fix(test): bind scanner evidence to execution and build identity

Co-Authored-By: heihutu <heihutu@gmail.com>
Co-Authored-By: zhi22915 <qiuzgang@gmail.com>

crates/e2e_test/build.rs

@@ -0,0 +1,74 @@
+// Copyright 2024 RustFS Team
+// Licensed under the Apache License, Version 2.0.
+
+use std::path::Path;
+use std::process::Command;
+
+fn git(root: &Path, args: &[&str]) -> Option<String> {
+    let output = Command::new("git").args(args).current_dir(root).output().ok()?;
+    output
+        .status
+        .success()
+        .then(|| String::from_utf8_lossy(&output.stdout).trim().to_owned())
+}
+
+fn emit(name: &str, value: &str) {
+    let value = if value.contains(['\n', '\r']) { "unknown" } else { value };
+    println!("cargo:rustc-env=RUSTFS_E2E_BUILD_{name}={value}");
+}
+
+fn main() {
+    let manifest = std::env::var_os("CARGO_MANIFEST_DIR").unwrap_or_default();
+    let root = Path::new(&manifest).join("../..");
+    // Cover dependency/common sources as well as this crate. HEAD/ref/index
+    // changes must refresh identity even when no Rust source mtime changes.
+    for path in [
+        "crates",
+        "rustfs",
+        "Cargo.toml",
+        "Cargo.lock",
+        "rust-toolchain.toml",
+        ".cargo",
+        ".config",
+    ] {
+        println!("cargo:rerun-if-changed={}", root.join(path).display());
+    }
+    let mut git_paths = vec!["HEAD".to_owned(), "index".to_owned(), "packed-refs".to_owned()];
+    if let Some(reference) = git(&root, &["symbolic-ref", "-q", "HEAD"]) {
+        git_paths.push(reference);
+    }
+    for path in git_paths {
+        if let Some(path) = git(&root, &["rev-parse", "--git-path", &path]) {
+            let path = Path::new(&path);
+            let path = if path.is_absolute() {
+                path.to_owned()
+            } else {
+                root.join(path)
+            };
+            if path.exists() {
+                println!("cargo:rerun-if-changed={}", path.display());
+            }
+        }
+    }
+    let revision = git(&root, &["rev-parse", "HEAD"]).unwrap_or_else(|| "unknown".to_owned());
+    let dirty = git(&root, &["status", "--porcelain", "--untracked-files=normal"]).is_none_or(|status| !status.is_empty());
+    let lock = git(&root, &["hash-object", "Cargo.lock"]).unwrap_or_else(|| "unknown".to_owned());
+    let mut features = std::env::vars()
+        .filter_map(|(key, _)| {
+            key.strip_prefix("CARGO_FEATURE_")
+                .map(|name| name.to_ascii_lowercase().replace('_', "-"))
+        })
+        .collect::<Vec<_>>();
+    features.sort();
+    emit("COMMIT", &revision);
+    emit("DIRTY", if dirty { "true" } else { "false" });
+    emit("LOCK", &lock);
+    emit("FEATURES", &features.join(","));
+    for name in ["TARGET", "PROFILE"] {
+        emit(name, &std::env::var(name).unwrap_or_else(|_| "unknown".to_owned()));
+    }
+    println!("cargo:rerun-if-env-changed=CARGO_ENCODED_RUSTFLAGS");
+    let flags = std::env::var("CARGO_ENCODED_RUSTFLAGS").unwrap_or_default();
+    let flags: String = flags.as_bytes().iter().map(|byte| format!("{byte:02x}")).collect();
+    emit("RUSTFLAGS_HEX", &flags);
+}

crates/e2e_test/src/chaos.rs

@@ -61,6 +61,8 @@ pub(crate) struct VersionShardCensus {
     pub has_xl_meta: bool,
     pub data_dir: Option<String>,
     pub erasure_index: Option<usize>,
+    pub data_blocks: Option<usize>,
+    pub parity_blocks: Option<usize>,
     pub expected_part_numbers: BTreeSet<usize>,
     pub present_part_fingerprints: BTreeMap<usize, PartShardFingerprint>,
     pub inline_data_fingerprint: Option<PartShardFingerprint>,
@@ -88,6 +90,8 @@ impl VersionShardCensus {
             && manifest.is_complete()
             && self.data_dir == manifest.data_dir
             && self.erasure_index == manifest.erasure_index
+            && self.data_blocks == manifest.data_blocks
+            && self.parity_blocks == manifest.parity_blocks
             && self.expected_part_numbers == manifest.expected_part_numbers
             && self.present_part_fingerprints == manifest.present_part_fingerprints
             && self.inline_data_fingerprint == manifest.inline_data_fingerprint
@@ -313,6 +317,8 @@ pub(crate) fn census_object_version_on_disk(
             has_xl_meta: false,
             data_dir: None,
             erasure_index: None,
+            data_blocks: None,
+            parity_blocks: None,
             expected_part_numbers: BTreeSet::new(),
             present_part_fingerprints: BTreeMap::new(),
             inline_data_fingerprint: None,
@@ -360,6 +366,8 @@ pub(crate) fn census_object_version_on_disk(
         has_xl_meta: true,
         data_dir,
         erasure_index,
+        data_blocks: Some(file_info.erasure.data_blocks),
+        parity_blocks: Some(file_info.erasure.parity_blocks),
         expected_part_numbers,
         present_part_fingerprints,
         inline_data_fingerprint,
@@ -413,6 +421,8 @@ mod tests {
             has_xl_meta: true,
             data_dir: Some("data-dir".to_string()),
             erasure_index: Some(3),
+            data_blocks: Some(2),
+            parity_blocks: Some(2),
             expected_part_numbers: BTreeSet::from([1]),
             present_part_fingerprints: BTreeMap::from([(1, shard_fingerprint(b"part").unwrap())]),
             inline_data_fingerprint: None,

crates/e2e_test/src/heal_erasure_disk_rebuild_test.rs

@@ -44,9 +44,10 @@ mod tests {
 
     #[derive(serde::Deserialize)]
     struct RestartEvidenceRun {
+        schema: u32,
         run_id: String,
         source_revision: String,
-        test_sources_sha256: String,
+        test_build: serde_json::Value,
         binary: EvidenceBuild,
         test_binary: EvidenceBuild,
     }
@@ -75,14 +76,13 @@ mod tests {
             return Err("oversized scanner/heal execution receipt".into());
         }
         let run: RestartEvidenceRun = serde_json::from_slice(&std::fs::read(receipt)?)?;
-        if run.run_id.len() != 32 || run.source_revision.len() != 40 {
+        if run.schema != 1 || run.run_id.len() != 32 || run.source_revision.len() != 40 {
             return Err("invalid scanner/heal execution identity".into());
         }
-        let mut sources = Sha256::new();
-        sources.update(include_bytes!("heal_erasure_disk_rebuild_test.rs"));
-        sources.update(include_bytes!("chaos.rs"));
-        let source_digest: String = sources.finalize().iter().map(|byte| format!("{byte:02x}")).collect();
-        assert_eq!(source_digest, run.test_sources_sha256, "compiled test sources must match the receipt");
+        let built = compiled_test_identity();
+        for key in ["source_revision", "dirty", "lock_blob", "features"] {
+            assert_eq!(built[key], run.test_build[key], "compiled test identity differs for {key}");
+        }
         assert_eq!(file_sha256(binary)?, run.binary.sha256, "server binary must match the run receipt");
         assert_eq!(
             file_sha256(&std::env::current_exe()?)?,
@@ -95,6 +95,18 @@ mod tests {
         Ok(Some((directory, run)))
     }
 
+    fn compiled_test_identity() -> serde_json::Value {
+        serde_json::json!({
+            "source_revision": env!("RUSTFS_E2E_BUILD_COMMIT"),
+            "dirty": env!("RUSTFS_E2E_BUILD_DIRTY") != "false",
+            "lock_blob": env!("RUSTFS_E2E_BUILD_LOCK"),
+            "features": env!("RUSTFS_E2E_BUILD_FEATURES"),
+            "target": env!("RUSTFS_E2E_BUILD_TARGET"),
+            "profile": env!("RUSTFS_E2E_BUILD_PROFILE"),
+            "rustflags_hex": env!("RUSTFS_E2E_BUILD_RUSTFLAGS_HEX"),
+        })
+    }
+
     struct TcpPortBlackhole {
         port: u16,
         comment: String,
@@ -1530,7 +1542,7 @@ mod tests {
             let evidence = serde_json::json!({
                 "schema": 1, "case": "background-target-restart", "evidence": "process-restart",
                 "run_id": run.run_id, "source_revision": run.source_revision,
-                "test_sources_sha256": run.test_sources_sha256,
+                "test_build": compiled_test_identity(),
                 "binary_sha256": run.binary.sha256, "test_binary_sha256": run.test_binary.sha256,
                 "topology": {"nodes": cluster.nodes.len(), "drives_per_node": cluster.nodes[0].data_dirs.len()},
                 "pid_before": target_pid, "pid_after": restarted_pid,