search for: skb_array

...have ptr_ring_resize that gets an array of > > > >rings and resizes them both to same length? > > OK, here it is. Untested so far, and no skb wrapper. > > Pls let me know whether this is what you had in mind. > > Exactly what I want. > > Thanks Ok and this for skb_array --> skb_array: add wrappers for resizing Signed-off-by: Michael S. Tsirkin <mst at redhat.com> -- diff --git a/include/linux/skb_array.h b/include/linux/skb_array.h index c900708..7e01c1f 100644 --- a/include/linux/skb_array.h +++ b/include/linux/skb_array.h @@ -151,16 +151,24 @@ stati...

...have ptr_ring_resize that gets an array of > > > >rings and resizes them both to same length? > > OK, here it is. Untested so far, and no skb wrapper. > > Pls let me know whether this is what you had in mind. > > Exactly what I want. > > Thanks Ok and this for skb_array --> skb_array: add wrappers for resizing Signed-off-by: Michael S. Tsirkin <mst at redhat.com> -- diff --git a/include/linux/skb_array.h b/include/linux/skb_array.h index c900708..7e01c1f 100644 --- a/include/linux/skb_array.h +++ b/include/linux/skb_array.h @@ -151,16 +151,24 @@ stati...

Hi all: This series tries to implement rx batching for vhost-net. This is done by batching the dequeuing from skb_array which was exported by underlayer socket and pass the sbk back through msg_control to finish userspace copying. Tests shows at most 19% improvment on rx pps. Please review. Thanks Jason Wang (8): ptr_ring: introduce batch dequeuing skb_array: introduce batch dequeuing tun: export skb_array...

Hi all: This series tries to implement rx batching for vhost-net. This is done by batching the dequeuing from skb_array which was exported by underlayer socket and pass the sbk back through msg_control to finish userspace copying. Tests shows at most 19% improvment on rx pps. Please review. Thanks Jason Wang (8): ptr_ring: introduce batch dequeuing skb_array: introduce batch dequeuing tun: export skb_array...

On Fri, Jun 17, 2016 at 03:41:20AM +0300, Michael S. Tsirkin wrote: > Would it help to have ptr_ring_resize that gets an array of > rings and resizes them both to same length? OK, here it is. Untested so far, and no skb wrapper. Pls let me know whether this is what you had in mind. --> ptr_ring: support resizing multiple queues Signed-off-by: Michael S. Tsirkin <mst at

On Fri, Jun 17, 2016 at 03:41:20AM +0300, Michael S. Tsirkin wrote: > Would it help to have ptr_ring_resize that gets an array of > rings and resizes them both to same length? OK, here it is. Untested so far, and no skb wrapper. Pls let me know whether this is what you had in mind. --> ptr_ring: support resizing multiple queues Signed-off-by: Michael S. Tsirkin <mst at

...roduce a notifier which was triggered when tx_queue_len was changed for a netdev. - resize all queues during the tx_queue_len changing. Tests shows about 15% improvement on guest rx pps: Before: ~1300000pps After : ~1500000pps Changes from V2: - add multiple rings resizing support for ptr_ring/skb_array - add zero length ring support - introdce a NETDEV_CHANGE_TX_QUEUE_LEN - drop new flags Changes from V1: - switch to use skb array instead of a customized circular buffer - add non-blocking support - rename .peek to .peek_len - drop lockless peeking since test show very minor improvement Jason Wa...

...roduce a notifier which was triggered when tx_queue_len was changed for a netdev. - resize all queues during the tx_queue_len changing. Tests shows about 15% improvement on guest rx pps: Before: ~1300000pps After : ~1500000pps Changes from V2: - add multiple rings resizing support for ptr_ring/skb_array - add zero length ring support - introdce a NETDEV_CHANGE_TX_QUEUE_LEN - drop new flags Changes from V1: - switch to use skb array instead of a customized circular buffer - add non-blocking support - rename .peek to .peek_len - drop lockless peeking since test show very minor improvement Jason Wa...

...ize all queues during the tx_queue_len changing. Tests shows about 15% improvement on guest rx pps: Before: ~1300000pps After : ~1500000pps Changes from V3: - fix kbuild warnings - call NETDEV_CHANGE_TX_QUEUE_LEN on IFLA_TXQLEN Changes from V2: - add multiple rings resizing support for ptr_ring/skb_array - add zero length ring support - introdce a NETDEV_CHANGE_TX_QUEUE_LEN - drop new flags Changes from V1: - switch to use skb array instead of a customized circular buffer - add non-blocking support - rename .peek to .peek_len - drop lockless peeking since test show very minor improvement Jason Wa...

...ize all queues during the tx_queue_len changing. Tests shows about 15% improvement on guest rx pps: Before: ~1300000pps After : ~1500000pps Changes from V3: - fix kbuild warnings - call NETDEV_CHANGE_TX_QUEUE_LEN on IFLA_TXQLEN Changes from V2: - add multiple rings resizing support for ptr_ring/skb_array - add zero length ring support - introdce a NETDEV_CHANGE_TX_QUEUE_LEN - drop new flags Changes from V1: - switch to use skb array instead of a customized circular buffer - add non-blocking support - rename .peek to .peek_len - drop lockless peeking since test show very minor improvement Jason Wa...

...ngtest/noring.c @@ -0,0 +1,69 @@ +#define _GNU_SOURCE +#include "main.h" +#include <assert.h> + +/* stub implementation: useful for measuring overhead */ +void alloc_ring(void) +{ +} + +/* guest side */ +int add_inbuf(unsigned len, void *buf, void *datap) +{ + return 0; +} + +/* + * skb_array API provides no way for producer to find out whether a given + * buffer was consumed. Our tests merely require that a successful get_buf + * implies that add_inbuf succeed in the past, and that add_inbuf will succeed, + * fake it accordingly. + */ +void *get_buf(unsigned *lenp, void **bufp) +{ + r...

...ngtest/noring.c @@ -0,0 +1,69 @@ +#define _GNU_SOURCE +#include "main.h" +#include <assert.h> + +/* stub implementation: useful for measuring overhead */ +void alloc_ring(void) +{ +} + +/* guest side */ +int add_inbuf(unsigned len, void *buf, void *datap) +{ + return 0; +} + +/* + * skb_array API provides no way for producer to find out whether a given + * buffer was consumed. Our tests merely require that a successful get_buf + * implies that add_inbuf succeed in the past, and that add_inbuf will succeed, + * fake it accordingly. + */ +void *get_buf(unsigned *lenp, void **bufp) +{ + r...

On Tue, Mar 21, 2017 at 12:04:46PM +0800, Jason Wang wrote: > We used to dequeue one skb during recvmsg() from skb_array, this could > be inefficient because of the bad cache utilization and spinlock > touching for each packet. This patch tries to batch them by calling > batch dequeuing helpers explicitly on the exported skb array and pass > the skb back through msg_control for underlayer socket to finish...

On Tue, Mar 21, 2017 at 12:04:46PM +0800, Jason Wang wrote: > We used to dequeue one skb during recvmsg() from skb_array, this could > be inefficient because of the bad cache utilization and spinlock > touching for each packet. This patch tries to batch them by calling > batch dequeuing helpers explicitly on the exported skb array and pass > the skb back through msg_control for underlayer socket to finish...

...After : ~1500000pps Acked-by: Michael S. Tsirkin <mst at redhat.com> Acked-from-altitude: 34697 feet. > Changes from V3: > - fix kbuild warnings > - call NETDEV_CHANGE_TX_QUEUE_LEN on IFLA_TXQLEN > > Changes from V2: > - add multiple rings resizing support for ptr_ring/skb_array > - add zero length ring support > - introdce a NETDEV_CHANGE_TX_QUEUE_LEN > - drop new flags > > Changes from V1: > - switch to use skb array instead of a customized circular buffer > - add non-blocking support > - rename .peek to .peek_len > - drop lockless peeking sin...

We used to dequeue one skb during recvmsg() from skb_array, this could be inefficient because of the bad cache utilization and spinlock touching for each packet. This patch tries to batch them by calling batch dequeuing helpers explicitly on the exported skb array and pass the skb back through msg_control for underlayer socket to finish the userspace copyi...

On 2017?03?22? 22:16, Michael S. Tsirkin wrote: > On Tue, Mar 21, 2017 at 12:04:46PM +0800, Jason Wang wrote: >> We used to dequeue one skb during recvmsg() from skb_array, this could >> be inefficient because of the bad cache utilization and spinlock >> touching for each packet. This patch tries to batch them by calling >> batch dequeuing helpers explicitly on the exported skb array and pass >> the skb back through msg_control for underlayer...

...} > > > -static int sk_has_rx_data(struct sock *sk) > > > +static int sk_has_rx_data(struct vhost_net_virtqueue *rvq, struct sock *sk) > > > { > > > struct socket *sock = sk->sk_socket; > > > + if (rvq->rx_array) > > > + return !__skb_array_empty(rvq->rx_array); > > > + > > > if (sock->ops->peek_len) > > > return sock->ops->peek_len(sock); > > I don't see which patch adds __skb_array_empty. > > This is not something new, it was introduced by ad69f35d1dc0a ("skb_arr...

...} > > > -static int sk_has_rx_data(struct sock *sk) > > > +static int sk_has_rx_data(struct vhost_net_virtqueue *rvq, struct sock *sk) > > > { > > > struct socket *sock = sk->sk_socket; > > > + if (rvq->rx_array) > > > + return !__skb_array_empty(rvq->rx_array); > > > + > > > if (sock->ops->peek_len) > > > return sock->ops->peek_len(sock); > > I don't see which patch adds __skb_array_empty. > > This is not something new, it was introduced by ad69f35d1dc0a ("skb_arr...

We used to queue tx packets in sk_receive_queue, this is less efficient since it requires spinlocks to synchronize between producer and consumer. This patch tries to address this by: - switch from sk_receive_queue to a skb_array, and resize it when tx_queue_len was changed. - introduce a new proto_ops peek_len which was used for peeking the skb length. - implement a tun version of peek_len for vhost_net to use and convert vhost_net to use peek_len if possible. Pktgen test shows about 15.3% improvement on guest recei...