qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Kevin Wolf <kwolf@redhat.com>
To: Raphael Norwitz <raphael.norwitz@nutanix.com>
Cc: "den-plotnikov@yandex-team.ru" <den-plotnikov@yandex-team.ru>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	"qemu-block@nongnu.org" <qemu-block@nongnu.org>,
	"mst@redhat.com" <mst@redhat.com>
Subject: Re: [PATCH v2 2/6] vhost-user-blk: Don't reconnect during initialisation
Date: Tue, 4 May 2021 11:10:47 +0200	[thread overview]
Message-ID: <YJEPlyyvDguh46Bb@merkur.fritz.box> (raw)
In-Reply-To: <20210503170107.GA21884@raphael-debian-dev>

Am 03.05.2021 um 19:01 hat Raphael Norwitz geschrieben:
> So we're not going with the suggestion to retry once or a fixed number
> of times? Any reason why not?

I thought we agreed that we'd add reconnection back in a follow-up
series that also addresses the different kinds of errors and retries
only when it makes sense?

Kevin

> On Thu, Apr 29, 2021 at 07:13:12PM +0200, Kevin Wolf wrote:
> > This is a partial revert of commits 77542d43149 and bc79c87bcde.
> > 
> > Usually, an error during initialisation means that the configuration was
> > wrong. Reconnecting won't make the error go away, but just turn the
> > error condition into an endless loop. Avoid this and return errors
> > again.
> > 
> > Additionally, calling vhost_user_blk_disconnect() from the chardev event
> > handler could result in use-after-free because none of the
> > initialisation code expects that the device could just go away in the
> > middle. So removing the call fixes crashes in several places.
> > 
> > For example, using a num-queues setting that is incompatible with the
> > backend would result in a crash like this (dereferencing dev->opaque,
> > which is already NULL):
> > 
> >  #0  0x0000555555d0a4bd in vhost_user_read_cb (source=0x5555568f4690, condition=(G_IO_IN | G_IO_HUP), opaque=0x7fffffffcbf0) at ../hw/virtio/vhost-user.c:313
> >  #1  0x0000555555d950d3 in qio_channel_fd_source_dispatch (source=0x555557c3f750, callback=0x555555d0a478 <vhost_user_read_cb>, user_data=0x7fffffffcbf0) at ../io/channel-watch.c:84
> >  #2  0x00007ffff7b32a9f in g_main_context_dispatch () at /lib64/libglib-2.0.so.0
> >  #3  0x00007ffff7b84a98 in g_main_context_iterate.constprop () at /lib64/libglib-2.0.so.0
> >  #4  0x00007ffff7b32163 in g_main_loop_run () at /lib64/libglib-2.0.so.0
> >  #5  0x0000555555d0a724 in vhost_user_read (dev=0x555557bc62f8, msg=0x7fffffffcc50) at ../hw/virtio/vhost-user.c:402
> >  #6  0x0000555555d0ee6b in vhost_user_get_config (dev=0x555557bc62f8, config=0x555557bc62ac "", config_len=60) at ../hw/virtio/vhost-user.c:2133
> >  #7  0x0000555555d56d46 in vhost_dev_get_config (hdev=0x555557bc62f8, config=0x555557bc62ac "", config_len=60) at ../hw/virtio/vhost.c:1566
> >  #8  0x0000555555cdd150 in vhost_user_blk_device_realize (dev=0x555557bc60b0, errp=0x7fffffffcf90) at ../hw/block/vhost-user-blk.c:510
> >  #9  0x0000555555d08f6d in virtio_device_realize (dev=0x555557bc60b0, errp=0x7fffffffcff0) at ../hw/virtio/virtio.c:3660
> > 
> > Signed-off-by: Kevin Wolf <kwolf@redhat.com>
> > ---
> >  hw/block/vhost-user-blk.c | 59 +++++++++++----------------------------
> >  1 file changed, 17 insertions(+), 42 deletions(-)
> > 
> > diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
> > index 7c85248a7b..c0b9958da1 100644
> > --- a/hw/block/vhost-user-blk.c
> > +++ b/hw/block/vhost-user-blk.c
> > @@ -50,6 +50,8 @@ static const int user_feature_bits[] = {
> >      VHOST_INVALID_FEATURE_BIT
> >  };
> >  
> > +static void vhost_user_blk_event(void *opaque, QEMUChrEvent event);
> > +
> >  static void vhost_user_blk_update_config(VirtIODevice *vdev, uint8_t *config)
> >  {
> >      VHostUserBlk *s = VHOST_USER_BLK(vdev);
> > @@ -362,19 +364,6 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
> >      vhost_dev_cleanup(&s->dev);
> >  }
> >  
> > -static void vhost_user_blk_event(void *opaque, QEMUChrEvent event,
> > -                                 bool realized);
> > -
> > -static void vhost_user_blk_event_realize(void *opaque, QEMUChrEvent event)
> > -{
> > -    vhost_user_blk_event(opaque, event, false);
> > -}
> > -
> > -static void vhost_user_blk_event_oper(void *opaque, QEMUChrEvent event)
> > -{
> > -    vhost_user_blk_event(opaque, event, true);
> > -}
> > -
> >  static void vhost_user_blk_chr_closed_bh(void *opaque)
> >  {
> >      DeviceState *dev = opaque;
> > @@ -382,12 +371,11 @@ static void vhost_user_blk_chr_closed_bh(void *opaque)
> >      VHostUserBlk *s = VHOST_USER_BLK(vdev);
> >  
> >      vhost_user_blk_disconnect(dev);
> > -    qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL,
> > -            vhost_user_blk_event_oper, NULL, opaque, NULL, true);
> > +    qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL, vhost_user_blk_event,
> > +                             NULL, opaque, NULL, true);
> >  }
> >  
> > -static void vhost_user_blk_event(void *opaque, QEMUChrEvent event,
> > -                                 bool realized)
> > +static void vhost_user_blk_event(void *opaque, QEMUChrEvent event)
> >  {
> >      DeviceState *dev = opaque;
> >      VirtIODevice *vdev = VIRTIO_DEVICE(dev);
> > @@ -401,17 +389,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent event,
> >          }
> >          break;
> >      case CHR_EVENT_CLOSED:
> > -        /*
> > -         * Closing the connection should happen differently on device
> > -         * initialization and operation stages.
> > -         * On initalization, we want to re-start vhost_dev initialization
> > -         * from the very beginning right away when the connection is closed,
> > -         * so we clean up vhost_dev on each connection closing.
> > -         * On operation, we want to postpone vhost_dev cleanup to let the
> > -         * other code perform its own cleanup sequence using vhost_dev data
> > -         * (e.g. vhost_dev_set_log).
> > -         */
> > -        if (realized && !runstate_check(RUN_STATE_SHUTDOWN)) {
> > +        if (!runstate_check(RUN_STATE_SHUTDOWN)) {
> >              /*
> >               * A close event may happen during a read/write, but vhost
> >               * code assumes the vhost_dev remains setup, so delay the
> > @@ -431,8 +409,6 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent event,
> >               * knowing its type (in this case vhost-user).
> >               */
> >              s->dev.started = false;
> > -        } else {
> > -            vhost_user_blk_disconnect(dev);
> >          }
> >          break;
> >      case CHR_EVENT_BREAK:
> > @@ -489,33 +465,32 @@ static void vhost_user_blk_device_realize(DeviceState *dev, Error **errp)
> >      s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
> >      s->connected = false;
> >  
> > -    qemu_chr_fe_set_handlers(&s->chardev,  NULL, NULL,
> > -                             vhost_user_blk_event_realize, NULL, (void *)dev,
> > -                             NULL, true);
> > -
> > -reconnect:
> >      if (qemu_chr_fe_wait_connected(&s->chardev, errp) < 0) {
> >          goto virtio_err;
> >      }
> >  
> > -    /* check whether vhost_user_blk_connect() failed or not */
> > -    if (!s->connected) {
> > -        goto reconnect;
> > +    if (vhost_user_blk_connect(dev) < 0) {
> > +        error_setg(errp, "vhost-user-blk: could not connect");
> > +        qemu_chr_fe_disconnect(&s->chardev);
> > +        goto virtio_err;
> >      }
> > +    assert(s->connected);
> >  
> >      ret = vhost_dev_get_config(&s->dev, (uint8_t *)&s->blkcfg,
> >                                 sizeof(struct virtio_blk_config));
> >      if (ret < 0) {
> > -        error_report("vhost-user-blk: get block config failed");
> > -        goto reconnect;
> > +        error_setg(errp, "vhost-user-blk: get block config failed");
> > +        goto vhost_err;
> >      }
> >  
> > -    /* we're fully initialized, now we can operate, so change the handler */
> > +    /* we're fully initialized, now we can operate, so add the handler */
> >      qemu_chr_fe_set_handlers(&s->chardev,  NULL, NULL,
> > -                             vhost_user_blk_event_oper, NULL, (void *)dev,
> > +                             vhost_user_blk_event, NULL, (void *)dev,
> >                               NULL, true);
> >      return;
> >  
> > +vhost_err:
> > +    vhost_dev_cleanup(&s->dev);
> >  virtio_err:
> >      g_free(s->vhost_vqs);
> >      s->vhost_vqs = NULL;
> > -- 
> > 2.30.2
> >
> 



  reply	other threads:[~2021-05-04  9:14 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-29 17:13 [PATCH v2 0/6] vhost-user-blk: Error handling fixes during initialistion Kevin Wolf
2021-04-29 17:13 ` [PATCH v2 1/6] vhost-user-blk: Make sure to set Error on realize failure Kevin Wolf
2021-05-03 17:12   ` Eric Blake
2021-05-03 17:24   ` Raphael Norwitz
2021-04-29 17:13 ` [PATCH v2 2/6] vhost-user-blk: Don't reconnect during initialisation Kevin Wolf
2021-05-03 17:01   ` Raphael Norwitz
2021-05-04  9:10     ` Kevin Wolf [this message]
2021-05-04  8:59   ` Michael S. Tsirkin
2021-05-04  9:27     ` Kevin Wolf
2021-05-04  9:44       ` Michael S. Tsirkin
2021-05-04 10:57         ` Kevin Wolf
2021-05-04 11:08           ` Michael S. Tsirkin
2021-04-29 17:13 ` [PATCH v2 3/6] vhost-user-blk: Improve error reporting in realize Kevin Wolf
2021-04-29 17:13 ` [PATCH v2 4/6] vhost-user-blk: Get more feature flags from vhost device Kevin Wolf
2021-04-29 17:13 ` [PATCH v2 5/6] virtio: Fail if iommu_platform is requested, but unsupported Kevin Wolf
2021-04-29 17:13 ` [PATCH v2 6/6] vhost-user-blk: Check that num-queues is supported by backend Kevin Wolf
2021-05-14 12:20 ` [PATCH v2 0/6] vhost-user-blk: Error handling fixes during initialistion Michael S. Tsirkin
2021-05-14 16:24   ` Kevin Wolf

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=YJEPlyyvDguh46Bb@merkur.fritz.box \
    --to=kwolf@redhat.com \
    --cc=den-plotnikov@yandex-team.ru \
    --cc=mst@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=raphael.norwitz@nutanix.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).