__check_pf: Add a cancellation cleanup handler [BZ #20975]

There are reports for hang in __check_pf:

https://github.com/JoeDog/siege/issues/4

It is reproducible only under specific configurations:

1. Large number of cores (>= 64) and large number of threads (> 3X of
the number of cores) with long lived socket connection.
2. Low power (frequency) mode.
3. Power management is enabled.

While holding lock, __check_pf calls make_request which calls __sendto
and __recvmsg.  Since __sendto and __recvmsg are cancellation points,
lock held by __check_pf won't be released and can cause deadlock when
thread cancellation happens in __sendto or __recvmsg.  Add a cancellation
cleanup handler for __check_pf to unlock the lock when cancelled by
another thread.  This fixes BZ #20975 and the siege hang issue.

(cherry picked from commit a443bd3fb233186038b8b483959ecb7978d1abea)

diff --git a/sysdeps/unix/sysv/linux/Makefile b/sysdeps/unix/sysv/linux/Makefile
index 9531641f8222547c3eec956647fa6c86a1b7c295..410d5c19f672946651e23434c057e96232bbe9ed 100644 (file)
--- a/sysdeps/unix/sysv/linux/Makefile
+++ b/sysdeps/unix/sysv/linux/Makefile
@@ -256,6 +256,8 @@ sysdep_headers += netinet/if_fddi.h netinet/if_tr.h \
                  netrom/netrom.h netpacket/packet.h netrose/rose.h \
                  neteconet/ec.h netiucv/iucv.h
 sysdep_routines += netlink_assert_response
+
+CFLAGS-check_pf.c += -fexceptions
 endif
 
 # Don't compile the ctype glue code, since there is no old non-GNU C library.

diff --git a/sysdeps/unix/sysv/linux/check_pf.c b/sysdeps/unix/sysv/linux/check_pf.c
index 9fef28b7b46223e3b0d0cb35935aeec21d48b5b0..863eeadbc73ad6070598e58b5424def295e1fa67 100644 (file)
--- a/sysdeps/unix/sysv/linux/check_pf.c
+++ b/sysdeps/unix/sysv/linux/check_pf.c
@@ -292,6 +292,14 @@ make_request (int fd, pid_t pid)
   return NULL;
 }
 
+#ifdef __EXCEPTIONS
+static void
+cancel_handler (void *arg __attribute__((unused)))
+{
+  /* Release the lock.  */
+  __libc_lock_unlock (lock);
+}
+#endif
 
 void
 attribute_hidden
@@ -304,6 +312,10 @@ __check_pf (bool *seen_ipv4, bool *seen_ipv6,
   struct cached_data *olddata = NULL;
   struct cached_data *data = NULL;
 
+#ifdef __EXCEPTIONS
+  /* Make sure that lock is released when the thread is cancelled.  */
+  __libc_cleanup_push (cancel_handler, NULL);
+#endif
   __libc_lock_lock (lock);
 
   if (cache_valid_p ())
@@ -338,6 +350,9 @@ __check_pf (bool *seen_ipv4, bool *seen_ipv6,
        }
     }
 
+#ifdef __EXCEPTIONS
+  __libc_cleanup_pop (0);
+#endif
   __libc_lock_unlock (lock);
 
   if (data != NULL)