Quoting Peter Chen (2016-06-29 18:18:27)

On Wed, Jun 29, 2016 at 11:59:21AM -0700, Stephen Boyd wrote:

quoted

Quoting Peter Chen (2016-06-28 23:32:11)

quoted

On Sun, Jun 26, 2016 at 12:28:27AM -0700, Stephen Boyd wrote:

quoted

The core framework already handles setting this parameter with a
platform quirk. Add the appropriate flag so that we always set
AHBBURST to 0. Technically DT should be doing this, but we always
do it for msm chipidea devices so setting the flag in the driver
works just as well.

You still need to set AHB burst value at dts, this flag is just for
override, see below:

ahb-burst-config = <0x0>;

Right, I have added that to dts now, but the CI_HDRC_OVERRIDE_AHB_BURST
flag allows us to specify it from the platdata structure in the
ci_hdrc_msm.c file. As the value is zero for msm type controllers, I
left it out of the static definition of platdata because all the
non-initialized members of that structure are going to be zero anyway. I
can explicitly set it to zero to make it more clear if you like.

I suggest setting it explicitly at dts, at current code, it is set
as zero explicitly too:)

Yes I'm making sure to always set it in dts as well. Should we drop this
patch and require dts to always have it set if needed? I was setting it
in the driver in case we had some older dts lying around, but I suppose
this is changing significantly enough where the qcom,ci_hdrc_msm binding
needs to be redone anyway. I'd still like to remove writing the value in
the reset callback though.