mbox series

[v4,0/3] thermal/netlink/intel_hfi: Enable HFI feature only when required

Message ID 20240223155942.60813-1-stanislaw.gruszka@linux.intel.com (mailing list archive)
Headers show
Series thermal/netlink/intel_hfi: Enable HFI feature only when required | expand

Message

Stanislaw Gruszka Feb. 23, 2024, 3:59 p.m. UTC
The patchset is based on

 https://git.kernel.org/pub/scm/linux/kernel/git/kuba/linux.git
 for-thermal-genetlink-family-bind-unbind-callbacks

and implement genetlink family bind/unbind callbacks in thermal-netlink and
add notifications to thermal subdrivers. Those willh allow drivers to send
netlink multicast events based on the presence of actual user-space consumers.
This functionality optimizes resource usage by allowing disabling
of features when not needed.

Then implement the notification mechanism in the intel_hif driver,
it is utilized to disable the Hardware Feedback Interface (HFI)
dynamically. By implementing a thermal genl notify callback, the driver
can now enable or disable the HFI based on actual demand, particularly
when user-space applications like intel-speed-select or Intel Low Power
daemon utilize events related to performance and energy efficiency
capabilities.

On machines where Intel HFI is present, but there are no user-space
components installed, we can save tons of CPU cycles.

Changes v4 -> v5:

- Fix hif vs. hfi in the changelog
- Rename thermal_gnl_chain
- Add new patch with rename of thermal_gnl_family
- Fix syscore concurrency comment
- Remove unneeded hfi_instance variable
- Rename hfi_do_enable/disable
- Avoid multiple enabling 
- Add comment about registering for events before they can be generated
- Rename hfi_thermal_clients_num since later there will be KVM clients

Changes v3 -> v4:

- Add 'static inline' in patch2

Changes v2 -> v3:

- Fix unused variable compilation warning
- Add missed Suggested by tag to patch2
 
Changes v1 -> v2:

- Rewrite using netlink_bind/netlink_unbind callbacks.

- Minor changelog tweaks.

- Add missing check in intel hfi syscore resume (had it on my testing,
but somehow missed in post).

- Do not use netlink_has_listeners() any longer, use custom counter instead.
To keep using netlink_has_listners() would be required to rearrange 
netlink_setsockopt() and possibly netlink_bind() functions, to call 
nlk->netlink_bind() after listeners are updated. So I decided to custom
counter. This have potential issue as thermal netlink registers before
intel_hif, so theoretically intel_hif can miss events. But since both
are required to be kernel build-in (if CONFIG_INTEL_HFI_THERMAL is
configured), they start before any user-space.

v1: https://lore.kernel.org/linux-pm/20240131120535.933424-1-stanislaw.gruszka@linux.intel.com//
v2: https://lore.kernel.org/linux-pm/20240206133605.1518373-1-stanislaw.gruszka@linux.intel.com/
v3: https://lore.kernel.org/linux-pm/20240209120625.1775017-1-stanislaw.gruszka@linux.intel.com/
v4: https://lore.kernel.org/linux-pm/20240212161615.161935-1-stanislaw.gruszka@linux.intel.com/


*** BLURB HERE ***

Stanislaw Gruszka (3):
  thermal: netlink: Add genetlink bind/unbind notifications
  thermal: netlink: Rename thermal_gnl_family
  thermal: intel: hfi: Enable interface only when required

 drivers/thermal/intel/intel_hfi.c | 97 ++++++++++++++++++++++++++++---
 drivers/thermal/thermal_netlink.c | 62 +++++++++++++++-----
 drivers/thermal/thermal_netlink.h | 26 +++++++++
 3 files changed, 161 insertions(+), 24 deletions(-)

Comments

srinivas pandruvada Feb. 23, 2024, 6:26 p.m. UTC | #1
Your heading version is not correct:

It should be
 "[PATCH v5 0/3] thermal/netlink/intel_hfi: Enable HFI feature only
when required"

Not "[PATCH v4.."

Thanks,
Srinivas

On Fri, 2024-02-23 at 16:59 +0100, Stanislaw Gruszka wrote:
> The patchset is based on
> 
>  https://git.kernel.org/pub/scm/linux/kernel/git/kuba/linux.git
>  for-thermal-genetlink-family-bind-unbind-callbacks
> 
> and implement genetlink family bind/unbind callbacks in thermal-
> netlink and
> add notifications to thermal subdrivers. Those willh allow drivers to
> send
> netlink multicast events based on the presence of actual user-space
> consumers.
> This functionality optimizes resource usage by allowing disabling
> of features when not needed.
> 
> Then implement the notification mechanism in the intel_hif driver,
> it is utilized to disable the Hardware Feedback Interface (HFI)
> dynamically. By implementing a thermal genl notify callback, the
> driver
> can now enable or disable the HFI based on actual demand,
> particularly
> when user-space applications like intel-speed-select or Intel Low
> Power
> daemon utilize events related to performance and energy efficiency
> capabilities.
> 
> On machines where Intel HFI is present, but there are no user-space
> components installed, we can save tons of CPU cycles.
> 
> Changes v4 -> v5:
> 
> - Fix hif vs. hfi in the changelog
> - Rename thermal_gnl_chain
> - Add new patch with rename of thermal_gnl_family
> - Fix syscore concurrency comment
> - Remove unneeded hfi_instance variable
> - Rename hfi_do_enable/disable
> - Avoid multiple enabling 
> - Add comment about registering for events before they can be
> generated
> - Rename hfi_thermal_clients_num since later there will be KVM
> clients
> 
> Changes v3 -> v4:
> 
> - Add 'static inline' in patch2
> 
> Changes v2 -> v3:
> 
> - Fix unused variable compilation warning
> - Add missed Suggested by tag to patch2
>  
> Changes v1 -> v2:
> 
> - Rewrite using netlink_bind/netlink_unbind callbacks.
> 
> - Minor changelog tweaks.
> 
> - Add missing check in intel hfi syscore resume (had it on my
> testing,
> but somehow missed in post).
> 
> - Do not use netlink_has_listeners() any longer, use custom counter
> instead.
> To keep using netlink_has_listners() would be required to rearrange 
> netlink_setsockopt() and possibly netlink_bind() functions, to call 
> nlk->netlink_bind() after listeners are updated. So I decided to
> custom
> counter. This have potential issue as thermal netlink registers
> before
> intel_hif, so theoretically intel_hif can miss events. But since both
> are required to be kernel build-in (if CONFIG_INTEL_HFI_THERMAL is
> configured), they start before any user-space.
> 
> v1:
> https://lore.kernel.org/linux-pm/20240131120535.933424-1-stanislaw.gruszka@linux.intel.com//
> v2:
> https://lore.kernel.org/linux-pm/20240206133605.1518373-1-stanislaw.gruszka@linux.intel.com/
> v3:
> https://lore.kernel.org/linux-pm/20240209120625.1775017-1-stanislaw.gruszka@linux.intel.com/
> v4:
> https://lore.kernel.org/linux-pm/20240212161615.161935-1-stanislaw.gruszka@linux.intel.com/
> 
> 
> *** BLURB HERE ***
> 
> Stanislaw Gruszka (3):
>   thermal: netlink: Add genetlink bind/unbind notifications
>   thermal: netlink: Rename thermal_gnl_family
>   thermal: intel: hfi: Enable interface only when required
> 
>  drivers/thermal/intel/intel_hfi.c | 97 ++++++++++++++++++++++++++++-
> --
>  drivers/thermal/thermal_netlink.c | 62 +++++++++++++++-----
>  drivers/thermal/thermal_netlink.h | 26 +++++++++
>  3 files changed, 161 insertions(+), 24 deletions(-)
>