On Mon, 22 Aug 2016, Peter Chen wrote:

On Mon, Aug 15, 2016 at 05:13:14PM +0800, Peter Chen wrote:

quoted

Some hard-wired USB devices need to do power sequence to let the
device work normally, the typical power sequence like: enable USB
PHY clock, toggle reset pin, etc. But current Linux USB driver
lacks of such code to do it, it may cause some hard-wired USB devices
works abnormal or can't be recognized by controller at all.

In this patch, it calls power sequence library APIs to finish
the power sequence events. At first, it calls pwrseq_alloc_generic
to create a generic power sequence instance, then it will do power
on sequence at hub's probe for all devices under this hub
(includes root hub).

At hub_disconnect, it will do power off sequence which is at powered
on list.

Signed-off-by: Peter Chen <redacted>
Tested-by Joshua Clayton [off-list ref]

Hi Alan,

The power sequence library code has been reviewed several rounds, would
you please help to review the change for USB part?

Some small problems...  See below.

quoted

--- a/drivers/usb/core/Makefile
+++ b/drivers/usb/core/Makefile

@@ -9,5 +9,6 @@ usbcore-y += port.o of.o
 
 usbcore-$(CONFIG_PCI)		+= hcd-pci.o
 usbcore-$(CONFIG_ACPI)		+= usb-acpi.o
+usbcore-$(CONFIG_PWRSEQ_GENERIC) += pwrseq.o

Please use a tab to align this with the lines above.

quoted

--- /dev/null
+++ b/drivers/usb/core/pwrseq.c

@@ -0,0 +1,100 @@
+/*
+ * pwrseq.c	USB device power sequence management
+ *
+ * Copyright (C) 2016 Freescale Semiconductor, Inc.
+ * Author: Peter Chen <peter.chen@nxp.com>
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2  of
+ * the License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/list.h>
+#include <linux/of.h>
+#include <linux/power/pwrseq.h>
+#include <linux/slab.h>
+#include <linux/usb.h>
+#include <linux/usb/hcd.h>
+
+#include "hub.h"
+
+struct usb_pwrseq_node {
+	struct pwrseq *pwrseq;
+	struct list_head list;
+};
+
+static int hub_of_pwrseq_on(struct device_node *np, struct usb_hub *hub)
+{
+	struct pwrseq *pwrseq;
+	struct usb_pwrseq_node *pwrseq_node;
+	int ret;
+
+	pwrseq = pwrseq_alloc_generic();
+	if (IS_ERR(pwrseq))
+		return PTR_ERR(pwrseq);
+
+	ret = pwrseq_get(np, pwrseq);
+	if (ret)
+		goto pwr_free;
+
+	ret = pwrseq_on(np, pwrseq);
+	if (ret)
+		goto pwr_put;
+
+	pwrseq_node = kzalloc(sizeof(*pwrseq_node), GFP_KERNEL);

You need to check that the kzalloc succeeded.

quoted

+	pwrseq_node->pwrseq = pwrseq;
+	list_add(&pwrseq_node->list, &hub->pwrseq_on_list);
+
+	return 0;
+
+pwr_put:
+	pwrseq_put(pwrseq);
+pwr_free:
+	pwrseq_free(pwrseq);
+	return ret;
+}

This subroutine looks very generic.  The only place where it cares that
it was called for a USB hub is the list_add(), and that could easily
become generic also (make this function take &hub->pwrseq_on_list as
its second argument, instead of hub).

It looks like this routine really belongs in the power sequence 
library.

quoted

+
+int hub_pwrseq_on(struct usb_hub *hub)
+{

The names of this routine and the subroutine above are too similar.  
After all, both functions require OF support.

quoted

+	struct device *parent;
+	struct usb_device *hdev = hub->hdev;
+	struct device_node *np;
+	int ret;
+
+	if (hdev->parent)
+		parent = &hdev->dev;
+	else
+		parent = bus_to_hcd(hdev->bus)->self.controller;
+
+	for_each_child_of_node(parent->of_node, np) {
+		ret = hub_of_pwrseq_on(np, hub);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+void hub_pwrseq_off(struct usb_hub *hub)
+{
+	struct pwrseq *pwrseq;
+	struct usb_pwrseq_node *pwrseq_node, *tmp_node;
+
+	list_for_each_entry_safe(pwrseq_node, tmp_node,
+			&hub->pwrseq_on_list, list) {
+		pwrseq = pwrseq_node->pwrseq;
+		pwrseq_off(pwrseq);
+		pwrseq_put(pwrseq);
+		pwrseq_free(pwrseq);
+		list_del(&pwrseq_node->list);
+		kfree(pwrseq_node);
+	}
+}

This routine also looks like it belongs in a library.

Alan Stern