similar to: [PATCH RFC] vhost: fix zcopy reference counting

This adds experimental zero copy support in vhost-net, disabled by default. To enable, set the zerocopytx module option to 1. This patch maintains the outstanding userspace buffers in the sequence it is delivered to vhost. The outstanding userspace buffers will be marked as done once the lower device buffers DMA has finished. This is monitored through last reference of kfree_skb callback. Two

This adds experimental zero copy support in vhost-net, disabled by default. To enable, set the zerocopytx module option to 1. This patch maintains the outstanding userspace buffers in the sequence it is delivered to vhost. The outstanding userspace buffers will be marked as done once the lower device buffers DMA has finished. This is monitored through last reference of kfree_skb callback. Two

>From: Shirley Ma <mashirle at us.ibm.com> This adds experimental zero copy support in vhost-net, disabled by default. To enable, set experimental_zcopytx module option to 1. This patch maintains the outstanding userspace buffers in the sequence it is delivered to vhost. The outstanding userspace buffers will be marked as done once the lower device buffers DMA has finished. This is

>From: Shirley Ma <mashirle at us.ibm.com> This adds experimental zero copy support in vhost-net, disabled by default. To enable, set experimental_zcopytx module option to 1. This patch maintains the outstanding userspace buffers in the sequence it is delivered to vhost. The outstanding userspace buffers will be marked as done once the lower device buffers DMA has finished. This is

From: Shirley Ma <mashirle at us.ibm.com> This adds experimental zero copy support in vhost-net, disabled by default. To enable, set the zerocopytx module option to 1. This patch maintains the outstanding userspace buffers in the sequence it is delivered to vhost. The outstanding userspace buffers will be marked as done once the lower device buffers DMA has finished. This is monitored

From: Shirley Ma <mashirle at us.ibm.com> This adds experimental zero copy support in vhost-net, disabled by default. To enable, set the zerocopytx module option to 1. This patch maintains the outstanding userspace buffers in the sequence it is delivered to vhost. The outstanding userspace buffers will be marked as done once the lower device buffers DMA has finished. This is monitored

On top of 'vhost: Allow device specific fields per vq', we can move device specific fields to device virt queue from vhost virt queue. Signed-off-by: Asias He <asias at redhat.com> --- drivers/vhost/net.c | 164 +++++++++++++++++++++++++++++++++++++++++++------- drivers/vhost/vhost.c | 57 +----------------- drivers/vhost/vhost.h | 22 ------- 3 files changed, 142

On top of 'vhost: Allow device specific fields per vq', we can move device specific fields to device virt queue from vhost virt queue. Signed-off-by: Asias He <asias at redhat.com> --- drivers/vhost/net.c | 164 +++++++++++++++++++++++++++++++++++++++++++------- drivers/vhost/vhost.c | 57 +----------------- drivers/vhost/vhost.h | 22 ------- 3 files changed, 142

vhost_net_ubuf_put_and_wait has a confusing name: it will actually also free it's argument. Thus since commit 1280c27f8e29acf4af2da914e80ec27c3dbd5c01 vhost_net_flush tries to use the argument after passing it to vhost_net_ubuf_put_and_wait, this results in use after free. To fix, don't free the argument in vhost_net_ubuf_put_and_wait, add an new API for callers that want to free ubufs.

vhost_net_ubuf_put_and_wait has a confusing name: it will actually also free it's argument. Thus since commit 1280c27f8e29acf4af2da914e80ec27c3dbd5c01 vhost_net_flush tries to use the argument after passing it to vhost_net_ubuf_put_and_wait, this results in use after free. To fix, don't free the argument in vhost_net_ubuf_put_and_wait, add an new API for callers that want to free ubufs.

Currently, vhost-net is the only consumer of vhost infrastructure. So vhost infrastructure and vhost-net driver are in a single module. Separating this as a vhost.ko module and a vhost-net.ko module makes it is easier to share code with other vhost drivers, e.g. vhost-blk.ko, tcm-vhost.ko. Signed-off-by: Asias He <asias at redhat.com> --- drivers/vhost/Kconfig | 10 +++++++++-

Currently, vhost-net is the only consumer of vhost infrastructure. So vhost infrastructure and vhost-net driver are in a single module. Separating this as a vhost.ko module and a vhost-net.ko module makes it is easier to share code with other vhost drivers, e.g. vhost-blk.ko, tcm-vhost.ko. Signed-off-by: Asias He <asias at redhat.com> --- drivers/vhost/Kconfig | 10 +++++++++-

vhost checked the counter within the refcnt before decrementing. It really wanted to know that it is the one that has the last reference, as a way to batch freeing resources a bit more efficiently. Note: we only let refcount go to 0 on device release. This works well but we now access the ref counter twice so there's a race: all users might see a high count and decide to defer freeing

vhost checked the counter within the refcnt before decrementing. It really wanted to know that it is the one that has the last reference, as a way to batch freeing resources a bit more efficiently. Note: we only let refcount go to 0 on device release. This works well but we now access the ref counter twice so there's a race: all users might see a high count and decide to defer freeing

As we now only update used ring after enabling the backend, we can write flags with __put_user: as that's done on data path, it matters. Signed-off-by: Michael S. Tsirkin <mst at redhat.com> --- It seems an obvious optimization so I queued this up already. drivers/vhost/vhost.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/vhost/vhost.c

As we now only update used ring after enabling the backend, we can write flags with __put_user: as that's done on data path, it matters. Signed-off-by: Michael S. Tsirkin <mst at redhat.com> --- It seems an obvious optimization so I queued this up already. drivers/vhost/vhost.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/vhost/vhost.c

On 09/04/2013 07:59 PM, Michael S. Tsirkin wrote: > On Mon, Sep 02, 2013 at 04:40:59PM +0800, Jason Wang wrote: >> Currently, even if the packet length is smaller than VHOST_GOODCOPY_LEN, if >> upend_idx != done_idx we still set zcopy_used to true and rollback this choice >> later. This could be avoided by determining zerocopy once by checking all >> conditions at one

On Thu, Sep 05, 2013 at 10:54:44AM +0800, Jason Wang wrote: > On 09/04/2013 07:59 PM, Michael S. Tsirkin wrote: > > On Mon, Sep 02, 2013 at 04:40:59PM +0800, Jason Wang wrote: > >> Currently, even if the packet length is smaller than VHOST_GOODCOPY_LEN, if > >> upend_idx != done_idx we still set zcopy_used to true and rollback this choice > >> later. This could

On Thu, Sep 05, 2013 at 10:54:44AM +0800, Jason Wang wrote: > On 09/04/2013 07:59 PM, Michael S. Tsirkin wrote: > > On Mon, Sep 02, 2013 at 04:40:59PM +0800, Jason Wang wrote: > >> Currently, even if the packet length is smaller than VHOST_GOODCOPY_LEN, if > >> upend_idx != done_idx we still set zcopy_used to true and rollback this choice > >> later. This could

On Mon, Sep 02, 2013 at 04:40:59PM +0800, Jason Wang wrote: > Currently, even if the packet length is smaller than VHOST_GOODCOPY_LEN, if > upend_idx != done_idx we still set zcopy_used to true and rollback this choice > later. This could be avoided by determining zerocopy once by checking all > conditions at one time before. > > Signed-off-by: Jason Wang <jasowang at