similar to: [PATCH RFC v8 00/11] vhost: ring format independence

This intentionally leaves "fixup" changes separate - hopefully that is enough to fix vhost-net crashes reported here, but it helps me keep track of what changed. I will naturally squash them later when we are done. This adds infrastructure required for supporting multiple ring formats. The idea is as follows: we convert descriptors to an independent format first, and process that

This intentionally leaves "fixup" changes separate - hopefully that is enough to fix vhost-net crashes reported here, but it helps me keep track of what changed. I will naturally squash them later when we are done. This adds infrastructure required for supporting multiple ring formats. The idea is as follows: we convert descriptors to an independent format first, and process that

We let the specifics of the ring format seep through to vhost API callers - mostly because there was only one format so it was hard to imagine what an independent API would look like. Now that there's an alternative in form of the packed ring, it's easier to see the issues, and fixing them is perhaps the cleanest way to add support for more formats. This patchset does this by indtroducing

This adds infrastructure required for supporting multiple ring formats. The idea is as follows: we convert descriptors to an independent format first, and process that converting to iov later. Used ring is similar: we fetch into an independent struct first, convert that to IOV later. The point is that we have a tight loop that fetches descriptors, which is good for cache utilization. This will

This adds infrastructure required for supporting multiple ring formats. The idea is as follows: we convert descriptors to an independent format first, and process that converting to iov later. Used ring is similar: we fetch into an independent struct first, convert that to IOV later. The point is that we have a tight loop that fetches descriptors, which is good for cache utilization. This will

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(-)

On 2020/6/2 ??9:06, Michael S. Tsirkin wrote: > Add a new API that doesn't assume used ring, heads, etc. > For now, we keep the old APIs around to make it easier > to convert drivers. > > Signed-off-by: Michael S. Tsirkin <mst at redhat.com> > --- > drivers/vhost/vhost.c | 52 ++++++++++++++++++++++++++++++++++--------- > drivers/vhost/vhost.h | 17

On 2020/6/2 ??9:06, Michael S. Tsirkin wrote: > Add a new API that doesn't assume used ring, heads, etc. > For now, we keep the old APIs around to make it easier > to convert drivers. > > Signed-off-by: Michael S. Tsirkin <mst at redhat.com> > --- > drivers/vhost/vhost.c | 52 ++++++++++++++++++++++++++++++++++--------- > drivers/vhost/vhost.h | 17

Hi all: This RFC implement packed ring layout. The code were tested with pmd implement by Jens at http://dpdk.org/ml/archives/dev/2018-January/089417.html. Minor change was needed for pmd codes to kick virtqueue since it assumes a busy polling backend. Test were done between localhost and guest. Testpmd (rxonly) in guest reports 2.4Mpps. Testpmd (txonly) repots about 2.1Mpps. Notes: The event

Hi all: This RFC implement packed ring layout. The code were tested with pmd implement by Jens at http://dpdk.org/ml/archives/dev/2018-January/089417.html. Minor change was needed for pmd codes to kick virtqueue since it assumes a busy polling backend. Test were done between localhost and guest. Testpmd (rxonly) in guest reports 2.4Mpps. Testpmd (txonly) repots about 2.1Mpps. Notes: The event

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 V2: - Don't enable rx vq if we meet an err or rx vq is empty Changes from V1:

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 V2: - Don't enable rx vq if we meet an err or rx vq is empty Changes from V1:

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. -

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

This adds infrastructure required for supporting multiple ring formats. The idea is as follows: we convert descriptors to an independent format first, and process that converting to iov later. The point is that we have a tight loop that fetches descriptors, which is good for cache utilization. This will also allow all kind of batching tricks - e.g. it seems possible to keep SMAP disabled while

Hi all: We do not support XDP batching for TUN since it can only receive one packet a time from vhost_net. This series tries to remove this limitation by: - introduce a TUN specific msg_control that can hold a pointer to an array of XDP buffs - try copy and build XDP buff in vhost_net - store XDP buffs in an array and submit them once for every N packets from vhost_net - since TUN can only

On Thu, Sep 06, 2018 at 12:05:26PM +0800, Jason Wang wrote: > This patch implements XDP batching for vhost_net. The idea is first to > try to do userspace copy and build XDP buff directly in vhost. Instead > of submitting the packet immediately, vhost_net will batch them in an > array and submit every 64 (VHOST_NET_BATCH) packets to the under layer > sockets through msg_control of

On Thu, Sep 06, 2018 at 12:05:26PM +0800, Jason Wang wrote: > This patch implements XDP batching for vhost_net. The idea is first to > try to do userspace copy and build XDP buff directly in vhost. Instead > of submitting the packet immediately, vhost_net will batch them in an > array and submit every 64 (VHOST_NET_BATCH) packets to the under layer > sockets through msg_control of