linux-next: manual merge of the vhost tree with the net-next tree

linux-next: manual merge of the vhost tree with the net-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1535 bytes --]

Hi all,

Today's linux-next merge of the vhost tree got a conflict in:

  drivers/net/virtio_net.c

between commit:

  6b16f9ee89b8 ("net: move skb->xmit_more hint to softnet data")

from the net-next tree and commit:

  916faccfe56d ("virtio-net: bql support")

from the vhost tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

-- 
Cheers,
Stephen Rothwell

diff --cc drivers/net/virtio_net.c
index ba246fc475ae,955b3e76eb8d..000000000000
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@@ -1568,13 -1571,15 +1572,15 @@@ static netdev_tx_t start_xmit(struct sk
  	struct send_queue *sq = &vi->sq[qnum];
  	int err;
  	struct netdev_queue *txq = netdev_get_tx_queue(dev, qnum);
- 	bool kick = !netdev_xmit_more();
 -	bool more = skb->xmit_more;
++	bool more = netdev_xmit_more();
  	bool use_napi = sq->napi.weight;
+ 	unsigned int bytes = skb->len;
+ 	bool kick;
  
  	/* Free up any pending old buffers before queueing new ones. */
- 	free_old_xmit_skbs(sq, false);
+ 	free_old_xmit_skbs(sq, txq, false);
  
- 	if (use_napi && kick)
+ 	if (use_napi && !more)
  		virtqueue_enable_cb_delayed(sq->vq);
  
  	/* timestamp packet in software */

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

linux-next: manual merge of the vhost tree with the net-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1507 bytes --]

Hi all,

Today's linux-next merge of the vhost tree got a conflict in:

  drivers/net/virtio_net.c

between commit:

  308d7982dcdc ("virtio_net: extract interrupt coalescing settings to a structure")

from the net-next tree and commit:

  1a08d66726dc ("virtio_net: merge dma operations when filling mergeable buffers")

from the vhost tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

-- 
Cheers,
Stephen Rothwell

diff --cc drivers/net/virtio_net.c
index edfd5d8833e4,c9cab3f966db..000000000000
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@@ -127,11 -126,14 +127,19 @@@ static const struct virtnet_stat_desc v
  #define VIRTNET_SQ_STATS_LEN	ARRAY_SIZE(virtnet_sq_stats_desc)
  #define VIRTNET_RQ_STATS_LEN	ARRAY_SIZE(virtnet_rq_stats_desc)
  
 +struct virtnet_interrupt_coalesce {
 +	u32 max_packets;
 +	u32 max_usecs;
 +};
 +
+ /* The dma information of pages allocated at a time. */
+ struct virtnet_rq_dma {
+ 	dma_addr_t addr;
+ 	u32 ref;
+ 	u16 len;
+ 	u16 need_sync;
+ };
+ 
  /* Internal representation of a send virtqueue */
  struct send_queue {
  	/* Virtqueue associated with this send _queue */

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: manual merge of the vhost tree with the net-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 2080 bytes --]

Hi all,

On Fri, 23 Jun 2023 13:04:43 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> Today's linux-next merge of the vhost tree got a conflict in:
> 
>   drivers/net/virtio_net.c
> 
> between commit:
> 
>   80f50f918c6e ("virtio_net: separate the logic of freeing the rest mergeable buf")
> 
> from the net-next tree and commit:
> 
>   21081476b808 ("virtio_net: support dma premapped")

This is now commit

  c8dc5b9adff9 ("virtio_net: support dma premapped")

> from the vhost tree.
> 
> I fixed it up (see below) and can carry the fix as necessary. This
> is now fixed as far as linux-next is concerned, but any non trivial
> conflicts should be mentioned to your upstream maintainer when your tree
> is submitted for merging.  You may also want to consider cooperating
> with the maintainer of the conflicting tree to minimise any particularly
> complex conflicts.
> 
> diff --cc drivers/net/virtio_net.c
> index 0db14f6b87d3,d67b36fdba0d..000000000000
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@@ -1128,28 -1136,6 +1233,28 @@@ err
>   	return NULL;
>   }
>   
>  +static void mergeable_buf_free(struct receive_queue *rq, int num_buf,
>  +			       struct net_device *dev,
>  +			       struct virtnet_rq_stats *stats)
>  +{
>  +	struct page *page;
>  +	void *buf;
>  +	int len;
>  +
>  +	while (num_buf-- > 1) {
> - 		buf = virtqueue_get_buf(rq->vq, &len);
> ++		buf = virtnet_rq_get_buf(rq, &len, NULL);
>  +		if (unlikely(!buf)) {
>  +			pr_debug("%s: rx error: %d buffers missing\n",
>  +				 dev->name, num_buf);
>  +			dev->stats.rx_length_errors++;
>  +			break;
>  +		}
>  +		stats->bytes += len;
>  +		page = virt_to_head_page(buf);
>  +		put_page(page);
>  +	}
>  +}
>  +
>   /* Why not use xdp_build_skb_from_frame() ?
>    * XDP core assumes that xdp frags are PAGE_SIZE in length, while in
>    * virtio-net there are 2 points that do not match its requirements:

This is now a conflict between the vhost tree and Linus' tree.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

linux-next: manual merge of the vhost tree with the net-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1742 bytes --]

Hi all,

Today's linux-next merge of the vhost tree got a conflict in:

  drivers/net/virtio_net.c

between commit:

  80f50f918c6e ("virtio_net: separate the logic of freeing the rest mergeable buf")

from the net-next tree and commit:

  21081476b808 ("virtio_net: support dma premapped")

from the vhost tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

-- 
Cheers,
Stephen Rothwell

diff --cc drivers/net/virtio_net.c
index 0db14f6b87d3,d67b36fdba0d..000000000000
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@@ -1128,28 -1136,6 +1233,28 @@@ err
  	return NULL;
  }
  
 +static void mergeable_buf_free(struct receive_queue *rq, int num_buf,
 +			       struct net_device *dev,
 +			       struct virtnet_rq_stats *stats)
 +{
 +	struct page *page;
 +	void *buf;
 +	int len;
 +
 +	while (num_buf-- > 1) {
- 		buf = virtqueue_get_buf(rq->vq, &len);
++		buf = virtnet_rq_get_buf(rq, &len, NULL);
 +		if (unlikely(!buf)) {
 +			pr_debug("%s: rx error: %d buffers missing\n",
 +				 dev->name, num_buf);
 +			dev->stats.rx_length_errors++;
 +			break;
 +		}
 +		stats->bytes += len;
 +		page = virt_to_head_page(buf);
 +		put_page(page);
 +	}
 +}
 +
  /* Why not use xdp_build_skb_from_frame() ?
   * XDP core assumes that xdp frags are PAGE_SIZE in length, while in
   * virtio-net there are 2 points that do not match its requirements:

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 484 bytes --]

linux-next: manual merge of the vhost tree with the net-next tree

[Stephen Rothwell]

Hi Michael,

Today's linux-next merge of the vhost tree got a conflict in:

  tools/virtio/ringtest/Makefile

between commit:

  9fb6bc5b4a78 ("ptr_ring: ring test")

from the net-next tree and commit:

  139ab4d4e68b ("tools/virtio: add noring tool")

from the vhost tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

-- 
Cheers,
Stephen Rothwell

diff --cc tools/virtio/ringtest/Makefile
index 50e086c6a7b6,6173adae9f08..000000000000
--- a/tools/virtio/ringtest/Makefile
+++ b/tools/virtio/ringtest/Makefile
@@@ -1,6 -1,6 +1,6 @@@
  all:
  
- all: ring virtio_ring_0_9 virtio_ring_poll virtio_ring_inorder ptr_ring
 -all: ring virtio_ring_0_9 virtio_ring_poll virtio_ring_inorder noring
++all: ring virtio_ring_0_9 virtio_ring_poll virtio_ring_inorder ptr_ring noring
  
  CFLAGS += -Wall
  CFLAGS += -pthread -O2 -ggdb
@@@ -16,13 -15,13 +16,15 @@@ ring: ring.o main.
  virtio_ring_0_9: virtio_ring_0_9.o main.o
  virtio_ring_poll: virtio_ring_poll.o main.o
  virtio_ring_inorder: virtio_ring_inorder.o main.o
 +ptr_ring: ptr_ring.o main.o
+ noring: noring.o main.o
  clean:
  	-rm main.o
  	-rm ring.o ring
  	-rm virtio_ring_0_9.o virtio_ring_0_9
  	-rm virtio_ring_poll.o virtio_ring_poll
  	-rm virtio_ring_inorder.o virtio_ring_inorder
 +	-rm ptr_ring.o ptr_ring
+ 	-rm noring.o noring
  
  .PHONY: all clean

linux-next: manual merge of the vhost tree with the net-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 5412 bytes --]

Hi Michael,

Today's linux-next merge of the vhost tree got a conflict in
drivers/vhost/net.c between commit 70181d51209c ("vhost_net: remove tx
polling state") from the net-next tree and commits 11569c7119a6 ("vhost:
Allow device specific fields per vq"), 15a51679a0a1 ("vhost: Move
vhost-net zerocopy support fields to net.c") and 0cc4c2bd142b ("vhost:
move per-vq net specific fields out to net") from the vhost tree.

I fixed it up (I think - see below - there is probably a better way) and
can carry the fix as necessary (no action is required).

diff --cc drivers/vhost/net.c
index 87c216c,661d9be..0000000
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@@ -64,10 -64,46 +64,36 @@@ enum 
  	VHOST_NET_VQ_MAX = 2,
  };
  
 -enum vhost_net_poll_state {
 -	VHOST_NET_POLL_DISABLED = 0,
 -	VHOST_NET_POLL_STARTED = 1,
 -	VHOST_NET_POLL_STOPPED = 2,
 -};
 -
+ struct vhost_ubuf_ref {
+ 	struct kref kref;
+ 	wait_queue_head_t wait;
+ 	struct vhost_virtqueue *vq;
+ };
+ 
+ struct vhost_net_virtqueue {
+ 	struct vhost_virtqueue vq;
+ 	/* hdr is used to store the virtio header.
+ 	 * Since each iovec has >= 1 byte length, we never need more than
+ 	 * header length entries to store the header. */
+ 	struct iovec hdr[sizeof(struct virtio_net_hdr_mrg_rxbuf)];
+ 	size_t vhost_hlen;
+ 	size_t sock_hlen;
+ 	/* vhost zerocopy support fields below: */
+ 	/* last used idx for outstanding DMA zerocopy buffers */
+ 	int upend_idx;
+ 	/* first used idx for DMA done zerocopy buffers */
+ 	int done_idx;
+ 	/* an array of userspace buffers info */
+ 	struct ubuf_info *ubuf_info;
+ 	/* Reference counting for outstanding ubufs.
+ 	 * Protected by vq mutex. Writers must also take device mutex. */
+ 	struct vhost_ubuf_ref *ubufs;
+ };
+ 
  struct vhost_net {
  	struct vhost_dev dev;
- 	struct vhost_virtqueue vqs[VHOST_NET_VQ_MAX];
+ 	struct vhost_net_virtqueue vqs[VHOST_NET_VQ_MAX];
  	struct vhost_poll poll[VHOST_NET_VQ_MAX];
 -	/* Tells us whether we are polling a socket for TX.
 -	 * We only do this when socket buffer fills up.
 -	 * Protected by tx vq lock. */
 -	enum vhost_net_poll_state tx_poll_state;
  	/* Number of TX recently submitted.
  	 * Protected by tx vq lock. */
  	unsigned tx_packets;
@@@ -224,8 -369,18 +337,8 @@@ static void handle_tx(struct vhost_net 
  	mutex_lock(&vq->mutex);
  	vhost_disable_notify(&net->dev, vq);
  
- 	hdr_size = vq->vhost_hlen;
- 	zcopy = vq->ubufs;
 -	if (wmem < sock->sk->sk_sndbuf / 2)
 -		tx_poll_stop(net);
+ 	hdr_size = nvq->vhost_hlen;
+ 	zcopy = nvq->ubufs;
  
  	for (;;) {
  		/* Release DMAs done buffers first */
@@@ -246,11 -401,21 +359,12 @@@
  			/* If more outstanding DMAs, queue the work.
  			 * Handle upend_idx wrap around
  			 */
- 			num_pends = likely(vq->upend_idx >= vq->done_idx) ?
- 				    (vq->upend_idx - vq->done_idx) :
- 				    (vq->upend_idx + UIO_MAXIOV - vq->done_idx);
+ 			num_pends = likely(nvq->upend_idx >= nvq->done_idx) ?
+ 				    (nvq->upend_idx - nvq->done_idx) :
+ 				    (nvq->upend_idx + UIO_MAXIOV -
+ 				     nvq->done_idx);
 -			if (unlikely(num_pends > VHOST_MAX_PEND)) {
 -				tx_poll_start(net, sock);
 -				set_bit(SOCK_ASYNC_NOSPACE, &sock->flags);
 +			if (unlikely(num_pends > VHOST_MAX_PEND))
  				break;
 -			}
  			if (unlikely(vhost_enable_notify(&net->dev, vq))) {
  				vhost_disable_notify(&net->dev, vq);
  				continue;
@@@ -309,10 -474,12 +423,10 @@@
  			if (zcopy_used) {
  				if (ubufs)
  					vhost_ubuf_put(ubufs);
- 				vq->upend_idx = ((unsigned)vq->upend_idx - 1) %
- 					UIO_MAXIOV;
+ 				nvq->upend_idx = ((unsigned)nvq->upend_idx - 1)
+ 					% UIO_MAXIOV;
  			}
  			vhost_discard_vq_desc(vq, 1);
 -			if (err == -EAGAIN || err == -ENOBUFS)
 -				tx_poll_start(net, sock);
  			break;
  		}
  		if (err != len)
@@@ -584,24 -770,32 +716,31 @@@ static int vhost_net_open(struct inode 
  static void vhost_net_disable_vq(struct vhost_net *n,
  				 struct vhost_virtqueue *vq)
  {
- 	struct vhost_poll *poll = n->poll + (vq - n->vqs);
++	struct vhost_net_virtqueue *vnq;
++	struct vhost_poll *poll;
++
  	if (!vq->private_data)
  		return;
 -	if (vq == &n->vqs[VHOST_NET_VQ_TX].vq) {
 -		tx_poll_stop(n);
 -		n->tx_poll_state = VHOST_NET_POLL_DISABLED;
 -	} else
 -		vhost_poll_stop(n->poll + VHOST_NET_VQ_RX);
++	vnq = container_of(vq, struct vhost_net_virtqueue, vq);
++	poll = n->poll + (vnq - n->vqs);
 +	vhost_poll_stop(poll);
  }
  
  static int vhost_net_enable_vq(struct vhost_net *n,
  				struct vhost_virtqueue *vq)
  {
- 	struct vhost_poll *poll = n->poll + (vq - n->vqs);
++	struct vhost_net_virtqueue *vnq;
++	struct vhost_poll *poll;
  	struct socket *sock;
 -	int ret;
  
  	sock = rcu_dereference_protected(vq->private_data,
  					 lockdep_is_held(&vq->mutex));
  	if (!sock)
  		return 0;
 -	if (vq == &n->vqs[VHOST_NET_VQ_TX].vq) {
 -		n->tx_poll_state = VHOST_NET_POLL_STOPPED;
 -		ret = tx_poll_start(n, sock);
 -	} else
 -		ret = vhost_poll_start(n->poll + VHOST_NET_VQ_RX, sock->file);
++	vnq = container_of(vq, struct vhost_net_virtqueue, vq);
++	poll = n->poll + (vnq - n->vqs);
  
 -	return ret;
 +	return vhost_poll_start(poll, sock->file);
  }
  
  static struct socket *vhost_net_stop_vq(struct vhost_net *n,

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the vhost tree with the net-next tree

[Michael S. Tsirkin]

On Thu, Apr 19, 2012 at 03:35:47PM +1000, Stephen Rothwell wrote:
> Hi Michael,
> 
> Today's linux-next merge of the vhost tree got a conflict in
> drivers/net/virtio_net.c between commit 586d17c5a01b ("virtio-net: send
> gratuitous packets when needed") from the net-next tree and commit
> 66fdfb7bc203 ("virtio-net: remove useless disable on freeze") from the
> vhost tree.
> 
> Just context changes.  I fixed it up (I think - see below) and can carry
> the fix as necessary.

The fix looks right, thanks a lot.

> -- 
> Cheers,
> Stephen Rothwell                    sfr@canb.auug.org.au
> 
> diff --cc drivers/net/virtio_net.c
> index fa58c78,3e4767f..0000000
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@@ -1224,16 -1183,6 +1224,11 @@@ static int virtnet_freeze(struct virtio
>   {
>   	struct virtnet_info *vi = vdev->priv;
>   
>  +	/* Prevent config work handler from accessing the device */
>  +	mutex_lock(&vi->config_lock);
>  +	vi->config_enable = false;
>  +	mutex_unlock(&vi->config_lock);
>  +
> - 	virtqueue_disable_cb(vi->rvq);
> - 	virtqueue_disable_cb(vi->svq);
> - 	if (virtio_has_feature(vi->vdev, VIRTIO_NET_F_CTRL_VQ))
> - 		virtqueue_disable_cb(vi->cvq);
> - 
>   	netif_device_detach(vi->dev);
>   	cancel_delayed_work_sync(&vi->refill);
>   

linux-next: manual merge of the vhost tree with the net-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1154 bytes --]

Hi Michael,

Today's linux-next merge of the vhost tree got a conflict in
drivers/net/virtio_net.c between commit 586d17c5a01b ("virtio-net: send
gratuitous packets when needed") from the net-next tree and commit
66fdfb7bc203 ("virtio-net: remove useless disable on freeze") from the
vhost tree.

Just context changes.  I fixed it up (I think - see below) and can carry
the fix as necessary.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

diff --cc drivers/net/virtio_net.c
index fa58c78,3e4767f..0000000
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@@ -1224,16 -1183,6 +1224,11 @@@ static int virtnet_freeze(struct virtio
  {
  	struct virtnet_info *vi = vdev->priv;
  
 +	/* Prevent config work handler from accessing the device */
 +	mutex_lock(&vi->config_lock);
 +	vi->config_enable = false;
 +	mutex_unlock(&vi->config_lock);
 +
- 	virtqueue_disable_cb(vi->rvq);
- 	virtqueue_disable_cb(vi->svq);
- 	if (virtio_has_feature(vi->vdev, VIRTIO_NET_F_CTRL_VQ))
- 		virtqueue_disable_cb(vi->cvq);
- 
  	netif_device_detach(vi->dev);
  	cancel_delayed_work_sync(&vi->refill);
  

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]