Messages in this thread | | | Date | Wed, 13 May 2015 13:58:05 +0100 | From | Daniel Thompson <> | Subject | Re: [PATCH v8 14/16] ARM: dts: Introduce STM32F429 MCU |
| |
On 13/05/15 12:45, Maxime Coquelin wrote: > 2015-05-12 23:21 GMT+02:00 Arnd Bergmann <arnd@arndb.de>: >> On Saturday 09 May 2015 09:53:56 Maxime Coquelin wrote: >>> +#include <dt-bindings/mfd/stm32f4-rcc.h> >>> + >>> >> >> Can you find a way to avoid this dependency? >> >> Maybe you can change the bindings so that the numbers you pass as >> arguments to the reset and clock specifiers reflect the numbers that >> the hardware use? > > If I understand correctly, you prefer the way I did in v7 [0]? > > I don't have a strong opinion on this. Either way is fine to me. > > I changed the bindings in the v8 after discussions with Daniel > Thompson, who is implementing the clock driver part of the RCC IP. > > He proposed we used common defines, because each peripheral has a > reset line and a clock gate. > Both reset and clock are represented as a single bit, with only the > base offset differing between clock and reset. > You can have a look at chapter 6 of the reference manual [1] if you > find some time. > > Having common defines between clocks and reset make sense to me, but I > also understand your point of avoiding dependencies. > > Maybe I can revert back to v7 bindings for now, and then we can > reconsider using common defines when Daniel will send the clock > patches. > Note that doing that won't break the DT binary compatibility, as the > raw reset values, or the ones from defines are the same. > > Daniel, could you share an example of the bindings you would use for the clocks?
For the most cases, where there is a clock gate just before the peripheral it looks pretty much like the reset driver and I use the bit offset of the clock gating bit as the index.
However there are a couple of clocks without gating just before the clock reaches the peripheral:
1. A hard coded /8. I think this will have to be given a synthetic number.
2. Ungated dividers. For these I am using the bit offset of the LSB of the mux field.
So I think there is only one value that is completely unrelated to the hardware and will use a magic constant instead.
I had planned to macros similar to the STM32F4_AxB_RESET() family of macros in both clk driver and DT in order to reuse the bit layouts from dt-bindings/mfd/stm32f4-rcc.h .
Normal case would have looked like this:
timer3: timer@40000000 { compatible = "st,stm32-timer"; reg = <0x40000000 0x400>; interrupts = <28>; resets = <&rcc STM32F4_APB1_RESET(TIM3)>; clocks = <&rcc STM32F4_APB1_CLK(TIM3)>; status = "disabled"; };
Without the macros it looks like this:
timer3: timer@40000000 { compatible = "st,stm32-timer"; reg = <0x40000000 0x400>; interrupts = <28>; resets = <&rcc 257>; clocks = <&rcc 513>; status = "disabled"; };
However we could perhaps be more literate even if we don't use the macros?
timer3: timer@40000000 { compatible = "st,stm32-timer"; reg = <0x40000000 0x400>; interrupts = <28>; resets = <&rcc ((0x20*8) + 1)>; clocks = <&rcc ((0x40*8) + 1)>; status = "disabled"; };
Daniel.
> Kind regards, > Maxime > >> >> Arnd > > [0]: http://lkml.iu.edu/hypermail/linux/kernel/1504.3/04523.html > [1]: http://www.st.com/web/en/resource/technical/document/reference_manual/DM00031020.pdf >
| |