[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH 03/18] vhost_net: do not assume nvqs is always 2
From: |
Eli Cohen |
Subject: |
Re: [PATCH 03/18] vhost_net: do not assume nvqs is always 2 |
Date: |
Thu, 24 Jun 2021 09:22:14 +0300 |
User-agent: |
Mutt/1.9.5 (bf161cf53efb) (2018-04-13) |
On Mon, Jun 21, 2021 at 12:16:35PM +0800, Jason Wang wrote:
> This patch switches to initialize dev.nvqs from the VhostNetOptions
> instead of assuming it was 2. This is useful for implementing control
> virtqueue support which will be a single vhost_net structure with a
> single cvq.
Maybe worth mentioning in the changelog that nvqs is still set to 2 for
all users and this patch does not change functionality.
Reviewed-by: Eli Cohen <elic@nvidia.com>
> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
> hw/net/vhost_net.c | 2 +-
> include/net/vhost_net.h | 1 +
> net/tap.c | 1 +
> net/vhost-user.c | 1 +
> net/vhost-vdpa.c | 1 +
> 5 files changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/hw/net/vhost_net.c b/hw/net/vhost_net.c
> index 6bd4184f96..ef1370bd92 100644
> --- a/hw/net/vhost_net.c
> +++ b/hw/net/vhost_net.c
> @@ -163,9 +163,9 @@ struct vhost_net *vhost_net_init(VhostNetOptions *options)
> goto fail;
> }
> net->nc = options->net_backend;
> + net->dev.nvqs = options->nvqs;
>
> net->dev.max_queues = 1;
> - net->dev.nvqs = 2;
> net->dev.vqs = net->vqs;
>
> if (backend_kernel) {
> diff --git a/include/net/vhost_net.h b/include/net/vhost_net.h
> index 172b0051d8..fba40cf695 100644
> --- a/include/net/vhost_net.h
> +++ b/include/net/vhost_net.h
> @@ -14,6 +14,7 @@ typedef struct VhostNetOptions {
> VhostBackendType backend_type;
> NetClientState *net_backend;
> uint32_t busyloop_timeout;
> + unsigned int nvqs;
> void *opaque;
> } VhostNetOptions;
>
> diff --git a/net/tap.c b/net/tap.c
> index f5686bbf77..f716be3e3f 100644
> --- a/net/tap.c
> +++ b/net/tap.c
> @@ -749,6 +749,7 @@ static void net_init_tap_one(const NetdevTapOptions *tap,
> NetClientState *peer,
> qemu_set_nonblock(vhostfd);
> }
> options.opaque = (void *)(uintptr_t)vhostfd;
> + options.nvqs = 2;
>
> s->vhost_net = vhost_net_init(&options);
> if (!s->vhost_net) {
> diff --git a/net/vhost-user.c b/net/vhost-user.c
> index ffbd94d944..b93918c5a4 100644
> --- a/net/vhost-user.c
> +++ b/net/vhost-user.c
> @@ -85,6 +85,7 @@ static int vhost_user_start(int queues, NetClientState
> *ncs[],
> options.net_backend = ncs[i];
> options.opaque = be;
> options.busyloop_timeout = 0;
> + options.nvqs = 2;
> net = vhost_net_init(&options);
> if (!net) {
> error_report("failed to init vhost_net for queue %d", i);
> diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c
> index 19187dce8c..18b45ad777 100644
> --- a/net/vhost-vdpa.c
> +++ b/net/vhost-vdpa.c
> @@ -105,6 +105,7 @@ static int vhost_vdpa_add(NetClientState *ncs, void *be)
> options.net_backend = ncs;
> options.opaque = be;
> options.busyloop_timeout = 0;
> + options.nvqs = 2;
>
> net = vhost_net_init(&options);
> if (!net) {
> --
> 2.25.1
>
- [PATCH 00/18] vhost-vDPA multiqueue, Jason Wang, 2021/06/21
- [PATCH 01/18] vhost_net: remove the meaningless assignment in vhost_net_start_one(), Jason Wang, 2021/06/21
- [PATCH 02/18] vhost: use unsigned int for nvqs, Jason Wang, 2021/06/21
- [PATCH 03/18] vhost_net: do not assume nvqs is always 2, Jason Wang, 2021/06/21
- [PATCH 04/18] vhost-vdpa: remove the unnecessary check in vhost_vdpa_add(), Jason Wang, 2021/06/21
- [PATCH 05/18] vhost-vdpa: don't cleanup twice in vhost_vdpa_add(), Jason Wang, 2021/06/21
- [PATCH 06/18] vhost-vdpa: fix leaking of vhost_net in vhost_vdpa_add(), Jason Wang, 2021/06/21