Re: [Qemu-devel] [PATCH] vhost-user: delay vhost_user_stop

[Marc-André Lureau]

Hi

On Sat, Feb 25, 2017 at 12:25 AM Marc-André Lureau <
address@hidden> wrote:

> Since commit b0a335e351103bf92f3f9d0bd5759311be8156ac, a socket write
> may trigger a disconnect events, calling vhost_user_stop() and clearing
> all the vhost_dev strutures holding data that vhost.c functions expect
> to remain valid. Delay the cleanup to keep the vhost_dev structure
> valid during the vhost.c functions.
>
> Signed-off-by: Marc-André Lureau <address@hidden>
> ---
>  net/vhost-user.c | 49 +++++++++++++++++++++++++++++++++++++++++++------
>  1 file changed, 43 insertions(+), 6 deletions(-)
>
> diff --git a/net/vhost-user.c b/net/vhost-user.c
> index 77b8110f8c..00573c8ac8 100644
> --- a/net/vhost-user.c
> +++ b/net/vhost-user.c
> @@ -25,6 +25,7 @@ typedef struct VhostUserState {
>      guint watch;
>      uint64_t acked_features;
>      bool started;
> +    QEMUBH *chr_closed_bh;
>  } VhostUserState;
>
>  VHostNetState *vhost_user_get_vhost_net(NetClientState *nc)
> @@ -190,9 +191,40 @@ static gboolean net_vhost_user_watch(GIOChannel
> *chan, GIOCondition cond,
>
>      qemu_chr_fe_disconnect(&s->chr);
>
> +    s->watch = 0;
>      return FALSE;
>  }
>
> +static void chr_closed_bh(void *opaque)
> +{
> +    const char *name = opaque;
> +    NetClientState *ncs[MAX_QUEUE_NUM];
> +    VhostUserState *s;
> +    Error *err = NULL;
> +    int queues;
> +
> +    queues = qemu_find_net_clients_except(name, ncs,
> +                                          NET_CLIENT_DRIVER_NIC,
> +                                          MAX_QUEUE_NUM);
> +    assert(queues < MAX_QUEUE_NUM);
> +
> +    s = DO_UPCAST(VhostUserState, nc, ncs[0]);
> +
> +    qmp_set_link(name, false, &err);
> +    vhost_user_stop(queues, ncs);
> +    if (s->watch) {
> +        g_source_remove(s->watch);
> +    }
> +    s->watch = 0;
> +
> +    qemu_bh_delete(s->chr_closed_bh);
> +    s->chr_closed_bh = NULL;
> +
> +    if (err) {
> +        error_report_err(err);
> +    }
> +}
> +
>  static void net_vhost_user_event(void *opaque, int event)
>  {
>      const char *name = opaque;
> @@ -212,20 +244,25 @@ static void net_vhost_user_event(void *opaque, int
> event)
>      trace_vhost_user_event(chr->label, event);
>      switch (event) {
>      case CHR_EVENT_OPENED:
> -        s->watch = qemu_chr_fe_add_watch(&s->chr, G_IO_HUP,
> -                                         net_vhost_user_watch, s);
>          if (vhost_user_start(queues, ncs, &s->chr) < 0) {
>              qemu_chr_fe_disconnect(&s->chr);
>              return;
>          }
> +        s->watch = qemu_chr_fe_add_watch(&s->chr, G_IO_HUP,
> +                                         net_vhost_user_watch, s);
>          qmp_set_link(name, true, &err);
> +        s->chr_closed_bh = qemu_bh_new(chr_closed_bh, opaque);
>          s->started = true;
>          break;
>      case CHR_EVENT_CLOSED:
> -        qmp_set_link(name, false, &err);
> -        vhost_user_stop(queues, ncs);
> -        g_source_remove(s->watch);
> -        s->watch = 0;
> +        /* a close event may happen during a read/write, but vhost
> +         * code assumes the vhost_dev remains setup, so delay the
> +         * stop & clear to idle.
> +         * FIXME: better handle failure in vhost code, remove bh
> +         */
> +        if (s->chr_closed_bh) {
> +            qemu_bh_schedule(s->chr_closed_bh);
>

And this is also racy if it gets a CHR_EVENT_OPENED before the bh is run.
/me not sure we want to go there.


> +        }
>          break;
>      }
>
> --
> 2.12.0.rc2.3.gc93709801
>
>
> --
Marc-André Lureau