This patch fixes data race on CacheFileHandle::mFileSize when accessed on main thread by nsHttpChannel::ReportNetVSCacheTelemetry(). All other usages are on cache I/O thread.
Differential Revision: https://phabricator.services.mozilla.com/D60407
--HG--
extra : moz-landing-system : lando
We still need the front and the spec files to talk to older servers.
However we don't maintain forward compatibility, which means the server is not supposed to work with older clients.
So the emulation actor can be removed.
Differential Revision: https://phabricator.services.mozilla.com/D60294
--HG--
extra : moz-landing-system : lando
In `MediaPipelineTransmit::SetTrack()` when a track of different MTG is replaced, the transmission is stopped and restarted asynchronously. This can create a problem if a new stop has arrived in the meantime. The transmission should not be restarted in that case.
This change adds a boolean, to check when an asynchronous start is expected and ignores it if not needed.
Differential Revision: https://phabricator.services.mozilla.com/D59750
--HG--
extra : moz-landing-system : lando
C++: Remove reference to JavaScript typeof from 'Operators' section and specifically refer to sizeof only.
JavaScript: Added remainder of the guideline referring to typeof.
Differential Revision: https://phabricator.services.mozilla.com/D60261
--HG--
extra : moz-landing-system : lando
`viewport-fit` is hint that browser application can use cutout area. So we should expose it to GeckoView application to set `layoutInDisplayCutoutMode`.
When meta element is found or changed, `ContentDelegate.onMetaviewportFitChange` is called. Even if nothing, it will be called after DOMContentLoaded is fired.
Depends on D57398
Differential Revision: https://phabricator.services.mozilla.com/D55610
--HG--
extra : moz-landing-system : lando
Actually, we have `DOMMeta*` event for modifying meta element. But GeckoView wants viewport-fit change only.
@hiro suggests new event that is fired by viewport-fit will be changed. This event doesn't need current viewport-fit value since GeckoView doesn't need current value immediately.
Depends on D57397
Differential Revision: https://phabricator.services.mozilla.com/D57398
--HG--
extra : moz-landing-system : lando
Per discussing with @hiro, we should add new method to get viewport-fit
instead of adding parameter of `nsIDOMWindowUtils.getViewportInfo`.
Depends on D55609
Differential Revision: https://phabricator.services.mozilla.com/D57397
--HG--
extra : moz-landing-system : lando
For safe area insets (cutout) support, CSS Round Display Level 1 (https://drafts.csswg.org/css-round-display/#viewport-fit-descriptor) has new viewport value as `viewport-fit`.
To support safe area insets that is notch on display, CSS Environment Variables Module Level 1 (https://drafts.csswg.org/css-env-1/#safe-area-insets) adds `safearea-insets-*` (left, top, right and bottom). Also, `meta` element has `viewport-fit` enum value. (ex `<meta name="viewport" content="viewport-fit=cover>`) whether web browser window cover notch area.
`viewport-fit` has 3 enum value, `auto`, `cover` and `contain`. GeckoView wants to expose this value to browser application such Fenix. Because Android API (https://developer.android.com/guide/topics/display-cutout) uses window root layout (It is set by Application) to cover notch on display.
Differential Revision: https://phabricator.services.mozilla.com/D55609
--HG--
extra : moz-landing-system : lando
This patch will check the exit code of the ProcessHandler() that
calls the tooltool script, and will also retry when it fails.
Differential Revision: https://phabricator.services.mozilla.com/D60397
--HG--
extra : moz-landing-system : lando
A third-party application can modify the export directory, the export address/name/ordinal
tables, or an entry in those tables. If that happens, we will see an RVA is located outside
the mapped image and `RVAToPtr` returns null. This patch makes sure we don't hit null AV
when modification is detected.
`FindExportAddressTableEntry` should not return a pointer to the modified table entry because
we dereference it in another process to cross-process detour.
Differential Revision: https://phabricator.services.mozilla.com/D59738
--HG--
extra : moz-landing-system : lando
Instead of showing percentages, show actual numbers for memory usage,
with caveats:
- On macOS, psutil doesn't seem to be getting anything that would sum up
to the total it reports, so we keep the "percent".
- While the sum of all the fields shown on Linux does sum up to the total,
the visible usage doesn't quite match what the "percent" look like for
the same dataset.
- On Windows, only "used" and "free" are available. They do sum up to
the total.
Differential Revision: https://phabricator.services.mozilla.com/D60226
--HG--
extra : moz-landing-system : lando
Changes:
Bug 1552563 introduced the notion that `compiz` should only be run where it was required. This was an attempt to save some seconds through the CI system as `compiz` takes 15 seconds to return an exit code.
Now that majority of tests are running under ubuntu1804, which runs the GNOME desktop environment, the flag `NEED_COMPIZ` is no longer necessary.
Differential Revision: https://phabricator.services.mozilla.com/D60342
--HG--
extra : moz-landing-system : lando
This allows us to compare a `UniquePtr` with it's raw pointer type. This allows `nsTArray`'s default comparator to function properly and lets us find/remove items from an nsTArray<UniquePtr<T>> with just a raw pointer.
Differential Revision: https://phabricator.services.mozilla.com/D60339
--HG--
extra : moz-landing-system : lando