From d51654822fa0ad4327c7465fe03ea9427be9e150 Mon Sep 17 00:00:00 2001
From: Michael Bolin <mbolin@openai.com>
Date: Tue, 22 Jul 2025 00:41:27 -0700
Subject: [PATCH] fix: use PR_SET_PDEATHSIG so to ensure child processes are
 killed in a timely manner (#1626)

Some users have reported issues where child processes are not cleaned up
after Codex exits (e.g., https://github.com/openai/codex/issues/1570).

This is generally a tricky issue on operating systems: if a parent
process receives `SIGKILL`, then it terminates immediately and cannot
communicate with the child.

**It only helps on Linux**, but this PR introduces the use of `prctl(2)`
so that if the parent process dies, `SIGTERM` will be delivered to the
child process. Whereas previously, I believe that if Codex spawned a
long-running process (like `tsc --watch`) and the Codex process received
`SIGKILL`, the `tsc --watch` process would be reparented to the init
process and would never be killed. Now with the use of `prctl(2)`, the
`tsc --watch` process should receive `SIGTERM` in that scenario.

We still need to come up with a solution for macOS. I've started to look
at `launchd`, but I'm researching a number of options.
---
 codex-rs/Cargo.lock       |  1 +
 codex-rs/core/Cargo.toml  |  1 +
 codex-rs/core/src/exec.rs | 25 +++++++++++++++++++++++++
 3 files changed, 27 insertions(+)

diff --git a/codex-rs/Cargo.lock b/codex-rs/Cargo.lock
index 9c604e79..9b4a4e32 100644
--- a/codex-rs/Cargo.lock
+++ b/codex-rs/Cargo.lock
@@ -669,6 +669,7 @@ dependencies = [
  "fs2",
  "futures",
  "landlock",
+ "libc",
  "maplit",
  "mcp-types",
  "mime_guess",
diff --git a/codex-rs/core/Cargo.toml b/codex-rs/core/Cargo.toml
index e192a71f..a87894bc 100644
--- a/codex-rs/core/Cargo.toml
+++ b/codex-rs/core/Cargo.toml
@@ -22,6 +22,7 @@ env-flags = "0.1.1"
 eventsource-stream = "0.2.3"
 fs2 = "0.4.3"
 futures = "0.3"
+libc = "0.2.174"
 mcp-types = { path = "../mcp-types" }
 mime_guess = "2.0"
 rand = "0.9"
diff --git a/codex-rs/core/src/exec.rs b/codex-rs/core/src/exec.rs
index 3b37cb53..4b33b0b3 100644
--- a/codex-rs/core/src/exec.rs
+++ b/codex-rs/core/src/exec.rs
@@ -384,6 +384,31 @@ async fn spawn_child_async(
         cmd.env(CODEX_SANDBOX_NETWORK_DISABLED_ENV_VAR, "1");
     }
 
+    // If this Codex process dies (including being killed via SIGKILL), we want
+    // any child processes that were spawned as part of a `"shell"` tool call
+    // to also be terminated.
+
+    // This relies on prctl(2), so it only works on Linux.
+    #[cfg(target_os = "linux")]
+    unsafe {
+        cmd.pre_exec(|| {
+            // This prctl call effectively requests, "deliver SIGTERM when my
+            // current parent dies."
+            if libc::prctl(libc::PR_SET_PDEATHSIG, libc::SIGTERM) == -1 {
+                return Err(io::Error::last_os_error());
+            }
+
+            // Though if there was a race condition and this pre_exec() block is
+            // run _after_ the parent (i.e., the Codex process) has already
+            // exited, then the parent is the _init_ process (which will never
+            // die), so we should just terminate the child process now.
+            if libc::getppid() == 1 {
+                libc::raise(libc::SIGTERM);
+            }
+            Ok(())
+        });
+    }
+
     match stdio_policy {
         StdioPolicy::RedirectForShellTool => {
             // Do not create a file descriptor for stdin because otherwise some