Re: [PATCH v4] extcon: Palmas Extcon Driver

From: Kishon Vijay Abraham I
Date: Tue May 07 2013 - 01:21:35 EST


Hi,

On Tuesday 07 May 2013 06:13 AM, myungjoo.ham wrote:
From: Graeme Gregory <gg@xxxxxxxxxxxxxxx>

This is the driver for the USB comparator built into the palmas chip. It
handles the various USB OTG events that can be generated by cable
insertion/removal.

Signed-off-by: Graeme Gregory <gg@xxxxxxxxxxxxxxx>
Signed-off-by: Moiz Sonasath <m-sonasath@xxxxxx>
Signed-off-by: Ruchika Kharwar <ruchika@xxxxxx>
Signed-off-by: Kishon Vijay Abraham I <kishon@xxxxxx>
[kishon@xxxxxx: adapted palmas usb driver to use the extcon framework]
Signed-off-by: Sebastien Guiriec <s-guiriec@xxxxxx>

Here goes some comments in the code:

[]

diff --git a/drivers/extcon/extcon-palmas.c
b/drivers/extcon/extcon-palmas.c
new file mode 100644
index 0000000..3ef042f
--- /dev/null
+++ b/drivers/extcon/extcon-palmas.c
@@ -0,0 +1,389 @@
+/*
+ * Palmas USB transceiver driver
+ *
+ * Copyright (C) 2013 Texas Instruments Incorporated - http://www.ti.com
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * Author: Graeme Gregory <gg@...mlogic.co.uk>
+ * Author: Kishon Vijay Abraham I <kishon@...com>
+ *
+ * Based on twl6030_usb.c
+ *
+ * Author: Hema HK <hemahk@...com>
+ *
+ * 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.
+ */

Why the email addresses are masked in the source code?
(I'm just curious as this is the first time to see such in kernel source)

That was not intentional. Took the previous version from the net. My bad.

+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/usb/phy_companion.h>
+#include <linux/regulator/consumer.h>
+#include <linux/err.h>
+#include <linux/notifier.h>
+#include <linux/slab.h>
+#include <linux/delay.h>
+#include <linux/mfd/palmas.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+#include <linux/extcon/extcon_palmas.h>
+
+static const char *palmas_extcon_cable[] = {
+ [0] = "USB",
+ [1] = "USB-HOST",

[1] = "USB-Host",

+ NULL,
+};
+
+static const int mutually_exclusive[] = {0x3, 0x0};
+
+static void palmas_usb_wakeup(struct palmas *palmas, int enable)
+{
+ if (enable)
+ palmas_write(palmas, PALMAS_USB_OTG_BASE, PALMAS_USB_WAKEUP,
+ PALMAS_USB_WAKEUP_ID_WK_UP_COMP);
+ else
+ palmas_write(palmas, PALMAS_USB_OTG_BASE, PALMAS_USB_WAKEUP,
0);
+}
+
+static ssize_t palmas_usb_vbus_show(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ unsigned long flags;
+ int ret = -EINVAL;
+ struct palmas_usb *palmas_usb = dev_get_drvdata(dev);
+
+ spin_lock_irqsave(&palmas_usb->lock, flags);

This spinlock is used in this sysfs-show function only.
Is this really needed?
The only protected value here is "linkstat", which is "readonly"
in this protected context.
Thus, removing the spin_lock and updating like the following should
be the same with less overhead:

int linkstat = palmas_usb->linkstat;
switch(linkstat) {

hmm.. ok.


+
+ switch (palmas_usb->linkstat) {
+ case PALMAS_USB_STATE_VBUS:
+ ret = snprintf(buf, PAGE_SIZE, "vbus\n");
+ break;
+ case PALMAS_USB_STATE_ID:
+ ret = snprintf(buf, PAGE_SIZE, "id\n");
+ break;
+ case PALMAS_USB_STATE_DISCONNECT:
+ ret = snprintf(buf, PAGE_SIZE, "none\n");
+ break;
+ default:
+ ret = snprintf(buf, PAGE_SIZE, "UNKNOWN\n");
+ }
+ spin_unlock_irqrestore(&palmas_usb->lock, flags);
+
+ return ret;
+}
+static DEVICE_ATTR(vbus, 0444, palmas_usb_vbus_show, NULL);
+

[]

+
+static void palmas_set_vbus_work(struct work_struct *data)
+{
+ int ret;
+ struct palmas_usb *palmas_usb = container_of(data, struct
palmas_usb,
+
set_vbus_work);
+
+ if (IS_ERR_OR_NULL(palmas_usb->vbus_reg)) {
+ dev_err(palmas_usb->dev, "invalid regulator\n");
+ return;
+ }
+
+ /*
+ * Start driving VBUS. Set OPA_MODE bit in CHARGERUSB_CTRL1
+ * register. This enables boost mode.
+ */
+
+ if (palmas_usb->vbus_enable) {
+ ret = regulator_enable(palmas_usb->vbus_reg);
+ if (ret)
+ dev_dbg(palmas_usb->dev, "regulator enable
failed\n");
+ } else {
+ regulator_disable(palmas_usb->vbus_reg);
+ }
+}
+
+static int palmas_set_vbus(struct phy_companion *comparator, bool
enabled)
+{
+ struct palmas_usb *palmas_usb = comparator_to_palmas(comparator);
+
+ palmas_usb->vbus_enable = enabled;
+ schedule_work(&palmas_usb->set_vbus_work);
+
+ return 0;
+}
+
+static int palmas_start_srp(struct phy_companion *comparator)
+{
+ struct palmas_usb *palmas_usb = comparator_to_palmas(comparator);
+
+ palmas_write(palmas_usb->palmas, PALMAS_USB_OTG_BASE,
+ PALMAS_USB_VBUS_CTRL_SET,
PALMAS_USB_VBUS_CTRL_SET_VBUS_DISCHRG
+ | PALMAS_USB_VBUS_CTRL_SET_VBUS_IADP_SINK);
+ palmas_write(palmas_usb->palmas, PALMAS_USB_OTG_BASE,
+ PALMAS_USB_VBUS_CTRL_SET,
+ PALMAS_USB_VBUS_CTRL_SET_VBUS_CHRG_VSYS |
+ PALMAS_USB_VBUS_CTRL_SET_VBUS_IADP_SINK);
+
+ mdelay(100);

Why not msleep()? It's long enough to consider sleep.
Is this in an atomic context? (if so, 100msec seems even more awful)

I guess it shouldn't be called from atomic context. Actually srp hasn't been tested because the controller driver we use with palmas dont have the infrastructure yet.

+
+ palmas_write(palmas_usb->palmas, PALMAS_USB_OTG_BASE,
+ PALMAS_USB_VBUS_CTRL_CLR,
+ PALMAS_USB_VBUS_CTRL_SET_VBUS_CHRG_VSYS |
+ PALMAS_USB_VBUS_CTRL_SET_VBUS_CHRG_VSYS);
+
+ return 0;
+}
+
+static void palmas_dt_to_pdata(struct device_node *node,
+ struct palmas_usb_platform_data *pdata)
+{
+ pdata->no_control_vbus = of_property_read_bool(node,
+ "ti,no_control_vbus");
+ pdata->wakeup = of_property_read_bool(node, "ti,wakeup");
+}
+
+static int palmas_usb_probe(struct platform_device *pdev)
+{

[]

+ /* init spinlock for workqueue */
+ spin_lock_init(&palmas_usb->lock);

[]

+ /* init spinlock for workqueue */
+ spin_lock_init(&palmas_usb->lock);

Why this lock is initialized again?

Will remove it.

Thanks
Kishon
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/