Updated to more descriptive title and button text.
The default message dialog will also show the suspending app name.
Test: atest \
com.android.server.pm.SuspendPackagesTest#testInterceptorActivity
Bug: 75332201
Change-Id: Ica77fb04874a8e32fe0d82ac27a9babd3b186f8d
New permissions system for slices. Its up a little disk space at the
optimization of CPU/Mem usage, and most importantly doesn't have a
limit on the number that can exist.
Test: runtest --path frameworks/base/services/tests/uiservicestests
Test: runtest --path cts/tests/tests/slices
Test: manual grant through SliceBrowser
Bug: 77323760
Change-Id: Ie601bca042a7b28113a511383ca7e3a35c7c8c71
Also make these configurable so we have the flexibility to change it if
necessary.
Setting the policy inside ActivityManagerService is not ideal, as that
means that AMS is the only place where the policy in ApplicationInfo is
correct. It should really be set inside PackageManagerService. However,
if it's set there, it would get out of date when the settings change, and
we'd have to update inside AMS anyway. So putting it only here seems ok
for now.
Test: $ adb shell settings put global hidden_api_policy_pre_p_apps 2
Test: $ adb shell settings put global hidden_api_policy_p_apps 2
Bug: 64382372
Change-Id: Ic4cbbb1e6464623e90c17ae08c0b6cbbe0dfa125
Add flags to let callers specify if they want IEC (power-of-two) or
SI (power-of-ten) units when formatting bytes.
Continue using SI units by default, since certain folks seem to have
strong opinions about that.
Bug: 76159924
Test: atest android.text.format.FormatterTest
Exempt-From-Owner-Approval: previous PS approved
Change-Id: I0074bb2578c2230e938b3f39c2564b1083feb825
Earlier setPackagesSuspended ignored the rest of the paramters when
suspend state did not change. This was a problem because then there was
no good way to update the other parameters without unsuspending the app,
which is not desirable.
Removed setSuspendedPackageAppExtras as now they can be update with this
api.
Also sending broadcasts when packages get unsuspended due to suspending
package removed.
Test: Existing tests pass:
atest com.android.server.pm.PackageUserStateTest
atest com.android.server.pm.SuspendPackagesTest
atest com.android.server.pm.PackageManagerSettingsTests
Bug: 77522553
Change-Id: I72a3c228d3d65c430e242da97b2bc6997ec6a135
We've traced down all code paths where NetworkInfo.isRoaming() and
NET_CAPABILITY_NOT_ROAMING can disagree, so we're only left with
noise that happens from NetworkAgent race conditions when someone
asks for a NetworkState between pending NetworkInfo and
NetworkCapabilities updates.
We can look at adding a way for a NetworkAgent to update these
two objects atomically, but not for this release.
Bug: 70174865
Test: builds, boots
Change-Id: I830e1fc6dc922e9eb7f8c2698d75181e00aaf7fb
Prior to this change there was a chance that an updating app would not
exist in mPackages and cause a permission check for that app to fail.
This change moves all permission checks to use mSettings and the cached
package it contains to do the checks.
Change-Id: I0717bddbb08b1d0dbab3ea79fa0d2067aa858753
Fixes: 76228188
Test: Manual - system starts, permission checks work before / after update
Address API review by describing relationship between the
PackageManager.hasSigningCertificate() methods and the PackageInfo
GET_SIGNING_CERTIFICATES method, as well as differentiating the
UID documentation from the package-name based one.
Bug: 74831566
Test: None, doc change.
Change-Id: I11c556325f9b2efbc2e5e1cf896b9c58db092ae8
adb shell settings put global hide_error_dialogs 1
This forces the same behavior as other devices that hide all
dialogs.
Bug: 77721907
Test: manual
Change-Id: I93e03d55f885f1874c33f46cad8cede0c759063b
Use the common rethrowFromSystemServer() pattern. Carefully only
throws for calls going to system_server; leaves existing behavior
intact when calling a ContentProvider.
Bug: 77671218
Test: builds, boots
Change-Id: Ie5e0763fb5e62b832f2b6a03c8f9d72dab3bf89a
This required advancing the IpManager to IpClient refactoring
(Bluetooth used IpManager and friends). Most importantly, the
Bluetooth code used WaitForProvisioningCallback, so this is
moved into IpClient proper now. Also: some more renaming
cleanup.
Test: as follows
- built
- flashed
- booted
- runtest frameworks-net passes
Bug: 62476366
Bug: 72663849
Merged-In: I5e5b2c59ad3ea9ad574e8e683bcab3b09cfe0791
Merged-In: I89b45310181d952129fb2294db63654da9b4057f
Change-Id: I9eee2a9d5c31cdd48a054a7edc3403584efb1864
(cherry picked from commit 8d1fe54be1)
