[RFC][PATCH v6 00/19] Provide a zero-copy method on KVM virtio-net.

From: xiaohui . xin
Date: Fri May 21 2010 - 05:28:55 EST

We provide an zero-copy method which driver side may get external
buffers to DMA. Here external means driver don't use kernel space
to allocate skb buffers. Currently the external buffer can be from
guest virtio-net driver.

The idea is simple, just to pin the guest VM user space and then
let host NIC driver has the chance to directly DMA to it.
The patches are based on vhost-net backend driver. We add a device
which provides proto_ops as sendmsg/recvmsg to vhost-net to
send/recv directly to/from the NIC driver. KVM guest who use the
vhost-net backend may bind any ethX interface in the host side to
get copyless data transfer thru guest virtio-net frontend.

patch 01-13: net core changes.
patch 14-18: new device as interface to mantpulate external buffers.
patch 19: for vhost-net.

The guest virtio-net driver submits multiple requests thru vhost-net
backend driver to the kernel. And the requests are queued and then
completed after corresponding actions in h/w are done.

For read, user space buffers are dispensed to NIC driver for rx when
a page constructor API is invoked. Means NICs can allocate user buffers
from a page constructor. We add a hook in netif_receive_skb() function
to intercept the incoming packets, and notify the zero-copy device.

For write, the zero-copy deivce may allocates a new host skb and puts
payload on the skb_shinfo(skb)->frags, and copied the header to skb->data.
The request remains pending until the skb is transmitted by h/w.

Here, we have ever considered 2 ways to utilize the page constructor
API to dispense the user buffers.

One: Modify __alloc_skb() function a bit, it can only allocate a
structure of sk_buff, and the data pointer is pointing to a
user buffer which is coming from a page constructor API.
Then the shinfo of the skb is also from guest.
When packet is received from hardware, the skb->data is filled
directly by h/w. What we have done is in this way.

Pros: We can avoid any copy here.
Cons: Guest virtio-net driver needs to allocate skb as almost
the same method with the host NIC drivers, say the size
of netdev_alloc_skb() and the same reserved space in the
head of skb. Many NIC drivers are the same with guest and
ok for this. But some lastest NIC drivers reserves special
room in skb head. To deal with it, we suggest to provide
a method in guest virtio-net driver to ask for parameter
we interest from the NIC driver when we know which device
we have bind to do zero-copy. Then we ask guest to do so.

Two: Modify driver to get user buffer allocated from a page constructor
API(to substitute alloc_page()), the user buffer are used as payload
buffers and filled by h/w directly when packet is received. Driver
should associate the pages with skb (skb_shinfo(skb)->frags). For
the head buffer side, let host allocates skb, and h/w fills it.
After that, the data filled in host skb header will be copied into
guest header buffer which is submitted together with the payload buffer.

Pros: We could less care the way how guest or host allocates their
Cons: We still need a bit copy here for the skb header.

We are not sure which way is the better here. This is the first thing we want
to get comments from the community. We wish the modification to the network
part will be generic which not used by vhost-net backend only, but a user
application may use it as well when the zero-copy device may provides async
read/write operations later.

We have got comments from Michael. And he said the first method will break
the compatiblity of virtio-net driver and may complicate the qemu live
migration. Currently, we tried to ignore the skb_reserve() if the device
is doing zero-copy. Then guest virtio-net driver wil not changed. So we now
continue to go with the first way.
But comments about the two ways are still appreicated.

We provide multiple submits and asynchronous notifiicaton to
vhost-net too.

Our goal is to improve the bandwidth and reduce the CPU usage.
Exact performance data will be provided later. But for simple
test with netperf, we found bindwidth up and CPU % up too,
but the bindwidth up ratio is much more than CPU % up ratio.

What we have not done yet:
packet split support
To support GRO
Performance tuning

what we have done in v1:
polish the RCU usage
deal with write logging in asynchroush mode in vhost
add notifier block for mp device
rename page_ctor to mp_port in netdevice.h to make it looks generic
add mp_dev_change_flags() for mp device to change NIC state
add CONIFG_VHOST_MPASSTHRU to limit the usage when module is not load
a small fix for missing dev_put when fail
using dynamic minor instead of static minor number
a __KERNEL__ protect to mp_get_sock()

what we have done in v2:

remove most of the RCU usage, since the ctor pointer is only
changed by BIND/UNBIND ioctl, and during that time, NIC will be
stopped to get good cleanup(all outstanding requests are finished),
so the ctor pointer cannot be raced into wrong situation.

Remove the struct vhost_notifier with struct kiocb.
Let vhost-net backend to alloc/free the kiocb and transfer them
via sendmsg/recvmsg.

use get_user_pages_fast() and set_page_dirty_lock() when read.

Add some comments for netdev_mp_port_prep() and handle_mpassthru().

what we have done in v3:
the async write logging is rewritten
a drafted synchronous write function for qemu live migration
a limit for locked pages from get_user_pages_fast() to prevent Dos

what we have done in v4:
add iocb completion callback from vhost-net to queue iocb in mp device
replace vq->receiver by mp_sock_data_ready()
remove stuff in mp device which access structures from vhost-net
modify skb_reserve() to ignore host NIC driver reserved space
rebase to the latest vhost tree
split large patches into small pieces, especially for net core part.

what we have done in v5:
address Arnd Bergmann's comments
-remove IFF_MPASSTHRU_EXCL flag in mp device
-remove mp_release ops
move dev_is_mpassthru() as inline func
fix a bug in memory relinquish
Apply to current git (2.6.34-rc6) tree.

what we have done in v6:
move create_iocb() out of page_dtor which may happen in interrupt context
-This remove the potential issues which lock called in interrupt context
make the cache used by mp, vhost as static, and created/destoryed during
modules init/exit functions.
-This makes multiple mp guest created at the same time.

using netperf with GSO/TSO disabled, 10G NIC,
disabled packet split mode, with raw socket case compared to vhost.

bindwidth will be from 1.1Gbps to 1.7Gbps
CPU % from 120%-140% to 140%-160%

We have retested the performance based on 2.6.34-rc6 in above situtation.
vhost 1.4Gbps 120% ~ 130%
vhost + zero-copy 2.7Gbps 160% ~ 180%
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/