.4", default-features = false }
 rayon = { version = "1.3.0", optional = true }
-serde = { version = "1.0.106", features = ["derive"] }
-bincode = "1.2.1"
-subtle = { version = "2.4", default-features = false }
-rand_core = { version = "0.5", default-features = false }
-zeroize = { version = "1", default-features = false }
-itertools = "0.10.0"
-colored = "2.0.0"
-flate2 = "1.0.14"
-thiserror = "1.0"
+serde = { version = "1.0.106", features = ["derive"], default-features = false }
+bincode = { version = "1.3.3", default-features = false }
+subtle = { version = "2.4", features = ["i128"], default-features = false }
+zeroize = { version = "1.5", default-features = false }
+itertools = { version = "0.10.0", default-features = false }
+colored = { version = "2.0.0", default-features = false, optional = true }
+flate2 = { version = "1.0.14" }
 
 [dev-dependencies]
 criterion = "0.3.1"
@@ -38,20 +40,38 @@ path = "src/lib.rs"
 [[bin]]
 name = "snark"
 path = "profiler/snark.rs"
+required-features = ["std"]
 
 [[bin]]
 name = "nizk"
 path = "profiler/nizk.rs"
+required-features = ["std"]
 
 [[bench]]
 name = "snark"
 harness = false
+required-features = ["std"]
 
 [[bench]]
 name = "nizk"
 harness = false
+required-features = ["std"]
 
 [features]
-default = ["curve25519-dalek/simd_backend"]
+default = ["std", "simd_backend"]
+std = [
+    "curve25519-dalek/std",
+    "digest/std",
+    "merlin/std",
+    "rand/std",
+    "sha3/std",
+    "byteorder/std",
+    "serde/std",
+    "subtle/std",
+    "zeroize/std",
+    "itertools/use_std",
+    "flate2/rust_backend",
+]
+simd_backend = ["curve25519-dalek/simd_backend"]
 multicore = ["rayon"]
-profile = []
+profile = ["colored"]
diff --git a/README.md b/README.md
@@ -1,33 +1,37 @@
 # Spartan: High-speed zkSNARKs without trusted setup
 
 ![Rust](https://github.com/microsoft/Spartan/workflows/Rust/badge.svg)
-[![](https://img.shields.io/crates/v/spartan.svg)]((https://crates.io/crates/spartan))
+[![](https://img.shields.io/crates/v/spartan.svg)](<(https://crates.io/crates/spartan)>)
 
 Spartan is a high-speed zero-knowledge proof system, a cryptographic primitive that enables a prover to prove a mathematical statement to a verifier without revealing anything besides the validity of the statement. This repository provides `libspartan,` a Rust library that implements a zero-knowledge succinct non-interactive argument of knowledge (zkSNARK), which is a type of zero-knowledge proof system with short proofs and fast verification times. The details of the Spartan proof system are described in our [paper](https://eprint.iacr.org/2019/550) published at [CRYPTO 2020](https://crypto.iacr.org/2020/). The security of the Spartan variant implemented in this library is based on the discrete logarithm problem in the random oracle model.
 
 A simple example application is proving the knowledge of a secret s such that H(s) == d for a public d, where H is a cryptographic hash function (e.g., SHA-256, Keccak). A more complex application is a database-backed cloud service that produces proofs of correct state machine transitions for auditability. See this [paper](https://eprint.iacr.org/2020/758.pdf) for an overview and this [paper](https://eprint.iacr.org/2018/907.pdf) for details.
 
-Note that this library has *not* received a security review or audit.
+Note that this library has _not_ received a security review or audit.
 
 ## Highlights
+
 We now highlight Spartan's distinctive features.
 
-* **No "toxic" waste:** Spartan is a *transparent* zkSNARK and does not require a trusted setup. So, it does not involve any trapdoors that must be kept secret or require a multi-party ceremony to produce public parameters.
+- **No "toxic" waste:** Spartan is a _transparent_ zkSNARK and does not require a trusted setup. So, it does not involve any trapdoors that must be kept secret or require a multi-party ceremony to produce public parameters.
 
-* **General-purpose:** Spartan produces proofs for arbitrary NP statements. `libspartan` supports NP statements expressed as rank-1 constraint satisfiability (R1CS) instances, a popular language for which there exists efficient transformations and compiler toolchains from high-level programs of interest.
+- **General-purpose:** Spartan produces proofs for arbitrary NP statements. `libspartan` supports NP statements expressed as rank-1 constraint satisfiability (R1CS) instances, a popular language for which there exists efficient transformations and compiler toolchains from high-level programs of interest.
 
-* **Sub-linear verification costs:** Spartan is the first transparent proof system with sub-linear verification costs for arbitrary NP statements (e.g., R1CS).
+- **Sub-linear verification costs:** Spartan is the first transparent proof system with sub-linear verification costs for arbitrary NP statements (e.g., R1CS).
 
-* **Standardized security:** Spartan's security relies on the hardness of computing discrete logarithms (a standard cryptographic assumption) in the random oracle model. `libspartan` uses `ristretto255`, a prime-order group abstraction atop `curve25519` (a high-speed elliptic curve). We use [`curve25519-dalek`](https://docs.rs/curve25519-dalek) for arithmetic over `ristretto255`. 
+- **Standardized security:** Spartan's security relies on the hardness of computing discrete logarithms (a standard cryptographic assumption) in the random oracle model. `libspartan` uses `ristretto255`, a prime-order group abstraction atop `curve25519` (a high-speed elliptic curve). We use [`curve25519-dalek`](https://docs.rs/curve25519-dalek) for arithmetic over `ristretto255`.
 
-* **State-of-the-art performance:** 
-Among transparent SNARKs, Spartan offers the fastest prover with speedups of 36–152× depending on the baseline, produces proofs that are shorter by 1.2–416×, and incurs the lowest verification times with speedups of 3.6–1326×. The only exception is proof sizes under Bulletproofs, but Bulletproofs incurs slower verification both asymptotically and concretely. When compared to the state-of-the-art zkSNARK with trusted setup, Spartan’s prover is 2× faster for arbitrary R1CS instances and 16× faster for data-parallel workloads.
+- **State-of-the-art performance:**
+  Among transparent SNARKs, Spartan offers the fastest prover with speedups of 36–152× depending on the baseline, produces proofs that are shorter by 1.2–416×, and incurs the lowest verification times with speedups of 3.6–1326×. The only exception is proof sizes under Bulletproofs, but Bulletproofs incurs slower verification both asymptotically and concretely. When compared to the state-of-the-art zkSNARK with trusted setup, Spartan’s prover is 2× faster for arbitrary R1CS instances and 16× faster for data-parallel workloads.
 
 ### Implementation details
-`libspartan` uses [`merlin`](https://docs.rs/merlin/) to automate the Fiat-Shamir transform. We also introduce a new type called `RandomTape` that extends a `Transcript` in `merlin` to allow the prover's internal methods to produce private randomness using its private transcript without having to create `OsRng` objects throughout the code. An object of type `RandomTape` is initialized with a new random seed from `OsRng` for each proof produced by the library. 
+
+`libspartan` uses [`merlin`](https://docs.rs/merlin/) to automate the Fiat-Shamir transform. We also introduce a new type called `RandomTape` that extends a `Transcript` in `merlin` to allow the prover's internal methods to produce private randomness using its private transcript without having to create `OsRng` objects throughout the code. An object of type `RandomTape` is initialized with a new random seed from `OsRng` for each proof produced by the library.
 
 ## Examples
+
 To import `libspartan` into your Rust project, add the following dependency to `Cargo.toml`:
+
 ```text
 spartan = "0.7.1"
 ```
@@ -36,11 +40,11 @@ The following example shows how to use `libspartan` to create and verify a SNARK
 Some of our public APIs' style is inspired by the underlying crates we use.
 
 ```rust
-# extern crate libspartan;
-# extern crate merlin;
-# use libspartan::{Instance, SNARKGens, SNARK};
-# use merlin::Transcript;
-# fn main() {
+extern crate libspartan;
+extern crate merlin;
+use libspartan::{Instance, SNARKGens, SNARK};
+use merlin::Transcript;
+fn main() {
     // specify the size of an R1CS instance
     let num_vars = 1024;
     let num_cons = 1024;
@@ -66,16 +70,17 @@ Some of our public APIs' style is inspired by the underlying crates we use.
       .verify(&comm, &inputs, &mut verifier_transcript, &gens)
       .is_ok());
     println!("proof verification successful!");
-# }
+}
 ```
 
 Here is another example to use the NIZK variant of the Spartan proof system:
+
 ```rust
-# extern crate libspartan;
-# extern crate merlin;
-# use libspartan::{Instance, NIZKGens, NIZK};
-# use merlin::Transcript;
-# fn main() {
+extern crate libspartan;
+extern crate merlin;
+use libspartan::{Instance, NIZKGens, NIZK};
+use merlin::Transcript;
+fn main() {
     // specify the size of an R1CS instance
     let num_vars = 1024;
     let num_cons = 1024;
@@ -97,20 +102,22 @@ Here is another example to use the NIZK variant of the Spartan proof system:
       .verify(&inst, &inputs, &mut verifier_transcript, &gens)
       .is_ok());
     println!("proof verification successful!");
-# }
+}
 ```
 
 Finally, we provide an example that specifies a custom R1CS instance instead of using a synthetic instance
+
 ```rust
 #![allow(non_snake_case)]
-# extern crate curve25519_dalek;
-# extern crate libspartan;
-# extern crate merlin;
-# use curve25519_dalek::scalar::Scalar;
-# use libspartan::{InputsAssignment, Instance, SNARKGens, VarsAssignment, SNARK};
-# use merlin::Transcript;
-# use rand::rngs::OsRng;
-# fn main() {
+extern crate curve25519_dalek;
+extern crate libspartan;
+extern crate merlin;
+use curve25519_dalek::scalar::Scalar;
+use libspartan::{InputsAssignment, Instance, SNARKGens, VarsAssignment, SNARK};
+use merlin::Transcript;
+use rand::rngs::OsRng;
+
+fn main() {
   // produce a tiny instance
   let (
     num_cons,
@@ -146,17 +153,17 @@ Finally, we provide an example that specifies a custom R1CS instance instead of
     .verify(&comm, &assignment_inputs, &mut verifier_transcript, &gens)
     .is_ok());
   println!("proof verification successful!");
-# }
-
-# fn produce_tiny_r1cs() -> (
-#  usize,
-#  usize,
-#  usize,
-#  usize,
-#  Instance,
-#  VarsAssignment,
-#  InputsAssignment,
-# ) {
+}
+
+fn produce_tiny_r1cs() -> (
+ usize,
+ usize,
+ usize,
+ usize,
+ Instance,
+ VarsAssignment,
+ InputsAssignment,
+) {
   // We will use the following example, but one could construct any R1CS instance.
   // Our R1CS instance is three constraints over five variables and two public inputs
   // (Z0 + Z1) * I0 - Z2 = 0
@@ -182,7 +189,7 @@ Finally, we provide an example that specifies a custom R1CS instance instead of
   // a variable that holds a byte representation of 1
   let one = Scalar::one().to_bytes();
 
-  // R1CS is a set of three sparse matrices A B C, where is a row for every 
+  // R1CS is a set of three sparse matrices A B C, where is a row for every
   // constraint and a column for every entry in z = (vars, 1, inputs)
   // An R1CS instance is satisfiable iff:
   // Az \circ Bz = Cz, where z = (vars, 1, inputs)
@@ -233,7 +240,7 @@ Finally, we provide an example that specifies a custom R1CS instance instead of
   inputs[0] = i0.to_bytes();
   inputs[1] = i1.to_bytes();
   let assignment_inputs = InputsAssignment::new(&inputs).unwrap();
-  
+
   // check if the instance we created is satisfiable
   let res = inst.is_sat(&assignment_vars, &assignment_inputs);
   assert_eq!(res.unwrap(), true);
@@ -247,64 +254,92 @@ Finally, we provide an example that specifies a custom R1CS instance instead of
     assignment_vars,
     assignment_inputs,
   )
-# }
+ }
 ```
 
 For more examples, see [`examples/`](examples) directory in this repo.
 
 ## Building `libspartan`
+
 Install [`rustup`](https://rustup.rs/)
 
 Switch to nightly Rust using `rustup`:
+
 ```text
 rustup default nightly
 ```
 
 Clone the repository:
+
 ```text
 git clone https://github.com/Microsoft/Spartan
 cd Spartan
 ```
 
 To build docs for public APIs of `libspartan`:
+
 ```text
 cargo doc
 ```
 
 To run tests:
+
 ```text
 RUSTFLAGS="-C target_cpu=native" cargo test
 ```
 
 To build `libspartan`:
+
 ```text
 RUSTFLAGS="-C target_cpu=native" cargo build --release
 ```
 
 > NOTE: We enable SIMD instructions in `curve25519-dalek` by default, so if it fails to build remove the "simd_backend" feature argument in `Cargo.toml`.
 
 ### Supported features
-* `profile`: enables fine-grained profiling information (see below for its use)
+
+- `std`: enables std features (enabled by default)
+- `simd_backend`: enables `curve25519-dalek`'s simd feature (enabled by default)
+- `profile`: enables fine-grained profiling information (see below for its use)
+
+### WASM Support
+
+`libspartan` depends upon `rand::OsRng` (internally uses `getrandom` crate), it has out of box support for `wasm32-wasi`.
+
+For the target `wasm32-unknown-unknown` disable default features for spartan
+and add direct dependency on `getrandom` with `wasm-bindgen` feature enabled.
+
+```toml
+[dependencies]
+spartan = { version = "0.7", default-features = false }
+# since spartan uses getrandom(rand's OsRng), we need to enable 'wasm-bindgen'
+# feature to make it feed rand seed from js/nodejs env
+# https://docs.rs/getrandom/0.1.16/getrandom/index.html#support-for-webassembly-and-asmjs
+getrandom = { version = "0.1", features = ["wasm-bindgen"] }
+```
 
 ## Performance
 
 ### End-to-end benchmarks
+
 `libspartan` includes two benches: `benches/nizk.rs` and `benches/snark.rs`. If you report the performance of Spartan in a research paper, we recommend using these benches for higher accuracy instead of fine-grained profiling (listed below).
 
 To run end-to-end benchmarks:
-```text 
+
+```text
 RUSTFLAGS="-C target_cpu=native" cargo bench
 ```
 
 ### Fine-grained profiling
+
 Build `libspartan` with `profile` feature enabled. It creates two profilers: `./target/release/snark` and `./target/release/nizk`.
 
 These profilers report performance as depicted below (for varying R1CS instance sizes). The reported
 performance is from running the profilers on a Microsoft Surface Laptop 3 on a single CPU core of Intel Core i7-1065G7 running Ubuntu 20.04 (atop WSL2 on Windows 10).
 See Section 9 in our [paper](https://eprint.iacr.org/2019/550) to see how this compares with other zkSNARKs in the literature.
 
 ```text
-$ ./target/release/snark 
+$ ./target/release/snark
 Profiler:: SNARK
   * number_of_constraints 1048576
   * number_of_variables 1048576
@@ -355,7 +390,7 @@ Profiler:: SNARK
 ```
 
 ```text
-$ ./target/release/nizk 
+$ ./target/release/nizk
 Profiler:: NIZK
   * number_of_constraints 1048576
   * number_of_variables 1048576