On Tue, Mar 06, 2018 at 09:31:29PM +0000, Mark Brown wrote:

On Tue, Mar 06, 2018 at 09:25:25AM -0800, Greg KH wrote:

quoted

On Tue, Mar 06, 2018 at 02:26:34PM +0000, Mark Brown wrote:
Really?  Like what?  Last I looked it's only about 300 or so patches.
Something like less than .5% of the normal SoC backport size for any ARM
system recently.  There were some numbers published a few months ago
about the real count, I can dig them up if you are curious.

Really.  

The Android tree is making non-trivial modifications adding new features
in core bits of the kernel like the scheduler - that's got an impact
which will have follow on validation costs if it's not introduced early
on in the process.

Ah crap, I forgot all about the scheduler mess.  Sorry about that, I've
been focusing on device-specific stuff too much (nothing I much care
lives outside of drivers/...)

So you are right, for some types of devices, the andoid-common tree
isn't the best idea.  However, the validation "costs" better be starting
all over based on the intrusiveness of these patches, they are
non-trivial :(

thanks,

greg k-h