Re: [Xen-devel] [RFC 21/23] net/xen-netback: Make it running on 64KB page granularity

On Fri, May 15, 2015 at 01:35:42PM +0100, Julien Grall wrote:

Hi Wei,

Thanks you for the review.

On 15/05/15 03:35, Wei Liu wrote:

quoted

On Thu, May 14, 2015 at 06:01:01PM +0100, Julien Grall wrote:

quoted

The PV network protocol is using 4KB page granularity. The goal of this
patch is to allow a Linux using 64KB page granularity working as a
network backend on a non-modified Xen.

It's only necessary to adapt the ring size and break skb data in small
chunk of 4KB. The rest of the code is relying on the grant table code.

Although only simple workload is working (dhcp request, ping). If I try
to use wget in the guest, it will stall until a tcpdump is started on
the vif interface in DOM0. I wasn't able to find why.

I think in wget workload you're more likely to break down 64K pages to
4K pages. Some of your calculation of mfn, offset might be wrong.

If so, why tcpdump on the vif interface would make wget suddenly
working? Does it make netback use a different path?

No, but if might make core network component behave differently, this is
only my suspicion.

Do you see malformed packets with tcpdump?

quoted

quoted

I have not modified XEN_NETBK_RX_SLOTS_MAX because I wasn't sure what
it's used for (I have limited knowledge on the network driver).

This is the maximum slots a guest packet can use. AIUI the protocol
still works on 4K granularity (you break 64K page to a bunch of 4K
pages), you don't need to change this.

1 slot = 1 grant right? If so, XEN_NETBK_RX_SLOTS_MAX is based on the
number of Linux page. So we would have to get the number for Xen page.

Yes, 1 slot = 1 grant. I see what you're up to now. Yes, you need to
change this constant to match underlying HV page.

Although, I gave a try to multiple by XEN_PFN_PER_PAGE (4KB/64KB = 16)
but it get stuck in the loop.

I don't follow. What is the new #define? Which loop does it get stuck?

quoted

quoted

---
 drivers/net/xen-netback/common.h  |  7 ++++---
 drivers/net/xen-netback/netback.c | 27 ++++++++++++++-------------
 2 files changed, 18 insertions(+), 16 deletions(-)

diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h
index 8a495b3..0eda6e9 100644
--- a/drivers/net/xen-netback/common.h
+++ b/drivers/net/xen-netback/common.h

@@ -44,6 +44,7 @@
 #include <xen/interface/grant_table.h>
 #include <xen/grant_table.h>
 #include <xen/xenbus.h>
+#include <xen/page.h>
 #include <linux/debugfs.h>
 
 typedef unsigned int pending_ring_idx_t;

@@ -64,8 +65,8 @@ struct pending_tx_info {
 	struct ubuf_info callback_struct;
 };
 
-#define XEN_NETIF_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, PAGE_SIZE)
-#define XEN_NETIF_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, PAGE_SIZE)
+#define XEN_NETIF_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, XEN_PAGE_SIZE)
+#define XEN_NETIF_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, XEN_PAGE_SIZE)
 
 struct xenvif_rx_meta {
 	int id;

@@ -80,7 +81,7 @@ struct xenvif_rx_meta {
 /* Discriminate from any valid pending_idx value. */
 #define INVALID_PENDING_IDX 0xFFFF
 
-#define MAX_BUFFER_OFFSET PAGE_SIZE
+#define MAX_BUFFER_OFFSET XEN_PAGE_SIZE
 
 #define MAX_PENDING_REQS XEN_NETIF_TX_RING_SIZE
 

diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
index 9ae1d43..ea5ce84 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c

@@ -274,7 +274,7 @@ static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb
 {
 	struct gnttab_copy *copy_gop;
 	struct xenvif_rx_meta *meta;
-	unsigned long bytes;
+	unsigned long bytes, off_grant;
 	int gso_type = XEN_NETIF_GSO_TYPE_NONE;
 
 	/* Data must not cross a page boundary. */

@@ -295,7 +295,8 @@ static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb
 		if (npo->copy_off == MAX_BUFFER_OFFSET)
 			meta = get_next_rx_buffer(queue, npo);
 
-		bytes = PAGE_SIZE - offset;
+		off_grant = offset & ~XEN_PAGE_MASK;
+		bytes = XEN_PAGE_SIZE - off_grant;
 		if (bytes > size)
 			bytes = size;
 

@@ -314,9 +315,9 @@ static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb
 		} else {
 			copy_gop->source.domid = DOMID_SELF;
 			copy_gop->source.u.gmfn =
-				virt_to_mfn(page_address(page));
+				virt_to_mfn(page_address(page) + offset);
 		}
-		copy_gop->source.offset = offset;
+		copy_gop->source.offset = off_grant;
 
 		copy_gop->dest.domid = queue->vif->domid;
 		copy_gop->dest.offset = npo->copy_off;

@@ -747,7 +748,7 @@ static int xenvif_count_requests(struct xenvif_queue *queue,
 		first->size -= txp->size;
 		slots++;
 
-		if (unlikely((txp->offset + txp->size) > PAGE_SIZE)) {
+		if (unlikely((txp->offset + txp->size) > XEN_PAGE_SIZE)) {
 			netdev_err(queue->vif->dev, "Cross page boundary, txp->offset: %x, size: %u\n",
 				 txp->offset, txp->size);
 			xenvif_fatal_tx_err(queue->vif);

@@ -1241,11 +1242,11 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue,
 		}
 
 		/* No crossing a page as the payload mustn't fragment. */
-		if (unlikely((txreq.offset + txreq.size) > PAGE_SIZE)) {
+		if (unlikely((txreq.offset + txreq.size) > XEN_PAGE_SIZE)) {
 			netdev_err(queue->vif->dev,
 				   "txreq.offset: %x, size: %u, end: %lu\n",
 				   txreq.offset, txreq.size,
-				   (txreq.offset&~PAGE_MASK) + txreq.size);
+				   (txreq.offset&~XEN_PAGE_MASK) + txreq.size);
 			xenvif_fatal_tx_err(queue->vif);
 			break;
 		}

@@ -1287,7 +1288,7 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue,
 			virt_to_mfn(skb->data);

You didn't change the calculation of MFN here. I think it returns the
MFN of the first 4K sub-page of that 64K page.  Do I miss anything?

There is no change required. On ARM virt_to_mfn is implemented with:

pfn_to_mfn(virt_to_phys(v) >> XEN_PAGE_SHIFT)

which will return a 4KB PFN (see patch #23).

OK. I missed that patch.

quoted

quoted

 		queue->tx_copy_ops[*copy_ops].dest.domid = DOMID_SELF;
 		queue->tx_copy_ops[*copy_ops].dest.offset =
-			offset_in_page(skb->data);
+			offset_in_page(skb->data) & ~XEN_PAGE_MASK;
 
 		queue->tx_copy_ops[*copy_ops].len = data_len;
 		queue->tx_copy_ops[*copy_ops].flags = GNTCOPY_source_gref;

@@ -1366,8 +1367,8 @@ static int xenvif_handle_frag_list(struct xenvif_queue *queue, struct sk_buff *s

This function is to coalesce frag_list to a new SKB. It's completely
fine to use the natural granularity of backend domain. The way you
modified it can lead to waste of memory, i.e. you only use first 4K of a
64K page.

Thanks for explaining. I wasn't sure how the function works so I change
it for safety. I will redo the change.

FWIW, I'm sure there is other place in netback where we waste memory
with 64KB page granularity (such as grant table). I need to track them.

Let me know if you have some place in mind where the memory usage can be
improved.

I was about to say the mmap_pages array is an array of pages. But that
probably belongs to grant table driver.

Wei.

quoted

quoted

 			return -ENOMEM;
 		}
 
-		if (offset + PAGE_SIZE < skb->len)
-			len = PAGE_SIZE;
+		if (offset + XEN_PAGE_SIZE < skb->len)
+			len = XEN_PAGE_SIZE;
 		else
 			len = skb->len - offset;
 		if (skb_copy_bits(skb, offset, page_address(page), len))

@@ -1396,7 +1397,7 @@ static int xenvif_handle_frag_list(struct xenvif_queue *queue, struct sk_buff *s
 	/* Fill the skb with the new (local) frags. */
 	memcpy(skb_shinfo(skb)->frags, frags, i * sizeof(skb_frag_t));
 	skb_shinfo(skb)->nr_frags = i;
-	skb->truesize += i * PAGE_SIZE;
+	skb->truesize += i * XEN_PAGE_SIZE;

The true size accounts for the actual memory occupied by this SKB. Since
the page is allocated with alloc_page, the granularity should be
PAGE_SIZE not XEN_PAGE_SIZE.

Ok. I will replace with PAGE_SIZE.

Regards,

-- 
Julien Grall