* Péter Ujfalusi [off-list ref] [201231 12:55]:

On 12/30/20 10:43 AM, Tony Lindgren wrote:

quoted

@@ -58,7 +58,8 @@ omap4430_adc_to_temp[OMAP4430_ADC_END_VALUE - OMAP4430_ADC_START_VALUE + 1] = {
 const struct ti_bandgap_data omap4430_data = {
 	.features = TI_BANDGAP_FEATURE_MODE_CONFIG |
 			TI_BANDGAP_FEATURE_CLK_CTRL |
-			TI_BANDGAP_FEATURE_POWER_SWITCH,
+			TI_BANDGAP_FEATURE_POWER_SWITCH |
+			TI_BANDGAP_FEATURE_CONT_MODE_ONLY,

Can we add a comment with the observations?

Sure, and I also noticed that the timeout triggers also on dra7
too. I need to recheck what all are affected.. At least we now
see warnings on the SoCs affected.

quoted

@@ -605,8 +606,10 @@ ti_bandgap_force_single_read(struct ti_bandgap *bgp, int id)
 	u32 counter = 1000;
 	struct temp_sensor_registers *tsr;
 
-	/* Select single conversion mode */
-	if (TI_BANDGAP_HAS(bgp, MODE_CONFIG))
+	/* Select continuous or single conversion mode */
+	if (TI_BANDGAP_HAS(bgp, CONT_MODE_ONLY))
+		RMW_BITS(bgp, id, bgap_mode_ctrl, mode_ctrl_mask, 1);
+	else if (TI_BANDGAP_HAS(bgp, MODE_CONFIG))
 		RMW_BITS(bgp, id, bgap_mode_ctrl, mode_ctrl_mask, 0);

Would not be better to:
if (TI_BANDGAP_HAS(bgp, MODE_CONFIG)) {
	if (TI_BANDGAP_HAS(bgp, CONT_MODE_ONLY))
		RMW_BITS(bgp, id, bgap_mode_ctrl, mode_ctrl_mask, 1);
	else
		RMW_BITS(bgp, id, bgap_mode_ctrl, mode_ctrl_mask, 0);
}

One can only switch to cont/single mode if the mode config is possible.

Yup makes sense thanks for spotting that.

Regards,

Tony