Has anyone android working with andy-tracking?

Michael Trimarchi trimarchi at gandalf.sssup.it
Sun Feb 8 17:39:48 CET 2009


Andy Green wrote:
> Hash: SHA1
> Somebody in the thread at some point said:
> | Michael, if this wakelock stuff is just a local addition you made and
> | not part of android, it's better we try to get rid of it.
> |
> | If Google are creating it though, maybe they have a larger plan and we
> | need to deal with that.
> |
> | Can you clarify where the wakelock code came from?
> |
> |> Wakelock are google code :).
> Thanks.
> Do we know if Google add the wakelock code because they will have
> complex multistep transactions in kernel space?
> If they do plan this we should find how to live with it, because the
> wakelocks are going to be needed then to protect the kernel-side
> transactions.
> If it's only working around stuff we know we can do by cooperation in
> userspace, basically by filtering before issuing the suspend action, and
> you can arrange that in Android, then we don't need it (although I
> wonder if all the potential races there are really take care of).
I give a full answer on wedsday because I have a deadline in an European 
> - -Andy
> Version: GnuPG v1.4.9 (GNU/Linux)
> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
> iEYEARECAAYFAkmO+4EACgkQOjLpvpq7dMrHXQCeKXOfrzY52Xb3FRfbOFqzzt2z
> qvYAn0SDrDCE6ZNrBwi/qoOPLlQpx9ML
> =4bGH

More information about the openmoko-kernel mailing list