dpcm_fe_dai_startup() (= A) calls dpcm_set_fe_runtime() (= B) to setup
DPCM runtime. From *naming point of view*, it sounds like setup function
for FE.

(A)	static int dpcm_fe_dai_startup(...)
	{
		...
(B)		dpcm_set_fe_runtime(...);
		...
	}

But in dpcm_set_fe_runtime() (= B),
It  setups FE by for_each loop (= X),
and setups BE by dpcm_runtime_merge_xxx() (= Y).

(B)	static void dpcm_set_fe_runtime(...)
	{
		...
 ^		for_each_rtd_cpu_dais(...) {
 |			...
(X)			soc_pcm_hw_update_rate(...);
 |			soc_pcm_hw_update_chan(...);
 |			soc_pcm_hw_update_format(...);
 v		}

 ^		dpcm_runtime_merge_format(...);
(Y)		dpcm_runtime_merge_chan(...);
 v		dpcm_runtime_merge_rate(...);
	}

These means that the function which is called as xxx_fe_xxx()
is setups both FE and BE. This is confusable and can be hot bed for bug.

To tidyup it, as 1st step, this patch adds new dpcm_runtime_setup_fe()
for (X).
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87pn0tvsgx.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: Mark Brown <broonie@kernel.org>