net: Fix percpu counters deadlock

When we converted the protocol atomic counters such as the orphan
count and the total socket count deadlocks were introduced due to
the mismatch in BH status of the spots that used the percpu counter
operations.

Based on the diagnosis and patch by Peter Zijlstra, this patch
fixes these issues by disabling BH where we may be in process
context.

Reported-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Tested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 9 insertions, 4 deletions

diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c
index 614958b7c276..eb62e58bff79 100644
--- a/net/ipv4/proc.c
+++ b/net/ipv4/proc.c
@@ -38,6 +38,7 @@
 #include <net/tcp.h>
 #include <net/udp.h>
 #include <net/udplite.h>
+#include <linux/bottom_half.h>
 #include <linux/inetdevice.h>
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
@@ -50,13 +51,17 @@
 static int sockstat_seq_show(struct seq_file *seq, void *v)
 {
         struct net *net = seq->private;
+        int orphans, sockets;
+
+        local_bh_disable();
+        orphans = percpu_counter_sum_positive(&tcp_orphan_count),
+        sockets = percpu_counter_sum_positive(&tcp_sockets_allocated),
+        local_bh_enable();
 
         socket_seq_show(seq);
         seq_printf(seq, "TCP: inuse %d orphan %d tw %d alloc %d mem %d\n",
-                   sock_prot_inuse_get(net, &tcp_prot),
-                   (int)percpu_counter_sum_positive(&tcp_orphan_count),
-                   tcp_death_row.tw_count,
-                   (int)percpu_counter_sum_positive(&tcp_sockets_allocated),
+                   sock_prot_inuse_get(net, &tcp_prot), orphans,
+                   tcp_death_row.tw_count, sockets,
                    atomic_read(&tcp_memory_allocated));
         seq_printf(seq, "UDP: inuse %d mem %d\n",
                    sock_prot_inuse_get(net, &udp_prot),