From: Jason Gunthorpe <jgg@mellanox.com>
Date: Wed, 6 May 2020 10:46:57 +0300
Subject: RDMA/cm: Add a note explaining how the timewait is eventually freed
Patch-mainline: v5.8-rc1
Git-commit: 09fb406a569b2f4b596048e1cfa2c92a35b8fc9b
References: jsc#SLE-15176

The way the cm_timewait_info is converted into a work and then freed
is very subtle and surprising, add a note clarifying the lifetime
here.

Link: https://lore.kernel.org/r/20200506074701.9775-7-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Acked-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
---
 drivers/infiniband/core/cm.c |    5 +++++
 1 file changed, 5 insertions(+)

--- a/drivers/infiniband/core/cm.c
+++ b/drivers/infiniband/core/cm.c
@@ -1025,6 +1025,11 @@ static void cm_enter_timewait(struct cm_
 				   msecs_to_jiffies(wait_time));
 	spin_unlock_irqrestore(&cm.lock, flags);
 
+	/*
+	 * The timewait_info is converted into a work and gets freed during
+	 * cm_free_work() in cm_timewait_handler().
+	 */
+	BUILD_BUG_ON(offsetof(struct cm_timewait_info, work) != 0);
 	cm_id_priv->timewait_info = NULL;
 }