1 files changed, 71 insertions, 11 deletions

diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c
index 374f8afbf8a5..e621f8d9b436 100644
--- a/drivers/hv/ring_buffer.c
+++ b/drivers/hv/ring_buffer.c
@@ -181,7 +181,7 @@ void hv_ringbuffer_pre_init(struct vmbus_channel *channel)
 
 /* Initialize the ring buffer. */
 int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info,
-		       struct page *pages, u32 page_cnt)
+		       struct page *pages, u32 page_cnt, u32 max_pkt_size)
 {
 	int i;
 	struct page **pages_wraparound;
@@ -223,6 +223,14 @@ int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info,
 		sizeof(struct hv_ring_buffer);
 	ring_info->priv_read_index = 0;
 
+	/* Initialize buffer that holds copies of incoming packets */
+	if (max_pkt_size) {
+		ring_info->pkt_buffer = kzalloc(max_pkt_size, GFP_KERNEL);
+		if (!ring_info->pkt_buffer)
+			return -ENOMEM;
+		ring_info->pkt_buffer_size = max_pkt_size;
+	}
+
 	spin_lock_init(&ring_info->ring_lock);
 
 	return 0;
@@ -235,6 +243,9 @@ void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info)
 	vunmap(ring_info->ring_buffer);
 	ring_info->ring_buffer = NULL;
 	mutex_unlock(&ring_info->ring_buffer_mutex);
+
+	kfree(ring_info->pkt_buffer);
+	ring_info->pkt_buffer_size = 0;
 }
 
 /* Write to the ring buffer. */
@@ -375,7 +386,7 @@ int hv_ringbuffer_read(struct vmbus_channel *channel,
 	memcpy(buffer, (const char *)desc + offset, packetlen);
 
 	/* Advance ring index to next packet descriptor */
-	__hv_pkt_iter_next(channel, desc);
+	__hv_pkt_iter_next(channel, desc, true);
 
 	/* Notify host of update */
 	hv_pkt_iter_close(channel);
@@ -402,6 +413,22 @@ static u32 hv_pkt_iter_avail(const struct hv_ring_buffer_info *rbi)
 }
 
 /*
+ * Get first vmbus packet without copying it out of the ring buffer
+ */
+struct vmpacket_descriptor *hv_pkt_iter_first_raw(struct vmbus_channel *channel)
+{
+	struct hv_ring_buffer_info *rbi = &channel->inbound;
+
+	hv_debug_delay_test(channel, MESSAGE_DELAY);
+
+	if (hv_pkt_iter_avail(rbi) < sizeof(struct vmpacket_descriptor))
+		return NULL;
+
+	return (struct vmpacket_descriptor *)(hv_get_ring_buffer(rbi) + rbi->priv_read_index);
+}
+EXPORT_SYMBOL_GPL(hv_pkt_iter_first_raw);
+
+/*
  * Get first vmbus packet from ring buffer after read_index
  *
  * If ring buffer is empty, returns NULL and no other action needed.
@@ -409,17 +436,49 @@ static u32 hv_pkt_iter_avail(const struct hv_ring_buffer_info *rbi)
 struct vmpacket_descriptor *hv_pkt_iter_first(struct vmbus_channel *channel)
 {
 	struct hv_ring_buffer_info *rbi = &channel->inbound;
-	struct vmpacket_descriptor *desc;
+	struct vmpacket_descriptor *desc, *desc_copy;
+	u32 bytes_avail, pkt_len, pkt_offset;
 
-	hv_debug_delay_test(channel, MESSAGE_DELAY);
-	if (hv_pkt_iter_avail(rbi) < sizeof(struct vmpacket_descriptor))
+	desc = hv_pkt_iter_first_raw(channel);
+	if (!desc)
 		return NULL;
 
-	desc = hv_get_ring_buffer(rbi) + rbi->priv_read_index;
-	if (desc)
-		prefetch((char *)desc + (desc->len8 << 3));
+	bytes_avail = min(rbi->pkt_buffer_size, hv_pkt_iter_avail(rbi));
+
+	/*
+	 * Ensure the compiler does not use references to incoming Hyper-V values (which
+	 * could change at any moment) when reading local variables later in the code
+	 */
+	pkt_len = READ_ONCE(desc->len8) << 3;
+	pkt_offset = READ_ONCE(desc->offset8) << 3;
+
+	/*
+	 * If pkt_len is invalid, set it to the smaller of hv_pkt_iter_avail() and
+	 * rbi->pkt_buffer_size
+	 */
+	if (pkt_len < sizeof(struct vmpacket_descriptor) || pkt_len > bytes_avail)
+		pkt_len = bytes_avail;
+
+	/*
+	 * If pkt_offset is invalid, arbitrarily set it to
+	 * the size of vmpacket_descriptor
+	 */
+	if (pkt_offset < sizeof(struct vmpacket_descriptor) || pkt_offset > pkt_len)
+		pkt_offset = sizeof(struct vmpacket_descriptor);
+
+	/* Copy the Hyper-V packet out of the ring buffer */
+	desc_copy = (struct vmpacket_descriptor *)rbi->pkt_buffer;
+	memcpy(desc_copy, desc, pkt_len);
+
+	/*
+	 * Hyper-V could still change len8 and offset8 after the earlier read.
+	 * Ensure that desc_copy has legal values for len8 and offset8 that
+	 * are consistent with the copy we just made
+	 */
+	desc_copy->len8 = pkt_len >> 3;
+	desc_copy->offset8 = pkt_offset >> 3;
 
-	return desc;
+	return desc_copy;
 }
 EXPORT_SYMBOL_GPL(hv_pkt_iter_first);
 
@@ -431,7 +490,8 @@ EXPORT_SYMBOL_GPL(hv_pkt_iter_first);
  */
 struct vmpacket_descriptor *
 __hv_pkt_iter_next(struct vmbus_channel *channel,
-		   const struct vmpacket_descriptor *desc)
+		   const struct vmpacket_descriptor *desc,
+		   bool copy)
 {
 	struct hv_ring_buffer_info *rbi = &channel->inbound;
 	u32 packetlen = desc->len8 << 3;
@@ -444,7 +504,7 @@ __hv_pkt_iter_next(struct vmbus_channel *channel,
 		rbi->priv_read_index -= dsize;
 
 	/* more data? */
-	return hv_pkt_iter_first(channel);
+	return copy ? hv_pkt_iter_first(channel) : hv_pkt_iter_first_raw(channel);
 }
 EXPORT_SYMBOL_GPL(__hv_pkt_iter_next);