Hello,

I posted about this issue on android-platform some time ago
and got no reply. Sorry for crossposting, but this is quite
important for me right now. Perhaps someone here can bail me
out!

I am still [1] trying to evaluate Android's IPC mechanism. For this
I use a small test app calculating random numbers once in one
Activity and a second time via a Service.

I would like to answer the question, how much overhead IPC via Binder
in Android causes and which parts exactly.

These were my steps:
- call android.os.Debug.enableEmulatorTraceOutput() in the method
  implementing the service and in the client calling it (necessary?)
- start the emulator with the -trace flag
- enable tracing in QEMU just before I call my test app by pressing F9
- convert the QEMU trace so I can open it with traceview or create
html
  statistics (Fadden's tip from the same thread [1] helped here)

My Service lives in server.dom.da and the client in client.dom.da.
When I grep through the trace I can only find these package names
mentioned
once at the beginning of the dmtrace file. Why do they show up in
traceview, but
not once in the html statistics?

I also tried getting trace information with
android.os.Debug.startMethodTracing(),
but as expected all CPU-time spent in the Binder is not logged here.

I uploaded my trace, if anyone would like to take look, you can grab
the files separately or as a tar from [2]. My test app can be found
in
RNS_14.01.09.zip.

Thanks for any help!


[1]
http://groups.google.com/group/android-platform/browse_thread/thread/96290e48ca18ae65/0f09283f9cd7abb1?lnk=gst&q=qtools#0f09283f9cd7abb1
[2] http://debworld.de/~don/mytrace/
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"android-framework" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/android-framework?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to