Every JS plugin is assigned a unique ID. When an instance of a JS plugin is created the
frame loader that we use to load the plugin's handler URI will create a special
TabContext. This TabContext causes the ContentParent to use the process for this specific
JS plugin (creating one if it hasn't already) when it creates the PBrowser actors.
This causes the iframes for all the instances of a specific JS plugin to be grouped in the
same process.
Since we now want to support APZ for remote frameloaders in popups, but do not
want to needlessly enable it in simple popup widgets which should never host
remote content, we need to prevent remote content from being loaded into those
popups.
MozReview-Commit-ID: WfjMC2p2eK
Currently, we only correctly support remote layer trees for frameloaders that
use the same layer manager as their document. Since we need to be able to host
remote <browser> content in popup widgets for remote WebExtensions, we need to
tie the frameloaders to the layer manager of their host element, rather than
the root layer manager for the document.
MozReview-Commit-ID: 4RCsamFBiQw
Since we now want to support APZ for remote frameloaders in popups, but do not
want to needlessly enable it in simple popup widgets which should never host
remote content, we need to prevent remote content from being loaded into those
popups.
MozReview-Commit-ID: WfjMC2p2eK
Currently, we only correctly support remote layer trees for frameloaders that
use the same layer manager as their document. Since we need to be able to host
remote <browser> content in popup widgets for remote WebExtensions, we need to
tie the frameloaders to the layer manager of their host element, rather than
the root layer manager for the document.
MozReview-Commit-ID: 4RCsamFBiQw
This patch replaces the usage of sNextTabParent pointer to store the next
PBrowser parent actor to be used by the next frame loader with the
following information:
* In the case where the content JS has requested a new tab, the ID of the
next TabParent will be stored on the <xul:browser> element.
* In the case where the content JS has requested a new window, the ID of
the next TabParent will be stored on the created nsXULWindow.
In bug 1147911, the concept of a remoteType was added to a xul:browser.
This was an attribute which would control the type of remote process was
intended to be used to load the page.
In order to swap two frameLoaders, it has always been necessary for them
to either both contain remote content, or both contain non-remote
content. This check is made in nsFrameLoader::SwapWithOtherLoader, by
checking `IsRemoteFrame() != aOther->IsRemoteFrame()`, and then
returning `NS_ERROR_NOT_IMPLEMENTED` if that was not the case.
In the follow-up bug 1317293, the check which is being removed here was
added to ensure that the remoteType of two frameLoaders which are being
swapped also matched. This was not a technical limitation, but rather
something which "seemed to make sense to do".
This bug removes that check as it is not a technical limitation and
causes problems in edge cases around Large-Allocation processes now that
the remoteType is being used to denote a Large-Allocation process.
Namely, it means that attempting to drag a Large-Allocation window into
a new window when at the Large-Allocation process cap will fail, due to
being unable to create a new Large-Allocation process in the new window
to swap with.
The new swapping of the attributes which is added below is done with the
intent that the `remoteType` attribute of the xul:browser element should
match the `remoteType` attribute of the frameLoader inside of it at all
times. As the swap can now occur between two different `remoteType`s,
this is necessary to keep that constraint.
MozReview-Commit-ID: BHFgjwRGrge
This changes the `relatedBrowser` property which held a <xul:browser> to the
more explicit `sameProcessAsFrameLoader` which takes an nsIFrameLoader.
This clarifies the purpose of the property and also (by switching to the frame
loader) makes it easier to set in some contexts.
MozReview-Commit-ID: LnEvSP8zkto
With GroupedSHistory, history navigations may now require the browser to
change which frameloader is stored internally from within Core. This
patch adds a mechanism to allow for chrome code to respond to these
changes and both delay the change, or respond once the change is
performed.
Delaying the change is accomplished through the BrowserWillChangeProcess
event, which is fired when it is determined that a process change will
happen for the given browser, but the change has not occured yet. During
this time the nsIFrameLoader::AddProcessChangeBlockingPromise method may
be called on the target browser's frameloader. Any promises passed to
this method will be waited on, and the process change will not occur
until they have all been fulfiled.
Once that has occured, the process change occurs, and the
BrowserChangedProcess event is fired.
This is useful for chrome code which needs to flush state from the
original process before the change, and then which needs to connect
state in the new process with state in the chrome process.
MozReview-Commit-ID: C0Xn6pfruB2
This adds a single flag, SWAP_KEEP_PERMANENT_KEY, which tells the
browser that when it performs the swap, the permanent key should stick
with the browser, rather than following the frameLoader.
This patch also adds the implementation of tabbrowser.swapBrowsers,
which was previously absent, despite being referenced.
MozReview-Commit-ID: CLwJYzpY8Pp
This patch removes support for mozapp iframes, leaving support for
mozbrowser iframes intact. Some of the code has been rewritten in order
to phrase things in terms of mozbrowser only, as opposed to mozbrowser
or app. In some places, code that was only useful with apps has been
completely removed, so that the APIs consumed can also be removed. In
some places where the notion of appId was bleeding out of this API, now
we use NO_APP_ID. Other notions of appId which were restricted to this
API have been removed.
