[Spice-devel] [PATCH spice-server v2 08/12] cursor-channel: Update some declarations and documentation

[Jonathon Jongsma]

This also looks like a patch that should get squashed (maybe parts
should be squashed into two different previous commits, see below), but
I agree with all of the changes.

Acked-by: Jonathon Jongsma <jjongsma at redhat.com>


On Tue, 2019-03-26 at 19:10 +0000, Frediano Ziglio wrote:
> Signed-off-by: Frediano Ziglio <fziglio at redhat.com>
> ---
>  server/cursor-channel-client.h |  6 ++++++
>  server/cursor-channel.c        |  2 --
>  server/cursor-channel.h        | 17 ++---------------
>  3 files changed, 8 insertions(+), 17 deletions(-)
> 
> diff --git a/server/cursor-channel-client.h b/server/cursor-channel-
> client.h
> index 56b3b312..4deae535 100644
> --- a/server/cursor-channel-client.h
> +++ b/server/cursor-channel-client.h
> @@ -75,6 +75,12 @@ enum {
>      RED_PIPE_ITEM_TYPE_INVAL_CURSOR_CACHE,
>  };
>  
> +/**
> + * Migrate a client channel from a CursorChannel.
> + * This is the equivalent of RedChannel client migrate callback.
> + */
> +void                 cursor_channel_client_migrate(RedChannelClient
> *client);
> +
>  G_END_DECLS

It seems that this hunk and the last should get squashed with patch
06/12


>  
>  #endif /* CURSOR_CHANNEL_CLIENT_H_ */
> diff --git a/server/cursor-channel.c b/server/cursor-channel.c
> index c88e5cd1..d936b791 100644
> --- a/server/cursor-channel.c
> +++ b/server/cursor-channel.c
> @@ -337,8 +337,6 @@ void cursor_channel_set_mouse_mode(CursorChannel
> *cursor, uint32_t mode)
>  
>  /**
>   * Connect a new client to CursorChannel.
> - * This is the equivalent of RedChannel client connect callback.
> - * See comment on cursor_channel_new.
>   */
>  static void
>  cursor_channel_connect(CursorChannel *cursor, RedClient *client,
> RedStream *stream,
> diff --git a/server/cursor-channel.h b/server/cursor-channel.h
> index ce1b92cc..dc48279a 100644
> --- a/server/cursor-channel.h
> +++ b/server/cursor-channel.h
> @@ -48,14 +48,8 @@ GType cursor_channel_get_type(void) G_GNUC_CONST;
>  /**
>   * Create CursorChannel.
>   * Since CursorChannel is intended to be run in a separate thread,
> - * it does not register its own client callbacks since they would
> - * be called from a different thread. Therefore users of this
> - * class are responsible for registering their own client callbacks
> - * for CursorChannel. These 'wrapper' client callbacks must forward
> - * execution on to the CursorChannel thread.
> - * cursor_channel_client_migrate() and cursor_channel_connect() are
> - * provided as helper functions and should only be called from the
> - * CursorChannel thread.
> + * the function accepts a dispatcher parameter to allows some
> + * operations to be executed in the channel thread.
>   */
>  CursorChannel* cursor_channel_new(RedsState *server, int id,
>                                    const SpiceCoreInterfaceInternal
> *core,

I think these hunks should be squashed into patch 02/12


> @@ -66,13 +60,6 @@
> void                 cursor_channel_do_init     (CursorChannel
> *cursor);
>  void                 cursor_channel_process_cmd (CursorChannel
> *cursor, RedCursorCmd *cursor_cmd);
>  void                 cursor_channel_set_mouse_mode(CursorChannel
> *cursor, uint32_t mode);
>  
> -/**
> - * Migrate a client channel from a CursorChannel.
> - * This is the equivalent of RedChannel client migrate callback.
> - * See comment on cursor_channel_new.
> - */
> -void                 cursor_channel_client_migrate(RedChannelClient
> *client);
> -
>  G_END_DECLS
>  
>  #endif /* CURSOR_CHANNEL_H_ */


Patch 06/12