Enable the vaux regulator at probe time and keep it enabled for the
entire PCIe controller lifecycle. This ensures support for outbound
wake-up mechanism such as WAKE# signaling.

Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
---
 drivers/pci/controller/dwc/pci-imx6.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
index 5a38cfaf989b1..1c1dce2d87e44 100644
--- a/drivers/pci/controller/dwc/pci-imx6.c
+++ b/drivers/pci/controller/dwc/pci-imx6.c
@@ -159,6 +159,7 @@ struct imx_pcie {
 	u32			tx_deemph_gen2_6db;
 	u32			tx_swing_full;
 	u32			tx_swing_low;
+	struct regulator	*vaux;
 	struct regulator	*vpcie;
 	struct regulator	*vph;
 	void __iomem		*phy_base;
@@ -1739,6 +1740,20 @@ static int imx_pcie_probe(struct platform_device *pdev)
 	pci->max_link_speed = 1;
 	of_property_read_u32(node, "fsl,max-link-speed", &pci->max_link_speed);
 
+	/*
+	 * Refer to PCIe CEM r6.0, sec 2.3 WAKE# Signal, WAKE# signal is only
+	 * asserted by the Add-in Card when all its functions are in D3Cold
+	 * state and at least one of its functions is enabled for wakeup
+	 * generation. The 3.3V auxiliary power (+3.3Vaux) must be present and
+	 * used for wakeup process. Since the main power supply would be gated
+	 * off to let Add-in Card to be in D3Cold, get the vaux and keep it
+	 * enabled to power up WAKE# circuit for the entire PCIe controller
+	 * lifecycle when WAKE# is supported.
+	 */
+	ret = devm_regulator_get_enable_optional(&pdev->dev, "vaux");
+	if (ret < 0 && ret != -ENODEV)
+		return dev_err_probe(dev, ret, "failed to enable vaux");
+
 	imx_pcie->vpcie = devm_regulator_get_optional(&pdev->dev, "vpcie");
 	if (IS_ERR(imx_pcie->vpcie)) {
 		if (PTR_ERR(imx_pcie->vpcie) != -ENODEV)
-- 
2.37.1

On Tue, Aug 19, 2025 at 03:16:30PM GMT, Richard Zhu wrote:
> Enable the vaux regulator at probe time and keep it enabled for the
> entire PCIe controller lifecycle. This ensures support for outbound
> wake-up mechanism such as WAKE# signaling.
> 
> Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
> ---
>  drivers/pci/controller/dwc/pci-imx6.c | 15 +++++++++++++++
>  1 file changed, 15 insertions(+)
> 
> diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
> index 5a38cfaf989b1..1c1dce2d87e44 100644
> --- a/drivers/pci/controller/dwc/pci-imx6.c
> +++ b/drivers/pci/controller/dwc/pci-imx6.c
> @@ -159,6 +159,7 @@ struct imx_pcie {
>  	u32			tx_deemph_gen2_6db;
>  	u32			tx_swing_full;
>  	u32			tx_swing_low;
> +	struct regulator	*vaux;
>  	struct regulator	*vpcie;
>  	struct regulator	*vph;
>  	void __iomem		*phy_base;
> @@ -1739,6 +1740,20 @@ static int imx_pcie_probe(struct platform_device *pdev)
>  	pci->max_link_speed = 1;
>  	of_property_read_u32(node, "fsl,max-link-speed", &pci->max_link_speed);
>  
> +	/*
> +	 * Refer to PCIe CEM r6.0, sec 2.3 WAKE# Signal, WAKE# signal is only
> +	 * asserted by the Add-in Card when all its functions are in D3Cold
> +	 * state and at least one of its functions is enabled for wakeup
> +	 * generation. The 3.3V auxiliary power (+3.3Vaux) must be present and
> +	 * used for wakeup process. Since the main power supply would be gated
> +	 * off to let Add-in Card to be in D3Cold, get the vaux and keep it
> +	 * enabled to power up WAKE# circuit for the entire PCIe controller
> +	 * lifecycle when WAKE# is supported.
> +	 */

This comment implies that the presence of Vaux is dependent on WAKE#. But there
is no such check present in the code. Maybe you are referring to the fact that
the platform will only supply Vaux if it intends to support WAKE#?

But I guess you can just drop the comment altogether and move it to patch
description.

- Mani

-- 
மணிவண்ணன் சதாசிவம்