Re: [Linaro-open-discussions] Linaro : Meeting : Linux Kernel Discussion : Wednesday Meeting notes
Hi Ulf,
Sorry for the late reply, please see my comments inline.
On 2020/11/6 22:37, Ulf Hansson via Linaro-open-discussions wrote:
On Fri, 6 Nov 2020 at 09:53, Mike Holmes via Linaro-open-discussions linaro-open-discussions@op-lists.linaro.org wrote:
All
Thanks for a good discussion yesterday, Hanjun's Unconre DVFS slides are now attached to the notes at https://collaborate.linaro.org/display/LOD/2020-11-04+Meeting+Meeting+notes
Just to make sure people know about it...
For DT based platforms we have recently enabled a generic solution in the Linux kernel to support DVFS. From a top level point of view, it consists of letting a power-domain provider act as the backend driver for changing the performance state (voltage/frequency). The mapping of what state to select, is done through OPP tables.
I am not an expert of ACPI specifications, but perhaps the DT bindings can be used as inspiration for what is missing!?
Could you point me to the patches or documentation?
Thanks Hanjun
+ Viresh
Hi Hanjun,
On Thu, 12 Nov 2020 at 02:32, Hanjun Guo guohanjun@huawei.com wrote:
Hi Ulf,
Sorry for the late reply, please see my comments inline.
On 2020/11/6 22:37, Ulf Hansson via Linaro-open-discussions wrote:
On Fri, 6 Nov 2020 at 09:53, Mike Holmes via Linaro-open-discussions linaro-open-discussions@op-lists.linaro.org wrote:
All
Thanks for a good discussion yesterday, Hanjun's Unconre DVFS slides are now attached to the notes at https://collaborate.linaro.org/display/LOD/2020-11-04+Meeting+Meeting+notes
Just to make sure people know about it...
For DT based platforms we have recently enabled a generic solution in the Linux kernel to support DVFS. From a top level point of view, it consists of letting a power-domain provider act as the backend driver for changing the performance state (voltage/frequency). The mapping of what state to select, is done through OPP tables.
I am not an expert of ACPI specifications, but perhaps the DT bindings can be used as inspiration for what is missing!?
Could you point me to the patches or documentation?
In the kernel git these DT docs should be the most important, I think:
Documentation/devicetree/bindings/power/power-domain.yaml Documentation/devicetree/bindings/power/power_domain.txt Documentation/devicetree/bindings/opp/opp.txt
I have looped in Viresh, maybe he has some additional OPP docs to point to?
Kind regards Uffe
On 13-11-20, 10:02, Ulf Hansson wrote:
- Viresh
Hi Hanjun,
On Thu, 12 Nov 2020 at 02:32, Hanjun Guo guohanjun@huawei.com wrote:
Hi Ulf,
Sorry for the late reply, please see my comments inline.
On 2020/11/6 22:37, Ulf Hansson via Linaro-open-discussions wrote:
On Fri, 6 Nov 2020 at 09:53, Mike Holmes via Linaro-open-discussions linaro-open-discussions@op-lists.linaro.org wrote:
All
Thanks for a good discussion yesterday, Hanjun's Unconre DVFS slides are now attached to the notes at https://collaborate.linaro.org/display/LOD/2020-11-04+Meeting+Meeting+notes
Just to make sure people know about it...
For DT based platforms we have recently enabled a generic solution in the Linux kernel to support DVFS. From a top level point of view, it consists of letting a power-domain provider act as the backend driver for changing the performance state (voltage/frequency). The mapping of what state to select, is done through OPP tables.
I am not an expert of ACPI specifications, but perhaps the DT bindings can be used as inspiration for what is missing!?
Could you point me to the patches or documentation?
In the kernel git these DT docs should be the most important, I think:
Documentation/devicetree/bindings/power/power-domain.yaml Documentation/devicetree/bindings/power/power_domain.txt Documentation/devicetree/bindings/opp/opp.txt
That's all what I have as well.
I have looped in Viresh, maybe he has some additional OPP docs to point to?
linaro-open-discussions@op-lists.linaro.org
-
Hanjun Guo -
Ulf Hansson -
Viresh Kumar