lkml.org 
[lkml]   [2023]   [Oct]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH 2/5] dt-bindings: nvmem: Convert xlnx,zynqmp-nvmem.txt to yaml
From


On 10/13/23 14:54, Krzysztof Kozlowski wrote:
> On 13/10/2023 14:08, Michal Simek wrote:
>>
>>
>> On 10/13/23 13:58, Krzysztof Kozlowski wrote:
>>> On 13/10/2023 13:51, Michal Simek wrote:
>>>>
>>>>
>>>> On 10/13/23 13:46, Krzysztof Kozlowski wrote:
>>>>> On 13/10/2023 13:22, Michal Simek wrote:
>>>>>>>
>>>>>>>> +
>>>>>>>> +required:
>>>>>>>> + - compatible
>>>>>>>
>>>>>>> required: block goes after patternProperties: block
>>>>>>>
>>>>>>>> +
>>>>>>>> +patternProperties:
>>>>>>>> + "^soc_revision@0$":
>>>>>>>
>>>>>>> Why do you define individual memory cells? Is this part of a binding?
>>>>>>> IOW, OS/Linux requires this?
>>>>>>
>>>>>> nvmem has in kernel interface where you can reference to nodes. nvmem_cell_get()
>>>>>> calls. It means you should be able to describe internal layout that's why names
>>>>>> are used. And address in name is there because of reg property is used to
>>>>>> describe base offset and size.
>>>>>
>>>>> That's not really what I am asking. Why internal layout of memory must
>>>>> be part of the bindings?
>>>>
>>>> It doesn't need to be but offsets are hardcoded inside the driver itself and
>>>> they can't be different.
>>>
>>> Hm, where? I opened drivers/nvmem/zynqmp_nvmem.c and I do not see any
>>> hard-coded offsets.
>>
>> Current driver supports only soc revision from offset 0.
>> But if you look at 5/5 you need to define offsets where information is present.
>> +#define SOC_VERSION_OFFSET 0x0
>> +#define EFUSE_START_OFFSET 0xC
>> +#define EFUSE_END_OFFSET 0xFC
>> +#define EFUSE_PUF_START_OFFSET 0x100
>> +#define EFUSE_PUF_MID_OFFSET 0x140
>> +#define EFUSE_PUF_END_OFFSET 0x17F
>
> There is nothing like this in existing driver, so the argument that "I
> am adding this to the binding during conversion because driver needs it"
> is not true. Conversion is only a conversion.

Conversion in 2/5 is adding only soc revision which is already there. It is
starting from 0 and world size is 1. And 0 is not listed because that's start
all the time.

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/nvmem/zynqmp_nvmem.c?h=v6.6-rc5#n39

And soc revision was also listed in origin binding example.

> Now, if you want to add something new to the binding because of new
> driver changes, that's separate topic.

Functionality in firmware is there for quite a long time but as I said I am fine
if map is not going to be inside dt binding spec.

> And since it is new change in the driver I can comment: please don't.
> Your nvmem driver should not depend on it. nvmem is only the provider.

Let's see what Srinivas says about implementation. If driver should be just
provider then pretty much current driver should be completely rewritten to
different style. I mean to have just transport via SMCs with offset/size and
then providing functionality in firmware.

Thanks,
Michal

\
 
 \ /
  Last update: 2023-10-13 15:07    [W:0.125 / U:1.148 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site