Re: Patch: Another AWT patch

[Bryce McKinlay <bryce at albatross dot co dot nz>]

Anthony Green wrote:

> Index: java/awt/Component.java
> ===================================================================
> RCS file: /cvs/java/libgcj/libjava/java/awt/Component.java,v
> retrieving revision 1.7
> diff -u -r1.7 Component.java
> --- Component.java      2000/07/31 02:03:51     1.7
> +++ Component.java      2000/07/31 02:29:22
> @@ -71,6 +71,8 @@
>    transient Container parent;
>    transient java.awt.peer.ComponentPeer peer;
>
> +  transient Object treeLock;
> +
>    transient ComponentListener componentListener;

I'm not 100% sure how the treelock is supposed to work, but there
shouldn't be one for every component. I think making it a static field
with an initializer should be sufficient for now.

FYI, Rolf's implementation goes like this:

    /* FIXME: Only one global tree lock for the all component?
       Considering that there can exist completely seperate component
       trees on different graphics devices, it doesn't seem right that
       actions on one graphics device should defer layout on other
       graphics devives. */
    /* Make the treelock a String so that it can easily be identified
       in debug dumps. */
    private static Object AWT_TREE_LOCK = "AWT_TREE_LOCK";

regards

  [ bryce ]