On 07/02/2023 18:03, lucas.tanure@xxxxxxxxxxxxx wrote: > On 2/7/23 4:48 PM, Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx> wrote: >> On 07/02/2023 17:34, Lucas Tanure wrote: >>> On 07-02-2023 16:13, Krzysztof Kozlowski wrote: >>>> On 07/02/2023 16:46, Lucas Tanure wrote: >>>>>>> + Shared boost allows two amplifiers to share a single boost circuit by >>>>>>> + communicating on the MDSYNC bus. The passive amplifier does not control >>>>>>> + the boost and receives data from the active amplifier. GPIO1 should be >>>>>>> + configured for Sync when shared boost is used. Shared boost is not >>>>>>> + compatible with External boost. Active amplifier requires >>>>>>> + boost-peak-milliamp, boost-ind-nanohenry and boost-cap-microfarad. >>>>>>> 0 = Internal Boost >>>>>>> 1 = External Boost >>>>>>> + 2 = Reserved >>>>>> >>>>>> How binding can be reserved? For what and why? Drop. 2 is shared active, >>>>>> 3 is shared passive. >>>>> 2 Is shared boost without VSPK switch, a mode not supported for new >>>>> system designs. But there is laptops using it, so we need to keep >>>>> supporting in the driver. >>>> >>>> That's not the answer. 2 is nothing here, so it cannot be reserved. >>>> Aren't you mixing now some register value with bindings? >>>> >>>> Best regards, >>>> Krzysztof >>>> >>>> >>> I have added a new patch with propper documentation. >>> And I would like to use 3 and 4 for shared boost as >>> CS35L41_EXT_BOOST_NO_VSPK_SWITCH already exist as 2 and is used in the >>> current driver. >> >> I don't see CS35L41_EXT_BOOST_NO_VSPK_SWITCH in the bindings. >> >>> The laptop that uses CS35L41_EXT_BOOST_NO_VSPK_SWITCH doesn't have the >>> property "cirrus,boost-type", but to make everything consistent I would >>> prefer to use 3 and 4 for the new boost types. >>> Is that ok with you? >> >> I don't see how it is related. The value does not exist, so whether >> laptop has that property or not, is not really related, right? >> >> Best regards, >> Krzysztof >> >> > The value does exist in the code, but no device should have that in ACPI/DTB, so yes the value doesn't exist for ACPI/DTB purposes. > I can change CS35L41_EXT_BOOST_NO_VSPK_SWITCH to another value, like 99, and use 2 and 3 for shared boost. > I will re-submit that with v3. > Is that ok with you? I guess we still talk about different things. The code does not have a binding for the boost, therefore it does not use boost binding. Whatever it does with CS35L41_EXT_BOOST_NO_VSPK_SWITCH outside of DT, is not my topic and I don't care. That's why I asked folks to use strings for such enumerations, not register values - to avoid any confusion between the code and bindings (and also make it more readable for humans). Best regards, Krzysztof
