Auto merge of #114560 - RalfJung:miri, r=RalfJung

update Miri

.github/workflows/ci.yml

@@ -130,7 +130,7 @@ jobs:
       - name: clippy
         run: ./miri clippy -- -D warnings
       - name: rustdoc
-        run: RUSTDOCFLAGS="-Dwarnings" cargo doc --document-private-items
+        run: RUSTDOCFLAGS="-Dwarnings" ./miri cargo doc --document-private-items
 
   # These jobs doesn't actually test anything, but they're only used to tell
   # bors the build completed, as there is no practical way to detect when a

CONTRIBUTING.md

@@ -107,7 +107,7 @@ evaluation error was originally raised.
 ### UI testing
 
 We use ui-testing in Miri, meaning we generate `.stderr` and `.stdout` files for the output
-produced by Miri. You can use `./miri bless` to automatically (re)generate these files when
+produced by Miri. You can use `./miri test --bless` to automatically (re)generate these files when
 you add new tests or change how Miri presents certain output.
 
 Note that when you also use `MIRIFLAGS` to change optimizations and similar, the ui output

cargo-miri/src/main.rs

@@ -56,7 +56,7 @@ fn main() {
         return;
     }
 
-    // The way rustdoc invokes rustc is indistuingishable from the way cargo invokes rustdoc by the
+    // The way rustdoc invokes rustc is indistinguishable from the way cargo invokes rustdoc by the
     // arguments alone. `phase_cargo_rustdoc` sets this environment variable to let us disambiguate.
     if env::var_os("MIRI_CALLED_FROM_RUSTDOC").is_some() {
         // ...however, we then also see this variable when rustdoc invokes us as the testrunner!

cargo-miri/src/phases.rs

@@ -94,7 +94,7 @@ pub fn phase_cargo_miri(mut args: impl Iterator<Item = String>) {
     let target = target.as_ref().unwrap_or(host);
 
     // We always setup.
-    setup(&subcommand, target, &rustc_version, verbose);
+    let miri_sysroot = setup(&subcommand, target, &rustc_version, verbose);
 
     // Invoke actual cargo for the job, but with different flags.
     // We re-use `cargo test` and `cargo run`, which makes target and binary handling very easy but
@@ -159,6 +159,8 @@ pub fn phase_cargo_miri(mut args: impl Iterator<Item = String>) {
     // Forward all further arguments (not consumed by `ArgSplitFlagValue`) to cargo.
     cmd.args(args);
 
+    // Let it know where the Miri sysroot lives.
+    cmd.env("MIRI_SYSROOT", miri_sysroot);
     // Set `RUSTC_WRAPPER` to ourselves.  Cargo will prepend that binary to its usual invocation,
     // i.e., the first argument is `rustc` -- which is what we use in `main` to distinguish
     // the two codepaths. (That extra argument is why we prefer this over setting `RUSTC`.)
@@ -519,7 +521,7 @@ pub fn phase_runner(mut binary_args: impl Iterator<Item = String>, phase: Runner
     // `.rmeta`.
     // We also need to remove `--error-format` as cargo specifies that to be JSON,
     // but when we run here, cargo does not interpret the JSON any more. `--json`
-    // then also nees to be dropped.
+    // then also needs to be dropped.
     let mut args = info.args.into_iter();
     let error_format_flag = "--error-format";
     let json_flag = "--json";
@@ -538,8 +540,7 @@ pub fn phase_runner(mut binary_args: impl Iterator<Item = String>, phase: Runner
     }
     // Respect `MIRIFLAGS`.
     if let Ok(a) = env::var("MIRIFLAGS") {
-        // This code is taken from `RUSTFLAGS` handling in cargo.
-        let args = a.split(' ').map(str::trim).filter(|s| !s.is_empty()).map(str::to_string);
+        let args = flagsplit(&a);
         cmd.args(args);
     }
 

cargo-miri/src/setup.rs

@@ -13,13 +13,20 @@ use crate::util::*;
 /// Performs the setup required to make `cargo miri` work: Getting a custom-built libstd. Then sets
 /// `MIRI_SYSROOT`. Skipped if `MIRI_SYSROOT` is already set, in which case we expect the user has
 /// done all this already.
-pub fn setup(subcommand: &MiriCommand, target: &str, rustc_version: &VersionMeta, verbose: usize) {
+pub fn setup(
+    subcommand: &MiriCommand,
+    target: &str,
+    rustc_version: &VersionMeta,
+    verbose: usize,
+) -> PathBuf {
     let only_setup = matches!(subcommand, MiriCommand::Setup);
     let ask_user = !only_setup;
     let print_sysroot = only_setup && has_arg_flag("--print-sysroot"); // whether we just print the sysroot path
-    if !only_setup && std::env::var_os("MIRI_SYSROOT").is_some() {
-        // Skip setup step if MIRI_SYSROOT is explicitly set, *unless* we are `cargo miri setup`.
-        return;
+    if !only_setup {
+        if let Some(sysroot) = std::env::var_os("MIRI_SYSROOT") {
+            // Skip setup step if MIRI_SYSROOT is explicitly set, *unless* we are `cargo miri setup`.
+            return sysroot.into();
+        }
     }
 
     // Determine where the rust sources are located.  The env var trumps auto-detection.
@@ -92,6 +99,8 @@ pub fn setup(subcommand: &MiriCommand, target: &str, rustc_version: &VersionMeta
             command.env("RUSTC", &cargo_miri_path);
         }
         command.env("MIRI_CALLED_FROM_SETUP", "1");
+        // Miri expects `MIRI_SYSROOT` to be set when invoked in target mode. Even if that directory is empty.
+        command.env("MIRI_SYSROOT", &sysroot_dir);
         // Make sure there are no other wrappers getting in our way (Cc
         // https://github.com/rust-lang/miri/issues/1421,
         // https://github.com/rust-lang/miri/issues/2429). Looks like setting
@@ -105,7 +114,7 @@ pub fn setup(subcommand: &MiriCommand, target: &str, rustc_version: &VersionMeta
                 command.arg("-v");
             }
         } else {
-            // Supress output.
+            // Suppress output.
             command.stdout(process::Stdio::null());
             command.stderr(process::Stdio::null());
         }
@@ -117,8 +126,6 @@ pub fn setup(subcommand: &MiriCommand, target: &str, rustc_version: &VersionMeta
     // the user might have set, which is consistent with normal `cargo build` that does
     // not apply `RUSTFLAGS` to the sysroot either.
     let rustflags = &["-Cdebug-assertions=off", "-Coverflow-checks=on"];
-    // Make sure all target-level Miri invocations know their sysroot.
-    std::env::set_var("MIRI_SYSROOT", &sysroot_dir);
 
     // Do the build.
     if print_sysroot {
@@ -159,4 +166,6 @@ pub fn setup(subcommand: &MiriCommand, target: &str, rustc_version: &VersionMeta
         // Print just the sysroot and nothing else to stdout; this way we do not need any escaping.
         println!("{}", sysroot_dir.display());
     }
+
+    sysroot_dir
 }

cargo-miri/src/util.rs

@@ -1,7 +1,5 @@
-use std::collections::HashMap;
 use std::env;
 use std::ffi::OsString;
-use std::fmt::Write as _;
 use std::fs::File;
 use std::io::{self, BufWriter, Read, Write};
 use std::ops::Not;
@@ -114,6 +112,11 @@ pub fn cargo() -> Command {
     Command::new(env::var_os("CARGO").unwrap_or_else(|| OsString::from("cargo")))
 }
 
+pub fn flagsplit(flags: &str) -> Vec<String> {
+    // This code is taken from `RUSTFLAGS` handling in cargo.
+    flags.split(' ').map(str::trim).filter(|s| !s.is_empty()).map(str::to_string).collect()
+}
+
 /// Execute the `Command`, where possible by replacing the current process with a new process
 /// described by the `Command`. Then exit this process with the exit code of the new process.
 pub fn exec(mut cmd: Command) -> ! {
@@ -242,46 +245,10 @@ pub fn local_crates(metadata: &Metadata) -> String {
     local_crates
 }
 
-fn env_vars_from_cmd(cmd: &Command) -> Vec<(String, String)> {
-    let mut envs = HashMap::new();
-    for (key, value) in std::env::vars() {
-        envs.insert(key, value);
-    }
-    for (key, value) in cmd.get_envs() {
-        if let Some(value) = value {
-            envs.insert(key.to_string_lossy().to_string(), value.to_string_lossy().to_string());
-        } else {
-            envs.remove(&key.to_string_lossy().to_string());
-        }
-    }
-    let mut envs: Vec<_> = envs.into_iter().collect();
-    envs.sort();
-    envs
-}
-
 /// Debug-print a command that is going to be run.
 pub fn debug_cmd(prefix: &str, verbose: usize, cmd: &Command) {
     if verbose == 0 {
         return;
     }
-    // We only do a single `eprintln!` call to minimize concurrency interactions.
-    let mut out = prefix.to_string();
-    writeln!(out, " running command: env \\").unwrap();
-    if verbose > 1 {
-        // Print the full environment this will be called in.
-        for (key, value) in env_vars_from_cmd(cmd) {
-            writeln!(out, "{key}={value:?} \\").unwrap();
-        }
-    } else {
-        // Print only what has been changed for this `cmd`.
-        for (var, val) in cmd.get_envs() {
-            if let Some(val) = val {
-                writeln!(out, "{}={val:?} \\", var.to_string_lossy()).unwrap();
-            } else {
-                writeln!(out, "--unset={}", var.to_string_lossy()).unwrap();
-            }
-        }
-    }
-    write!(out, "{cmd:?}").unwrap();
-    eprintln!("{out}");
+    eprintln!("{prefix} running command: {cmd:?}");
 }