On Sat, May 9, 2015 at 7:44 PM, Maxime Ripard
[off-list ref] wrote:

On Wed, May 06, 2015 at 10:47:33PM +0200, Jens Kuske wrote:

quoted

quoted

quoted

+ *     You should have received a copy of the GNU General Public
+ *     License along with this file; if not, write to the Free
+ *     Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston,
+ *     MA 02110-1301 USA

Could you remove that last paragraph?
It generates a checkpatch warning.

Sure, will be removed. Just copied it from some other sunxi dtsi.

Yeah, I know, I'm even the one that introduced this in the first place
:)

I sent a patch earlier this week to remove it from the other DT.

quoted

quoted

quoted

+          ahb12_rst: reset@01c202c0 {
+                  #reset-cells = <1>;
+                  compatible = "allwinner,sun6i-a31-clock-reset";
+                  reg = <0x01c202c0 0xc>;
+          };

This reset controller also resets the timers, it should be initialised
much earlier.

What about having an allwinner,sun8i-h3-bus-reset, and adding it to
the list of compatibles to initialise earlier in
drivers/reset/reset-sunxi.c?

Of course, it would cover the other reset controllers that you have
below.

You mean using a single bus_rst instead of the three?

Yes.

quoted

Or, why not using allwinner,sun6i-a31-ahb1-reset for ahb12_rst

Strictly speaking, they do not control the same set of devices. I'd
prefer to have a different compatible in case we need to setup a
particular behaviour on a given SoC (for example, force out of reset a
particular device, even if no driver is actually using it), without
impacting the other.

quoted

and adding a .init_time = sun6i_timer_init to the sun8i machine.

But we will need to do that yes.

quoted

I'm a bit confused here now, because for A23, which is almost
identical, it got removed after your comment:
http://lists.infradead.org/pipermail/linux-arm-kernel/2014-June/265064.html

Hmmmm, I think I somehow overlooked the fact that the timer was there,
even though Chen-Yu said it. My bad :/

On the A23, the high speed timer block only has 1 timer. The sun5i-hrtimer
driver requires 2, and turns out we weren't using them anyway, so I just
dropped sun5i-hrtimer support on A23.

If the other sun8i SoCs have 2 or more timers, feel free to support them.

ChenYu