canvas: Move to shared memory for images and canvas backing stores.

The idea here is to land this before making images and canvas IPC-safe,
because this will shake out bugs relating to the shared memory. There
are currently test timeouts that are preventing multiprocess images and
canvas from landing, and I believe those are due to the inefficiency of
sending large amounts of data in the unoptimized builds we test with. By
moving to shared memory, this should drastically reduce the number of
copies and `serde` serialization.

Under the hood, this uses Mach OOL messages on Mac and temporary
memory-mapped files on Linux.

components/net_traits/image/base.rs

@@ -2,6 +2,7 @@
  * License, v. 2.0. If a copy of the MPL was not distributed with this
  * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
 
+use ipc_channel::ipc::IpcSharedMemory;
 use png;
 use stb_image::image as stb_image2;
 use std::mem;
@@ -23,7 +24,7 @@ pub struct Image {
     pub width: u32,
     pub height: u32,
     pub format: PixelFormat,
-    pub bytes: Vec<u8>,
+    pub bytes: IpcSharedMemory,
 }
 
 // TODO(pcwalton): Speed up with SIMD, or better yet, find some way to not do this.
@@ -66,7 +67,7 @@ pub fn load_from_memory(buffer: &[u8]) -> Option<Image> {
                 };
 
                 let bytes = mem::replace(bytes, Vec::new());
-
+                let bytes = IpcSharedMemory::from_bytes(&bytes[..]);
                 let image = Image {
                     width: png_image.width,
                     height: png_image.height,
@@ -96,7 +97,7 @@ pub fn load_from_memory(buffer: &[u8]) -> Option<Image> {
                     width: image.width as u32,
                     height: image.height as u32,
                     format: PixelFormat::RGBA8,
-                    bytes: image.data,
+                    bytes: IpcSharedMemory::from_bytes(&image.data[..]),
                 })
             }
             stb_image2::LoadResult::ImageF32(_image) => {