feat: experimental env var: CODEX_SANDBOX_NETWORK_DISABLED (#879)

When using Codex to develop Codex itself, I noticed that sometimes it
would try to add `#[ignore]` to the following tests:

```
keeps_previous_response_id_between_tasks()
retries_on_early_close()
```

Both of these tests start a `MockServer` that launches an HTTP server on
an ephemeral port and requires network access to hit it, which the
Seatbelt policy associated with `--full-auto` correctly denies. If I
wasn't paying attention to the code that Codex was generating, one of
these `#[ignore]` annotations could have slipped into the codebase,
effectively disabling the test for everyone.

To that end, this PR enables an experimental environment variable named
`CODEX_SANDBOX_NETWORK_DISABLED` that is set to `1` if the
`SandboxPolicy` used to spawn the process does not have full network
access. I say it is "experimental" because I'm not convinced this API is
quite right, but we need to start somewhere. (It might be more
appropriate to have an env var like `CODEX_SANDBOX=full-auto`, but the
challenge is that our newer `SandboxPolicy` abstraction does not map to
a simple set of enums like in the TypeScript CLI.)

We leverage this new functionality by adding the following code to the
aforementioned tests as a way to "dynamically disable" them:

```rust
if std::env::var(CODEX_SANDBOX_NETWORK_DISABLED_ENV_VAR).is_ok() {
    println!(
        "Skipping test because it cannot execute when network is disabled in a Codex sandbox."
    );
    return;
}
```

We can use the `debug seatbelt --full-auto` command to verify that
`cargo test` fails when run under Seatbelt prior to this change:

```
$ cargo run --bin codex -- debug seatbelt --full-auto -- cargo test
---- keeps_previous_response_id_between_tasks stdout ----

thread 'keeps_previous_response_id_between_tasks' panicked at /Users/mbolin/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/wiremock-0.6.3/src/mock_server/builder.rs:107:46:
Failed to bind an OS port for a mock server.: Os { code: 1, kind: PermissionDenied, message: "Operation not permitted" }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace


failures:
    keeps_previous_response_id_between_tasks

test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

error: test failed, to rerun pass `-p codex-core --test previous_response_id`
```

Though after this change, the above command succeeds! This means that,
going forward, when Codex operates on Codex itself, when it runs `cargo
test`, only "real failures" should cause the command to fail.

As part of this change, I decided to tighten up the codepaths for
running `exec()` for shell tool calls. In particular, we do it in `core`
for the main Codex business logic itself, but we also expose this logic
via `debug` subcommands in the CLI in the `cli` crate. The logic for the
`debug` subcommands was not quite as faithful to the true business logic
as I liked, so I:

* refactored a bit of the Linux code, splitting `linux.rs` into
`linux_exec.rs` and `landlock.rs` in the `core` crate.
* gating less code behind `#[cfg(target_os = "linux")]` because such
code does not get built by default when I develop on Mac, which means I
either have to build the code in Docker or wait for CI signal
* introduced `macro_rules! configure_command` in `exec.rs` so we can
have both sync and async versions of this code. The synchronous version
seems more appropriate for straight threads or potentially fork/exec.
This commit is contained in:
Michael Bolin
2025-05-09 18:29:34 -07:00
committed by GitHub
parent 7795272282
commit fde48aaa0d
12 changed files with 275 additions and 128 deletions

View File

@@ -0,0 +1,23 @@
#[cfg(unix)]
pub(crate) fn handle_exit_status(status: std::process::ExitStatus) -> ! {
use std::os::unix::process::ExitStatusExt;
// Use ExitStatus to derive the exit code.
if let Some(code) = status.code() {
std::process::exit(code);
} else if let Some(signal) = status.signal() {
std::process::exit(128 + signal);
} else {
std::process::exit(1);
}
}
#[cfg(windows)]
pub(crate) fn handle_exit_status(status: std::process::ExitStatus) -> ! {
if let Some(code) = status.code() {
std::process::exit(code);
} else {
// Rare on Windows, but if it happens: use fallback code.
std::process::exit(1);
}
}

View File

@@ -3,12 +3,14 @@
//! On Linux the command is executed inside a Landlock + seccomp sandbox by
//! calling the low-level `exec_linux` helper from `codex_core::linux`.
use codex_core::exec::StdioPolicy;
use codex_core::exec::spawn_child_sync;
use codex_core::exec_linux::apply_sandbox_policy_to_current_thread;
use codex_core::protocol::SandboxPolicy;
use std::os::unix::process::ExitStatusExt;
use std::process;
use std::process::Command;
use std::process::ExitStatus;
use crate::exit_status::handle_exit_status;
/// Execute `command` in a Linux sandbox (Landlock + seccomp) the way Codex
/// would.
pub fn run_landlock(command: Vec<String>, sandbox_policy: SandboxPolicy) -> anyhow::Result<()> {
@@ -19,20 +21,15 @@ pub fn run_landlock(command: Vec<String>, sandbox_policy: SandboxPolicy) -> anyh
// Spawn a new thread and apply the sandbox policies there.
let handle = std::thread::spawn(move || -> anyhow::Result<ExitStatus> {
let cwd = std::env::current_dir()?;
codex_core::linux::apply_sandbox_policy_to_current_thread(sandbox_policy, &cwd)?;
let status = Command::new(&command[0]).args(&command[1..]).status()?;
apply_sandbox_policy_to_current_thread(&sandbox_policy, &cwd)?;
let mut child = spawn_child_sync(command, cwd, &sandbox_policy, StdioPolicy::Inherit)?;
let status = child.wait()?;
Ok(status)
});
let status = handle
.join()
.map_err(|e| anyhow::anyhow!("Failed to join thread: {e:?}"))??;
// Use ExitStatus to derive the exit code.
if let Some(code) = status.code() {
process::exit(code);
} else if let Some(signal) = status.signal() {
process::exit(128 + signal);
} else {
process::exit(1);
}
handle_exit_status(status);
}

View File

@@ -1,4 +1,5 @@
#[cfg(target_os = "linux")]
mod exit_status;
#[cfg(unix)]
pub mod landlock;
pub mod proto;
pub mod seatbelt;

View File

@@ -82,7 +82,7 @@ async fn main() -> anyhow::Result<()> {
let sandbox_policy = create_sandbox_policy(full_auto, sandbox);
seatbelt::run_seatbelt(command, sandbox_policy).await?;
}
#[cfg(target_os = "linux")]
#[cfg(unix)]
DebugCommand::Landlock(LandlockCommand {
command,
sandbox,
@@ -91,7 +91,7 @@ async fn main() -> anyhow::Result<()> {
let sandbox_policy = create_sandbox_policy(full_auto, sandbox);
codex_cli::landlock::run_landlock(command, sandbox_policy)?;
}
#[cfg(not(target_os = "linux"))]
#[cfg(not(unix))]
DebugCommand::Landlock(_) => {
anyhow::bail!("Landlock is only supported on Linux.");
}

View File

@@ -1,18 +1,16 @@
use codex_core::exec::create_seatbelt_command;
use codex_core::exec::StdioPolicy;
use codex_core::exec::spawn_command_under_seatbelt;
use codex_core::protocol::SandboxPolicy;
use crate::exit_status::handle_exit_status;
pub async fn run_seatbelt(
command: Vec<String>,
sandbox_policy: SandboxPolicy,
) -> anyhow::Result<()> {
let cwd = std::env::current_dir().expect("failed to get cwd");
let seatbelt_command = create_seatbelt_command(command, &sandbox_policy, &cwd);
let status = tokio::process::Command::new(seatbelt_command[0].clone())
.args(&seatbelt_command[1..])
.spawn()
.map_err(|e| anyhow::anyhow!("Failed to spawn command: {}", e))?
.wait()
.await
.map_err(|e| anyhow::anyhow!("Failed to wait for command: {}", e))?;
std::process::exit(status.code().unwrap_or(1));
let cwd = std::env::current_dir()?;
let mut child =
spawn_command_under_seatbelt(command, &sandbox_policy, cwd, StdioPolicy::Inherit).await?;
let status = child.wait().await?;
handle_exit_status(status);
}