fix(702): narrow all PrivRequest handlers to specific ops

This commit is contained in:
damocles 2026-06-01 17:05:31 +02:00 committed by mara
commit ec12ba4b1a

View file

@ -9,13 +9,14 @@
//! **Security model**: every request is validated against a strict
//! container-name allowlist before any filesystem or process operation.
//! Only containers whose names match the hive convention (`h-*`,
//! the manager container, or known sibling services) are accepted.
//! the manager container, or known sibling service containers) are
//! accepted. Every variant maps to a single known operation — no
//! arbitrary command pass-through.
//!
//! **Socket activation**: when systemd passes the listener socket via
//! `LISTEN_FDS=1` + `LISTEN_PID=<self>`, the inherited fd 3 is used
//! instead of binding a fresh socket.
use std::os::unix::fs::PermissionsExt as _;
use std::path::{Path, PathBuf};
use anyhow::{Context as _, Result, bail};
@ -28,13 +29,13 @@ use tokio::process::Command;
const AGENT_PREFIX: &str = "h-";
/// Manager container name (mirrors `lifecycle::MANAGER_NAME`).
const MANAGER_NAME: &str = "hm1nd";
const MANAGER_NAME: &str = "root";
/// Sibling service containers managed by hive-c0re.
const SIBLING_CONTAINERS: &[&str] = &["hive-forge", "hive-matrix", "hive-gateway"];
/// Allowed path prefixes for `Chown` / `Chmod` operations.
const SAFE_PATH_PREFIXES: &[&str] = &["/run/hive-agent/", "/var/lib/hyperhive/"];
/// Root of the per-agent unix-socket dirs on the host.
const SOCKET_DIR_ROOT: &str = "/run/hive-agent";
#[tokio::main]
async fn main() -> Result<()> {
@ -94,9 +95,9 @@ fn socket_listener() -> Result<UnixListener> {
.with_context(|| format!("create {}", parent.display()))?;
}
let _ = std::fs::remove_file(path);
let listener =
UnixListener::bind(path).with_context(|| format!("bind {PRIV_SOCK}"))?;
let listener = UnixListener::bind(path).with_context(|| format!("bind {PRIV_SOCK}"))?;
// Mode 0660: only the hive-core group can connect.
use std::os::unix::fs::PermissionsExt as _;
std::fs::set_permissions(path, std::fs::Permissions::from_mode(0o660))
.context("chmod priv.sock")?;
tracing::info!(path = PRIV_SOCK, "bound priv socket");
@ -147,32 +148,69 @@ async fn dispatch(line: &str) -> PrivResponse {
/// Execute a validated `PrivRequest`. Returns `(stdout, stderr)` on success.
async fn exec(req: PrivRequest) -> Result<(String, String)> {
match req {
PrivRequest::ContainerRun { ref args } => {
validate_container_args(args)?;
let out = Command::new("nixos-container")
.args(args)
.output()
.await
.context("invoke nixos-container")?;
let stdout = String::from_utf8_lossy(&out.stdout).into_owned();
let stderr = String::from_utf8_lossy(&out.stderr).into_owned();
// Log each line so progress is visible in journald even
// without a streaming protocol.
for line in stdout.lines() {
tracing::info!(target: "nixos-container", "{line}");
PrivRequest::StartContainer { ref name } => {
validate_container_name(name)?;
container_run(&["start", &container_system_name(name)]).await
}
PrivRequest::StopContainer { ref name } => {
validate_container_name(name)?;
container_run(&["stop", &container_system_name(name)]).await
}
PrivRequest::KillContainer { ref name } => {
validate_container_name(name)?;
container_run(&["kill", &container_system_name(name)]).await
}
PrivRequest::UpdateContainer { ref name, ref flake_ref } => {
validate_container_name(name)?;
validate_flake_ref(flake_ref)?;
container_run(&["update", &container_system_name(name), "--flake", flake_ref]).await
}
PrivRequest::CreateContainer { ref name, ref flake_ref } => {
validate_container_name(name)?;
validate_flake_ref(flake_ref)?;
container_run(&["create", &container_system_name(name), "--flake", flake_ref]).await
}
PrivRequest::DestroyContainer { ref name } => {
validate_container_name(name)?;
container_run(&["destroy", &container_system_name(name)]).await
}
PrivRequest::ListContainers => container_run(&["list"]).await,
PrivRequest::WriteNspawnConf { ref container, ref content } => {
validate_container_system_name(container)?;
let path = format!("/etc/nixos-containers/{container}.conf");
std::fs::write(&path, content).with_context(|| format!("write {path}"))?;
Ok((String::new(), String::new()))
}
PrivRequest::WriteResourceLimits {
ref container,
ref memory_max,
ref cpu_quota,
} => {
validate_container_system_name(container)?;
let dir = format!("/run/systemd/system/container@{container}.service.d");
std::fs::create_dir_all(&dir).with_context(|| format!("create {dir}"))?;
let path = format!("{dir}/hyperhive-limits.conf");
let content = format!("[Service]\nMemoryMax={memory_max}\nCPUQuota={cpu_quota}\n");
std::fs::write(&path, content).with_context(|| format!("write {path}"))?;
Ok((String::new(), String::new()))
}
PrivRequest::RemoveServiceDropin { ref container } => {
validate_container_system_name(container)?;
let dir = format!("/run/systemd/system/container@{container}.service.d");
if Path::new(&dir).exists() {
std::fs::remove_dir_all(&dir)
.with_context(|| format!("remove {dir}"))?;
}
for line in stderr.lines() {
tracing::warn!(target: "nixos-container", "{line}");
}
if !out.status.success() {
bail!(
"nixos-container {} failed ({}): {}",
args.join(" "),
out.status,
stderr.trim()
);
}
Ok((stdout, stderr))
Ok((String::new(), String::new()))
}
PrivRequest::DaemonReload => {
@ -191,58 +229,6 @@ async fn exec(req: PrivRequest) -> Result<(String, String)> {
Ok((String::new(), String::new()))
}
PrivRequest::WriteNspawnConf { ref container, ref content } => {
validate_container_name(container)?;
let path = format!("/etc/nixos-containers/{container}.conf");
std::fs::write(&path, content).with_context(|| format!("write {path}"))?;
Ok((String::new(), String::new()))
}
PrivRequest::WriteSystemdDropin {
ref container,
ref filename,
ref content,
} => {
validate_container_name(container)?;
validate_dropin_filename(filename)?;
let dir =
format!("/run/systemd/system/container@{container}.service.d");
std::fs::create_dir_all(&dir)
.with_context(|| format!("create {dir}"))?;
let path = format!("{dir}/{filename}");
std::fs::write(&path, content).with_context(|| format!("write {path}"))?;
Ok((String::new(), String::new()))
}
PrivRequest::RemoveSystemdDropin { ref container } => {
validate_container_name(container)?;
let dir =
format!("/run/systemd/system/container@{container}.service.d");
if Path::new(&dir).exists() {
std::fs::remove_dir_all(&dir)
.with_context(|| format!("remove {dir}"))?;
}
Ok((String::new(), String::new()))
}
PrivRequest::Chown { ref path, uid, gid } => {
validate_safe_path(path)?;
std::os::unix::fs::chown(path, Some(uid), Some(gid))
.with_context(|| {
format!("chown {} to {uid}:{gid}", path.display())
})?;
Ok((String::new(), String::new()))
}
PrivRequest::Chmod { ref path, mode } => {
validate_safe_path(path)?;
std::fs::set_permissions(path, std::fs::Permissions::from_mode(mode))
.with_context(|| {
format!("chmod {:o} {}", mode, path.display())
})?;
Ok((String::new(), String::new()))
}
PrivRequest::ReloadGatewayNginx => {
let out = Command::new("systemd-run")
.args(["--machine=hive-gateway", "--quiet", "--", "nginx", "-s", "reload"])
@ -258,27 +244,78 @@ async fn exec(req: PrivRequest) -> Result<(String, String)> {
}
Ok((String::new(), String::new()))
}
PrivRequest::ChownSocketDir { ref agent_name, uid, gid } => {
validate_agent_name(agent_name)?;
let path = socket_dir_path(agent_name);
std::os::unix::fs::chown(&path, Some(uid), Some(gid))
.with_context(|| format!("chown {} to {uid}:{gid}", path.display()))?;
Ok((String::new(), String::new()))
}
PrivRequest::ChmodSocketDir { ref agent_name, mode } => {
validate_agent_name(agent_name)?;
let path = socket_dir_path(agent_name);
use std::os::unix::fs::PermissionsExt as _;
std::fs::set_permissions(&path, std::fs::Permissions::from_mode(mode))
.with_context(|| format!("chmod {:o} {}", mode, path.display()))?;
Ok((String::new(), String::new()))
}
}
}
/// Validate that nixos-container args reference only hive-managed containers.
fn validate_container_args(args: &[String]) -> Result<()> {
if args.is_empty() {
bail!("ContainerRun: args must not be empty");
/// Invoke `nixos-container` with the given args, log output to journald.
async fn container_run(args: &[&str]) -> Result<(String, String)> {
let out = Command::new("nixos-container")
.args(args)
.output()
.await
.context("invoke nixos-container")?;
let stdout = String::from_utf8_lossy(&out.stdout).into_owned();
let stderr = String::from_utf8_lossy(&out.stderr).into_owned();
for line in stdout.lines() {
tracing::info!(target: "nixos-container", "{line}");
}
// Verbs whose second positional argument is a container name.
const CONTAINER_ARG_VERBS: &[&str] =
&["start", "stop", "kill", "restart", "destroy", "update", "run", "create"];
if CONTAINER_ARG_VERBS.contains(&args[0].as_str()) {
if let Some(name) = args.get(1) {
validate_container_name(name)?;
}
for line in stderr.lines() {
tracing::warn!(target: "nixos-container", "{line}");
}
// `list` and unknown verbs take no container arg - no further validation needed.
if !out.status.success() {
bail!(
"nixos-container {} failed ({}): {}",
args.join(" "),
out.status,
stderr.trim()
);
}
Ok((stdout, stderr))
}
/// Return the system container name for a logical agent name.
/// Manager (`MANAGER_NAME`) passes through; sub-agents get `h-` prefix.
fn container_system_name(name: &str) -> String {
if name == MANAGER_NAME {
name.to_owned()
} else {
format!("{AGENT_PREFIX}{name}")
}
}
/// Path of the per-agent unix-socket dir on the host.
fn socket_dir_path(agent_name: &str) -> PathBuf {
PathBuf::from(format!("{SOCKET_DIR_ROOT}/{agent_name}"))
}
/// Validate a logical agent name (the name hive-c0re uses internally,
/// before the `h-` container prefix is applied).
fn validate_agent_name(name: &str) -> Result<()> {
if name == MANAGER_NAME {
return Ok(());
}
validate_name_chars(name)?;
Ok(())
}
/// Validate that a container name is managed by hive.
/// Validate a logical agent name and check it maps to a hive-managed container.
fn validate_container_name(name: &str) -> Result<()> {
if name == MANAGER_NAME {
return Ok(());
@ -286,30 +323,40 @@ fn validate_container_name(name: &str) -> Result<()> {
if SIBLING_CONTAINERS.contains(&name) {
return Ok(());
}
if name.starts_with(AGENT_PREFIX) {
let suffix = &name[AGENT_PREFIX.len()..];
if !suffix.is_empty()
&& suffix.chars().all(|c| c.is_ascii_lowercase() || c.is_ascii_digit() || c == '-')
{
return Ok(());
}
validate_name_chars(name)?;
Ok(())
}
/// Validate a system-level container name (already has `h-` prefix for
/// sub-agents, or is the manager name / sibling service name).
fn validate_container_system_name(name: &str) -> Result<()> {
if name == MANAGER_NAME {
return Ok(());
}
if SIBLING_CONTAINERS.contains(&name) {
return Ok(());
}
if let Some(suffix) = name.strip_prefix(AGENT_PREFIX) {
validate_name_chars(suffix)?;
return Ok(());
}
bail!("container name {name:?} is not managed by hive");
}
fn validate_dropin_filename(filename: &str) -> Result<()> {
if filename.is_empty() || filename.contains('/') || filename.contains("..") {
bail!("invalid drop-in filename: {filename:?}");
fn validate_name_chars(name: &str) -> Result<()> {
if name.is_empty()
|| !name
.chars()
.all(|c| c.is_ascii_lowercase() || c.is_ascii_digit() || c == '-')
{
bail!("invalid name {name:?}: must be non-empty lowercase ascii + digits + hyphens");
}
Ok(())
}
fn validate_safe_path(path: &Path) -> Result<()> {
let s = path.to_string_lossy();
for prefix in SAFE_PATH_PREFIXES {
if s.starts_with(prefix) && !s.contains("..") {
return Ok(());
}
fn validate_flake_ref(flake_ref: &str) -> Result<()> {
if flake_ref.is_empty() || flake_ref.contains('\n') || flake_ref.contains('\0') {
bail!("invalid flake_ref {flake_ref:?}");
}
bail!("path {} is outside allowed prefixes", path.display());
Ok(())
}