On 03/03, Dwivedi, Avaneesh Kumar (avani) wrote:
> On 2/28/2017 4:18 AM, Stephen Boyd wrote:
> >On 01/30, Avaneesh Kumar Dwivedi wrote:
> >>@@ -1213,6 +1299,47 @@ static int q6v5_remove(struct platform_device *pdev)
> >> return 0;
> >> }
> >>+static const struct rproc_hexagon_res msm8996_mss = {
> >>+ .hexagon_mba_image = "mba.mbn",
> >>+ .proxy_supply = (struct qcom_mss_reg_res[]) {
> >>+ {
> >>+ .supply = "vdd_mx",
> >>+ .uV = 6,
> >>+ },
> >>+ {
> >>+ .supply = "vdd_cx",
> >>+ .uV = 7,
> >>+ .uA = 100000,
> >>+ },
> >vdd cx and vdd mx are corners. The plan is to _not_ use the
> >regulator framework for those, so treating them like supplies is
> >incorrect here.
> vdd cx and mx though in downstream are voted for corner but they are
> always ON domain upstream as per regulator team when i discussed
> with them.
> should i drop them altogether?
I would say yes, drop them. The on/off state doesn't matter here.
This code wants to max out the corner for a period of time until
the remote processor has booted far enough to make their own vote
on these RPM resources.
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
