When reading from the parcel, if the number of colors is invalid, early
exit.
Add two more checks: setInfo must return true, and Parcel::readInplace
must return non-NULL. The former ensures that the previously read values
(width, height, etc) were valid, and the latter checks that the Parcel
had enough data even if the number of colors was reasonable.
Also use an auto-deleter to handle deletion of the SkBitmap.
Cherry pick from change-Id: Icbd562d6d1f131a723724883fd31822d337cf5a6
BUG=19666945
Change-Id: Iab0d218c41ae0c39606e333e44cda078eef32291
Check dalvik.vm.dex2oat-threads in AndroidRuntime and pass to ART
with "-j" as a compiler option, if found.
Check dalvik.vm.image-dex2oat-threads in AndroidRuntime and pass to
ART with "-j" as an image compiler option, if found.
Bug: 19992386
(cherry picked from commit e0352388238875ef7e7b31d914fc903b08ff0f6b)
Change-Id: I5e7806cf560607d31a1d6901dffb14bee538c9cc
The GL layer gets exclusive use of the direct producer dimensions;
we should only set the user dimensions.
Bug: 16404608
Change-Id: Ie5b720104bdbee6b096b62520baa1c6b518c6c1c
Changes proposed by Ben Hawkes of Google Project Zero.
NinePatchPeeker.cpp:
Instead of asserting, return false for bad data.
ResourceTypes.h:
Store ninepatch values as unsigned.
BUG:19151999
Change-Id: Ibe35e7569f632c6bb8a34a7701e26bb6ed547ec2
(cherry picked from commit a730ef3f77fc495bc90199b4d45efab26d609782)
Changes proposed by Ben Hawkes of Google Project Zero.
NinePatchPeeker.cpp:
Instead of asserting, return false for bad data.
ResourceTypes.h:
Store ninepatch values as unsigned.
BUG:19151999
Change-Id: Ibe35e7569f632c6bb8a34a7701e26bb6ed547ec2
bug:18921087
This fixes issues where bitmaps are cached as textures, and not
reuploaded when reused by BitmapRegionDecoder, showing stale content.
Change-Id: I7626fd6acd36d8e38723689734855ce9154eb8c8
This exception may be generated through MediaRecorder method calls
as well, so clarify that the issue is an invalid camera object.
Bug: 18414095
Change-Id: I189ab1a71ab0d6a432b87c45a82c90fe11d22ef8
Bug: 18521508
Fix an issue where an RNA's native object was destroyed
before the java-side object was started
Change-Id: I487fb476e0ecdf7000515f4f7320e8cfbc50a52b
You can now do "adb shell dumpsys procstats --start-testing" to
enable high frequency pss sampling.
Also improved the low on RAM mem reporting to separate out RAM
from memtrack, in case the data we are getting from that is bad.
And fixed meminfo --oom to work correctly again.
Change-Id: I7af17eab110a82298bd7b0ce381f8fa5c96c1f6a
Bug: 18203577
The issue occurs as a result of performTraversals() both doing
a window relayout call *and* early-returning because it's not dirty.
To fix this pauseSurface() returns whether or not the RT-side is
"dirty" to force ViewRootImpl to do a draw even if mDirty is
otherwise empty.
Change-Id: I534f367e75d18d273ebf14df3927f5c464ef6bef
Bug: 18204974
Even if we are canceling or ending an animation
nStart() *must* be called, otherwise the native-side listener
is not attached (lazy-attached for JNI cyclic reference reasons),
and then Animator::callOnFinishedListener() no-ops as there's
no listener set
Add a lifecycle verifier to ensure that nStart is always
called on animators that are attached and get finished
Change-Id: Ibc345b5be97b6d3f95a11c361ebe020d030fd3b6
Let AndroidRuntime check for /system/etc/compiled-classes and push
it to the runtime for boot image creation.
Bug: 18410571
(cherry picked from commit ca775941f3b7981aabf3a6a3b84d6c94f4f76aff)
Change-Id: I04d195c12def4bf1136ed8c6b3dc439b640de7b2
Let AndroidRuntime check for /system/etc/compiled-classes and push
it to the runtime for boot image creation.
Bug: 18410571
Change-Id: I2510316381f2661166af24d7e14b013f4e045556
