Refactored image cache task - details below.

* Simpler image cache API for clients to use.
 * Significantly fewer threads.
   * One thread for image cache task (multiplexes commands, decoder threads and async resource requests).
   * 4 threads for decoder worker tasks.
 * Removed ReflowEvent hacks in script and layout tasks.
   * Image elements pass a Trusted<T> to image cache, which is used to dirty nodes via script task. Previous use of Untrusted addresses was unsafe.
   * Image requests such as background-image on layout / paint threads trigger repaint only rather than full reflow.
 * Add reflow batching for when multiple images load quickly.
   * Reduces the number of paints loading wikipedia from ~95 to ~35.
 * Reasonably simple to add proper prefetch support in a follow up PR.
 * Async loaded images always construct Image fragments now, instead of generic.
   * Image fragments support the image not being present.
 * Simpler implementation of synchronous image loading for reftests.
 * Removed image holder.
 * image.onload support.
 * image NaturalWidth and NaturalHeight support.
 * Updated WPT expectations.
This commit is contained in:
Glenn Watson 2015-04-20 13:34:26 +10:00
parent e278e5b9a2
commit d8aef7208e
33 changed files with 2785 additions and 1679 deletions

View file

@ -13,17 +13,18 @@ use gfx::display_list::OpaqueNode;
use gfx::font_cache_task::FontCacheTask;
use gfx::font_context::FontContext;
use msg::constellation_msg::ConstellationChan;
use net_traits::image::base::Image;
use net_traits::image_cache_task::{ImageCacheChan, ImageCacheTask, ImageState};
use script::layout_interface::{Animation, LayoutChan};
use script_traits::UntrustedNodeAddress;
use net_traits::local_image_cache::LocalImageCache;
use std::boxed;
use std::cell::Cell;
use std::ptr;
use std::sync::mpsc::Sender;
use std::sync::{Arc, Mutex};
use std::sync::Arc;
use std::sync::mpsc::{channel, Sender};
use style::selector_matching::Stylist;
use url::Url;
use util::geometry::Au;
use util::opts;
struct LocalLayoutContext {
font_context: FontContext,
@ -55,8 +56,11 @@ fn create_or_get_local_context(shared_layout_context: &SharedLayoutContext)
/// Layout information shared among all workers. This must be thread-safe.
pub struct SharedLayoutContext {
/// The local image cache.
pub image_cache: Arc<Mutex<LocalImageCache<UntrustedNodeAddress>>>,
/// The shared image cache task.
pub image_cache_task: ImageCacheTask,
/// A channel for the image cache to send responses to.
pub image_cache_sender: ImageCacheChan,
/// The current screen size.
pub screen_size: Size2D<Au>,
@ -139,4 +143,42 @@ impl<'a> LayoutContext<'a> {
&mut cached_context.style_sharing_candidate_cache
}
}
pub fn get_or_request_image(&self, url: Url) -> Option<Arc<Image>> {
// See if the image is already available
let result = self.shared.image_cache_task.get_image_if_available(url.clone());
match result {
Ok(image) => Some(image),
Err(state) => {
// If we are emitting an output file, then we need to block on
// image load or we risk emitting an output file missing the image.
let is_sync = opts::get().output_file.is_some();
match (state, is_sync) {
// Image failed to load, so just return nothing
(ImageState::LoadError, _) => None,
// Not loaded, test mode - load the image synchronously
(_, true) => {
let (sync_tx, sync_rx) = channel();
self.shared.image_cache_task.request_image(url,
ImageCacheChan(sync_tx),
None);
sync_rx.recv().unwrap().image
}
// Not yet requested, async mode - request image from the cache
(ImageState::NotRequested, false) => {
self.shared.image_cache_task.request_image(url,
self.shared.image_cache_sender.clone(),
None);
None
}
// Image has been requested, is still pending. Return no image
// for this paint loop. When the image loads it will trigger
// a reflow and/or repaint.
(ImageState::Pending, false) => None,
}
}
}
}
}