[PATCH] I/OAT: remove CPU hotplug lock from net_dma_rebalance

[PATCH] I/OAT: remove CPU hotplug lock from net_dma_rebalance

[Chris Leech]

Remove the lock_cpu_hotplug()/unlock_cpu_hotplug() calls from net_dma_rebalance

The lock_cpu_hotplug()/unlock_cpu_hotplug() sequence in net_dma_rebalance
is both incorrect (as pointed out by David Miller) because lock_cpu_hotplug()
may sleep while the net_dma_event_lock spinlock is held, and unnecessary (as
pointed out by Andrew Morton) as spin_lock() disables preemption which
protects from CPU hotplug events.

Signed-off-by: Chris Leech <christopher.leech@intel.com>
---

  net/core/dev.c |    5 -----
 1 files changed, 0 insertions(+), 5 deletions(-)

diff --git a/net/core/dev.c b/net/core/dev.c
index 4d2b516..780d770 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3429,12 +3429,9 @@ static void net_dma_rebalance(void)
        unsigned int cpu, i, n;
        struct dma_chan *chan;

-       lock_cpu_hotplug();
-
        if (net_dma_count == 0) {
                for_each_online_cpu(cpu)

rcu_assign_pointer(per_cpu(softnet_data.net_dma, cpu), NULL);
-               unlock_cpu_hotplug();
                return;
        }

@@ -3454,8 +3451,6 @@ static void net_dma_rebalance(void)
                i++;
        }
        rcu_read_unlock();
-
-       unlock_cpu_hotplug();
 }

 /**

Re: [PATCH] I/OAT: remove CPU hotplug lock from net_dma_rebalance

[David Miller]

From: "Chris Leech" <christopher.leech@intel.com>
Date: Tue, 1 Aug 2006 11:33:02 -0700

>         if (net_dma_count == 0) {
>                 for_each_online_cpu(cpu)
> 
> rcu_assign_pointer(per_cpu(softnet_data.net_dma, cpu), NULL);
> -               unlock_cpu_hotplug();

Why is proper patch submission so damn difficult for people?
This patch is corrupted severely.

I'm fixing this up since it's such an obvious patch, but this
issue is getting really rediculious.  I can't believe how many
people submit line-wrapped, tab destroyed, patches these days.