We were not passing the length of the UTF-16 string to
String16::setTo. As a result, it was copying the contents of
the text up to the first null it found.
First problem, these strings are not typically null terminated!
Second problem, if the string contained a null character, then
we might truncate it. However, we only truncated the string
when the copy constructor was invoked (say, when we called
get() on the cache) but not in internalTextCopy() (before
adding the key to the cache).
As a result of the second problem, we would first search
the cache for a key that matched a partially copied truncated
string (potentially reading uninitialized memory that followed it).
Finding none, we would add the entry to the cache using
the correct key.
If the cache already had a value associated with the correct key,
then the put would fail, returning false. Charging ever onwards,
we would add the size of the entry to the cache size.
Proceeding in this manner, it was possible for the cache to
believe it had less remaining space than it really did. At that
point, it was possible for the cache to evict all entries and
yet still not think it had room to add a new one, so it would
continue trying to make space indefinitely.
Bug: 5576812
Change-Id: I05251594f6b2da0a5dc09f7200f04fe9100ec766
LayoutTransition was making an incorrect assumption that there could
only be one transition animation on a child of a transitioning container.
But if multiple children are added/removed to/from that container, there would
be multiple calls to set up changing animations for each existing child
of that container. This meant that the child would have multiple, new
OnLayoutChangeListeners added to it as part of the setup process.
Meanwhile, we would cache only the latest listener in a hashmap that used
the child as a key for the listener. Then when we cleaned up the hashmap later,
we would remove only the latest listener from the child, leaving the rest there
for eternity.
The fix is to skip the setup entirely for children that already have listeners
set on them; they must, if that's the case, already have been set up and are
already listening for layout changes. Setting up the animation is redundant,
and adding another listener is a leak.
issue #5588509: memory leak in systemui
Change-Id: Ie2192593d84702be7243c18760dfdb3a027b761c
1. The dismiss method in a PopupWindow is first removing the
popup View from the WindowManager and then if the content
View differs from the popup View the former is removed from
the latter. Then dismiss() clears the dismissed flag so a
subsequent call to dismiss() is a NOP. However, removing
a View both from the WindowManager and from its parent
ViewGroup triggers stuff wich may lead to a subsequent call
to dismiss(). This leads to a stack overflow exception.
bug:5598944
Change-Id: I2aeeda591be3e9aa98fec1ee17ea8f0e746e6992
Initially set to 3 minutes this raised the standby current
by 12.5% so changing to 6 minutes.
Bug: 5534004
Change-Id: I602f5fe4de35d0db2dbacf0c615c300c57dd2d0d
1. Fix the ClassCastException while handling spec out 'Setup menu'
-Ghana MTN simcard and JDI simcard sends abnormal 'setup menu'cmd.
-Those 'setup menu' is spec out.
-At the end of the proactive cmd,extra bytes '0x00 0x00 0x00 0x00' is
followed.
- That cause ClassCastException and phone crash.
Bug: 5574160
Change-Id: Ieafb6c4efd94bb4e2a39a04612a6761c958654bb
Signed-off-by: dujin.cha <dujin.cha@samsung.com>
The phone app needs a way to distinguish between (a) numbers that are
definitely emergency numbers, and (b) numbers that *might* result in an
emergency call being dialed, but aren't specifically emergency numbers
themselves.
(The phone app needs this distinction in order to enforce the restriction
that 3rd party apps should not be allowed to make emergency calls using
the ACTION_CALL intent, while still making sure that the in-call UI only
displays the "emergency call" state for numbers that are *definitely*
emergency numbers. See bug 5493790 for the full details;)
So this change adds a full set of "isPotentialEmergencyNumber()" methods
to go along with the "isEmergencyNumber()" methods we've had all along.
The "potential" variants behave identically to the original methods,
*except* that they ultimately use number.startsWith() rather than
number.equals() when comparing against the list of emergency numbers.
TESTED:
- Unit test 'PhoneNumberUtilsTest#testIsEmergencyNumber' passes.
(The PhoneNumberUtilsTest class doesn't pass in its entirety, but it was
broken before this change also.)
- Also see the commit description of change
Ib949fea3c0ce6b341a90e617a03ba3f22c69018b for the exact tests I ran
against the phone app.
This change should be submitted along with
Change-Id: Ib949fea3c0ce6b341a90e617a03ba3f22c69018b
in apps/Phone (but this change must go in first to avoid breaking the
build.)
Bug: 5493790
Change-Id: Ic528cfcc555734cdaf4ca8a18a50199771ba49b1
- Precondition: config_sms_enabled_single_shift_tables is configured as
1 (Turkish) in frameworks/base/core/res/res/values/config.xml
- Cause: There is no consideration for National Language Shift Tables in
SmsMessage::fragmentText function.
- Solution: The header length is calculated properly according to
National Language Shift Table
- modified to add test cases and fix calculation bug (jhamby@google.com)
Bug: 5553544
Change-Id: I9eaefbbd6b3d75f8c41cbf9d0cb03a701cfa1cb3
Outsiders asking for this list may cause the list to change on another thread.
Fixing general synchronization issues.
bug:5531630
Change-Id: I7a3ee0bba3db40f45bcb0159491942fa4cf38c37
Sometimes the lockscreen view is recreated even though it has already
been created (and therefore Face Unlock is already running). One
example of this is when the lockscreen momentarily displays in
landscape mode and then corrects itself into portrait mode.
When lockscreen recreates itself, it removes the view and then later
re-adds the view. During that time the window token is null and may
be passed to Face Unlock when it tries to restart. The reason this
doesn't happen *every* time the view is recreated is because the
onServiceConnected() callback starts Face Unlock, and usually it
runs after the view is re-added, but sometimes it runs before the view
is re-added, resulting in this bug.
An earlier fix put null-token checking inside of the Face Unlock
service, which prevented crashing but Face Unlock was still unable
to run. This fix makes the null token case not happen so Face Unlock
can run every time. It accomplishes this by simply not restarting
Face Unlock until the view has been re-added.
This fix also replaces checking two flags everywhere to see if
Face Unlock is being used and instead uses a single function call.
Change-Id: Ib46f25f2a58ab2e70470337861c25ee81a858873
Was mistakenly assuming that Parcel::writeFileDescriptor took
ownership of the fd that was passed in. It does not!
Added some comments and a default parameter to allow the caller
to specify whether it wishes the Parcel to take ownership.
Bug: 5563374
Change-Id: I5a12f51d582bf246ce90133cce7690bb9bca93f6
Also use the AlarmManager instead of messages so the delays
are consistent whether sleeping or not.
Bug: 5534004
Change-Id: I24118b30214dddf8183c1200a89555d6c528e606
