Patchwork [net-next,v1] xen-netback: make copy batch size configurable

login
register
mail settings
Submitter Joao Martins
Date Nov. 13, 2017, 4:34 p.m.
Message ID <20171113163401.eia4pdyimysfg4h6@paddy>
Download mbox | patch
Permalink /patch/383545/
State New
Headers show

Comments

Joao Martins - Nov. 13, 2017, 4:34 p.m.
On Mon, Nov 13, 2017 at 11:58:03AM +0000, Paul Durrant wrote:
> On Mon, Nov 13, 2017 at 11:54:00AM +0000, Joao Martins wrote:
> > On 11/13/2017 10:33 AM, Paul Durrant wrote:
> > > On 11/10/2017 19:35 PM, Joao Martins wrote:

[snip]

> > >> diff --git a/drivers/net/xen-netback/rx.c b/drivers/net/xen-netback/rx.c
> > >> index b1cf7c6f407a..793a85f61f9d 100644
> > >> --- a/drivers/net/xen-netback/rx.c
> > >> +++ b/drivers/net/xen-netback/rx.c
> > >> @@ -168,11 +168,14 @@ static void xenvif_rx_copy_add(struct
> > >> xenvif_queue *queue,
> > >>  			       struct xen_netif_rx_request *req,
> > >>  			       unsigned int offset, void *data, size_t len)
> > >>  {
> > >> +	unsigned int batch_size;
> > >>  	struct gnttab_copy *op;
> > >>  	struct page *page;
> > >>  	struct xen_page_foreign *foreign;
> > >>
> > >> -	if (queue->rx_copy.num == COPY_BATCH_SIZE)
> > >> +	batch_size = min(xenvif_copy_batch_size, queue->rx_copy.size);
> > >
> > > Surely queue->rx_copy.size and xenvif_copy_batch_size are always
> > > identical? Why do you need this statement (and hence stack variable)?
> > >
> > This statement was to allow to be changed dynamically and would
> > affect all newly created guests or running guests if value happened
> > to be smaller than initially allocated. But I suppose I should make
> > behaviour more consistent with the other params we have right now
> > and just look at initially allocated one `queue->rx_copy.batch_size` ?
> 
> Yes, that would certainly be consistent but I can see value in
> allowing it to be dynamically tuned, so perhaps adding some re-allocation
> code to allow the batch to be grown as well as shrunk might be nice.

The shrink one we potentially risk losing data, so we need to gate the
reallocation whenever `rx_copy.num` is less than the new requested
batch. Worst case means guestrx_thread simply uses the initial
allocated value.

Anyhow, something like the below scissored diff (on top of your comments):
Paul Durrant - Nov. 13, 2017, 4:39 p.m.
> -----Original Message-----
> From: Joao Martins [mailto:joao.m.martins@oracle.com]
> Sent: 13 November 2017 16:34
> To: Paul Durrant <Paul.Durrant@citrix.com>
> Cc: netdev@vger.kernel.org; Wei Liu <wei.liu2@citrix.com>; xen-
> devel@lists.xenproject.org
> Subject: Re: [PATCH net-next v1] xen-netback: make copy batch size
> configurable
> 
> On Mon, Nov 13, 2017 at 11:58:03AM +0000, Paul Durrant wrote:
> > On Mon, Nov 13, 2017 at 11:54:00AM +0000, Joao Martins wrote:
> > > On 11/13/2017 10:33 AM, Paul Durrant wrote:
> > > > On 11/10/2017 19:35 PM, Joao Martins wrote:
> 
> [snip]
> 
> > > >> diff --git a/drivers/net/xen-netback/rx.c b/drivers/net/xen-
> netback/rx.c
> > > >> index b1cf7c6f407a..793a85f61f9d 100644
> > > >> --- a/drivers/net/xen-netback/rx.c
> > > >> +++ b/drivers/net/xen-netback/rx.c
> > > >> @@ -168,11 +168,14 @@ static void xenvif_rx_copy_add(struct
> > > >> xenvif_queue *queue,
> > > >>  			       struct xen_netif_rx_request *req,
> > > >>  			       unsigned int offset, void *data, size_t len)
> > > >>  {
> > > >> +	unsigned int batch_size;
> > > >>  	struct gnttab_copy *op;
> > > >>  	struct page *page;
> > > >>  	struct xen_page_foreign *foreign;
> > > >>
> > > >> -	if (queue->rx_copy.num == COPY_BATCH_SIZE)
> > > >> +	batch_size = min(xenvif_copy_batch_size, queue-
> >rx_copy.size);
> > > >
> > > > Surely queue->rx_copy.size and xenvif_copy_batch_size are always
> > > > identical? Why do you need this statement (and hence stack variable)?
> > > >
> > > This statement was to allow to be changed dynamically and would
> > > affect all newly created guests or running guests if value happened
> > > to be smaller than initially allocated. But I suppose I should make
> > > behaviour more consistent with the other params we have right now
> > > and just look at initially allocated one `queue->rx_copy.batch_size` ?
> >
> > Yes, that would certainly be consistent but I can see value in
> > allowing it to be dynamically tuned, so perhaps adding some re-allocation
> > code to allow the batch to be grown as well as shrunk might be nice.
> 
> The shrink one we potentially risk losing data, so we need to gate the
> reallocation whenever `rx_copy.num` is less than the new requested
> batch. Worst case means guestrx_thread simply uses the initial
> allocated value.

Can't you just re-alloc immediately after the flush (when num is guaranteed to be zero)?

  Paul
Joao Martins - Nov. 13, 2017, 4:53 p.m.
On Mon, Nov 13, 2017 at 04:39:09PM +0000, Paul Durrant wrote:
> > -----Original Message-----
> > From: Joao Martins [mailto:joao.m.martins@oracle.com]
> > Sent: 13 November 2017 16:34
> > To: Paul Durrant <Paul.Durrant@citrix.com>
> > Cc: netdev@vger.kernel.org; Wei Liu <wei.liu2@citrix.com>; xen-
> > devel@lists.xenproject.org
> > Subject: Re: [PATCH net-next v1] xen-netback: make copy batch size
> > configurable
> > 
> > On Mon, Nov 13, 2017 at 11:58:03AM +0000, Paul Durrant wrote:
> > > On Mon, Nov 13, 2017 at 11:54:00AM +0000, Joao Martins wrote:
> > > > On 11/13/2017 10:33 AM, Paul Durrant wrote:
> > > > > On 11/10/2017 19:35 PM, Joao Martins wrote:
> > 
> > [snip]
> > 
> > > > >> diff --git a/drivers/net/xen-netback/rx.c b/drivers/net/xen-netback/rx.c
> > > > >> index b1cf7c6f407a..793a85f61f9d 100644
> > > > >> --- a/drivers/net/xen-netback/rx.c
> > > > >> +++ b/drivers/net/xen-netback/rx.c
> > > > >> @@ -168,11 +168,14 @@ static void xenvif_rx_copy_add(struct xenvif_queue *queue,
> > > > >>  			       struct xen_netif_rx_request *req,
> > > > >>  			       unsigned int offset, void *data, size_t len)
> > > > >>  {
> > > > >> +	unsigned int batch_size;
> > > > >>  	struct gnttab_copy *op;
> > > > >>  	struct page *page;
> > > > >>  	struct xen_page_foreign *foreign;
> > > > >>
> > > > >> -	if (queue->rx_copy.num == COPY_BATCH_SIZE)
> > > > >> +	batch_size = min(xenvif_copy_batch_size, queue->rx_copy.size);
> > > > >
> > > > > Surely queue->rx_copy.size and xenvif_copy_batch_size are always
> > > > > identical? Why do you need this statement (and hence stack variable)?
> > > > >
> > > > This statement was to allow to be changed dynamically and would
> > > > affect all newly created guests or running guests if value happened
> > > > to be smaller than initially allocated. But I suppose I should make
> > > > behaviour more consistent with the other params we have right now
> > > > and just look at initially allocated one `queue->rx_copy.batch_size` ?
> > >
> > > Yes, that would certainly be consistent but I can see value in
> > > allowing it to be dynamically tuned, so perhaps adding some re-allocation
> > > code to allow the batch to be grown as well as shrunk might be nice.
> > 
> > The shrink one we potentially risk losing data, so we need to gate the
> > reallocation whenever `rx_copy.num` is less than the new requested
> > batch. Worst case means guestrx_thread simply uses the initial
> > allocated value.
> 
> Can't you just re-alloc immediately after the flush (when num is
> guaranteed to be zero)?

/facepalm

Yes, after the flush should make things much simpler.

Joao

Patch

diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h
index a165a4123396..8e4eaf3a507d 100644
--- a/drivers/net/xen-netback/common.h
+++ b/drivers/net/xen-netback/common.h
@@ -359,6 +359,7 @@  irqreturn_t xenvif_ctrl_irq_fn(int irq, void *data);
 
 void xenvif_rx_action(struct xenvif_queue *queue);
 void xenvif_rx_queue_tail(struct xenvif_queue *queue, struct sk_buff *skb);
+int xenvif_rx_copy_realloc(struct xenvif_queue *queue, unsigned int size);
 
 void xenvif_carrier_on(struct xenvif *vif);
 
diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index 1892bf9327e4..14613b5fcccb 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -516,20 +516,13 @@  struct xenvif *xenvif_alloc(struct device *parent, domid_t domid,
 
 int xenvif_init_queue(struct xenvif_queue *queue)
 {
-	unsigned int size = xenvif_copy_batch_size;
 	int err, i;
-	void *addr;
-
-	addr = vzalloc(size * sizeof(struct gnttab_copy));
-	if (!addr)
-		goto err;
-	queue->rx_copy.op = addr;
 
-	addr = vzalloc(size * sizeof(RING_IDX));
-	if (!addr)
+	err = xenvif_rx_copy_realloc(queue, xenvif_copy_batch_size);
+	if (err) {
+		netdev_err(queue->vif->dev, "Could not alloc rx_copy\n");
 		goto err;
-	queue->rx_copy.idx = addr;
-	queue->rx_copy.batch_size = size;
+	}
 
 	queue->credit_bytes = queue->remaining_credit = ~0UL;
 	queue->credit_usec  = 0UL;
diff --git a/drivers/net/xen-netback/rx.c b/drivers/net/xen-netback/rx.c
index be3946cdaaf6..f54bfe72188c 100644
--- a/drivers/net/xen-netback/rx.c
+++ b/drivers/net/xen-netback/rx.c
@@ -130,6 +130,51 @@  static void xenvif_rx_queue_drop_expired(struct xenvif_queue *queue)
 	}
 }
 
+int xenvif_rx_copy_realloc(struct xenvif_queue *queue, unsigned int size)
+{
+	void *op = NULL, *idx = NULL;
+
+	/* No reallocation if new size doesn't fit ongoing requests */
+	if (!size || queue->rx_copy.num > size)
+		return -EINVAL;
+
+	op = vzalloc(size * sizeof(struct gnttab_copy));
+	if (!op)
+		goto err;
+
+	idx = vzalloc(size * sizeof(RING_IDX));
+	if (!idx)
+		goto err;
+
+	/* Ongoing requests need copying */
+	if (queue->rx_copy.num) {
+		unsigned int tmp;
+
+		tmp = queue->rx_copy.num * sizeof(struct gnttab_copy);
+		memcpy(op, queue->rx_copy.op, tmp);
+
+		tmp = queue->rx_copy.num * sizeof(RING_IDX);
+		memcpy(idx, queue->rx_copy.idx, tmp);
+	}
+
+	if (queue->rx_copy.op || queue->rx_copy.idx) {
+		vfree(queue->rx_copy.op);
+		vfree(queue->rx_copy.idx);
+	}
+
+	queue->rx_copy.op = op;
+	queue->rx_copy.idx = idx;
+	queue->rx_copy.batch_size = size;
+	netdev_dbg(queue->vif->dev, "Reallocated rx_copy for batch size %u\n",
+		   size);
+	return 0;
+
+err:
+	vfree(op);
+	vfree(idx);
+	return -ENOMEM;
+}
+
 static void xenvif_rx_copy_flush(struct xenvif_queue *queue)
 {
 	unsigned int i;
@@ -168,14 +213,14 @@  static void xenvif_rx_copy_add(struct xenvif_queue *queue,
 			       struct xen_netif_rx_request *req,
 			       unsigned int offset, void *data, size_t len)
 {
-	unsigned int batch_size;
 	struct gnttab_copy *op;
 	struct page *page;
 	struct xen_page_foreign *foreign;
 
-	batch_size = min(xenvif_copy_batch_size, queue->rx_copy.batch_size);
+	if (unlikely(xenvif_copy_batch_size != queue->rx_copy.batch_size))
+		xenvif_rx_copy_realloc(queue, xenvif_copy_batch_size);
 
-	if (queue->rx_copy.num == batch_size)
+	if (queue->rx_copy.num == queue->rx_copy.batch_size)
 		xenvif_rx_copy_flush(queue);
 
 	op = &queue->rx_copy.op[queue->rx_copy.num];