Instead of creating a display list for the entire page, only create one
for an area that expands around the viewport. On my machine this makes
incremental layout of http://timecube.com 50% faster.
This fixes issue #3630
A short summary of the changes:
* Use atomic generation id to cancel inflight requests
* Handles nested calls to abort, open, send inside handlers
* Adds XHRReleaseMsg to delay freeing XHR object till all
inflight events are processed
* Handles both timeout, errors and abort/open in a symmetric fashion
i.e All inflight events will be cancelled for timeouts, aborts,
errors and on calling open.
* Change the ErroredMsg enum to be more symmetric with the returned
Error enum
I noticed a few possible changes that could make the code for fetch task simpler:
* We can remove the additional timer task and let the fetch task manage
its own timer (or maybe the resource loader can do this.)
* The CORS related steps could also be moved into the resource loader.
* Right now upload events are not support. This requires some support
from resource loader.
Instead of looking at the display tree, have ContentBox(es)Query consult
the flow tree. This allow optimizing away parts of the display tree
later. To do this we need to be more careful about how we send reflow
requests, only querying the flow tree when possible.
Fixes#3790.
This fixes issue #3630
A short summary of the changes:
* Use generation id to cancel inflight requests
* Handles nested calls to abort, open, send inside handlers
* Adds XHRReleaseMsg to delay freeing XHR object till all
inflight events are processed
* Change the ErroredMsg enum to be more symmetric with the returned
Error enum
This implements the parts of the "prepare a script element" algorithm that are
required for synchronous scripts. It also adds some infrastructure for future
support of the `async` and `defer` attributes.
This patch puts in the initial framework for incremental reflow. Nodes' styles
are no longer recalculated unless the node has changed.
I've been hacking on the general problem of incremental reflow for the past
couple weeks, and I've yet to get a full implementation that actually passes all
the reftests + wikipedia + cnn. Therefore, I'm going to try to land the different
parts of it one by one.
This patch only does incremental style recalc, without incremental flow
construction, inline-size bubbling, reflow, or display lists. Those will be coming
in that order as I finish them.
At least with this strategy, I can land a working version of incremental reflow,
even if not yet complete.
r? @pcwalton
The ScriptTask is owned by the proc in ScriptTaskFactory::create.
We started using a managed box for what would eventually become ScriptTask in
d21d27e086.
The ScriptTask is owned by the proc in ScriptTaskFactory::create.
We started using a managed box for what would eventually become ScriptTask in
d21d27e086.
Since Page isn't reflected, it cannot be rooted, and pointers to JS-managed
objects inside it are not guaranteed to be traced. This may cause the pointer
to become dangling.
