sd-pam: Drop uid so parent signal arrives at child.

The PAM helper thread needs to capture the death signal from the
parent, but is prohibited from doing so since when the child dies
as normal user, the kernel won't allow it to send a TERM to the
PAM helper thread which is running as root.

This causes the PAM threads to never exit, accumulating after
user sessions exit.

There is however really no need to keep the PAM threads running as
root, so, we can just setresuid() to the same user as defined in the
unit file for the parent thread (User=). This makes the TERM signal
arrive as normal. In case setresuid() fails, we ignore the error, so
we at least fall back to the current behaviour.

1 files changed, 13 insertions, 5 deletions

diff --git a/src/core/execute.c b/src/core/execute.c
index 953cfa2baa..4d4091940f 100644
--- a/src/core/execute.c
+++ b/src/core/execute.c
@@ -703,6 +703,7 @@ static int null_conv(
 static int setup_pam(
                 const char *name,
                 const char *user,
+                uid_t uid,
                 const char *tty,
                 char ***pam_env,
                 int fds[], unsigned n_fds) {
@@ -781,10 +782,17 @@ static int setup_pam(
                 open here that have been opened by PAM. */
                 close_many(fds, n_fds);
 
-                /* Wait until our parent died. This will most likely
-                 * not work since the kernel does not allow
-                 * unprivileged parents kill their privileged children
-                 * this way. We rely on the control groups kill logic
+                /* Drop privileges - we don't need any to pam_close_session
+                 * and this will make PR_SET_PDEATHSIG work in most cases.
+                 * If this fails, ignore the error - but expect sd-pam threads
+                 * to fail to exit normally */
+                if (setresuid(uid, uid, uid) < 0)
+                        log_error("Error: Failed to setresuid() in sd-pam: %s", strerror(-r));
+
+                /* Wait until our parent died. This will only work if
+                 * the above setresuid() succeeds, otherwise the kernel
+                 * will not allow unprivileged parents kill their privileged
+                 * children this way. We rely on the control groups kill logic
                  * to do the rest for us. */
                 if (prctl(PR_SET_PDEATHSIG, SIGTERM) < 0)
                         goto child_finish;
@@ -1294,7 +1302,7 @@ int exec_spawn(ExecCommand *command,
 
 #ifdef HAVE_PAM
                 if (context->pam_name && username) {
-                        err = setup_pam(context->pam_name, username, context->tty_path, &pam_env, fds, n_fds);
+                        err = setup_pam(context->pam_name, username, uid, context->tty_path, &pam_env, fds, n_fds);
                         if (err < 0) {
                                 r = EXIT_PAM;
                                 goto fail_child;