On 4/20/22 14:21, Thomas Zimmermann wrote: > Hi > > Am 20.04.22 um 14:02 schrieb Javier Martinez Canillas: >> Hello Thomas, >> >> On 4/20/22 13:08, Thomas Zimmermann wrote: >> >> [snip] >> >>> --- a/drivers/gpu/drm/bridge/synopsys/Kconfig >>> +++ b/drivers/gpu/drm/bridge/synopsys/Kconfig >>> @@ -1,6 +1,8 @@ >>> # SPDX-License-Identifier: GPL-2.0-only >>> config DRM_DW_HDMI >>> tristate >>> + select DRM_DISPLAY_HDMI_HELPER >>> + select DRM_DISPLAY_HELPER >>> select DRM_KMS_HELPER >>> select REGMAP_MMIO >>> select CEC_CORE if CEC_NOTIFIER >> >> [snip] >> >>> +config DRM_DISPLAY_HDMI_HELPER >>> + bool >>> + select DRM_DISPLAY_HELPER >>> + help >>> + DRM display helpers for HDMI. >>> + >> >> Maybe you mentioned already when we discussed it in v1, but could >> you please remember me why we need to select both symbols for the >> drivers, even when DRM_DISPLAY_HELPER is already selected here ? >> >> Or are select operations not transitive ? > > DRM_DISPLAY_HELPER enables the module itself. It's tristate. > DRM_DISPLAY_HDMI_HELPER, DRM_DISPLAY_DP_HELPER, etc are the individual > features of the helper module. Each is a boolean. Drivers can enable > them according to their needs. This split allows linking correctly in > all cases, even if some drivers are built-in and others are modules. > This didn't always work with the original patches. > Right, but that wasn't my question. I wondered why for example DRM_DW_HDMI Kconfig needs to select both DRM_DISPLAY_HDMI_HELPER and DRM_DISPLAY_HELPER since DRM_DISPLAY_HDMI_HELPER already selects DRM_DISPLAY_HELPER. -- Best regards, Javier Martinez Canillas Linux Engineering Red Hat
