Understanding which storage class is used

I am trying to understand the bevaviour of the storage class assignment in Matlab.

Take the following example:

1. In the first scenario the storage class of 'lineSignal' is set to 'GetSet' and 'outputSignal' is set to 'auto'. This yields the following c-code:

/* Model step function */
void example_step(void)
{
  /* (no output/update code required) */
}

/* Model initialize function */
void example_initialize(void)
{
  /* ConstCode for Constant: '<Root>/Constant' */
  set_lineSignal(1.0);
}

2. In the second scenario the storage class of 'lineSignal' is set to 'auto' and 'outputSignal' is set to 'GetSet'. This yields the following c-code:

/* Model step function */
void example_step(void)
{
  /* (no output/update code required) */
}

/* Model initialize function */
void example_initialize(void)
{
  /* (no initialization code required) */
}

3. In the third scenario the storage class of 'lineSignal' and 'outputSignal' are both set to 'GetSet'. This yields the following c-code:

/* Model step function */
void example_step(void)
{
  /* (no output/update code required) */
}

/* Model initialize function */
void example_initialize(void)
{
  /* ConstCode for Constant: '<Root>/Constant' */
  set_lineSignal(1.0);
}

Based on this, I have two questions:

1. Why is the initialization code not generated when the GetSet storage class is used on a port?
2. Which storage class for which signal is used for code generation?