[DLM] fix softlockup in dlm_recv [54/54]

From: Steven Whitehouse
Date: Mon Feb 05 2007 - 09:42:36 EST


>From aee6c8244dea3199ff3d17fc3167660ea95299ec Mon Sep 17 00:00:00 2001
From: Patrick Caulfield <pcaulfie@xxxxxxxxxx>
Date: Thu, 1 Feb 2007 16:46:33 +0000
Subject: [PATCH] [DLM] fix softlockup in dlm_recv

This patch stops the dlm_recv workqueue from busy-waiting when a node
disconnects. This can cause soft lockup errors on debug systems and bad
performance generally.

Signed-Off-By: Patrick Caulfield <pcaulfie@xxxxxxxxxx>
Signed-off-by: Steven Whitehouse <swhiteho@xxxxxxxxxx>

diff --git a/fs/dlm/lowcomms-tcp.c b/fs/dlm/lowcomms-tcp.c
index 18ade44..f1efd17 100644
--- a/fs/dlm/lowcomms-tcp.c
+++ b/fs/dlm/lowcomms-tcp.c
@@ -2,7 +2,7 @@
*******************************************************************************
**
** Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
-** Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
+** Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
**
** This copyrighted material is made available to anyone wishing to use,
** modify, copy, or redistribute it subject to the terms and conditions
@@ -109,7 +109,6 @@ struct connection {
struct page *rx_page;
struct cbuf cb;
int retries;
- atomic_t waiting_requests;
#define MAX_CONNECT_RETRIES 3
struct connection *othercon;
struct work_struct rwork; /* Receive workqueue */
@@ -278,8 +277,11 @@ static int receive_from_sock(struct connection *con)

mutex_lock(&con->sock_mutex);

- if (con->sock == NULL)
- goto out;
+ if (con->sock == NULL) {
+ ret = -EAGAIN;
+ goto out_close;
+ }
+
if (con->rx_page == NULL) {
/*
* This doesn't need to be atomic, but I think it should
@@ -352,7 +354,6 @@ static int receive_from_sock(struct connection *con)
con->rx_page = NULL;
}

-out:
if (call_again_soon)
goto out_resched;
mutex_unlock(&con->sock_mutex);
@@ -370,6 +371,9 @@ out_close:
close_connection(con, false);
/* Reconnect when there is something to send */
}
+ /* Don't return success if we really got EOF */
+ if (ret == 0)
+ ret = -EAGAIN;

return ret;
}
@@ -847,7 +851,6 @@ int dlm_lowcomms_close(int nodeid)
if (con) {
clean_one_writequeue(con);
close_connection(con, true);
- atomic_set(&con->waiting_requests, 0);
}
return 0;

--
1.4.4.2



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/