Re: [PATCH spice-server 5/8] stream-device: Workaround Qemu bug closing device

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sun, Feb 18, 2018 at 06:58:11PM +0000, Frediano Ziglio wrote:
> Previous patch causes a bug into Qemu if the patch
> 46764fe09ca2e0f15c0981a672c166ed8cf57e72 ("virtio-serial: fix segfault
> on disconnect") is not include in that version of Qemu.
> This crash happens when device is closed inside a write operation.
> For SPICE character device a spice_server_char_device_wakeup is called
> to write data which handles both read and write pending operations.
> As we want to close the device but we can't do it inside the handler
> operation schedule a timer that will close the guest device outside
> this callback.
> The proper solution would be to patch Qemu but making sure this
> is not so easy so this workaround patch.
> Code is marked with some comments to remember to remove this
> hack in a safe future.
> 
> Signed-off-by: Frediano Ziglio <fziglio@xxxxxxxxxx>
> ---
>  server/stream-device.c | 34 ++++++++++++++++++++++++++++++----
>  1 file changed, 30 insertions(+), 4 deletions(-)
> 
> diff --git a/server/stream-device.c b/server/stream-device.c
> index 688a4e07..170c8637 100644
> --- a/server/stream-device.c
> +++ b/server/stream-device.c
> @@ -57,6 +57,7 @@ struct StreamDevice {
>      bool flow_stopped;
>      StreamChannel *stream_channel;
>      CursorChannel *cursor_channel;
> +    SpiceTimer *close_timer;
>  };
>  
>  struct StreamDeviceClass {
> @@ -68,6 +69,8 @@ static StreamDevice *stream_device_new(SpiceCharDeviceInstance *sin, RedsState *
>  
>  G_DEFINE_TYPE(StreamDevice, stream_device, RED_TYPE_CHAR_DEVICE)
>  
> +static void char_device_set_state(RedCharDevice *char_dev, int state);
> +
>  typedef bool StreamMsgHandler(StreamDevice *dev, SpiceCharDeviceInstance *sin)
>      SPICE_GNUC_WARN_UNUSED_RESULT;
>  
> @@ -77,6 +80,16 @@ static StreamMsgHandler handle_msg_format, handle_msg_data, handle_msg_cursor_se
>  static bool handle_msg_invalid(StreamDevice *dev, SpiceCharDeviceInstance *sin,
>                                 const char *error_msg) SPICE_GNUC_WARN_UNUSED_RESULT;
>  
> +static void
> +close_timer_func(void *opaque)
> +{
> +    StreamDevice *dev = opaque;
> +
> +    if (dev->opened && dev->has_error) {
> +        char_device_set_state(RED_CHAR_DEVICE(dev), 0);
> +    }
> +}
> +
>  static bool
>  stream_device_partial_read(StreamDevice *dev, SpiceCharDeviceInstance *sin)
>  {
> @@ -94,7 +107,17 @@ stream_device_partial_read(StreamDevice *dev, SpiceCharDeviceInstance *sin)
>          while (sif->read(sin, buf, sizeof(buf)) > 0) {
>              continue;
>          }
> -        sif->state(sin, 0);
> +
> +        // This code is a workaround for a Qemu bug, see patch
> +        // "stream-device: Workaround Qemu bug closing device"
> +        // as calling sif->state here can cause a crash schedule
> +        // a timer and do the call in it. Remove this code when
> +        // will be sure all Qemu versions has been patched.
> +        RedsState *reds = red_char_device_get_server(RED_CHAR_DEVICE(dev));
> +        if (!dev->close_timer) {
> +            dev->close_timer = reds_core_timer_add(reds, close_timer_func, dev);
> +        }
> +        reds_core_timer_start(reds, dev->close_timer, 0);
>          return false;
>      }
>  
> @@ -501,6 +524,9 @@ stream_device_dispose(GObject *object)
>  {
>      StreamDevice *dev = STREAM_DEVICE(object);
>  
> +    RedsState *reds = red_char_device_get_server(RED_CHAR_DEVICE(dev));
> +    reds_core_timer_remove(reds, dev->close_timer);
> +
>      if (dev->stream_channel) {
>          // close all current connections and drop the reference
>          red_channel_destroy(RED_CHANNEL(dev->stream_channel));
> @@ -562,7 +588,7 @@ reset_channels(StreamDevice *dev)
>  }
>  
>  static void
> -char_device_enable(RedCharDevice *char_dev)
> +char_device_set_state(RedCharDevice *char_dev, int state)
>  {
>      SpiceCharDeviceInstance *sin = NULL;
>      g_object_get(char_dev, "sin", &sin, NULL);
> @@ -570,7 +596,7 @@ char_device_enable(RedCharDevice *char_dev)
>  
>      SpiceCharDeviceInterface *sif = spice_char_device_get_interface(sin);
>      if (sif->state) {
> -        sif->state(sin, 1);
> +        sif->state(sin, state);
>      }
>  }
>  
> @@ -599,7 +625,7 @@ stream_device_port_event(RedCharDevice *char_dev, uint8_t event)
>      // this could prevent a possible race condition where we open the
>      // device from the guest and it take some time to enable causing
>      // temporary writing issues.
> -    char_device_enable(char_dev);
> +    char_device_set_state(char_dev, 1);

I would squash this part of the change in the previous commit.

Christophe

Attachment: signature.asc
Description: PGP signature

_______________________________________________
Spice-devel mailing list
Spice-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/spice-devel

[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]     [Monitors]