tipc: increase timeout in tipc_sk_enqueue()
authorHoang Le <hoang.h.le@dektech.com.au>
Mon, 13 Sep 2021 09:28:52 +0000 (16:28 +0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 22 Sep 2021 09:43:09 +0000 (11:43 +0200)
commitde9b2e5256951d542bf8721fdb216ce3d6c463ba
tree8928f2fdfb726d8207ba658de4526dd4b4182eab
parent88f3259b6828ad0b911b8cc7de659c1d454baf2b
tipc: increase timeout in tipc_sk_enqueue()

commit f4bb62e64c88c93060c051195d3bbba804e56945 upstream.

In tipc_sk_enqueue() we use hardcoded 2 jiffies to extract
socket buffer from generic queue to particular socket.
The 2 jiffies is too short in case there are other high priority
tasks get CPU cycles for multiple jiffies update. As result, no
buffer could be enqueued to particular socket.

To solve this, we switch to use constant timeout 20msecs.
Then, the function will be expired between 2 jiffies (CONFIG_100HZ)
and 20 jiffies (CONFIG_1000HZ).

Fixes: c637c1035534 ("tipc: resolve race problem at unicast message reception")
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/tipc/socket.c