This is to ensure that multiple completions cannot be attempted on the same `GeckoResult`, resulting in crashes.
Differential Revision: https://phabricator.services.mozilla.com/D36929
--HG--
extra : moz-landing-system : lando
This is caused by a race condition when the compositor is detached. Because the actual detachment happens in a new thread, `detach` can complete and release the lock on `mLayerViewSupport`, and `RecvScreenPixels` can obtain the lock, before `mLayerViewSupport` is properly cleaned up. We therefore check to ensure that `lvs` is not null before calling a method on it.
Differential Revision: https://phabricator.services.mozilla.com/D36928
--HG--
extra : moz-landing-system : lando
When long tapping on input element, the behavior of virtual keyboard is different of Chrome. Firefox opens virtual keyboard, but Chrome doesn't.
So I would like to change to same behavior of Chrome.
Also, after this fix, action bar (copy, paste and etc) won't be disappeared immediately after it appears by long tap. Before applying this fix, virtual keyboard is shown, then focused element is scrolled to center. So action bar may be disappeared. But after applying this fix, focused element isn't scrolled. So action bar keeps same position.
Depends on D35991
Differential Revision: https://phabricator.services.mozilla.com/D35992
--HG--
extra : moz-landing-system : lando
PowerOfTwo makes for a cleaner and more expressive interface, showing that the
profiler will use a power-of-2 storage size.
Using PowerOfTwoMask in ProfilerBuffer also makes it more obvious that we want
cheap modulo operations.
And we don't need to keep the original capacity, as it's only used once and can
easily be recomputed from the mask.
Differential Revision: https://phabricator.services.mozilla.com/D36027
--HG--
extra : moz-landing-system : lando
PowerOfTwo makes for a cleaner and more expressive interface, showing that the
profiler will use a power-of-2 storage size.
Using PowerOfTwoMask in ProfilerBuffer also makes it more obvious that we want
cheap modulo operations.
And we don't need to keep the original capacity, as it's only used once and can
easily be recomputed from the mask.
Differential Revision: https://phabricator.services.mozilla.com/D36027
--HG--
extra : moz-landing-system : lando
Adds GeckoChildProcessHost::GetAll() and use it in ChromeUtils::GetProcInfo()
Differential Revision: https://phabricator.services.mozilla.com/D33920
--HG--
extra : moz-landing-system : lando
Adds GeckoChildProcessHost::GetAll() and use it in ChromeUtils::GetProcInfo()
Differential Revision: https://phabricator.services.mozilla.com/D33920
--HG--
extra : moz-landing-system : lando
On android, android's nsWindow creates LayerManaer only in nsWindow::Create(). When WebRender error happened, gecko just stopped rendering by disabling Webrender.
The nsWindow needs to re-create LayerManager during disabling Webrender. Further, during disabling WebRender, All GeckoSurfaceTextures should not be attached to GLContext. It is for preventing a conflict with AttachToGLContext() call in SurfaceTextureHost::EnsureAttached().
Differential Revision: https://phabricator.services.mozilla.com/D26687
--HG--
extra : moz-landing-system : lando
Adds GeckoChildProcessHost::GetAll() and use it in ChromeUtils::GetProcInfo()
Differential Revision: https://phabricator.services.mozilla.com/D33920
--HG--
extra : moz-landing-system : lando
To run task_for_pid() on child processes, we need the child task port for
security reasons. This port can be obtained via a Mach IPC exchange.
This is what GeckoChildProcessHost::GetChildTask() provides, so we use it
in cocoa's version of GetProcInfo()
Differential Revision: https://phabricator.services.mozilla.com/D25927
--HG--
extra : moz-landing-system : lando
Only gtk returns failure ever, and nobody checks the result anyway.
Use an enum class so that it's clear from the caller what it means.
Differential Revision: https://phabricator.services.mozilla.com/D32353
--HG--
extra : moz-landing-system : lando
Only gtk returns failure ever, and nobody checks the result anyway.
Use an enum class so that it's clear from the caller what it means.
Differential Revision: https://phabricator.services.mozilla.com/D32353
--HG--
extra : moz-landing-system : lando
This should be an idempotent patch. The way to come up with this patch has been:
* Run the first script attached to the bug and pipe it to xclip, then paste it
in color.rs
* Add the relevant #[derive] annotations and remove the color.mako.rs
definition.
* Reorder the values to match the ColorID definition, on which some widget
prefs and caching stuff relies on.
* Manually port some documentation from nsLookAndFeel.h
* Run `rg 'eColorID_' | cut -d : -f 1 | sort | uniq >files`
* Run the second script attached to the bug.
* Manually fix usage of `LAST_COLOR` (adding the `End` variant), and adding
casts to integer as needed.
* Add an static assert so that people remember to update the prefs, rather than
a comment on the definition :)
Differential Revision: https://phabricator.services.mozilla.com/D32610
--HG--
extra : moz-landing-system : lando
Actually, all JNI Exceptions with `java.lang.OutOfMemoryError` call
`NS_ABORT_OOM(0)`. But `JNIEnv::NewString` in `StringParam::GetString` can know
OOM size when returning `nullptr`. So call `NS_ABORT_OOM` directly when
`NewString` is failure.
Differential Revision: https://phabricator.services.mozilla.com/D31026
--HG--
extra : moz-landing-system : lando
Profiling shows that we're spending a lot of time on startup inside
android.media.MediaCodecList.getCodecCount when GfxInfo::GetFeatureStatus calls
over to Java to determine whether hardware accelerated video encoding is
supported.
Looking at the Java stacks in the profile, Android is spending most of our time
creating a list of codecs. It doesn't look like there's a faster way to query
for hardware accelerated video support. So to speed this up we can cache the
value in the user's profile. We also store the OS version, which we can use to
detect when the OS is updated so we can invalidate the cache then.
Presumably an OS update is the only way a device can have its hardware acceleration
support status change.
With this change, the time we take figuring out the HW encode/decode status
goes from ~100ms on a cold run to ~0.01ms on a cache hit on my HD8 tablet.
Differential Revision: https://phabricator.services.mozilla.com/D31380
--HG--
extra : moz-landing-system : lando
Support using the Google Play-provided FIDO2 API for Web Authentication.
FIDO U2F API support is being handled subsequently in Bug 1550625.
This patch uses the privileged APIs and thus will only work on Fennec Nightly, Beta, and Release builds.
Differential Revision: https://phabricator.services.mozilla.com/D1148
--HG--
extra : moz-landing-system : lando
The GECKOBUNDLE macros are useful to more than just a11y code, so let's move them into the jni package so that all jni consumers may drink of their sweet nectar.
Differential Revision: https://phabricator.services.mozilla.com/D30585
--HG--
extra : moz-landing-system : lando
This reunifies the behaviour changed in bug 1294232 to ensure that the
vendor ID of GfxInfo is the same between graphics hardware. Vendor ID
should always represent Intel, Nvidia, ATI, etc such that callers can
reason about the performance characteristics without being exposed to
the driver implementation for that platform. Now we split off the more
detailed driver information into the "driver vendor" which will contain
more information, such as what implementation is being used (e.g.
mesa/i965 for modern Intel graphics cards). This field is exposed to the
blocklist and will be useful for allowing different rules for different
driver implementations.
We also now provide a default implementation for
GfxInfoBase::FindMonitors for platforms missing support. This will just
list the primary screen size used without listing secondary monitors,
refresh rate, and such.
Differential Revision: https://phabricator.services.mozilla.com/D29471
This reunifies the behaviour changed in bug 1294232 to ensure that the
vendor ID of GfxInfo is the same between graphics hardware. Vendor ID
should always represent Intel, Nvidia, ATI, etc such that callers can
reason about the performance characteristics without being exposed to
the driver implementation for that platform. Now we split off the more
detailed driver information into the "driver vendor" which will contain
more information, such as what implementation is being used (e.g.
mesa/i965 for modern Intel graphics cards). This field is exposed to the
blocklist and will be useful for allowing different rules for different
driver implementations.
We also now provide a default implementation for
GfxInfoBase::FindMonitors for platforms missing support. This will just
list the primary screen size used without listing secondary monitors,
refresh rate, and such.
Differential Revision: https://phabricator.services.mozilla.com/D29471
This reunifies the behaviour changed in bug 1294232 to ensure that the
vendor ID of GfxInfo is the same between graphics hardware. Vendor ID
should always represent Intel, Nvidia, ATI, etc such that callers can
reason about the performance characteristics without being exposed to
the driver implementation for that platform. Now we split off the more
detailed driver information into the "driver vendor" which will contain
more information, such as what implementation is being used (e.g.
mesa/i965 for modern Intel graphics cards). This field is exposed to the
blocklist and will be useful for allowing different rules for different
driver implementations.
We also now provide a default implementation for
GfxInfoBase::FindMonitors for platforms missing support. This will just
list the primary screen size used without listing secondary monitors,
refresh rate, and such.
Differential Revision: https://phabricator.services.mozilla.com/D29471
Eugen Sawin
Tom Schuster
Cosmin Sabou
Razvan Maries
Emily Toop
Bobby Holley
Coroiu Cristina
Jonathan Kingston
Sylvestre Ledru
Makoto Kato
Henri Sivonen
Gerald Squelart
sotaro
Andreea Pavel
Tarek Ziadé
Oana Pop Rus
Gurzau Raul
Boris Zbarsky
Emilio Cobos Álvarez
Brindusan Cristian
Valentin Gosu
Andrew Osmond
Chris Pearce
J.C. Jones
Mihai Alexandru Michis
Daniel Varga
Thomas Nguyen
Eitan Isaacson