Re: [PATCH v3 6/6] PCI: pwrctl: Add power control driver for qps615

[Bjorn Helgaas]

On Tue, Nov 12, 2024 at 08:31:38PM +0530, Krishna chaitanya chundru wrote:
> QPS615 is the PCIe switch which has one upstream and three downstream
> ports. To one of the downstream ports ethernet MAC is connected as endpoint
> device. Other two downstream ports are supposed to connect to external
> device. One Host can connect to QPS615 by upstream port. QPS615 switch
> needs to be configured after powering on and before PCIe link was up.

s/is the/is a/

> The PCIe controller driver already enables link training at the host side
> even before qps615 driver probe happens, due to this when driver enables
> power to the switch it participates in the link training and PCIe link
> may come up before configuring the switch through i2c. To prevent the
> host from participating in link training, disable link training on the
> host side to ensure the link does not come up before the switch is
> configured via I2C.
> 
> Based up on dt property and type of the port, qps615 is configured
> through i2c.

> +config PCI_PWRCTL_QPS615
> +	tristate "PCI Power Control driver for QPS615"
> +	select PCI_PWRCTL
> +	help
> +	  Say Y here to enable the pwrctl driver for Qualcomm
> +	  QPS615 PCIe switch which enables and configures it
> +	  through i2c.

Does this work if build as a module?  I guess maybe the QPS615 doesn't
even get powered on without this, so it would be configured and then
powered on after module-load time?

s/pwrctl/pwrctrl/ (also in filename, code, variables, etc below)

Will need to be rebased to account for the pwrctl -> pwrctrl rename in
v6.13-rc1.

> +static int qps615_pwrctl_power_on(struct qps615_pwrctl_ctx *ctx)
> +{
> +	struct qps615_pwrctl_cfg *cfg;
> +	int ret, i;
> +
> +	ret = regulator_bulk_enable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
> +	if (ret < 0)
> +		return dev_err_probe(ctx->pwrctl.dev, ret, "cannot enable regulators\n");
> +
> +	gpiod_set_value(ctx->reset_gpio, 0);
> +
> +	 /* wait for the internal osc frequency to stablise */

s/stablise/stabilise/

> +	usleep_range(10000, 10500);
> +
> +	ret = qps615_pwrctl_assert_deassert_reset(ctx, false);
> +	if (ret)
> +		goto out;
> +
> +	if (ctx->cfg[QPS615_USP].axi_freq_125) {
> +		ret = qps615_pwrctl_i2c_write(ctx->client, QPS615_BUS_CONTROL, BIT(16));
> +		if (ret)
> +			dev_err(ctx->pwrctl.dev, "Setting AXI clk freq failed %d\n", ret);
> +	}
> +
> +	for (i = 0; i < QPS615_MAX; i++) {
> +		cfg = &ctx->cfg[i];
> +		if (cfg->disable_port) {
> +			ret = qps615_pwrctl_disable_port(ctx, i);
> +			if (ret) {
> +				dev_err(ctx->pwrctl.dev, "Disabling port failed\n");
> +				goto out;
> +			}
> +		}

A helper function could remove a level of indentation for the body of
this loop.

> +static int qps615_pwrctl_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct pci_host_bridge *bridge;
> +	enum qps615_pwrctl_ports port;
> +	struct qps615_pwrctl_ctx *ctx;
> +	int ret, addr;
> +
> +	bridge = pci_find_host_bridge(to_pci_dev(dev->parent)->bus);

Move to where "bridge" is used.

> +	platform_set_drvdata(pdev, ctx);

Can this be moved to the point where we know we won't return an error?

Bjorn