daandemeyer / rpms / systemd

Forked from rpms/systemd 2 years ago
Clone
Blob Blame History Raw
From c34ec3588a40e595bb837f3c12e44808c7cd20fd Mon Sep 17 00:00:00 2001
From: Michal Sekletar <msekleta@redhat.com>
Date: Wed, 1 May 2019 15:52:42 +0200
Subject: [PATCH] udev: call poll() again after killing the spawned process

Later we check .revents of pfd. Hence we need to initialize it properly
and to do that we need to call poll() once again because previously it
exited twice with timeout and hence left pfd uninitialized. For the
third time it should return immediately since we killed the spawned
process with SIGKILL.

Related: #1697909
---
 src/udev/udev-event.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/src/udev/udev-event.c b/src/udev/udev-event.c
index 7fe64f04a4..07b82d093e 100644
--- a/src/udev/udev-event.c
+++ b/src/udev/udev-event.c
@@ -593,6 +593,10 @@ static int spawn_wait(struct udev_event *event,
                         if (fdcount == 0) {
                                 log_error("timeout: killing '%s' ["PID_FMT"]", cmd, pid);
                                 kill(pid, SIGKILL);
+
+                                fdcount = poll(pfd, 1, 1000);
+                                if (fdcount <= 0)
+                                        continue;
                         }
                 }