Hi Maxime,

2016-06-02 18:02 GMT+02:00 Maxime Coquelin [off-list ref]:

Hi Cedric,

2016-06-02 17:35 GMT+02:00 M'boumba Cedric Madianga [off-list ref]:

quoted

Hi,

quoted

quoted

+
+/**
+ * stm32f4_i2c_xfer() - Transfer combined I2C message
+ * @i2c_adap: Adapter pointer to the controller
+ * @msgs: Pointer to data to be written.
+ * @num: Number of messages to be executed
+ */
+static int stm32f4_i2c_xfer(struct i2c_adapter *i2c_adap, struct i2c_msg msgs[],
+                           int num)
+{
+       struct stm32f4_i2c_dev *i2c_dev = i2c_get_adapdata(i2c_adap);
+       int ret, i;
+
+       i2c_dev->busy = true;
+
+       ret = clk_prepare_enable(i2c_dev->clk);
+       if (ret) {
+               dev_err(i2c_dev->dev, "Failed to prepare_enable clock\n");
+               return ret;
+       }
+
+       stm32f4_i2c_hw_config(i2c_dev);

Maybe you could call this only at probe and resume time?
You would save some register accesses.

Some clarification about this point.
We need to call stm32f4_i2c_hw_config before each I2C transfer as at
the end of I2C communication the peripheral is automatically disabled
and configuration registers are reset.

Ok, but I wonder how the IP knows this is the last i2c message to be sent?

When the IP is enabled we could generated a STOP at any moment by
setting a bit in CR1 register.
At this moment, the IP knows that this is the last byte to be sent.
After STOP generation, I notice that all registers are automatically
set with their reset values

Or maybe it gets re-initialized as soon as the clk is disabled?

Not really because they are all re-initialized before disabling the clock

Thanks for the inputs,
Maxime