Re: [PATCH 1/2] i2c: exynos5: add High Speed I2C controller driver

From: Naveen Krishna Ch
Date: Fri Dec 28 2012 - 01:44:02 EST


Hello Balbi,

On 28 December 2012 04:27, Felipe Balbi <balbi@xxxxxx> wrote:
> Hi,
>
> On Tue, Dec 25, 2012 at 04:55:54PM +0530, Naveen Krishna Chatradhi wrote:
>> Adds support for High Speed I2C driver found in Exynos5 and later
>> SoCs from Samsung. This driver currently supports Auto mode.
>>
>> Driver only supports Device Tree method.
>>
>> Signed-off-by: Taekgyun Ko <taeggyun.ko@xxxxxxxxxxx>
>> Signed-off-by: Naveen Krishna Chatradhi <ch.naveen@xxxxxxxxxxx>
>> ---
>> Changes since v1:
>> Fixed the comments from Felipe Balbi and Thomas Abraham.
>>
>> drivers/i2c/busses/Kconfig | 7 +
>> drivers/i2c/busses/Makefile | 1 +
>> drivers/i2c/busses/i2c-exynos5.c | 652 ++++++++++++++++++++++++++++++++++++++
>> drivers/i2c/busses/i2c-exynos5.h | 102 ++++++
>> 4 files changed, 762 insertions(+)
>> create mode 100644 drivers/i2c/busses/i2c-exynos5.c
>> create mode 100644 drivers/i2c/busses/i2c-exynos5.h
>>
>> diff --git a/drivers/i2c/busses/Kconfig b/drivers/i2c/busses/Kconfig
>> index bdca511..4caea76 100644
>> --- a/drivers/i2c/busses/Kconfig
>> +++ b/drivers/i2c/busses/Kconfig
>> @@ -618,6 +618,13 @@ config I2C_S3C2410
>> Say Y here to include support for I2C controller in the
>> Samsung SoCs.
>>
>> +config I2C_EXYNOS5
>> + tristate "Exynos5 high-speed I2C driver"
>> + depends on ARCH_EXYNOS5
>> + help
>> + Say Y here to include support for High-speed I2C controller in the
>> + Exynos5 based Samsung SoCs.
>> +
>> config I2C_S6000
>> tristate "S6000 I2C support"
>> depends on XTENSA_VARIANT_S6000
>> diff --git a/drivers/i2c/busses/Makefile b/drivers/i2c/busses/Makefile
>> index 6181f3f..4b1548c 100644
>> --- a/drivers/i2c/busses/Makefile
>> +++ b/drivers/i2c/busses/Makefile
>> @@ -61,6 +61,7 @@ obj-$(CONFIG_I2C_PUV3) += i2c-puv3.o
>> obj-$(CONFIG_I2C_PXA) += i2c-pxa.o
>> obj-$(CONFIG_I2C_PXA_PCI) += i2c-pxa-pci.o
>> obj-$(CONFIG_I2C_S3C2410) += i2c-s3c2410.o
>> +obj-$(CONFIG_I2C_EXYNOS5) += i2c-exynos5.o
>> obj-$(CONFIG_I2C_S6000) += i2c-s6000.o
>> obj-$(CONFIG_I2C_SH7760) += i2c-sh7760.o
>> obj-$(CONFIG_I2C_SH_MOBILE) += i2c-sh_mobile.o
>> diff --git a/drivers/i2c/busses/i2c-exynos5.c b/drivers/i2c/busses/i2c-exynos5.c
>> new file mode 100644
>> index 0000000..7614f60
>> --- /dev/null
>> +++ b/drivers/i2c/busses/i2c-exynos5.c
>> @@ -0,0 +1,652 @@
>> +/* linux/drivers/i2c/busses/i2c-exynos5.c
>
> no need for the full path. Generally this would look like:
>
> * i2c-exynos5.c - Samsung Exynos 5 I2C Controller Driver
>
> no strong feelings however.
>
>> + * Copyright (C) 2012 Samsung Electronics Co., Ltd.
>> + *
>> + * High speed I2C controller driver
>> + * for Exynos5 and later SoCs from Samsung.
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License version 2 as
>> + * published by the Free Software Foundation.
>> +*/
>> +
>> +#include <linux/kernel.h>
>> +#include <linux/module.h>
>> +
>> +#include <linux/i2c.h>
>> +#include <linux/init.h>
>> +#include <linux/time.h>
>> +#include <linux/interrupt.h>
>> +#include <linux/delay.h>
>> +#include <linux/errno.h>
>> +#include <linux/err.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/pm_runtime.h>
>> +#include <linux/clk.h>
>> +#include <linux/slab.h>
>> +#include <linux/io.h>
>> +#include <linux/of_address.h>
>> +#include <linux/of_gpio.h>
>> +#include <linux/of_irq.h>
>> +#include <linux/of_i2c.h>
>> +
>> +#include "i2c-exynos5.h"
>
> it doesn't look like this header is even needed. All those macros could
> be defined here in the C-source file which is the only user.
>
>> +#define EXYNOS5_I2C_TIMEOUT (msecs_to_jiffies(1000))
>> +
>> +/* timeout for pm runtime autosuspend */
>> +#define EXYNOS5_I2C_PM_TIMEOUT 1000 /* ms */
>> +
>> +struct exynos5_i2c {
>> + struct i2c_adapter adap;
>> + unsigned int suspended:1;
>> +
>> + struct i2c_msg *msg;
>> + unsigned int msg_idx;
>> + struct completion msg_complete;
>> + unsigned int msg_ptr;
>> +
>> + unsigned int irq;
>> +
>> + void __iomem *regs;
>> + struct clk *clk;
>> + struct device *dev;
>> + int gpios[2];
>> +
>> + int bus_num;
>> + int speed_mode;
>> +};
>> +
>> +static const struct of_device_id exynos5_i2c_match[] = {
>> + { .compatible = "samsung,exynos5-hsi2c" },
>> + {},
>> +};
>> +MODULE_DEVICE_TABLE(of, exynos5_i2c_match);
>> +
>> +/* TODO: Should go to debugfs */
>> +static inline void dump_i2c_register(struct exynos5_i2c *i2c)
>> +{
>> + dev_vdbg(i2c->dev, "Register dump(%d) :\n %x\n %x\n %x\n %x\n"
>> + " %x\n %x\n %x\n %x\n %x\n"
>> + " %x\n %x\n %x\n %x\n %x\n"
>> + " %x\n %x\n %x\n %x\n %x\n"
>> + " %x\n %x\n %x\n %x\n %x\n",
>> + i2c->suspended,
>> + readl(i2c->regs + HSI2C_CTL),
>> + readl(i2c->regs + HSI2C_FIFO_CTL),
>> + readl(i2c->regs + HSI2C_TRAILIG_CTL),
>> + readl(i2c->regs + HSI2C_CLK_CTL),
>> + readl(i2c->regs + HSI2C_CLK_SLOT),
>> + readl(i2c->regs + HSI2C_INT_ENABLE),
>> + readl(i2c->regs + HSI2C_INT_STATUS),
>> + readl(i2c->regs + HSI2C_ERR_STATUS),
>> + readl(i2c->regs + HSI2C_FIFO_STATUS),
>> + readl(i2c->regs + HSI2C_TX_DATA),
>> + readl(i2c->regs + HSI2C_RX_DATA),
>> + readl(i2c->regs + HSI2C_CONF),
>> + readl(i2c->regs + HSI2C_AUTO_CONF),
>> + readl(i2c->regs + HSI2C_TIMEOUT),
>> + readl(i2c->regs + HSI2C_MANUAL_CMD),
>> + readl(i2c->regs + HSI2C_TRANS_STATUS),
>> + readl(i2c->regs + HSI2C_TIMING_HS1),
>> + readl(i2c->regs + HSI2C_TIMING_HS2),
>> + readl(i2c->regs + HSI2C_TIMING_HS3),
>> + readl(i2c->regs + HSI2C_TIMING_FS1),
>> + readl(i2c->regs + HSI2C_TIMING_FS2),
>> + readl(i2c->regs + HSI2C_TIMING_FS3),
>> + readl(i2c->regs + HSI2C_TIMING_SLA),
>> + readl(i2c->regs + HSI2C_ADDR));
>> +}
>
> NAK, please add debugfs support already. This will be here for a long
> time otherwise
Actually patch 2/2 does the same, posted them separately for RFC
>
>> +static inline void exynos5_i2c_stop(struct exynos5_i2c *i2c, int ret)
>
> no need for inline, GCC will handle that. Also, this prototype looks
> awkward, you're passing caller's return code to be checked here... it's
> weird, to say the least.
>
>> +{
>> + dev_dbg(i2c->dev, "STOP\n");
>
> I would call this dev_vdbg()
>
>> +
>> + i2c->msg_idx++;
>> + if (ret)
>> + i2c->msg_idx = ret;
>> +
>> + /* Disable interrrupts */
>> + writel(0, i2c->regs + HSI2C_INT_ENABLE);
>> + complete(&i2c->msg_complete);
>> +}
>> +
>> +static void exynos5_i2c_en_timeout(struct exynos5_i2c *i2c)
>> +{
>> + unsigned long i2c_timeout = readl(i2c->regs + HSI2C_TIMEOUT);
>> +
>> + /* Clear to enable Timeout */
>> + i2c_timeout &= ~HSI2C_TIMEOUT_EN;
>> + writel(i2c_timeout, i2c->regs + HSI2C_TIMEOUT);
>> +}
>> +
>> +static void exynos5_i2c_master_run(struct exynos5_i2c *i2c)
>> +{
>> + /* Start data transfer in Master mode */
>> + u32 i2c_auto_conf = readl(i2c->regs + HSI2C_AUTO_CONF);
>> + i2c_auto_conf |= HSI2C_MASTER_RUN;
>> + writel(i2c_auto_conf, i2c->regs + HSI2C_AUTO_CONF);
>> +}
>> +
>> +/* exynos5_i2c_set_bus
>> + *
>> + * get the i2c bus for a master/slave transaction
>> + */
>
> multiline comment style is wrong.
>
>> +static int exynos5_i2c_set_bus(struct exynos5_i2c *i2c, int master)
>
> this function is only called with master set to 1, do you even need that
> argument ? Do you have some out-of-tree code for supporting slave mode ?
>
> Please remove 'slave' support completely as that's not supported in
> mainline as of today, if you need slave support, then add it to the
> framework first.
Will remove slave stuff, I don't plans for the as of now.
>
>> +{
>> + unsigned long t_status;
>> + int timeout = 400;
>> +
>> + while (timeout-- > 0) {
>> + t_status = readl(i2c->regs + HSI2C_TRANS_STATUS);
>> +
>> + if (master) {
>> + if (!(t_status & HSI2C_MASTER_BUSY))
>> + return 0;
>> + } else {
>> + if (!(t_status & HSI2C_SLAVE_BUSY))
>> + return 0;
>> + }
>> +
>> + msleep(20);
>> + }
>> +
>> + return -ETIMEDOUT;
>> +}
>> +
>> +/* exynos5_i2c_irq
>> + *
>> + * top level IRQ servicing routine
>> + */
>
> multiline comment style is wrong.
>
>> +static irqreturn_t exynos5_i2c_irq(int irqno, void *dev_id)
>> +{
>> + struct exynos5_i2c *i2c = dev_id;
>> + unsigned long t_stat;
>> + unsigned char byte;
>> +
>> + t_stat = readl(i2c->regs + HSI2C_TRANS_STATUS);
>> +
>> + if (t_stat & HSI2C_TRANS_ABORT) {
>> + /* deal with arbitration loss */
>> + dev_err(i2c->dev, "deal with arbitration loss\n");
>> + goto out;
>> + }
>> + if (i2c->msg->flags & I2C_M_RD) {
>> + if (t_stat & HSI2C_TRANS_DONE) {
>> + dev_dbg(i2c->dev, "Device found.");
>> + while ((readl(i2c->regs + HSI2C_FIFO_STATUS) &
>> + HSI2C_RX_FIFO_EMPTY) == 0) {
>> + byte = readl(i2c->regs + HSI2C_RX_DATA);
>> + dev_dbg(i2c->dev, "read rx_data = %x", byte);
>> + i2c->msg->buf[i2c->msg_ptr++] = byte;
>> + }
>> + } else if (t_stat & HSI2C_NO_DEV) {
>> + dev_dbg(i2c->dev, "No device found.");
>> + exynos5_i2c_stop(i2c, -ENXIO);
>> + } else if (t_stat & HSI2C_NO_DEV_ACK &&
>> + !(i2c->msg->flags & I2C_M_IGNORE_NAK)) {
>> + dev_dbg(i2c->dev, "No device Ack.");
>> + exynos5_i2c_stop(i2c, -ENXIO);
>> + }
>> + } else {
>> + byte = i2c->msg->buf[i2c->msg_ptr++];
>> + dev_dbg(i2c->dev, "write tx_data = %x ", byte);
>> + writel(byte, i2c->regs + HSI2C_TX_DATA);
>> + }
>> +
>> + if (i2c->msg_ptr >= i2c->msg->len)
>> + exynos5_i2c_stop(i2c, 0);
>> +
>> + out:
>> + /* Set those bits to clear them */
>> + writel(readl(i2c->regs + HSI2C_INT_STATUS),
>> + i2c->regs + HSI2C_INT_STATUS);
>> +
>> + return IRQ_HANDLED;
>> +}
>> +
>> +static void exynos5_i2c_message_start(struct exynos5_i2c *i2c,
>> + struct i2c_msg *msgs)
>> +{
>> + unsigned long usi_ctl = HSI2C_FUNC_MODE_I2C | HSI2C_MASTER;
>> + unsigned long i2c_auto_conf;
>> + unsigned long i2c_addr = ((msgs->addr & 0x7f) << 10);
>> + unsigned long usi_int_en = 0;
>> +
>> + exynos5_i2c_en_timeout(i2c);
>> +
>> + if (msgs->flags & I2C_M_RD) {
>> + usi_ctl &= ~HSI2C_TXCHON;
>> + usi_ctl |= HSI2C_RXCHON;
>> +
>> + i2c_auto_conf |= HSI2C_READ_WRITE;
>> +
>> + usi_int_en |= (HSI2C_INT_RX_ALMOSTFULL_EN |
>> + HSI2C_INT_TRAILING_EN);
>> + } else {
>> + usi_ctl &= ~HSI2C_RXCHON;
>> + usi_ctl |= HSI2C_TXCHON;
>> +
>> + i2c_auto_conf &= ~HSI2C_READ_WRITE;
>> +
>> + usi_int_en |= HSI2C_INT_TX_ALMOSTEMPTY_EN;
>> + }
>> +
>> + writel(i2c_addr, i2c->regs + HSI2C_ADDR);
>> + writel(usi_ctl, i2c->regs + HSI2C_CTL);
>> +
>> + i2c_auto_conf |= i2c->msg->len;
>> + i2c_auto_conf |= HSI2C_STOP_AFTER_TRANS;
>> + writel(i2c_auto_conf, i2c->regs + HSI2C_AUTO_CONF);
>> +
>> + exynos5_i2c_master_run(i2c);
>> +
>> + /* Enable appropriate interrupts */
>> + writel(usi_int_en, i2c->regs + HSI2C_INT_ENABLE);
>> +}
>> +
>> +static int exynos5_i2c_doxfer(struct exynos5_i2c *i2c, struct i2c_msg *msgs)
>> +{
>> + unsigned long timeout;
>> +
>> + if (i2c->suspended) {
>> + dev_err(i2c->dev, "HS-I2C is not initialzed.\n");
>> + return -EIO;
>> + }
>> +
>> + if (exynos5_i2c_set_bus(i2c, 1)) {
>> + dev_err(i2c->dev, "cannot get bus, Master busy.\n");
>> + return -EAGAIN;
>> + }
>> +
>> + i2c->msg = msgs;
>> + i2c->msg_ptr = 0;
>> + i2c->msg_idx = 0;
>> +
>> + INIT_COMPLETION(i2c->msg_complete);
>> +
>> + exynos5_i2c_message_start(i2c, msgs);
>> +
>> + timeout = wait_for_completion_timeout(&i2c->msg_complete,
>> + EXYNOS5_I2C_TIMEOUT);
>> +
>> + if (timeout == 0)
>> + dev_dbg(i2c->dev, "timeout\n");
>> + else if (i2c->msg_idx != msgs->len)
>> + dev_dbg(i2c->dev, "incomplete xfer (%d)\n", i2c->msg_idx);
>> +
>> + return i2c->msg_idx;
>> +}
>> +
>> +static int exynos5_i2c_xfer(struct i2c_adapter *adap,
>> + struct i2c_msg *msgs, int num)
>> +{
>> + struct exynos5_i2c *i2c = (struct exynos5_i2c *)adap->algo_data;
>> + int retry, i;
>> + int ret;
>> +
>> + ret = pm_runtime_get_sync(i2c->dev);
>> + if (IS_ERR_VALUE(ret))
>> + goto out;
>> +
>> + clk_prepare_enable(i2c->clk);
>> +
>> + for (retry = 0; retry < adap->retries; retry++) {
>> + for (i = 0; i < num; i++) {
>> + ret = exynos5_i2c_doxfer(i2c, msgs);
>> + msgs++;
>> +
>> + if (ret == -EAGAIN)
>> + break;
>> + }
>> + if (i == num) {
>> + clk_disable_unprepare(i2c->clk);
>> + ret = i2c->msg_idx;
>> + goto out;
>> + }
>
> why don't you just move this to the out label ?
Moving this to label, din't look so clean to me..
>
>> + dev_dbg(i2c->dev, "retrying transfer (%d)\n", retry);
>> +
>> + udelay(100);
>> + }
>> +
>> + ret = -EREMOTEIO;
>> + clk_disable_unprepare(i2c->clk);
>> + out:
>> + pm_runtime_mark_last_busy(i2c->dev);
>> + pm_runtime_put_autosuspend(i2c->dev);
>> + return ret;
>> +}
>> +
>> +static u32 exynos5_i2c_func(struct i2c_adapter *adap)
>> +{
>> + return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
>> +}
>> +
>> +static const struct i2c_algorithm exynos5_i2c_algorithm = {
>> + .master_xfer = exynos5_i2c_xfer,
>> + .functionality = exynos5_i2c_func,
>> +};
>> +
>> +static int exynos5_i2c_set_timing(struct exynos5_i2c *i2c, int speed_mode)
>> +{
>> + unsigned long i2c_timing_s1;
>> + unsigned long i2c_timing_s2;
>> + unsigned long i2c_timing_s3;
>> + unsigned long i2c_timing_sla;
>> + unsigned int op_clk;
>> + unsigned int clkin = clk_get_rate(i2c->clk);
>> + unsigned int n_clkdiv;
>> + unsigned int t_start_su, t_start_hd;
>> + unsigned int t_stop_su;
>> + unsigned int t_data_su, t_data_hd;
>> + unsigned int t_scl_l, t_scl_h;
>> + unsigned int t_sr_release;
>> + unsigned int t_ftl_cycle;
>> + unsigned int i = 0, utemp0 = 0, utemp1 = 0, utemp2 = 0;
>> +
>> + if (speed_mode == HSI2C_HIGH_SPD)
>> + op_clk = HSI2C_HS_TX_CLOCK;
>> + else
>> + op_clk = HSI2C_FS_TX_CLOCK;
>> +
>> + /* FPCLK / FI2C =
>> + * (CLK_DIV + 1) * (TSCLK_L + TSCLK_H + 2) + 8 + 2 * FLT_CYCLE
>> + * uTemp0 = (CLK_DIV + 1) * (TSCLK_L + TSCLK_H + 2)
>> + * uTemp1 = (TSCLK_L + TSCLK_H + 2)
>> + * uTemp2 = TSCLK_L + TSCLK_H
>> + */
>
> wrong indentation
>
> (back to my vacations now)
Thanks for your time. (i was wondering if anyone would care during
this vacation season.)
>
> --
> balbi

Will post the v3, fixing all comments except the (moving to out label) one.

--
Shine bright,
(: Nav :)
--
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/