Fuzz Bucket CRDT (#1442)

Reviewed-on: https://git.deuxfleurs.fr/Deuxfleurs/garage/pulls/1442
Reviewed-by: Alex <lx@deuxfleurs.fr>

Cargo.lock

@@ -1551,6 +1551,7 @@ dependencies = [
  "garage_db",
  "garage_model",
  "garage_table",
+ "garage_util",
  "libfuzzer-sys",
 ]
 

fuzz/Cargo.toml

@@ -13,6 +13,7 @@ libfuzzer-sys = { workspace = true }
 
 garage_db.workspace = true
 garage_table.workspace = true
+garage_util.workspace = true
 garage_model = { workspace = true, default-features = false, features = ["arbitrary"] }
 
 [[bin]]
@@ -28,3 +29,10 @@ path = "fuzz_targets/mpu_crdt.rs"
 test = false
 doc = false
 bench = false
+
+[[bin]]
+name = "bucket_crdt"
+path = "fuzz_targets/bucket_crdt.rs"
+test = false
+doc = false
+bench = false

fuzz/fuzz_targets/bucket_crdt.rs

@@ -0,0 +1,22 @@
+#![no_main]
+
+use garage_fuzz::check_crdt_laws;
+use garage_model::bucket_table::{Bucket, BucketParams};
+use garage_util::crdt::{self, Deletable};
+use libfuzzer_sys::fuzz_target;
+
+fn make(state: Deletable<BucketParams>) -> Bucket {
+	Bucket {
+		id: [0u8; 32].into(),
+		state,
+	}
+}
+
+fuzz_target!(|inputs: (
+	crdt::Deletable<BucketParams>,
+	crdt::Deletable<BucketParams>,
+	crdt::Deletable<BucketParams>
+)| {
+	let (a, b, c) = inputs;
+	check_crdt_laws(make(a), make(b), make(c));
+});

src/model/bucket_table.rs

@@ -63,6 +63,7 @@ mod v08 {
 	}
 
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct CorsRule {
 		pub id: Option<String>,
 		pub max_age_seconds: Option<u64>,
@@ -74,6 +75,7 @@ mod v08 {
 
 	/// Lifecycle configuration rule
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct LifecycleRule {
 		/// The ID of the rule
 		pub id: Option<String>,
@@ -91,6 +93,7 @@ mod v08 {
 	/// For each condition, if it is None, it is not verified (always true),
 	/// and if it is Some(x), then it is verified for value x
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize, Default)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct LifecycleFilter {
 		/// If Some(x), object key has to start with prefix x
 		pub prefix: Option<String>,
@@ -101,6 +104,7 @@ mod v08 {
 	}
 
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub enum LifecycleExpiration {
 		/// Objects expire x days after they were created
 		AfterDays(usize),
@@ -109,6 +113,7 @@ mod v08 {
 	}
 
 	#[derive(Default, PartialEq, Eq, PartialOrd, Ord, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct BucketQuotas {
 		/// Maximum size in bytes (bucket size = sum of sizes of objects in the bucket)
 		pub max_size: Option<u64>,
@@ -139,6 +144,7 @@ mod v2 {
 
 	/// Configuration for a bucket
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct BucketParams {
 		/// Bucket's creation date
 		pub creation_date: u64,
@@ -168,6 +174,7 @@ mod v2 {
 	}
 
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct WebsiteConfig {
 		pub index_document: String,
 		pub error_document: Option<String>,
@@ -178,24 +185,28 @@ mod v2 {
 	}
 
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct RedirectAll {
 		pub hostname: String,
 		pub protocol: String,
 	}
 
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct RoutingRule {
 		pub condition: Option<RedirectCondition>,
 		pub redirect: Redirect,
 	}
 
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct RedirectCondition {
 		pub http_error_code: Option<u16>,
 		pub prefix: Option<String>,
 	}
 
 	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 	pub struct Redirect {
 		pub hostname: Option<String>,
 		pub http_redirect_code: u16,

src/model/permission.rs

@@ -6,6 +6,7 @@ use garage_util::crdt::*;
 
 /// Permission given to a key in a bucket
 #[derive(PartialOrd, Ord, PartialEq, Eq, Clone, Copy, Debug, Serialize, Deserialize)]
+#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 pub struct BucketKeyPerm {
 	/// Timestamp at which the permission was given
 	pub timestamp: u64,

src/util/crdt/deletable.rs

@@ -4,6 +4,7 @@ use crate::crdt::crdt::*;
 
 /// Deletable object (once deleted, cannot go back)
 #[derive(Clone, Copy, Debug, Serialize, Deserialize, PartialEq, Eq)]
+#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 pub enum Deletable<T> {
 	Present(T),
 	Deleted,

src/util/crdt/lww.rs

@@ -38,6 +38,7 @@ use crate::crdt::crdt::*;
 /// This scheme is used by AWS S3 or Soundcloud and often without knowing
 /// in enterprise when reconciliating databases with ad-hoc scripts.
 #[derive(Clone, Debug, Serialize, Deserialize, PartialEq, Eq)]
+#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 pub struct Lww<T> {
 	ts: u64,
 	v: T,

src/util/crdt/lww_map.rs

@@ -200,3 +200,19 @@ where
 		Self::new()
 	}
 }
+
+#[cfg(feature = "arbitrary")]
+impl<'a, K, V> arbitrary::Arbitrary<'a> for LwwMap<K, V>
+where
+	K: arbitrary::Arbitrary<'a> + Clone + Ord,
+	V: arbitrary::Arbitrary<'a> + Clone + Crdt,
+{
+	fn arbitrary(u: &mut arbitrary::Unstructured<'a>) -> arbitrary::Result<Self> {
+		let map: std::collections::BTreeMap<K, (u64, V)> = arbitrary::Arbitrary::arbitrary(u)?;
+		let mut result = LwwMap::new();
+		for (k, (ts, v)) in map {
+			result.merge_raw(&k, ts, &v);
+		}
+		Ok(result)
+	}
+}

src/util/crdt/map.rs

@@ -123,3 +123,15 @@ where
 		Self { vals }
 	}
 }
+
+#[cfg(feature = "arbitrary")]
+impl<'a, K, V> arbitrary::Arbitrary<'a> for Map<K, V>
+where
+	K: arbitrary::Arbitrary<'a> + Clone + Ord,
+	V: arbitrary::Arbitrary<'a> + Clone + Crdt,
+{
+	fn arbitrary(u: &mut arbitrary::Unstructured<'a>) -> arbitrary::Result<Self> {
+		let map: std::collections::BTreeMap<K, V> = arbitrary::Arbitrary::arbitrary(u)?;
+		Ok(map.into_iter().collect())
+	}
+}