It wasn't properly lazy-initializing the service binder, so it always
thought the backend service didn't exist, and so always returned false.
Also directly validated that every usage of sService in the module is
now correctly lazy-initialized.
Bug 16661321
Change-Id: If5fbb18aef81bfa8fd70eb40a1f6af54cc96d804
While we're in there also call listeners when they're added
so they know the state immediately.
Bug: 17258031
Change-Id: I5f1186314795f3fafd78e1b3e2d5102cdaec65d6
When META-INF/MANIFEST.MF is missing, treat as NO_CERTIFICATES
instead of CERTIFICATE_ENCODING. Also remove redundant layer of
debugging details when wrapping exceptions.
Bug: 15667982
Change-Id: I6e8216d5bf6e42da1feb70c89f991001380305be
Column weights specify the relative proportions of space distribution
within a *row*. With that in mind, I flipped the vertical and horizontal
axes in the docs - but that was a mistake. This CL flips the docs back.
The code works the right way around so doesn't need to be modified.
Change-Id: I1a36882eee8d1b9bd7b8ef10e03cbf66de0ac261
(cherry picked from commit 4ae5f5ee917b2014abe3b6e0b99451693cd2b896)
When users upgrade to L, don't restore into a state that would force
zen mode to follow suit. A user's first encounter with zen mode
should be via the UI.
Bug:16826161
Change-Id: I4eb63dce37d88f3a91b5034b3393eed2b768562c
ActionMenuViews work in two modes: hosting another Menu instance or
creating their own. The former is used when an action bar is
displaying a window's options menu. The latter is used when an
ActionMenuView (or Toolbar) is placed within an arbitrary layout and
the getMenu method is called.
When showing a window's options menu, ActionMenuPresenter calls into
the ActionBarPolicy to determine if we should reserve an overflow
button or if overflow will be presented by the window instead. Always
reserve overflow if the ActionMenuView is presenting its own menu.
Bug 17381966
Change-Id: I17c495986994d421bf6276ae39ba233243432e97
Simulates the relevant portions of a DOWN/UP event pair to request focus
and bring up the IME.
BUG: 8213791
Change-Id: Idb32d81552ecbbdefc64686c914eba6d8e28b0b8
This changes the PACKAGE_USAGE_STATS permission to be
a signature only permission, meaning that only apps
signed with the system signature have access to the
UsageStats API enabled by default.
Bug:17428130
Change-Id: I79ba54d7102c770ea2490f014a1f6736f0f46b3a
- When dozing, the first motion event triggers a wake up.
- Fade the backgrounds to ease the transition out of doze.
Bug:17167296
Change-Id: I5615ca0839dfa3ed2cf3001baf407c707f0676d5
bug:17441218
GLConsumer uses glBindTexture() directly instead of going through
Caches::bindTexture(). This can cause libhwui to draw with the wrong
texture bound in the following case which involves 2 TextureViews:
Frame 1:
GLConsumer::updateTexImage() calls glBindTexture(GL_TEXTURE_EXTERNAL_OES, 1)
HWUI renders TextureView A:
calls Caches::bindTexture(GL_TEXTURE_EXTERNAL_OES, 1) and draws
Frame 2:
GLConsumer::updateTexImage() calls glBindTexture(GL_TEXTURE_EXTERNAL_OES, 1)
GLConsumer::updateTexImage() calls glBindTexture(GL_TEXTURE_EXTERNAL_OES, 2)
HWUI renders TextureView A:
calls Caches::bindTexture(GL_TEXTURE_EXTERNAL_OES, 1) and draws
HWUI renders TextureView B:
calls Caches::bindTexture(GL_TEXTURE_EXTERNAL_OES, 2) and draws
In this case, HWUI will incorrectly draw TextureView A using texture 2 on
frame 2, because mBoundTextures[0]=1, even though the texture currently
bound to GL_TEXTURE_EXTERNAL_OES is 2.
Since GLConsumer is always used with a target of GL_TEXTURE_EXTERNAL_OES,
work around this problem by having mBoundTextures[] store only the
textures bound to the target GL_TEXTURE_2D. This is the common case
where the extra performance is needed. Since it's legal to have
different textures bound to GL_TEXTURE_2D and GL_TEXTURE_EXTERNAL_OES
on one texture unit, Caches::bindTexture() does not need to clear
mBoundTextures[mTextureUnit] when target != GL_TEXTURE_2D.
Change-Id: I8bc54ab8adcfacad7f3ed17a31236dc7a86c967a
Signed-off-by: Fred Fettinger <fred.fettinger@motorola.com>
Otherwise, cannot reliably match up capture progressed and failure callbacks
with the start callback.
Bug: 17421092
Change-Id: I91d92be70a15536b215bac330370ce37e426ec26
Neither strtoll() nor strtoull() can parse the %llx formatted tag
values coming from the kernel. We know the lowest 32 bits will never
be set, so fast-path the 3-character case as 0x0.
Bug: 17365163
Change-Id: I238bbd2830c9335e7ab7a53362d6e12b46e0bcb3
This can be controlled by MDMs via DPM.
Also fixes:
- javadoc for restrictions
- persisting of cross profile copy/paste restriction
Bug: 17387303
Change-Id: Ie148f56189181d2a4c6345c0823d417ab13a94a3
There was a race where ActivityManager would return null
for getCurrentUser() when switching between guest accounts.
This is because the Guest account was marked for deletion
while it was still active.
Bug:17290209
Change-Id: I224fb4b6836380e5acb7dbeb8f3343d74505f88a
(Noticed the difference in a javadoc diff between Notification and
NotificationCompat)
Bug: 17424399
Change-Id: I639a46c429ffebf8ca47118b2ea80f40ccdc1286
