Re: [PATCH v4 0/5] Thermal devfreq cooling improvements with Energy Model

[Lukasz Luba]

Hi Daniel,

Do you think it has chance to go to as material for v5.11?

Regards,
Lukasz


On 12/10/20 2:30 PM, Lukasz Luba wrote:
> Hi all,
>
> This patch set is a continuation of my previous work, which aimed
> to add Energy Model to all devices [1]. This series is a follow up
> for the patches which got merged to v5.9-rc1. It aims to change
> the thermal devfreq cooling and use the Energy Model instead of
> private power table and structures. The power model is now simplified,
> static power and dynamic power are removed. The new registration interface
> in the patch 3/5 helps to register devfreq cooling and the EM in one call.
> There is also small improvement, patch 2/5 is changing the way how
> thermal gets the device status (now uses a copy) and normalize the values.
> The last patch is here for consistency and will probably go through drm tree.
>
> The patch set should apply on top of thermal/testing. It does not depend on
> new EM API change which is queued in the pm/linux-next tree as v5.11 material.
> Thus, could go in parallel. That was the main motiviation for this v4.
>
> changes:
> v4:
> - patch 3/5 - removed dependency on the EM API change
> -- removed em_dev_register_perf_domain() and just use
>     dev_pm_opp_of_register_em() which API has not changed
> -- removed a helper registration function and renamed
>     devfreq_cooling_em_register_power() to devfreq_cooling_em_register()
>     (was actually suggested by Ionela during review)
> -- moved energy_model.h to include in devfreq_cooling.c not .h, since
>     there is no EM structure in there anymore
> - adjusted comments and commit messages
> v3 [4]:
> - dropped direct check of device status and used just a copy of 'status';
>    a separate patch set will be proposed to address this issue
> - modified _normalize_load() and used 1024 scale to handle ms, us, ns
> - removed 'em_registered' and called em_dev_unregister_perf_domain()
>    unconditionally, so the drivers will have to make sure the right order of
>    all unregister calls to frameworks which might use EM; this call must be last
>    one; a proper comment added
> - removed 'em' pointer from struct devfreq_cooling_device, 'dev->em_pd' is used
> - removed of_node_get/put(), since the code can handle it
> - removed dfc_em_get_requested_power() (as missed to do it in v2)
> - collected all Reviewed-by tags
> v2 [3]:
> - renamed freq_get_state() and related to perf_idx pattern as
>    suggested by Ionela
> v1 [2]
>
> Regards,
> Lukasz Luba
>
> Lukasz Luba (5):
>    thermal: devfreq_cooling: change tracing function and arguments
>    thermal: devfreq_cooling: use a copy of device status
>    thermal: devfreq_cooling: add new registration functions with Energy
>      Model
>    thermal: devfreq_cooling: remove old power model and use EM
>    drm/panfrost: Register devfreq cooling and attempt to add Energy Model
>
>   drivers/gpu/drm/panfrost/panfrost_devfreq.c |   2 +-
>   drivers/thermal/devfreq_cooling.c           | 391 +++++++++-----------
>   include/linux/devfreq_cooling.h             |  27 +-
>   include/trace/events/thermal.h              |  19 +-
>   4 files changed, 198 insertions(+), 241 deletions(-)