On Tue, 2015-06-02 at 10:12 -0500, "Stéphane Viau" wrote:
Hi Ivan,
Hi Stephane,
On Mon, 2015-06-01 at 16:28 -0400, Stephane Viau wrote:
Some targets (eg: msm8994) use the pinctrl framework to configure interface pins. This change adds support for initialization and pinctrl active/sleep state control for the HDMI driver.
Signed-off-by: Stephane Viau sviau@codeaurora.org
v2:
- Add devicetree binding documentation for pinctrl property [Ivan]
- Use pinctrl framework's PINCTRL_STATE_DEFAULT/SLEEP states [Ivan]
<snip>
static int hdmi_bind(struct device *dev, struct device *master, void *data) @@ -365,6 +379,7 @@ static int hdmi_bind(struct device *dev, struct device *master, void *data) #ifdef CONFIG_OF struct device_node *of_node = dev->of_node; const struct of_device_id *match;
struct pinctrl *pinctrl; match = of_match_node(dt_match, of_node); if (match && match->data) {@@ -383,6 +398,18 @@ static int hdmi_bind(struct device *dev, struct device *master, void *data) hdmi_cfg->mux_sel_gpio = get_gpio(dev, of_node, "qcom,hdmi-tx-mux-sel"); hdmi_cfg->mux_lpm_gpio = get_gpio(dev, of_node, "qcom,hdmi-tx-mux-lpm");
/* not all targets have pinctrl, do not fail in case of error:*/
pinctrl = devm_pinctrl_get(dev);Probably I have to be more explicit. Why not using pins binding handled in driver really_probe()? I have to admit that I am not familiar with DRM subsystem.
This would work, indeed, for default/sleep/idle states.
I actually had in mind that we'd need to keep track of HDMI pinctrl states because we may need to add a couple more in the near future in order to independently enable/disable certain parts of the HDMI controller (eg: HPD, DDC, CEC..).
Each of this HW sub-sections of the controller are driven by a different "pin" in the downstream driver... But since this is nowhere close to being upstream-ed yet, I'll go ahead with your idea of using the common pins binding (v3 to follow).
v3? Why we need a patch if we already have required support?
Regards, Ivan