test: faster test execution in codex-core (#2633)

this dramatically improves time to run `cargo test -p codex-core` (~25x
speedup).

before:
```
cargo test -p codex-core  35.96s user 68.63s system 19% cpu 8:49.80 total
```

after:
```
cargo test -p codex-core  5.51s user 8.16s system 63% cpu 21.407 total
```

both tests measured "hot", i.e. on a 2nd run with no filesystem changes,
to exclude compile times.

approach inspired by [Delete Cargo Integration
Tests](https://matklad.github.io/2021/02/27/delete-cargo-integration-tests.html),
we move all test cases in tests/ into a single suite in order to have a
single binary, as there is significant overhead for each test binary
executed, and because test execution is only parallelized with a single
binary.
This commit is contained in:
Jeremy Rose
2025-08-24 11:10:53 -07:00
committed by GitHub
parent c6a52d611c
commit 32bbbbad61
56 changed files with 78 additions and 3 deletions

View File

@@ -0,0 +1,339 @@
#![allow(clippy::expect_used, clippy::unwrap_used)]
use anyhow::Context;
use assert_cmd::prelude::*;
use codex_core::CODEX_APPLY_PATCH_ARG1;
use std::fs;
use std::process::Command;
use tempfile::tempdir;
/// While we may add an `apply-patch` subcommand to the `codex` CLI multitool
/// at some point, we must ensure that the smaller `codex-exec` CLI can still
/// emulate the `apply_patch` CLI.
#[test]
fn test_standalone_exec_cli_can_use_apply_patch() -> anyhow::Result<()> {
let tmp = tempdir()?;
let relative_path = "source.txt";
let absolute_path = tmp.path().join(relative_path);
fs::write(&absolute_path, "original content\n")?;
Command::cargo_bin("codex-exec")
.context("should find binary for codex-exec")?
.arg(CODEX_APPLY_PATCH_ARG1)
.arg(
r#"*** Begin Patch
*** Update File: source.txt
@@
-original content
+modified by apply_patch
*** End Patch"#,
)
.current_dir(tmp.path())
.assert()
.success()
.stdout("Success. Updated the following files:\nM source.txt\n")
.stderr(predicates::str::is_empty());
assert_eq!(
fs::read_to_string(absolute_path)?,
"modified by apply_patch\n"
);
Ok(())
}
#[cfg(not(target_os = "windows"))]
#[tokio::test]
async fn test_apply_patch_tool() -> anyhow::Result<()> {
use core_test_support::load_sse_fixture_with_id_from_str;
use tempfile::TempDir;
use wiremock::Mock;
use wiremock::MockServer;
use wiremock::ResponseTemplate;
use wiremock::matchers::method;
use wiremock::matchers::path;
const SSE_TOOL_CALL_ADD: &str = r#"[
{
"type": "response.output_item.done",
"item": {
"type": "function_call",
"name": "apply_patch",
"arguments": "{\n \"input\": \"*** Begin Patch\\n*** Add File: test.md\\n+Hello world\\n*** End Patch\"\n}",
"call_id": "__ID__"
}
},
{
"type": "response.completed",
"response": {
"id": "__ID__",
"usage": {
"input_tokens": 0,
"input_tokens_details": null,
"output_tokens": 0,
"output_tokens_details": null,
"total_tokens": 0
},
"output": []
}
}
]"#;
const SSE_TOOL_CALL_UPDATE: &str = r#"[
{
"type": "response.output_item.done",
"item": {
"type": "function_call",
"name": "apply_patch",
"arguments": "{\n \"input\": \"*** Begin Patch\\n*** Update File: test.md\\n@@\\n-Hello world\\n+Final text\\n*** End Patch\"\n}",
"call_id": "__ID__"
}
},
{
"type": "response.completed",
"response": {
"id": "__ID__",
"usage": {
"input_tokens": 0,
"input_tokens_details": null,
"output_tokens": 0,
"output_tokens_details": null,
"total_tokens": 0
},
"output": []
}
}
]"#;
const SSE_TOOL_CALL_COMPLETED: &str = r#"[
{
"type": "response.completed",
"response": {
"id": "__ID__",
"usage": {
"input_tokens": 0,
"input_tokens_details": null,
"output_tokens": 0,
"output_tokens_details": null,
"total_tokens": 0
},
"output": []
}
}
]"#;
// Start a mock model server
let server = MockServer::start().await;
// First response: model calls apply_patch to create test.md
let first = ResponseTemplate::new(200)
.insert_header("content-type", "text/event-stream")
.set_body_raw(
load_sse_fixture_with_id_from_str(SSE_TOOL_CALL_ADD, "call1"),
"text/event-stream",
);
Mock::given(method("POST"))
// .and(path("/v1/responses"))
.respond_with(first)
.up_to_n_times(1)
.mount(&server)
.await;
// Second response: model calls apply_patch to update test.md
let second = ResponseTemplate::new(200)
.insert_header("content-type", "text/event-stream")
.set_body_raw(
load_sse_fixture_with_id_from_str(SSE_TOOL_CALL_UPDATE, "call2"),
"text/event-stream",
);
Mock::given(method("POST"))
.and(path("/v1/responses"))
.respond_with(second)
.up_to_n_times(1)
.mount(&server)
.await;
let final_completed = ResponseTemplate::new(200)
.insert_header("content-type", "text/event-stream")
.set_body_raw(
load_sse_fixture_with_id_from_str(SSE_TOOL_CALL_COMPLETED, "resp3"),
"text/event-stream",
);
Mock::given(method("POST"))
.and(path("/v1/responses"))
.respond_with(final_completed)
.expect(1)
.mount(&server)
.await;
let tmp_cwd = TempDir::new().unwrap();
Command::cargo_bin("codex-exec")
.context("should find binary for codex-exec")?
.current_dir(tmp_cwd.path())
.env("CODEX_HOME", tmp_cwd.path())
.env("OPENAI_API_KEY", "dummy")
.env("OPENAI_BASE_URL", format!("{}/v1", server.uri()))
.arg("--skip-git-repo-check")
.arg("-s")
.arg("workspace-write")
.arg("foo")
.assert()
.success();
// Verify final file contents
let final_path = tmp_cwd.path().join("test.md");
let contents = std::fs::read_to_string(&final_path)
.unwrap_or_else(|e| panic!("failed reading {}: {e}", final_path.display()));
assert_eq!(contents, "Final text\n");
Ok(())
}
#[cfg(not(target_os = "windows"))]
#[tokio::test]
async fn test_apply_patch_freeform_tool() -> anyhow::Result<()> {
use core_test_support::load_sse_fixture_with_id_from_str;
use tempfile::TempDir;
use wiremock::Mock;
use wiremock::MockServer;
use wiremock::ResponseTemplate;
use wiremock::matchers::method;
use wiremock::matchers::path;
const SSE_TOOL_CALL_ADD: &str = r#"[
{
"type": "response.output_item.done",
"item": {
"type": "custom_tool_call",
"name": "apply_patch",
"input": "*** Begin Patch\n*** Add File: test.md\n+Hello world\n*** End Patch",
"call_id": "__ID__"
}
},
{
"type": "response.completed",
"response": {
"id": "__ID__",
"usage": {
"input_tokens": 0,
"input_tokens_details": null,
"output_tokens": 0,
"output_tokens_details": null,
"total_tokens": 0
},
"output": []
}
}
]"#;
const SSE_TOOL_CALL_UPDATE: &str = r#"[
{
"type": "response.output_item.done",
"item": {
"type": "custom_tool_call",
"name": "apply_patch",
"input": "*** Begin Patch\n*** Update File: test.md\n@@\n-Hello world\n+Final text\n*** End Patch",
"call_id": "__ID__"
}
},
{
"type": "response.completed",
"response": {
"id": "__ID__",
"usage": {
"input_tokens": 0,
"input_tokens_details": null,
"output_tokens": 0,
"output_tokens_details": null,
"total_tokens": 0
},
"output": []
}
}
]"#;
const SSE_TOOL_CALL_COMPLETED: &str = r#"[
{
"type": "response.completed",
"response": {
"id": "__ID__",
"usage": {
"input_tokens": 0,
"input_tokens_details": null,
"output_tokens": 0,
"output_tokens_details": null,
"total_tokens": 0
},
"output": []
}
}
]"#;
// Start a mock model server
let server = MockServer::start().await;
// First response: model calls apply_patch to create test.md
let first = ResponseTemplate::new(200)
.insert_header("content-type", "text/event-stream")
.set_body_raw(
load_sse_fixture_with_id_from_str(SSE_TOOL_CALL_ADD, "call1"),
"text/event-stream",
);
Mock::given(method("POST"))
// .and(path("/v1/responses"))
.respond_with(first)
.up_to_n_times(1)
.mount(&server)
.await;
// Second response: model calls apply_patch to update test.md
let second = ResponseTemplate::new(200)
.insert_header("content-type", "text/event-stream")
.set_body_raw(
load_sse_fixture_with_id_from_str(SSE_TOOL_CALL_UPDATE, "call2"),
"text/event-stream",
);
Mock::given(method("POST"))
.and(path("/v1/responses"))
.respond_with(second)
.up_to_n_times(1)
.mount(&server)
.await;
let final_completed = ResponseTemplate::new(200)
.insert_header("content-type", "text/event-stream")
.set_body_raw(
load_sse_fixture_with_id_from_str(SSE_TOOL_CALL_COMPLETED, "resp3"),
"text/event-stream",
);
Mock::given(method("POST"))
// .and(path("/v1/responses"))
.respond_with(final_completed)
.expect(1)
.mount(&server)
.await;
let tmp_cwd = TempDir::new().unwrap();
Command::cargo_bin("codex-exec")
.context("should find binary for codex-exec")?
.current_dir(tmp_cwd.path())
.env("CODEX_HOME", tmp_cwd.path())
.env("OPENAI_API_KEY", "dummy")
.env("OPENAI_BASE_URL", format!("{}/v1", server.uri()))
.arg("--skip-git-repo-check")
.arg("-s")
.arg("workspace-write")
.arg("foo")
.assert()
.success();
// Verify final file contents
let final_path = tmp_cwd.path().join("test.md");
let contents = std::fs::read_to_string(&final_path)
.unwrap_or_else(|e| panic!("failed reading {}: {e}", final_path.display()));
assert_eq!(contents, "Final text\n");
Ok(())
}

View File

@@ -0,0 +1,3 @@
// Aggregates all former standalone integration tests as modules.
mod apply_patch;
mod sandbox;

View File

@@ -0,0 +1,219 @@
#![cfg(unix)]
use codex_core::protocol::SandboxPolicy;
use codex_core::spawn::StdioPolicy;
use std::collections::HashMap;
use std::future::Future;
use std::io;
use std::path::PathBuf;
use std::process::ExitStatus;
use tokio::process::Child;
#[cfg(target_os = "macos")]
async fn spawn_command_under_sandbox(
command: Vec<String>,
sandbox_policy: &SandboxPolicy,
cwd: PathBuf,
stdio_policy: StdioPolicy,
env: HashMap<String, String>,
) -> std::io::Result<Child> {
use codex_core::seatbelt::spawn_command_under_seatbelt;
spawn_command_under_seatbelt(command, sandbox_policy, cwd, stdio_policy, env).await
}
#[cfg(target_os = "linux")]
async fn spawn_command_under_sandbox(
command: Vec<String>,
sandbox_policy: &SandboxPolicy,
cwd: PathBuf,
stdio_policy: StdioPolicy,
env: HashMap<String, String>,
) -> std::io::Result<Child> {
use codex_core::landlock::spawn_command_under_linux_sandbox;
let codex_linux_sandbox_exe = assert_cmd::cargo::cargo_bin("codex-exec");
spawn_command_under_linux_sandbox(
codex_linux_sandbox_exe,
command,
sandbox_policy,
cwd,
stdio_policy,
env,
)
.await
}
#[tokio::test]
async fn python_multiprocessing_lock_works_under_sandbox() {
#[cfg(target_os = "macos")]
let writable_roots = Vec::<PathBuf>::new();
// From https://man7.org/linux/man-pages/man7/sem_overview.7.html
//
// > On Linux, named semaphores are created in a virtual filesystem,
// > normally mounted under /dev/shm.
#[cfg(target_os = "linux")]
let writable_roots = vec![PathBuf::from("/dev/shm")];
let policy = SandboxPolicy::WorkspaceWrite {
writable_roots,
network_access: false,
exclude_tmpdir_env_var: false,
exclude_slash_tmp: false,
};
let python_code = r#"import multiprocessing
from multiprocessing import Lock, Process
def f(lock):
with lock:
print("Lock acquired in child process")
if __name__ == '__main__':
lock = Lock()
p = Process(target=f, args=(lock,))
p.start()
p.join()
"#;
let mut child = spawn_command_under_sandbox(
vec![
"python3".to_string(),
"-c".to_string(),
python_code.to_string(),
],
&policy,
std::env::current_dir().expect("should be able to get current dir"),
StdioPolicy::Inherit,
HashMap::new(),
)
.await
.expect("should be able to spawn python under sandbox");
let status = child.wait().await.expect("should wait for child process");
assert!(status.success(), "python exited with {status:?}");
}
fn unix_sock_body() {
unsafe {
let mut fds = [0i32; 2];
let r = libc::socketpair(libc::AF_UNIX, libc::SOCK_DGRAM, 0, fds.as_mut_ptr());
assert_eq!(
r,
0,
"socketpair(AF_UNIX, SOCK_DGRAM) failed: {}",
io::Error::last_os_error()
);
let msg = b"hello_unix";
// write() from one end (generic write is allowed)
let sent = libc::write(fds[0], msg.as_ptr() as *const libc::c_void, msg.len());
assert!(sent >= 0, "write() failed: {}", io::Error::last_os_error());
// recvfrom() on the other end. We dont need the address for socketpair,
// so we pass null pointers for src address.
let mut buf = [0u8; 64];
let recvd = libc::recvfrom(
fds[1],
buf.as_mut_ptr() as *mut libc::c_void,
buf.len(),
0,
std::ptr::null_mut(),
std::ptr::null_mut(),
);
assert!(
recvd >= 0,
"recvfrom() failed: {}",
io::Error::last_os_error()
);
let recvd_slice = &buf[..(recvd as usize)];
assert_eq!(
recvd_slice,
&msg[..],
"payload mismatch: sent {} bytes, got {} bytes",
msg.len(),
recvd
);
// Also exercise AF_UNIX stream socketpair quickly to ensure AF_UNIX in general works.
let mut sfds = [0i32; 2];
let sr = libc::socketpair(libc::AF_UNIX, libc::SOCK_STREAM, 0, sfds.as_mut_ptr());
assert_eq!(
sr,
0,
"socketpair(AF_UNIX, SOCK_STREAM) failed: {}",
io::Error::last_os_error()
);
let snt2 = libc::write(sfds[0], msg.as_ptr() as *const libc::c_void, msg.len());
assert!(
snt2 >= 0,
"write(stream) failed: {}",
io::Error::last_os_error()
);
let mut b2 = [0u8; 64];
let rcv2 = libc::recv(sfds[1], b2.as_mut_ptr() as *mut libc::c_void, b2.len(), 0);
assert!(
rcv2 >= 0,
"recv(stream) failed: {}",
io::Error::last_os_error()
);
// Clean up
let _ = libc::close(sfds[0]);
let _ = libc::close(sfds[1]);
let _ = libc::close(fds[0]);
let _ = libc::close(fds[1]);
}
}
#[tokio::test]
async fn allow_unix_socketpair_recvfrom() {
run_code_under_sandbox(
"allow_unix_socketpair_recvfrom",
&SandboxPolicy::ReadOnly,
|| async { unix_sock_body() },
)
.await
.expect("should be able to reexec");
}
const IN_SANDBOX_ENV_VAR: &str = "IN_SANDBOX";
#[expect(clippy::expect_used)]
pub async fn run_code_under_sandbox<F, Fut>(
test_selector: &str,
policy: &SandboxPolicy,
child_body: F,
) -> io::Result<Option<ExitStatus>>
where
F: FnOnce() -> Fut + Send + 'static,
Fut: Future<Output = ()> + Send + 'static,
{
if std::env::var(IN_SANDBOX_ENV_VAR).is_err() {
let exe = std::env::current_exe()?;
let mut cmds = vec![exe.to_string_lossy().into_owned(), "--exact".into()];
let mut stdio_policy = StdioPolicy::RedirectForShellTool;
// Allow for us to pass forward --nocapture / use the right stdio policy.
if std::env::args().any(|a| a == "--nocapture") {
cmds.push("--nocapture".into());
stdio_policy = StdioPolicy::Inherit;
}
cmds.push(test_selector.into());
// Your existing launcher:
let mut child = spawn_command_under_sandbox(
cmds,
policy,
std::env::current_dir().expect("should be able to get current dir"),
stdio_policy,
HashMap::from([("IN_SANDBOX".into(), "1".into())]),
)
.await?;
let status = child.wait().await?;
Ok(Some(status))
} else {
// Child branch: run the provided body.
child_body().await;
Ok(None)
}
}