From mboxrd@z Thu Jan 1 00:00:00 1970 From: broonie@kernel.org (Mark Brown) Date: Tue, 29 Jul 2014 18:18:57 +0100 Subject: [RFC 3/5] regulator: core: Only apply constraints if available on list voltage In-Reply-To: <1406651339-28901-4-git-send-email-javier.martinez@collabora.co.uk> References: <1406651339-28901-1-git-send-email-javier.martinez@collabora.co.uk> <1406651339-28901-4-git-send-email-javier.martinez@collabora.co.uk> Message-ID: <20140729171857.GV17528@sirena.org.uk> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Tue, Jul 29, 2014 at 06:28:57PM +0200, Javier Martinez Canillas wrote: > If a selector can't be used on a platform due to voltage constraints, > regulator_list_voltage() returns 0. Doing this unconditionally made > sense since constraints were set in machine_constraints_voltage() at > regulator registration time. > > But for load switches that don't define a voltage output, the parent > supply voltage is used so the constraints should only be applied if > they were defined for the child regulators. No, think about what you're doing here and why we're filtering out unsettable voltages - this causes problems for consumers on regulators that don't have any ability to vary voltages since they will now be able to list voltages that they can't select. I would also expect any regulator where the supplied devices are able to vary the voltage to explicitly provide a constraint even if the implementation is done in a parent regulator. There may be constraints on the child supply which aren't directly present on the parent supply and can be ignored if the child supply is turned off. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: Digital signature URL: