We see abort messages like this when using JavaPixelAllocator and JavaMemoryUsageReporter.
W/dalvikvm( 680): JNI WARNING: threadid=2 using env from threadid=10
W/dalvikvm( 680): in Landroid/graphics/LargeBitmap;.nativeClean (I)V (CallVoidMethodV)
To fix it, we keep JavaVM, rather than JNIEnv, in JavaPixelAllocator and JavaMemoryUsageReporter,
because JavaVM allows us to get the JNIEnv corresponds to the current thread.
Change-Id: Ibd4b394a53dd3fdccc5a442eeb0dedf836479575
A usecase is to use hipri to keep alive 3g connections when wifi
is about to be brought up. They are the same network, so don't
reset unless both are down.
Change-Id: I8dabc42f315943b28ea34721793b0ee330c07bf6
system_server can potentially monitor uevent devices that are
of no use to it. For instance, an accelerometer implementation
as uevents. This would cause the process to be busy when
unnecessary. If a device cannot be classified, don't monitor it.
Change-Id: Ib2c93105e7d746d9c1a7414bea8bab3fb8c0b70a
On single-touch devices, pointer up/down is signalled by a BTN_TOUCH
key event. Previously we handled BTN_TOUCH immediately but some drivers
may produce the sequence BTN_TOUCH, ABS_X, ABS_Y, SYN_REPORT on pointer down
which caused us to emit a bad initial pointer down location.
Now we wait for SYN_REPORT before reporting the up or down.
On multi-touch devices, pointer up can be signalled by as little as
the sequence SYN_MT_REPORT, SYN_REPORT. This change ensures that we
handle this case.
Added support for reading ABS_MT_PRESSURE when available.
Corrected mapping of touchMajor/touchMinor on single touch devices.
Minor code cleanup.
Change-Id: Ic7ec4811241ed85a06e59b8a839ca05180d491d4
Sometimes the wrong fd was accessed when the device was addressed
by device id.
The earlier implementation assumed that two arrays were in sync
but one of them was compacted when devices were removed. Instead
of that dependency the device now keeps track of it's file descriptor.
Change-Id: Ib0f320603aafb07ded354bc3687de9759c9068f2
Merge commit 'a34c9ca30ea25d0c4d8ae51da7858c4ea58b7f25' into gingerbread
* commit 'a34c9ca30ea25d0c4d8ae51da7858c4ea58b7f25':
Add more error checking for ndc
In NativeDaemonConnector.doCommand() calls, there was inconsistent error
checking. This change adds error checking for every call and makes it so
that any call to .doCommand() that gets an error code won't cause the
code to hang forever.
Change-Id: If714282b6642f278fb8137f652af1a012670253b
The previous implementation fails to work properly when the .apk
and installed versions of the binaries have the same size and date.
Change-Id: I063817a935da9ad459858d7eec8bb3d940607850
When in a call, different audio modes need to be applied based on phone type.
(For example, SIP call needs to be in MODE_NORMAL while GSM call in MODE_IN_CALL.)
Originally, it's handled in PhoneUtils.setAudioMode(). It makes more sense now
to handle the actual logic in CallManager.
Change-Id: I58d8f31d6b4afe22f88da442daac2010781de801
Previously, the input dispatcher assumed that the input channel's
receive pipe file descriptor was a sufficiently unique identifier for
looking up input channels in its various tables. However, it can happen
that an input channel is disposed and then a new input channel is
immediately created that reuses the same file descriptor. Ordinarily
this is not a problem, however there is a small opportunity for a race
to arise in InputQueue.
When InputQueue receives an input event from the dispatcher, it
generates a finishedToken that encodes the channel's receive pipe fd,
and a sequence number. The finishedToken is used by the ViewRoot
as a handle for the event so that it can tell the InputQueue when
the event has finished being processed.
Here is the race:
1. InputQueue receives an input event, assigns a new finishedToken.
2. ViewRoot begins processing the input event.
3. During processing, ViewRoot unregisters the InputChannel.
4. A new InputChannel is created and is registered with the Input Queue.
This InputChannel happens to have the same receive pipe fd as
the one previously registered.
5. ViewRoot tells the InputQueue that it has finished processing the
input event, passing along the original finishedToken.
6. InputQueue throws an exception because the finishedToken's receive
pipe fd is registered but the sequence number is incorrect so it
assumes that the client has called finish spuriously.
The fix is to include a unique connection id within the finishedToken so
that the InputQueue can accurately confirm that the token belongs to
the currently registered InputChannel rather than to an old one that
happened to have the same receive pipe fd. When it notices this, it
ignores the spurious finish.
I've also made a couple of other small changes to avoid similar races
elsewhere.
This patch set also includes a fix to synthesize a finished signal
when the input channel is unregistered on the client side to
help keep the server and client in sync.
Bug: 2834068
Change-Id: I1de34a36249ab74c359c2c67a57e333543400f7b
