similar to: [PATCH V3 0/2] vhost_net polling optimization

Hi: This series tries to optimize vhost_net polling at two points: - Stop rx polling for reduicng the unnecessary wakeups during handle_rx(). - Conditonally enable tx polling for reducing the unnecessary traversing and spinlock touching. Test shows about 17% improvement on rx pps. Please review Changes from V1: - use vhost_net_disable_vq()/vhost_net_enable_vq() instead of open coding. -

Hi: This series tries to optimize vhost_net polling at two points: - Stop rx polling for reduicng the unnecessary wakeups during handle_rx(). - Conditonally enable tx polling for reducing the unnecessary traversing and spinlock touching. Test shows about 17% improvement on rx pps. Please review Changes from V1: - use vhost_net_disable_vq()/vhost_net_enable_vq() instead of open coding. -

On Mon, May 30, 2016 at 02:47:53AM -0400, Jason Wang wrote: > We don't stop rx polling socket during rx processing, this will lead > unnecessary wakeups from under layer net devices (E.g > sock_def_readable() form tun). Rx will be slowed down in this > way. This patch avoids this by stop polling socket during rx > processing. A small drawback is that this introduces some

On Mon, May 30, 2016 at 02:47:53AM -0400, Jason Wang wrote: > We don't stop rx polling socket during rx processing, this will lead > unnecessary wakeups from under layer net devices (E.g > sock_def_readable() form tun). Rx will be slowed down in this > way. This patch avoids this by stop polling socket during rx > processing. A small drawback is that this introduces some

Under heavy load vhost tx busypoll tend not to suppress vq kicks, which causes poor guest tx performance. The detailed scenario is described in commitlog of patch 2. Rx seems not to have that serious problem, but for consistency I made a similar change on rx to avoid rx wakeups (patch 3). Additionary patch 4 is to avoid rx kicks under heavy load during busypoll. Tx performance is greatly improved

Hi: This series implement batch updating of used ring for TX. This help to reduce the cache contention on used ring. The idea is first split datacopy path from zerocopy, and do only batching for datacopy. This is because zercopy had already supported its own batching. TX PPS was increased 25.8% and Netperf TCP does not show obvious differences. The split of datapath will also be helpful for

Hi: This series implement batch updating of used ring for TX. This help to reduce the cache contention on used ring. The idea is first split datacopy path from zerocopy, and do only batching for datacopy. This is because zercopy had already supported its own batching. TX PPS was increased 25.8% and Netperf TCP does not show obvious differences. The split of datapath will also be helpful for

After commit e2b3b35eb989 ("vhost_net: batch used ring update in rx"), we tend to batch updating used heads. But it doesn't flush batched heads before trying to do busy polling, this will cause vhost to wait for guest TX which waits for the used RX. Fixing by flush batched heads before busy loop. 1 byte TCP_RR performance recovers from 13107.83 to 50402.65. Fixes: e2b3b35eb989

Under heavy load vhost busypoll may run without suppressing notification. For example tx zerocopy callback can push tx work while handle_tx() is running, then busyloop exits due to vhost_has_work() condition and enables notification but immediately reenters handle_tx() because the pushed work was tx. In this case handle_tx() tries to disable notification again, but when using event_idx it by

Under heavy load vhost busypoll may run without suppressing notification. For example tx zerocopy callback can push tx work while handle_tx() is running, then busyloop exits due to vhost_has_work() condition and enables notification but immediately reenters handle_tx() because the pushed work was tx. In this case handle_tx() tries to disable notification again, but when using event_idx it by

On 2020/6/2 ??9:06, Michael S. Tsirkin wrote: > Convert vhost net to use the new format-agnostic API. > In particular, don't poke at vq internals such as the > heads array. > > Signed-off-by: Michael S. Tsirkin <mst at redhat.com> > --- > drivers/vhost/net.c | 153 +++++++++++++++++++++++--------------------- > 1 file changed, 81 insertions(+), 72 deletions(-)

When the rx buffer is too small for a packet, we will discard the vq descriptor and retry it for the next packet: while ((sock_len = vhost_net_rx_peek_head_len(net, sock->sk, &busyloop_intr))) { ... /* On overrun, truncate and discard */ if (unlikely(headcount > UIO_MAXIOV)) { iov_iter_init(&msg.msg_iter, READ, vq->iov, 1, 1); err = sock->ops->recvmsg(sock,

When the rx buffer is too small for a packet, we will discard the vq descriptor and retry it for the next packet: while ((sock_len = vhost_net_rx_peek_head_len(net, sock->sk, &busyloop_intr))) { ... /* On overrun, truncate and discard */ if (unlikely(headcount > UIO_MAXIOV)) { iov_iter_init(&msg.msg_iter, READ, vq->iov, 1, 1); err = sock->ops->recvmsg(sock,

On 2019/4/26 ??1:52, Michael S. Tsirkin wrote: > On Thu, Apr 25, 2019 at 03:33:19AM -0400, Jason Wang wrote: >> When the rx buffer is too small for a packet, we will discard the vq >> descriptor and retry it for the next packet: >> >> while ((sock_len = vhost_net_rx_peek_head_len(net, sock->sk, >> &busyloop_intr))) { >> ... >> /* On

On 2019/4/26 ??1:52, Michael S. Tsirkin wrote: > On Thu, Apr 25, 2019 at 03:33:19AM -0400, Jason Wang wrote: >> When the rx buffer is too small for a packet, we will discard the vq >> descriptor and retry it for the next packet: >> >> while ((sock_len = vhost_net_rx_peek_head_len(net, sock->sk, >> &busyloop_intr))) { >> ... >> /* On

Asias He (4): vhost-net: Always access vq->private_data under vq mutex vhost-test: Always access vq->private_data under vq mutex vhost-scsi: Always access vq->private_data under vq mutex vhost: Remove custom vhost rcu usage drivers/vhost/net.c | 37 ++++++++++++++++--------------------- drivers/vhost/scsi.c | 17 ++++++----------- drivers/vhost/test.c | 20

Asias He (4): vhost-net: Always access vq->private_data under vq mutex vhost-test: Always access vq->private_data under vq mutex vhost-scsi: Always access vq->private_data under vq mutex vhost: Remove custom vhost rcu usage drivers/vhost/net.c | 37 ++++++++++++++++--------------------- drivers/vhost/scsi.c | 17 ++++++----------- drivers/vhost/test.c | 20

On Wed, Jun 01, 2016 at 01:56:34AM -0400, Jason Wang wrote: > We always poll tx for socket, this is sub optimal since: > > - it will be only used when we exceed the sndbuf of the socket. > - since we use two independent polls for tx and vq, this will slightly > increase the waitqueue traversing time and more important, vhost > could not benefit from commit >

On Wed, Jun 01, 2016 at 01:56:34AM -0400, Jason Wang wrote: > We always poll tx for socket, this is sub optimal since: > > - it will be only used when we exceed the sndbuf of the socket. > - since we use two independent polls for tx and vq, this will slightly > increase the waitqueue traversing time and more important, vhost > could not benefit from commit >

On Sun, May 05, 2019 at 12:20:24PM +0800, Jason Wang wrote: > > On 2019/4/26 ??3:35, Jason Wang wrote: > > > > On 2019/4/26 ??1:52, Michael S. Tsirkin wrote: > > > On Thu, Apr 25, 2019 at 03:33:19AM -0400, Jason Wang wrote: > > > > When the rx buffer is too small for a packet, we will discard the vq > > > > descriptor and retry it for the next