[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <5e35d79b-06e7-4b19-87fc-a2087a3434c2@kernel.org>
Date: Fri, 16 May 2025 14:50:32 +0200
From: Krzysztof Kozlowski <krzk@...nel.org>
To: zhangsenchuan@...incomputing.com, gregkh@...uxfoundation.org,
robh@...nel.org, krzk+dt@...nel.org, conor+dt@...nel.org,
linux-usb@...r.kernel.org, linux-kernel@...r.kernel.org,
devicetree@...r.kernel.org, Thinh.Nguyen@...opsys.com, p.zabel@...gutronix.de
Cc: ningyu@...incomputing.com, linmin@...incomputing.com,
yangwei1@...incomputing.com
Subject: Re: [PATCH v1 2/2] usb: dwc3: eic7700: Add EIC7700 usb driver
On 16/05/2025 11:54, zhangsenchuan@...incomputing.com wrote:
> +static ssize_t dwc3_mode_store(struct device *device,
> + struct device_attribute *attr, const char *buf,
> + size_t count)
> +{
> + struct dwc3_eswin *eswin = dev_get_drvdata(device);
> + struct dwc3 *dwc = eswin->dwc;
> + enum usb_role new_role;
> + struct usb_role_switch *role_sw = dwc->role_sw;
> +
> + if (!strncmp(buf, "1", 1) || !strncmp(buf, "host", 4)) {
> + new_role = USB_ROLE_HOST;
> + } else if (!strncmp(buf, "0", 1) || !strncmp(buf, "peripheral", 10)) {
> + new_role = USB_ROLE_DEVICE;
> + } else {
> + dev_info(eswin->dev, "illegal dr_mode\n");
> + return count;
> + }
> + eswin->force_mode = true;
> +
> + mutex_lock(&eswin->lock);
> + usb_role_switch_set_role(role_sw, new_role);
> + mutex_unlock(&eswin->lock);
> +
> + return count;
> +}
> +
> +static DEVICE_ATTR_RW(dwc3_mode);
Missing ABI documentation. Anyway, unlikely this will be accepted.
> +
> +static ssize_t dwc3_hub_rst_show(struct device *device,
> + struct device_attribute *attr, char *buf)
> +{
> + struct dwc3_eswin *eswin = dev_get_drvdata(device);
> +
> + if (!IS_ERR(eswin->hub_gpio))
> + return sprintf(buf, "%d", gpiod_get_raw_value(eswin->hub_gpio));
> +
> + return sprintf(buf, "UNKONWN");
> +}
> +
> +static ssize_t dwc3_hub_rst_store(struct device *device,
> + struct device_attribute *attr,
> + const char *buf, size_t count)
> +{
> + struct dwc3_eswin *eswin = dev_get_drvdata(device);
> +
> + if (!IS_ERR(eswin->hub_gpio)) {
> + if (!strncmp(buf, "0", 1))
> + gpiod_set_raw_value(eswin->hub_gpio, 0);
> + else
> + gpiod_set_raw_value(eswin->hub_gpio, 1);
> + }
> +
> + return count;
> +}
> +
> +static DEVICE_ATTR_RW(dwc3_hub_rst);
> +
> +static struct attribute *dwc3_eswin_attrs[] = {
> + &dev_attr_dwc3_mode.attr,
> + &dev_attr_dwc3_hub_rst.attr,
> + NULL,
> +};
> +
> +static struct attribute_group dwc3_eswin_attr_group = {
> + .name = NULL, /* we want them in the same directory */
> + .attrs = dwc3_eswin_attrs,
> +};
> +
> +static int dwc3_eswin_device_notifier(struct notifier_block *nb,
> + unsigned long event, void *ptr)
> +{
> + struct dwc3_eswin *eswin =
> + container_of(nb, struct dwc3_eswin, device_nb);
> +
> + mutex_lock(&eswin->lock);
> + eswin->new_usb_role = USB_ROLE_DEVICE;
> + mutex_unlock(&eswin->lock);
> + if (!eswin->suspended)
> + schedule_work(&eswin->otg_work);
> +
> + return NOTIFY_DONE;
> +}
> +
> +static int dwc3_eswin_host_notifier(struct notifier_block *nb,
> + unsigned long event, void *ptr)
> +{
> + struct dwc3_eswin *eswin = container_of(nb, struct dwc3_eswin, host_nb);
> +
> + mutex_lock(&eswin->lock);
> + eswin->new_usb_role = USB_ROLE_HOST;
> + mutex_unlock(&eswin->lock);
> + if (!eswin->suspended)
> + schedule_work(&eswin->otg_work);
> +
> + return NOTIFY_DONE;
> +}
> +
> +static void dwc3_eswin_otg_extcon_evt_work(struct work_struct *work)
> +{
> + struct dwc3_eswin *eswin =
> + container_of(work, struct dwc3_eswin, otg_work);
> + struct usb_role_switch *role_sw = eswin->dwc->role_sw;
> +
> + if (true == eswin->force_mode)
> + return;
> + mutex_lock(&eswin->lock);
> + usb_role_switch_set_role(role_sw, eswin->new_usb_role);
> + mutex_unlock(&eswin->lock);
> +}
> +
> +static int dwc3_eswin_get_extcon_dev(struct dwc3_eswin *eswin)
> +{
> + struct device *dev = eswin->dev;
> + struct extcon_dev *edev;
> + s32 ret = 0;
> +
> + if (device_property_read_bool(dev, "extcon")) {
extcon is not a bool. This is just wrong... plus undocumented ABI.
> + edev = extcon_get_edev_by_phandle(dev, 0);
> + if (IS_ERR(edev)) {
> + if (PTR_ERR(edev) != -EPROBE_DEFER)
> + dev_err(dev, "couldn't get extcon device\n");
> + return PTR_ERR(edev);
Do not open code dev_err_probe.
> + }
> + eswin->edev = edev;
> + eswin->device_nb.notifier_call = dwc3_eswin_device_notifier;
> + ret = devm_extcon_register_notifier(dev, edev, EXTCON_USB,
> + &eswin->device_nb);
> + if (ret < 0)
> + dev_err(dev, "failed to register notifier for USB\n");
> +
> + eswin->host_nb.notifier_call = dwc3_eswin_host_notifier;
> + ret = devm_extcon_register_notifier(dev, edev, EXTCON_USB_HOST,
> + &eswin->host_nb);
> + if (ret < 0)
> + dev_err(dev,
> + "failed to register notifier for USB-HOST\n");
> + }
> +
> + return 0;
> +}
> +
> +static int __init dwc3_eswin_deassert(struct dwc3_eswin *eswin)
That's wrong annotation. You did not build your kernel with DEBUG
SECTION MISMATCH.
> +{
> + int rc;
> +
> + if (eswin->vaux_rst) {
> + rc = reset_control_deassert(eswin->vaux_rst);
> + WARN_ON(rc != 0);
> + }
> +
> + return 0;
> +}
> +
> +static int dwc3_eswin_assert(struct dwc3_eswin *eswin)
> +{
> + int rc = 0;
> +
> + if (eswin->vaux_rst) {
> + rc = reset_control_assert(eswin->vaux_rst);
> + WARN_ON(rc != 0);
> + }
> +
> + return 0;
> +}
> +
> +static int dwc_usb_clk_init(struct device *dev)
> +{
> + struct regmap *regmap;
> + u32 hsp_usb_bus;
> + u32 hsp_usb_axi_lp;
> + u32 hsp_usb_vbus_freq;
> + u32 hsp_usb_mpll;
> + int ret;
> +
> + regmap = syscon_regmap_lookup_by_phandle(dev->of_node,
> + "eswin,hsp_sp_csr");
> + if (IS_ERR(regmap)) {
> + dev_dbg(dev, "No hsp_sp_csr phandle specified\n");
> + return -1;
> + }
> + ret = of_property_read_u32_index(dev->of_node, "eswin,hsp_sp_csr", 1,
> + &hsp_usb_bus);
> + if (ret) {
> + dev_err(dev, "can't get usb sid cfg reg offset (%d)\n", ret);
> + return ret;
> + }
> + ret = of_property_read_u32_index(dev->of_node, "eswin,hsp_sp_csr", 2,
> + &hsp_usb_axi_lp);
> + if (ret) {
> + dev_err(dev, "can't get usb sid cfg reg offset (%d)\n", ret);
> + return ret;
> + }
> + ret = of_property_read_u32_index(dev->of_node, "eswin,hsp_sp_csr", 3,
> + &hsp_usb_vbus_freq);
> + if (ret) {
> + dev_err(dev, "can't get usb sid cfg reg offset (%d)\n", ret);
> + return ret;
> + }
> + ret = of_property_read_u32_index(dev->of_node, "eswin,hsp_sp_csr", 4,
> + &hsp_usb_mpll);
> + if (ret) {
> + dev_err(dev, "can't get usb sid cfg reg offset (%d)\n", ret);
> + return ret;
> + }
> +
> + /*
> + * usb1 clock init
> + * ref clock is 24M, below need to be set to satisfy usb phy requirement(125M)
> + */
> + regmap_write(regmap, hsp_usb_vbus_freq, HSP_USB_VBUS_FSEL);
> + regmap_write(regmap, hsp_usb_mpll, HSP_USB_MPLL_DEFAULT);
> + /*
> + * reset usb core and usb phy
> + */
> + regmap_write(regmap, hsp_usb_bus,
> + HSP_USB_BUS_FILTER_EN | HSP_USB_BUS_CLKEN_GM |
> + HSP_USB_BUS_CLKEN_GS | HSP_USB_BUS_SW_RST |
> + HSP_USB_BUS_CLK_EN);
> + regmap_write(regmap, hsp_usb_axi_lp,
> + HSP_USB_AXI_LP_XM_CSYSREQ | HSP_USB_AXI_LP_XS_CSYSREQ);
> +
> + return 0;
> +}
> +
> +static int dwc3_eswin_probe(struct platform_device *pdev)
> +{
> + struct dwc3_eswin *eswin;
> + struct device *dev = &pdev->dev;
> + struct device_node *np = dev->of_node, *child;
> + struct platform_device *child_pdev;
> + unsigned int count;
> + int ret;
> + int i;
> + int err_desc = 0;
> +
> + eswin = devm_kzalloc(dev, sizeof(*eswin), GFP_KERNEL);
> + if (!eswin)
> + return -ENOMEM;
> + eswin->hub_gpio = devm_gpiod_get(dev, "hub-rst", GPIOD_OUT_HIGH);
> + err_desc = IS_ERR(eswin->hub_gpio);
> +
> + if (!err_desc)
> + gpiod_set_raw_value(eswin->hub_gpio, 1);
> +
> + count = of_clk_get_parent_count(np);
> + if (!count)
> + return -ENOENT;
> +
> + eswin->num_clocks = count;
> + eswin->force_mode = false;
> + eswin->clks = devm_kcalloc(dev, eswin->num_clocks, sizeof(struct clk *),
> + GFP_KERNEL);
> + if (!eswin->clks)
> + return -ENOMEM;
> +
> + platform_set_drvdata(pdev, eswin);
> +
> + mutex_init(&eswin->lock);
> +
> + eswin->dev = dev;
> +
> + mutex_lock(&eswin->lock);
I don't understand the point of it. Explain me, what are you protecting
here from what? How is it possible?
> +
> + for (i = 0; i < eswin->num_clocks; i++) {
> + struct clk *clk;
> +
> + clk = of_clk_get(np, i);
No, use devm_clk_Get
> + if (IS_ERR(clk)) {
> + ret = PTR_ERR(clk);
> + goto err0;
> + }
> + ret = clk_prepare_enable(clk);
> + if (ret < 0) {
> + clk_put(clk);
> + goto err0;
> + }
> +
> + eswin->clks[i] = clk;
Use get_enabled and bulk api.
> + }
> +
> + eswin->vaux_rst = devm_reset_control_get(dev, "vaux");
> + if (IS_ERR_OR_NULL(eswin->vaux_rst)) {
OR_NULL? Why?
> + dev_err(dev, "Failed to asic0_rst handle\n");
Syntax is always: retrun dev_err_probe
> + return -EFAULT;
No, return proper error codes.
> + }
> +
> + dwc3_eswin_deassert(eswin);
> + dwc_usb_clk_init(dev);
> +
> + pm_runtime_set_active(dev);
> + pm_runtime_enable(dev);
> + ret = pm_runtime_get_sync(dev);
> + if (ret < 0) {
> + dev_err(dev, "get_sync failed with err %d\n", ret);
> + goto err1;
> + }
> +
> + child = of_get_child_by_name(np, "dwc3");
> + if (!child) {
> + dev_err(dev, "failed to find dwc3 core node\n");
> + ret = -ENODEV;
> + goto err1;
> + }
> + /* Allocate and initialize the core */
> + ret = of_platform_populate(np, NULL, NULL, dev);
> + if (ret) {
> + dev_err(dev, "failed to create dwc3 core\n");
> + goto err1;
> + }
> +
> + INIT_WORK(&eswin->otg_work, dwc3_eswin_otg_extcon_evt_work);
> + child_pdev = of_find_device_by_node(child);
> + if (!child_pdev) {
> + dev_err(dev, "failed to find dwc3 core device\n");
> + ret = -ENODEV;
> + goto err2;
> + }
> + eswin->dwc = platform_get_drvdata(child_pdev);
> + if (!eswin->dwc) {
> + dev_err(dev, "failed to get drvdata dwc3\n");
> + ret = -EPROBE_DEFER;
> + goto err2;
> + }
> + eswin->child_dev = &child_pdev->dev;
> + ret = dwc3_eswin_get_extcon_dev(eswin);
> + if (ret < 0)
> + dev_err(dev, "couldn't get extcon device: %d\n", ret);
> +
> + mutex_unlock(&eswin->lock);
> + ret = sysfs_create_group(&dev->kobj, &dwc3_eswin_attr_group);
> + if (ret)
> + dev_err(dev, "failed to create sysfs group: %d\n", ret);
> +
> + return ret;
> +err2:
> + cancel_work_sync(&eswin->otg_work);
> + of_platform_depopulate(dev);
> +
> +err1:
> + pm_runtime_put_sync(dev);
> + pm_runtime_disable(dev);
> + dwc3_eswin_assert(eswin);
> +err0:
> + for (i = 0; i < eswin->num_clocks && eswin->clks[i]; i++) {
> + if (!pm_runtime_status_suspended(dev))
> + clk_disable(eswin->clks[i]);
> + clk_unprepare(eswin->clks[i]);
> + clk_put(eswin->clks[i]);
> + }
> +
> + mutex_unlock(&eswin->lock);
> +
> + return ret;
> +}
> +
> +static void dwc3_eswin_remove(struct platform_device *pdev)
> +{
> + struct dwc3_eswin *eswin = platform_get_drvdata(pdev);
> + struct device *dev = &pdev->dev;
> + int i = 0;
> +
> + cancel_work_sync(&eswin->otg_work);
> +
> + sysfs_remove_group(&dev->kobj, &dwc3_eswin_attr_group);
> +
> + /* Restore hcd state before unregistering xhci */
> + if (eswin->edev && !eswin->connected) {
> + struct usb_hcd *hcd = dev_get_drvdata(&eswin->dwc->xhci->dev);
> +
> + pm_runtime_get_sync(dev);
> +
> + /*
> + * The xhci code does not expect that HCDs have been removed.
> + * It will unconditionally call usb_remove_hcd() when the xhci
> + * driver is unloaded in of_platform_depopulate(). This results
> + * in a crash if the HCDs were already removed. To avoid this
> + * crash, add the HCDs here as dummy operation.
> + * This code should be removed after pm runtime support
> + * has been added to xhci.
> + */
> + if (hcd->state == HC_STATE_HALT) {
> + usb_add_hcd(hcd, hcd->irq, IRQF_SHARED);
> + usb_add_hcd(hcd->shared_hcd, hcd->irq, IRQF_SHARED);
> + }
> + }
> +
> + of_platform_depopulate(dev);
> +
> + pm_runtime_put_sync(dev);
> + pm_runtime_disable(dev);
> +
> + dwc3_eswin_assert(eswin);
> + for (i = 0; i < eswin->num_clocks; i++) {
> + if (!pm_runtime_status_suspended(dev))
> + clk_disable(eswin->clks[i]);
> + clk_unprepare(eswin->clks[i]);
> + clk_put(eswin->clks[i]);
> + }
> +}
> +
> +#ifdef CONFIG_PM
> +static int dwc3_eswin_runtime_suspend(struct device *dev)
> +{
> + struct dwc3_eswin *eswin = dev_get_drvdata(dev);
> + int i;
> +
> + for (i = 0; i < eswin->num_clocks; i++)
> + clk_disable(eswin->clks[i]);
> +
> + device_init_wakeup(dev, false);
> +
> + return 0;
> +}
> +
> +static int dwc3_eswin_runtime_resume(struct device *dev)
> +{
> + struct dwc3_eswin *eswin = dev_get_drvdata(dev);
> + int i;
> +
> + for (i = 0; i < eswin->num_clocks; i++)
> + clk_enable(eswin->clks[i]);
> +
> + device_init_wakeup(dev, true);
This feels odd. What is the point of wakeup if you disable it for the
sleeping periods?
> +
> + return 0;
> +}
> +
> +static int __maybe_unused dwc3_eswin_suspend(struct device *dev)
> +{
> + struct dwc3_eswin *eswin = dev_get_drvdata(dev);
> + struct dwc3 *dwc = eswin->dwc;
> +
> + eswin->suspended = true;
> + cancel_work_sync(&eswin->otg_work);
> +
> + /*
> + * The flag of is_phy_on is only true if
> + * the DWC3 is in Host mode.
> + */
> + if (eswin->is_phy_on) {
> + phy_power_off(dwc->usb2_generic_phy[0]);
> +
> + /*
> + * If link state is Rx.Detect, it means that
> + * no usb device is connecting with the DWC3
> + * Host, and need to power off the USB3 PHY.
> + */
> + dwc->link_state = dwc3_gadget_get_link_state(dwc);
> + if (dwc->link_state == DWC3_LINK_STATE_RX_DET)
> + phy_power_off(dwc->usb3_generic_phy[0]);
> + }
> +
> + return 0;
> +}
> +
> +static int __maybe_unused dwc3_eswin_resume(struct device *dev)
> +{
> + struct dwc3_eswin *eswin = dev_get_drvdata(dev);
> + struct dwc3 *dwc = eswin->dwc;
> +
> + eswin->suspended = false;
> +
> + if (eswin->is_phy_on) {
> + phy_power_on(dwc->usb2_generic_phy[0]);
> +
> + if (dwc->link_state == DWC3_LINK_STATE_RX_DET)
> + phy_power_on(dwc->usb3_generic_phy[0]);
> + }
> +
> + if (eswin->edev)
> + schedule_work(&eswin->otg_work);
> +
> + return 0;
> +}
> +
> +static const struct dev_pm_ops dwc3_eswin_dev_pm_ops = {
> + SET_SYSTEM_SLEEP_PM_OPS(dwc3_eswin_suspend, dwc3_eswin_resume)
> + SET_RUNTIME_PM_OPS(dwc3_eswin_runtime_suspend,
> + dwc3_eswin_runtime_resume, NULL)
> +};
> +
> +#define DEV_PM_OPS (&dwc3_eswin_dev_pm_ops)
> +#else
> +#define DEV_PM_OPS NULL
> +#endif /* CONFIG_PM */
> +
> +static const struct of_device_id eswin_dwc3_match[] = {
> + { .compatible = "eswin,eic7700-dwc3" },
> + { /* Sentinel */ }
> +};
> +
> +MODULE_DEVICE_TABLE(of, eswin_dwc3_match);
> +
> +static struct platform_driver dwc3_eswin_driver = {
> + .probe = dwc3_eswin_probe,
> + .remove = dwc3_eswin_remove,
> + .driver = {
> + .name = "eic7700-dwc3",
> + .pm = DEV_PM_OPS,
> + .of_match_table = eswin_dwc3_match,
> + },
> +};
> +
> +module_platform_driver(dwc3_eswin_driver);
> +
> +MODULE_ALIAS("platform:eic7700-dwc3");
Drop. You should not need MODULE_ALIAS() in normal cases. If you need
it, usually it means your device ID table is wrong (e.g. misses either
entries or MODULE_DEVICE_TABLE()). MODULE_ALIAS() is not a substitute
for incomplete ID table.
Both of your drivers - this and PCI - are in very poor shape. I suggest
redoing them based on latest upstream drivers, not pushing to us your
downstream code.
Best regards,
Krzysztof
Powered by blists - more mailing lists