With Fission enabled, remote types can have the form
webIsolated=<URI>. This is bad, because we store them in telemetry in
some places. This patch adds and uses a method that sanitizes remote
types.
Differential Revision: https://phabricator.services.mozilla.com/D36068
--HG--
extra : moz-landing-system : lando
I couldn't reproduce the crash, but there are a few places in
ContentChild where we grab a window off of a BC without checking if
the window exists, so I added null checks.
Differential Revision: https://phabricator.services.mozilla.com/D35692
--HG--
extra : moz-landing-system : lando
This change comes in two parts. First, the code in WindowGlobalChild was changed
to detect the in-process case, and instruct the nsFrameLoader to become a
non-remote nsFrameLoader, and second the logic in WindowGlobalParent was updated
to ensure that the OwnerProcessID is updated after the change.
Differential Revision: https://phabricator.services.mozilla.com/D35060
--HG--
extra : moz-landing-system : lando
This adds an API for fetching security info per frame, no matter if we have
a certificate error or a valid certificate.
I tried to make this work in a Fission-compatible way, let me know if this
is the right approach.
Differential Revision: https://phabricator.services.mozilla.com/D34354
--HG--
extra : moz-landing-system : lando
There was a strange bug when converting SharedStringMap to use shared
memory: on Android, some tests would fail because a pref wasn't set or
there was something wrong with an expected error message. The root
cause was that mapping ashmem with MAP_PRIVATE results in all zeroes
(see bug 1525803), but that was read as a valid SharedStringMap with
length 0.
To prevent any possible future issues like that, this patch adds a
nonzero magic number to the header. It fits into padding on 64-bit and
the cost of setting and checking it should be essentially free.
Depends on D26744
Differential Revision: https://phabricator.services.mozilla.com/D26745
--HG--
extra : moz-landing-system : lando
There was a strange bug when converting SharedStringMap to use shared
memory: on Android, some tests would fail because a pref wasn't set or
there was something wrong with an expected error message. The root
cause was that mapping ashmem with MAP_PRIVATE results in all zeroes
(see bug 1525803), but that was read as a valid SharedStringMap with
length 0.
To prevent any possible future issues like that, this patch adds a
nonzero magic number to the header. It fits into padding on 64-bit and
the cost of setting and checking it should be essentially free.
Depends on D26744
Differential Revision: https://phabricator.services.mozilla.com/D26745
--HG--
extra : moz-landing-system : lando
App units of a remote browser element in the parent process are
different from app units inside the remote content in the child
process. We should apply the appropriate conversions by exposing
the relevant data as LayoutDevicePixel.
Differential Revision: https://phabricator.services.mozilla.com/D35334
--HG--
extra : moz-landing-system : lando
Using a precise visible rect for a top-level browser can cause an
artifact when resizing the window.
I believe this is caused by the message to resize the browser, and
the following paint, to be processed before the message with the
new visible rect.
We can be smarter about this, but in the interest of getting a
quick fix, this commit reverts the behavior here only for top-level
browsers.
Differential Revision: https://phabricator.services.mozilla.com/D34935
--HG--
extra : moz-landing-system : lando
The `WebProgress#sendLoadCallResult` method only existed to send a empty async
message and was only called from the `WebNavigationChild`. Since
`WebNavigationChild` is in the process of being removed, it makes sense to
inline the replaced method into its call site.
Differential Revision: https://phabricator.services.mozilla.com/D34566
--HG--
extra : moz-landing-system : lando
When `TabParent::ReconstructWebProgressRequest` was introduced in
31b206e2046f63af31424489e3d61d7761805878, it mistakenly was unconditionally
constructing an `nsIRequest` from the received `RequestData`, even when the URI
was null.
Additionally `ReconstructWebProgressRequest` has been updated to use the Gecko
style for out parameters (accepting an `nsIFoo**` and passing
`getter_AddRefs(...)` instead of `nsCOMPtr<nsIFoo>&`).
Differential Revision: https://phabricator.services.mozilla.com/D34445
--HG--
extra : moz-landing-system : lando
The `WebProgress#sendLoadCallResult` method only existed to send a empty async
message and was only called from the `WebNavigationChild`. Since
`WebNavigationChild` is in the process of being removed, it makes sense to
inline the replaced method into its call site.
Differential Revision: https://phabricator.services.mozilla.com/D34566
--HG--
extra : moz-landing-system : lando
When `TabParent::ReconstructWebProgressRequest` was introduced in
31b206e2046f63af31424489e3d61d7761805878, it mistakenly was unconditionally
constructing an `nsIRequest` from the received `RequestData`, even when the URI
was null.
Additionally `ReconstructWebProgressRequest` has been updated to use the Gecko
style for out parameters (accepting an `nsIFoo**` and passing
`getter_AddRefs(...)` instead of `nsCOMPtr<nsIFoo>&`).
Differential Revision: https://phabricator.services.mozilla.com/D34445
--HG--
extra : moz-landing-system : lando
Waiting for docshells and frameloaders to destroy will leave attached
browsing contexts attached too long. In case the children of a
browsing contexts cannot be cached we want to detach all of them as
soon as possible.
Also normalizes the use of BrowsingContext::mGroup.
Differential Revision: https://phabricator.services.mozilla.com/D33602
--HG--
extra : moz-landing-system : lando
Waiting for docshells and frameloaders to destroy will leave attached
browsing contexts attached too long. In case the children of a
browsing contexts cannot be cached we want to detach all of them as
soon as possible.
Also normalizes the use of BrowsingContext::mGroup.
Differential Revision: https://phabricator.services.mozilla.com/D33602
--HG--
extra : moz-landing-system : lando
Previously, BrowserBridgeChild::SendSetEmbedderAccessible was only called when an OuterDocAccessible was constructed.
However, it's also possible that the BrowserBridgeChild is created *after* the OuterDocAccessible.
Therefore, we must also do this when a BrowserBridgeChild is created if the OuterDocAccessible already exists.
Differential Revision: https://phabricator.services.mozilla.com/D34474
--HG--
extra : moz-landing-system : lando
Tom Ritter
Andrew McCreight
Narcis Beleuzu
Nika Layzell
Dorel Luca
Garvan Keeley
Ciure Andrei
Erica Wright
Paul Adenot
Ryan Hunt
Johann Hofmann
Jed Davis
Henri Sivonen
Andreea Pavel
Haik Aftandilian
Bogdan Tara
Mike Conley
Barret Rennie
Andreas Farre
Noemi Erli
Oana Pop Rus
Jonathan Kingston
Andrea Marchesini
Peter Van der Beken
James Teh