1. Decode RDNSS options.
2. Keep track of how many times the program was updated.
3. Remove the leading / from the IPv4 address.
Change-Id: Ida0518a94ea7d952c82c8181b09044eff907b714
When an application cannot be started, and there is no profile/device
owner, still return a PackageSuspendedDialog.
BUG: 28042198
Change-Id: I5c30393f9481840a965bb815235af5181561a063
Default PIP location hides the upper part of the Settings,
so it should be moved a bit not to cover it.
Bug: 27891236
Change-Id: I6f13419cc9dc91522696c14a202e25652bd2232a
...based on network availability
There is a new light maintenance state "waiting for network"
that we go in to after idle if the network is not currently
available. We will stay in this state the same duration as idle,
so if we are continually without network access this effectively
doubles the light doze idle durations.
Get rid of some wrongly copy/pasted code to allow doze light
to still work even if we don't have an SMD. It doesn't need
one.
Also a bunch of improvements to the shell commands to make it
easier to test / debug.
Change-Id: Iad024840661479dbfd54b5b3db6ab96fefe59bc0
Bug: 28043974
One of the parameters to Settings class wrongly passes the secondaryCpuAbi
as CpuAbiOverrideString causing mayhem for multiarch libraries. Use the
correct param.
Further, for multiarch, the cpuabioverride flag is ignored. Remove the
check for cpuabioverride when checking for use32bitabi flag.
Change-Id: I7eff057031e3ddb18fcce6f380658a522af8b64d
Process PSS swapped is accounted as part of total process Pss to
get similar Pss values whether the device is using ZRAM of not.
But SwapPss is compressed in RAM so should not be accounted as
direct RAM contributor but though getZramTotalSizeKb already
taking into account.
Change-Id: I944850fb2a2b3a7342c850f8b9d69f304b53db46
- Don't generate private attributes in public R.java
- Strip out @SystemApi from comment when generating @android.annotation.SystemApi
- Only emit a single line (up to the first period) of an attribute's comment within
a styleable's attribute table.
Change-Id: Id6316a6861540325934133958939a12074ad4428
Renamed CircleFramedDrawable to UserIconDrawable and
added badge/frame support.
Replaced several instances of user icon drawing with
UserIconDrawable
Bug: 27483956
Change-Id: I10d665406ea8b085667aa8c71ffcedffe72dafbd
Bug: 27286867
WindowManager has committed to stopped state
controlling the lifecycle of the Surface, so
make that a first-class thing in HWUI as well.
This makes it more resistent to things like
a rogue updateSurface() happening while mStopped=true,
leading to bad things down the line. Instead let
the surface be changed/updated as often as desired,
and just block any attempt to draw on that surface.
Also removes some unnecessary makeCurrent()s, as
EglManager ensures that we *always* have a valid
GL context now (using a pbuffer surface if there is
no window surface set)
Change-Id: Iead78ddebc7997e8fdb0c9534836352f5e54b9bd
When we are resizing the task bounds might be smaller than
the stack bounds. We don't want to clip to the task bounds here,
so we use an expanded rectangle, making sure to expand it
large enough for giant surfaces.
Bug: 27676101
Change-Id: I1a324a474a89e4652ccd15ebd853b0b8815a48f5
ag/898112 added passing the window title to accessibility. To do that,
it also updated copy of the title in WindowManager.LayoutParams. That
was a behavior change, and the change broke cts tests that enforce that
the title in LayoutParams matches its expected format.
This change restores the previous behavior and adds a separate field to
LayoutParams to old an up-to-date title to pass to accessibility.
Bug: 28002185
Change-Id: Ia5b549113600b7c4fcc80b76c3f3a944dddaf483
