summaryrefslogtreecommitdiff
path: root/include
diff options
context:
space:
mode:
authorSebastian Andrzej Siewior <bigeasy@linutronix.de>2013-03-20 17:06:20 (GMT)
committerScott Wood <scottwood@freescale.com>2013-04-04 22:09:40 (GMT)
commit18b4ed55ef83e77e8df2c45f1a99ffc6e5c07db6 (patch)
treee8441a32ac099c5b1a44507bdc0c499ece030c18 /include
parent14e27282bb54af3db98aa29fcd83895438a3bc98 (diff)
downloadlinux-fsl-qoriq-18b4ed55ef83e77e8df2c45f1a99ffc6e5c07db6.tar.xz
net: make devnet_rename_seq a mutex
On RT write_seqcount_begin() disables preemption and device_rename() allocates memory with GFP_KERNEL and grabs later the sysfs_mutex mutex. Since I don't see a reason why this can't be a mutex, make it one. We probably don't have that much reads at the same time in the hot path. Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Diffstat (limited to 'include')
-rw-r--r--include/linux/netdevice.h2
1 files changed, 1 insertions, 1 deletions
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index b705426..041bc5b 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1579,7 +1579,7 @@ extern int call_netdevice_notifiers(unsigned long val, struct net_device *dev);
extern rwlock_t dev_base_lock; /* Device list lock */
-extern seqcount_t devnet_rename_seq; /* Device rename seq */
+extern struct mutex devnet_rename_mutex;
#define for_each_netdev(net, d) \