Re: [PATCH 1/3] drm/amd/display: replace STATIC_IFN_KUNIT with VISIBLE_IF_KUNIT

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] Missing prototypes and global namespace pollution for generic function names when CONFIG_KUNIT is enabled but CONFIG_DRM_AMD_DC_KUNIT_TEST is disabled.
--

commit 559b79e4c5aa67bc448a3cebcac53b59d32d0f66
Author: Albert Esteve <aesteve@redhat.com>

drm/amd/display: replace STATIC_IFN_KUNIT with VISIBLE_IF_KUNIT
    
This commit switches amdgpu_dm test helpers from the local STATIC_IFN_KUNIT
macro to the generic VISIBLE_IF_KUNIT macro. The helpers become non-static
whenever KUnit is enabled, not just when AMD DC tests are built.

> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
> index ec483276d7538..5b3bebde0b083 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
> @@ -167,7 +168,7 @@ static inline void amdgpu_dm_exit_ips_for_hw_access(struct dc *dc)
>   * @return
>   * Counter for vertical blanks
>   */
> -STATIC_IFN_KUNIT u32 dm_vblank_get_counter(struct amdgpu_device *adev, int crtc)
> +VISIBLE_IF_KUNIT u32 dm_vblank_get_counter(struct amdgpu_device *adev, int crtc)

[Severity: Low]
If CONFIG_KUNIT is enabled but CONFIG_DRM_AMD_DC_KUNIT_TEST is disabled, 
VISIBLE_IF_KUNIT will make this function globally visible. However, if its 
prototype in drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h is still 
guarded by #if IS_ENABLED(CONFIG_DRM_AMD_DC_KUNIT_TEST), won't this cause a 
-Wmissing-prototypes build warning?

[ ... ]

> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_connector.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_connector.c
> index c8a1ab8c3b169..fe310b6d9f603 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_connector.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_connector.c
> @@ -724,7 +724,7 @@ amdgpu_dm_convert_color_depth_from_display_info(const struct drm_connector *conn
>  }
>  EXPORT_IF_KUNIT(amdgpu_dm_convert_color_depth_from_display_info);
>  
> -STATIC_IFN_KUNIT enum dc_aspect_ratio
> +VISIBLE_IF_KUNIT enum dc_aspect_ratio
>  get_aspect_ratio(const struct drm_display_mode *mode_in)

[Severity: Low]
Since VISIBLE_IF_KUNIT makes this function globally visible when CONFIG_KUNIT 
is enabled, does the generic name get_aspect_ratio risk namespace collisions 
with other display or media drivers?

[ ... ]

> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
> index 47cbc826ed451..20c1ce8cda446 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
> @@ -187,7 +190,7 @@ int psp_set_srm(struct psp_context *psp,
>  }
>  EXPORT_IF_KUNIT(psp_set_srm);
>  
> -STATIC_IFN_KUNIT
> +VISIBLE_IF_KUNIT
>  void process_output(struct hdcp_workqueue *hdcp_work)

[Severity: Low]
With VISIBLE_IF_KUNIT, this generic name process_output becomes a global 
symbol in vmlinux. Is there a risk of link-time collisions with other 
drivers?

Also, similar to the missing prototype issue mentioned earlier, could this 
lead to a build warning if the prototype in amdgpu_dm_hdcp.h is still 
guarded by CONFIG_DRM_AMD_DC_KUNIT_TEST?

[ ... ]

> @@ -386,7 +389,7 @@ void hdcp_handle_cpirq(struct hdcp_workqueue *hdcp_work, unsigned int link_index
>  }
>  EXPORT_IF_KUNIT(hdcp_handle_cpirq);
>  
> -STATIC_IFN_KUNIT
> +VISIBLE_IF_KUNIT
>  void event_callback(struct work_struct *work)

[Severity: Low]
Similar to the above, does exposing the generic name event_callback to the 
global namespace risk collisions when CONFIG_KUNIT is enabled?

[ ... ]

> @@ -608,7 +611,7 @@ bool enable_assr(void *handle, struct dc_link *link)
>  }
>  EXPORT_IF_KUNIT(enable_assr);
>  
> -STATIC_IFN_KUNIT
> +VISIBLE_IF_KUNIT
>  void update_config(void *handle, struct cp_psp_stream_config *config)

[Severity: Low]
Could making update_config globally visible cause namespace collisions, as it 
lacks a subsystem prefix?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260907-visibility-unification-v1-0-6a962e5b3351@redhat.com?part=1