[PATCH v2] Drivers: hv: vmbus: fix the race when querying & updating the percpu list

[PATCH v2] Drivers: hv: vmbus: fix the race when querying & updating the percpu list

[Dexuan Cui]

There is a rare race when we remove an entry from the global list
hv_context.percpu_list[cpu] in hv_process_channel_removal() ->
percpu_channel_deq() -> list_del(): at this time, if vmbus_on_event() ->
process_chn_event() -> pcpu_relid2channel() is trying to query the list,
we can get the general protection fault:

general protection fault: 0000 [#1] SMP
...
RIP: 0010:[<ffffffff81461b6b>]  [<ffffffff81461b6b>] vmbus_on_event+0xc4/0x149

Similarly, we also have the issue in the code path: vmbus_process_offer() ->
percpu_channel_enq().

We can resolve the issue by disabling the tasklet when updating the list.

Reported-by: Rolf Neugebauer <rolf.neugebauer@docker.com>
Cc: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Dexuan Cui <decui@microsoft.com>
---

v2: added tasklet_schedule() after tasklet_enable(). Thanks, Vitaly!

 drivers/hv/channel.c      |  5 +++++
 drivers/hv/channel_mgmt.c | 24 +++++++++++++-----------
 include/linux/hyperv.h    |  3 +++
 3 files changed, 21 insertions(+), 11 deletions(-)

diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c
index 56dd261..17c4711 100644
--- a/drivers/hv/channel.c
+++ b/drivers/hv/channel.c
@@ -546,8 +546,11 @@ static int vmbus_close_internal(struct vmbus_channel *channel)
 		put_cpu();
 		smp_call_function_single(channel->target_cpu, reset_channel_cb,
 					 channel, true);
+		smp_call_function_single(channel->target_cpu,
+					 percpu_channel_deq, channel, true);
 	} else {
 		reset_channel_cb(channel);
+		percpu_channel_deq(channel);
 		put_cpu();
 	}
 
@@ -592,6 +595,8 @@ static int vmbus_close_internal(struct vmbus_channel *channel)
 
 out:
 	tasklet_enable(tasklet);
+	/* for possible pending event */
+	tasklet_schedule(tasklet);
 
 	return ret;
 }
diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c
index 38b682ba..8e251e3 100644
--- a/drivers/hv/channel_mgmt.c
+++ b/drivers/hv/channel_mgmt.c
@@ -21,6 +21,7 @@
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/kernel.h>
+#include <linux/interrupt.h>
 #include <linux/sched.h>
 #include <linux/wait.h>
 #include <linux/mm.h>
@@ -277,7 +278,7 @@ static void free_channel(struct vmbus_channel *channel)
 	kfree(channel);
 }
 
-static void percpu_channel_enq(void *arg)
+void percpu_channel_enq(void *arg)
 {
 	struct vmbus_channel *channel = arg;
 	int cpu = smp_processor_id();
@@ -285,7 +286,7 @@ static void percpu_channel_enq(void *arg)
 	list_add_tail(&channel->percpu_list, &hv_context.percpu_list[cpu]);
 }
 
-static void percpu_channel_deq(void *arg)
+void percpu_channel_deq(void *arg)
 {
 	struct vmbus_channel *channel = arg;
 
@@ -313,15 +314,6 @@ void hv_process_channel_removal(struct vmbus_channel *channel, u32 relid)
 	BUG_ON(!channel->rescind);
 	BUG_ON(!mutex_is_locked(&vmbus_connection.channel_mutex));
 
-	if (channel->target_cpu != get_cpu()) {
-		put_cpu();
-		smp_call_function_single(channel->target_cpu,
-					 percpu_channel_deq, channel, true);
-	} else {
-		percpu_channel_deq(channel);
-		put_cpu();
-	}
-
 	if (channel->primary_channel == NULL) {
 		list_del(&channel->listentry);
 
@@ -363,6 +355,7 @@ void vmbus_free_channels(void)
  */
 static void vmbus_process_offer(struct vmbus_channel *newchannel)
 {
+	struct tasklet_struct *tasklet;
 	struct vmbus_channel *channel;
 	bool fnew = true;
 	unsigned long flags;
@@ -409,6 +402,8 @@ static void vmbus_process_offer(struct vmbus_channel *newchannel)
 
 	init_vp_index(newchannel, dev_type);
 
+	tasklet = hv_context.event_dpc[newchannel->target_cpu];
+	tasklet_disable(tasklet);
 	if (newchannel->target_cpu != get_cpu()) {
 		put_cpu();
 		smp_call_function_single(newchannel->target_cpu,
@@ -418,6 +413,9 @@ static void vmbus_process_offer(struct vmbus_channel *newchannel)
 		percpu_channel_enq(newchannel);
 		put_cpu();
 	}
+	tasklet_enable(tasklet);
+	/* for possible pending event */
+	tasklet_schedule(tasklet);
 
 	/*
 	 * This state is used to indicate a successful open
@@ -469,6 +467,7 @@ err_deq_chan:
 	list_del(&newchannel->listentry);
 	mutex_unlock(&vmbus_connection.channel_mutex);
 
+	tasklet_disable(tasklet);
 	if (newchannel->target_cpu != get_cpu()) {
 		put_cpu();
 		smp_call_function_single(newchannel->target_cpu,
@@ -477,6 +476,9 @@ err_deq_chan:
 		percpu_channel_deq(newchannel);
 		put_cpu();
 	}
+	tasklet_enable(tasklet);
+	/* for possible pending event */
+	tasklet_schedule(tasklet);
 
 err_free_chan:
 	free_channel(newchannel);
diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h
index 7be7237..95aea09 100644
--- a/include/linux/hyperv.h
+++ b/include/linux/hyperv.h
@@ -1328,6 +1328,9 @@ extern bool vmbus_prep_negotiate_resp(struct icmsg_hdr *,
 					struct icmsg_negotiate *, u8 *, int,
 					int);
 
+void percpu_channel_enq(void *arg);
+void percpu_channel_deq(void *arg);
+
 void hv_process_channel_removal(struct vmbus_channel *channel, u32 relid);
 
 /*
-- 
2.7.4

RE: [PATCH v2] Drivers: hv: vmbus: fix the race when querying & updating the percpu list

[Dexuan Cui]

> From: devel [mailto:driverdev-devel-bounces@linuxdriverproject.org] On
> Behalf Of Dexuan Cui
> Sent: Wednesday, May 18, 2016 11:44
> To: gregkh@linuxfoundation.org; linux-kernel@vger.kernel.org; driverdev-
> devel@linuxdriverproject.org; olaf@aepfle.de; apw@canonical.com;
> jasowang@redhat.com; KY Srinivasan <kys@microsoft.com>;
> vkuznets@redhat.com
> Cc: Haiyang Zhang <haiyangz@microsoft.com>
> Subject: [PATCH v2] Drivers: hv: vmbus: fix the race when querying &
> updating the percpu list
> 
> There is a rare race when we remove an entry from the global list
> hv_context.percpu_list[cpu] in hv_process_channel_removal() ->
> percpu_channel_deq() -> list_del(): at this time, if vmbus_on_event() ->
> process_chn_event() -> pcpu_relid2channel() is trying to query the list,
> we can get the general protection fault:
> 
> general protection fault: 0000 [#1] SMP
> ...
> RIP: 0010:[<ffffffff81461b6b>]  [<ffffffff81461b6b>]
> vmbus_on_event+0xc4/0x149
> 
> Similarly, we also have the issue in the code path: vmbus_process_offer() ->
> percpu_channel_enq().
> 
> We can resolve the issue by disabling the tasklet when updating the list.
> 
> Reported-by: Rolf Neugebauer <rolf.neugebauer@docker.com>
> Cc: Vitaly Kuznetsov <vkuznets@redhat.com>
> Signed-off-by: Dexuan Cui <decui@microsoft.com>
> ---
> 
> v2: added tasklet_schedule() after tasklet_enable(). Thanks, Vitaly!

Please ignore the patch for now.

I found an issue with the patch:  after I moved percpu_channel_deq()
from  hv_process_channel_removal() to vmbus_close_internal(), the
channel couldn't be removed from the per-cpu list, if the channel state
was not CHANNEL_OPENED_STATE.

I'll have to think about this and fix the issue in the next version.

Thanks,
-- Dexuan