> On Thu, Jul 24, 2025 at 02:20:10PM +0000, Tarang Raval wrote: > > > > > > 2. In the regulator code, you can reduce boilerplate by using > > > > > > devm_regulator_bulk_get_enable(). > > > > > > > > > > devm_regulator_bulk_get_enable() doesn't seem to be a good idea. You > > > > > generally don't want to enable power everywhere unconditionally, and > > > > > sensors very often need a guaranteed power up sequence. -----(1) > > > > > > > > The regulators are optional, we supply power to the camera sensor directly > > > > through dedicated power rails and there is no strict enable sequence > > > > required in this case. > > > > > > What exactly do you mean by "this case" ? Are you talking about one > > > particular sensor ? One particular camera module ? > > > > Laurent, by “this case” I meant the common scenario where power to the > > camera sensor is supplied by a PMIC regulator that is always-on. In such > > setups, the regulator is fixed and cannot be enabled or disabled from the > > driver, the sensor is always powered. > > > > This is what I’ve seen in most platforms, where the CSI input connector > > provides fixed 3.3V/1.8V power rails directly to the camera module. > > > > Of course, if the camera supply comes from a dedicated regulator controlled > > via a GPIO, then the driver would need to handle enable/disable sequencing > > explicitly. But I’m specifically referring to the first case, where the power rails > > are always-on. > > How does the sensor driver know which of those two cases it is dealing > with ? The sensor driver typically determines this via the presence (or absence) of regulator supply entries in the Device Tree. If a supply is not defined, it's assumed to be always-on (e.g., provided by the board via fixed rails). When defined, the driver retrieves and manages the regulator. This approach allows a single driver to support both cases, by treating supplies as optional and only enabling them when explicitly defined. At comment (1): you gave two reasons why we cannot use devm_regulator_bulk_get_enable. What I’m trying to say is: You mentioned "generally don't want to enable power everywhere unconditionally," but on almost every platform, the power rails are always-on. And regarding the second point — "sensors very often need a guaranteed power-up sequence" I don’t understand why this would be an issue. Even if we use devm_regulator_bulk_get_enable, the power-up sequence remains the same. So how is it not a good option in this case? Best Regards, Tarang
