Re: [PATCH net-next v7 3/3] dpll: add SiTime SiT9531x DPLL clock driver

[Jiri Pirko <jiri@resnulli.us>]

Wed, Aug 19, 2026 at 11:16:18PM +0200, vadim.fedorenko@linux.dev wrote:
>On 15/08/2026 23:19, Ali Rouhi wrote:
>> Add a DPLL subsystem driver for the SiTime SiT95316 and SiT95317
>> clock generators. These devices provide low-jitter clock outputs
>> commonly used in telecom, networking, and data center timing
>> applications.
>> 
>> The driver exposes all inputs and outputs through the Linux DPLL
>> subsystem, supporting:
>>   - Lock status monitoring via register polling or optional INTRB IRQ
>>   - Input priority management for automatic reference switchover
>>   - Per-output frequency readback from hardware state
>>   - Phase offset measurement via TDC (time-to-digital converter)
>>   - Phase adjustment for fine output alignment
>>   - Embedded sync (esync) pulse control on outputs
>>   - Fractional frequency offset of the selected reference
>>   - Optional reset-gpios for hardware reset
>> 
>> The driver reads all configuration from the device's on-chip NVM
>> at probe time -- no firmware loading is required.
>> 
>> Co-developed-by: Oleg Zadorozhnyi <Oleg.Zadorozhnyi@devoxsoftware.com>
>> Signed-off-by: Oleg Zadorozhnyi <Oleg.Zadorozhnyi@devoxsoftware.com>
>> Assisted-by: Claude:claude-4-opus [chat]
>> Signed-off-by: Ali Rouhi <arouhi@sitime.com>
>> ---
>>   MAINTAINERS                    |    7 +
>>   drivers/dpll/Kconfig           |    1 +
>>   drivers/dpll/Makefile          |    1 +
>>   drivers/dpll/sit9531x/Kconfig  |   17 +
>>   drivers/dpll/sit9531x/Makefile |    4 +
>>   drivers/dpll/sit9531x/core.c   | 3111 ++++++++++++++++++++++++++++++++
>>   drivers/dpll/sit9531x/core.h   |  372 ++++
>>   drivers/dpll/sit9531x/dpll.c   | 1232 +++++++++++++
>>   drivers/dpll/sit9531x/dpll.h   |   69 +
>>   drivers/dpll/sit9531x/prop.c   |  397 ++++
>>   drivers/dpll/sit9531x/prop.h   |   39 +
>>   drivers/dpll/sit9531x/regs.h   |  371 ++++
>>   12 files changed, 5621 insertions(+)
>
>as the new version is need, could you please think of how it can be
>split. reviewing if 5.6k lines of code in one go is quite tough

It's a new driver. Linear review in one hunk. I think it is better that
way as you see the whole thing right away.