Auto merge of #105058 - Nilstrieb:no-merge-commits-for-you-only-bors-is-allowed-to-do-that, r=jyn514

Add tidy check to deny merge commits

This will prevent users with the pre-push hook from pushing a merge commit.

Exceptions are added for subtree updates. These exceptions are a little hacky and may be non-exhaustive but can be extended in the future.

I added a link to `@jyn514's` blog post for the error case because that's the best resource to solve merge commits. But it would probably be better if it was integrated into https://rustc-dev-guide.rust-lang.org/git.html#no-merge-policy, then we could link that instead.

r? `@jyn514`

.github/workflows/ci.yml

@@ -71,6 +71,11 @@ jobs:
         uses: actions/checkout@v3
         with:
           fetch-depth: 2
+      - name: "checkout the `master` branch for tidy"
+        uses: actions/checkout@v3
+        with:
+          ref: master
+          fetch-depth: 1
       - name: configure the PR in which the error message will be posted
         run: "echo \"[CI_PR_NUMBER=$num]\""
         env:
@@ -485,6 +490,11 @@ jobs:
         uses: actions/checkout@v3
         with:
           fetch-depth: 2
+      - name: "checkout the `master` branch for tidy"
+        uses: actions/checkout@v3
+        with:
+          ref: master
+          fetch-depth: 1
       - name: configure the PR in which the error message will be posted
         run: "echo \"[CI_PR_NUMBER=$num]\""
         env:
@@ -600,6 +610,11 @@ jobs:
         uses: actions/checkout@v3
         with:
           fetch-depth: 2
+      - name: "checkout the `master` branch for tidy"
+        uses: actions/checkout@v3
+        with:
+          ref: master
+          fetch-depth: 1
       - name: configure the PR in which the error message will be posted
         run: "echo \"[CI_PR_NUMBER=$num]\""
         env:

Cargo.lock

@@ -259,6 +259,10 @@ dependencies = [
  "toml",
 ]
 
+[[package]]
+name = "build_helper"
+version = "0.1.0"
+
 [[package]]
 name = "bump-stage0"
 version = "0.1.0"
@@ -5304,6 +5308,7 @@ dependencies = [
 name = "tidy"
 version = "0.1.0"
 dependencies = [
+ "build_helper",
  "cargo_metadata 0.14.0",
  "ignore",
  "lazy_static",

Cargo.toml

@@ -4,6 +4,7 @@ members = [
   "library/std",
   "library/test",
   "src/rustdoc-json-types",
+  "src/tools/build_helper",
   "src/tools/cargotest",
   "src/tools/clippy",
   "src/tools/clippy/clippy_dev",

src/bootstrap/Cargo.lock

@@ -36,6 +36,7 @@ dependencies = [
 name = "bootstrap"
 version = "0.0.0"
 dependencies = [
+ "build_helper",
  "cc",
  "cmake",
  "fd-lock",
@@ -70,6 +71,10 @@ dependencies = [
  "regex-automata",
 ]
 
+[[package]]
+name = "build_helper"
+version = "0.1.0"
+
 [[package]]
 name = "cc"
 version = "1.0.73"

src/bootstrap/Cargo.toml

@@ -30,6 +30,7 @@ path = "bin/sccache-plus-cl.rs"
 test = false
 
 [dependencies]
+build_helper = { path = "../tools/build_helper" }
 cmake = "0.1.38"
 fd-lock = "3.0.8"
 filetime = "0.2"

src/bootstrap/format.rs

@@ -2,6 +2,7 @@
 
 use crate::builder::Builder;
 use crate::util::{output, program_out_of_date, t};
+use build_helper::git::get_rust_lang_rust_remote;
 use ignore::WalkBuilder;
 use std::collections::VecDeque;
 use std::path::{Path, PathBuf};
@@ -100,35 +101,6 @@ fn get_modified_rs_files(build: &Builder<'_>) -> Option<Vec<String>> {
     )
 }
 
-/// Finds the remote for rust-lang/rust.
-/// For example for these remotes it will return `upstream`.
-/// ```text
-/// origin  https://github.com/Nilstrieb/rust.git (fetch)
-/// origin  https://github.com/Nilstrieb/rust.git (push)
-/// upstream        https://github.com/rust-lang/rust (fetch)
-/// upstream        https://github.com/rust-lang/rust (push)
-/// ```
-fn get_rust_lang_rust_remote() -> Result<String, String> {
-    let mut git = Command::new("git");
-    git.args(["config", "--local", "--get-regex", "remote\\..*\\.url"]);
-
-    let output = git.output().map_err(|err| format!("{err:?}"))?;
-    if !output.status.success() {
-        return Err("failed to execute git config command".to_owned());
-    }
-
-    let stdout = String::from_utf8(output.stdout).map_err(|err| format!("{err:?}"))?;
-
-    let rust_lang_remote = stdout
-        .lines()
-        .find(|remote| remote.contains("rust-lang"))
-        .ok_or_else(|| "rust-lang/rust remote not found".to_owned())?;
-
-    let remote_name =
-        rust_lang_remote.split('.').nth(1).ok_or_else(|| "remote name not found".to_owned())?;
-    Ok(remote_name.into())
-}
-
 #[derive(serde::Deserialize)]
 struct RustfmtConfig {
     ignore: Vec<String>,

src/bootstrap/lib.rs

@@ -113,6 +113,7 @@
 use std::process::Command;
 use std::str;
 
+use build_helper::ci::CiEnv;
 use channel::GitInfo;
 use config::{DryRun, Target};
 use filetime::FileTime;
@@ -121,7 +122,7 @@
 use crate::builder::Kind;
 use crate::config::{LlvmLibunwind, TargetSelection};
 use crate::util::{
-    exe, libdir, mtime, output, run, run_suppressed, symlink_dir, try_run_suppressed, CiEnv,
+    exe, libdir, mtime, output, run, run_suppressed, symlink_dir, try_run_suppressed,
 };
 
 mod bolt;

src/bootstrap/native.rs

@@ -24,6 +24,8 @@
 use crate::util::{self, exe, output, t, up_to_date};
 use crate::{CLang, GitRepo};
 
+use build_helper::ci::CiEnv;
+
 #[derive(Clone)]
 pub struct LlvmResult {
     /// Path to llvm-config binary.
@@ -217,7 +219,7 @@ pub(crate) fn is_ci_llvm_available(config: &Config, asserts: bool) -> bool {
         return false;
     }
 
-    if crate::util::CiEnv::is_ci() {
+    if CiEnv::is_ci() {
         // We assume we have access to git, so it's okay to unconditionally pass
         // `true` here.
         let llvm_sha = detect_llvm_sha(config, true);

src/bootstrap/util.rs

@@ -255,35 +255,6 @@ pub enum CiEnv {
     GitHubActions,
 }
 
-impl CiEnv {
-    /// Obtains the current CI environment.
-    pub fn current() -> CiEnv {
-        if env::var("TF_BUILD").map_or(false, |e| e == "True") {
-            CiEnv::AzurePipelines
-        } else if env::var("GITHUB_ACTIONS").map_or(false, |e| e == "true") {
-            CiEnv::GitHubActions
-        } else {
-            CiEnv::None
-        }
-    }
-
-    pub fn is_ci() -> bool {
-        Self::current() != CiEnv::None
-    }
-
-    /// If in a CI environment, forces the command to run with colors.
-    pub fn force_coloring_in_ci(self, cmd: &mut Command) {
-        if self != CiEnv::None {
-            // Due to use of stamp/docker, the output stream of rustbuild is not
-            // a TTY in CI, so coloring is by-default turned off.
-            // The explicit `TERM=xterm` environment is needed for
-            // `--color always` to actually work. This env var was lost when
-            // compiling through the Makefile. Very strange.
-            cmd.env("TERM", "xterm").args(&["--color", "always"]);
-        }
-    }
-}
-
 pub fn forcing_clang_based_tests() -> bool {
     if let Some(var) = env::var_os("RUSTBUILD_FORCE_CLANG_BASED_TESTS") {
         match &var.to_string_lossy().to_lowercase()[..] {

src/ci/github-actions/ci.yml

@@ -103,6 +103,12 @@ x--expand-yaml-anchors--remove:
         with:
           fetch-depth: 2
 
+      - name: checkout the `master` branch for tidy
+        uses: actions/checkout@v3
+        with:
+          ref: master
+          fetch-depth: 1
+
       # Rust Log Analyzer can't currently detect the PR number of a GitHub
       # Actions build on its own, so a hint in the log message is needed to
       # point it in the right direction.

src/tools/build_helper/Cargo.toml

@@ -0,0 +1,8 @@
+[package]
+name = "build_helper"
+version = "0.1.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]

src/tools/build_helper/src/ci.rs

@@ -0,0 +1,40 @@
+use std::process::Command;
+
+#[derive(Copy, Clone, PartialEq, Eq, Debug)]
+pub enum CiEnv {
+    /// Not a CI environment.
+    None,
+    /// The Azure Pipelines environment, for Linux (including Docker), Windows, and macOS builds.
+    AzurePipelines,
+    /// The GitHub Actions environment, for Linux (including Docker), Windows and macOS builds.
+    GitHubActions,
+}
+
+impl CiEnv {
+    /// Obtains the current CI environment.
+    pub fn current() -> CiEnv {
+        if std::env::var("TF_BUILD").map_or(false, |e| e == "True") {
+            CiEnv::AzurePipelines
+        } else if std::env::var("GITHUB_ACTIONS").map_or(false, |e| e == "true") {
+            CiEnv::GitHubActions
+        } else {
+            CiEnv::None
+        }
+    }
+
+    pub fn is_ci() -> bool {
+        Self::current() != CiEnv::None
+    }
+
+    /// If in a CI environment, forces the command to run with colors.
+    pub fn force_coloring_in_ci(self, cmd: &mut Command) {
+        if self != CiEnv::None {
+            // Due to use of stamp/docker, the output stream of rustbuild is not
+            // a TTY in CI, so coloring is by-default turned off.
+            // The explicit `TERM=xterm` environment is needed for
+            // `--color always` to actually work. This env var was lost when
+            // compiling through the Makefile. Very strange.
+            cmd.env("TERM", "xterm").args(&["--color", "always"]);
+        }
+    }
+}

src/tools/build_helper/src/git.rs

@@ -0,0 +1,30 @@
+use std::process::Command;
+
+/// Finds the remote for rust-lang/rust.
+/// For example for these remotes it will return `upstream`.
+/// ```text
+/// origin  https://github.com/Nilstrieb/rust.git (fetch)
+/// origin  https://github.com/Nilstrieb/rust.git (push)
+/// upstream        https://github.com/rust-lang/rust (fetch)
+/// upstream        https://github.com/rust-lang/rust (push)
+/// ```
+pub fn get_rust_lang_rust_remote() -> Result<String, String> {
+    let mut git = Command::new("git");
+    git.args(["config", "--local", "--get-regex", "remote\\..*\\.url"]);
+
+    let output = git.output().map_err(|err| format!("{err:?}"))?;
+    if !output.status.success() {
+        return Err("failed to execute git config command".to_owned());
+    }
+
+    let stdout = String::from_utf8(output.stdout).map_err(|err| format!("{err:?}"))?;
+
+    let rust_lang_remote = stdout
+        .lines()
+        .find(|remote| remote.contains("rust-lang"))
+        .ok_or_else(|| "rust-lang/rust remote not found".to_owned())?;
+
+    let remote_name =
+        rust_lang_remote.split('.').nth(1).ok_or_else(|| "remote name not found".to_owned())?;
+    Ok(remote_name.into())
+}

src/tools/build_helper/src/lib.rs

@@ -0,0 +1,2 @@
+pub mod ci;
+pub mod git;

src/tools/tidy/Cargo.toml

@@ -5,6 +5,7 @@ edition = "2021"
 autobins = false
 
 [dependencies]
+build_helper = { path = "../build_helper" }
 cargo_metadata = "0.14"
 regex = "1"
 miropt-test-tools = { path = "../miropt-test-tools" }

src/tools/tidy/src/lib.rs

@@ -48,6 +48,7 @@ macro_rules! tidy_error {
 pub mod extdeps;
 pub mod features;
 pub mod mir_opt_tests;
+pub mod no_merge;
 pub mod pal;
 pub mod primitive_docs;
 pub mod style;

src/tools/tidy/src/main.rs

@@ -107,6 +107,8 @@ macro_rules! check {
         check!(alphabetical, &compiler_path);
         check!(alphabetical, &library_path);
 
+        check!(no_merge, ());
+
         let collected = {
             drain_handles(&mut handles);
 

src/tools/tidy/src/no_merge.rs

@@ -0,0 +1,72 @@
+//! This check makes sure that no accidental merge commits are introduced to the repository.
+//! It forbids all merge commits that are not caused by rollups/bors or subtree syncs.
+
+use std::process::Command;
+
+use build_helper::ci::CiEnv;
+use build_helper::git::get_rust_lang_rust_remote;
+
+macro_rules! try_unwrap_in_ci {
+    ($expr:expr) => {
+        match $expr {
+            Ok(value) => value,
+            Err(err) if CiEnv::is_ci() => {
+                panic!("Encountered error while testing Git status: {:?}", err)
+            }
+            Err(_) => return,
+        }
+    };
+}
+
+pub fn check(_: (), bad: &mut bool) {
+    let remote = try_unwrap_in_ci!(get_rust_lang_rust_remote());
+    let merge_commits = try_unwrap_in_ci!(find_merge_commits(&remote));
+
+    let mut bad_merge_commits = merge_commits.lines().filter(|commit| {
+        !(
+            // Bors is the ruler of merge commits.
+            commit.starts_with("Auto merge of") || commit.starts_with("Rollup merge of")
+        )
+    });
+
+    if let Some(merge) = bad_merge_commits.next() {
+        tidy_error!(
+            bad,
+            "found a merge commit in the history: `{merge}`.
+To resolve the issue, see this: https://rustc-dev-guide.rust-lang.org/git.html#i-made-a-merge-commit-by-accident.
+If you're doing a subtree sync, add your tool to the list in the code that emitted this error."
+        );
+    }
+}
+
+/// Runs `git log --merges --format=%s $REMOTE/master..HEAD` and returns all commits
+fn find_merge_commits(remote: &str) -> Result<String, String> {
+    let mut git = Command::new("git");
+    git.args([
+        "log",
+        "--merges",
+        "--format=%s",
+        &format!("{remote}/master..HEAD"),
+        // Ignore subtree syncs. Add your new subtrees here.
+        ":!src/tools/miri",
+        ":!src/tools/rust-analyzer",
+        ":!compiler/rustc_smir",
+        ":!library/portable-simd",
+        ":!compiler/rustc_codegen_gcc",
+        ":!src/tools/rustfmt",
+        ":!compiler/rustc_codegen_cranelift",
+        ":!src/tools/clippy",
+    ]);
+
+    let output = git.output().map_err(|err| format!("{err:?}"))?;
+    if !output.status.success() {
+        return Err(format!(
+            "failed to execute git log command: {}",
+            String::from_utf8_lossy(&output.stderr)
+        ));
+    }
+
+    let stdout = String::from_utf8(output.stdout).map_err(|err| format!("{err:?}"))?;
+
+    Ok(stdout)
+}