Karol Herbst
2021-Jul-24 14:13 UTC
[Nouveau] [PATCH] nouveau: make backlight support non optional
On Sat, Jul 24, 2021 at 4:05 PM Arnd Bergmann <arnd at kernel.org> wrote:> > On Sat, Jul 24, 2021 at 2:52 PM Karol Herbst <kherbst at redhat.com> wrote: > > > > On Sat, Jul 24, 2021 at 2:10 PM Karol Herbst <kherbst at redhat.com> wrote: > > > > > > On Sat, Jul 24, 2021 at 1:56 PM Arnd Bergmann <arnd at kernel.org> wrote: > > > > > > > > On Sat, Jul 24, 2021 at 11:55 AM Karol Herbst <kherbst at redhat.com> wrote: > > > > > > > > - You run into dependency loops that prevent a successful build when some > > > > other driver has a 'depends on'. Preventing these loops was the main > > > > reason I said we should do this change. > > > > > > > > In theory we could change the other 85 drivers that use 'depends on' today, > > > > and make BACKLIGHT_CLASS_DEVICE a hidden symbol that only ever > > > > selected by the drivers that need it. This would avoid the third problem but > > > > not the other one. > > > > > > > I see. Yeah, I guess we can do it this way then. I just wasn't aware > > > of the bigger picture here. Thanks for explaining. > > > > yeah... that doesn't work. So the issue is, that X86_PLATFORM_DEVICES > > is a little bit in the way. If I remove the select > > X86_PLATFORM_DEVICES then I guess problems once ACPI is enabled, but > > if I keep it, I get cyclic dep errors :/ > > Right, this is the exact problem I explained: since all other drivers use > 'depends on X86_PLATFORM_DEVICES' instead of 'select', you get a > loop again. Prior to changing the BACKLIGHT_CLASS_DEVICE dependency, > nouveau was pretty much on top of everything else in the hierarchy, > changing part of it can result in a loop. > > I see that there are about ten more 'select' statements that look like > they should not be there, and almost all of them were added in order > to be able to 'select MXM_WMI'. > > I think we can go as far as the patch below, which I've put in my > randconfig build machine, on top of your patch. I'm not entirely > sure how strong the dependency on MXM_WMI is: does it cause > a build failure when that is not enabled, or was this select just > added for convenience so users don't get surprised when it's missing? > > Arnd >we use the MXM_WMI in code. We also have to keep arm in mind and not break stuff there. So I will try to play around with your changes and see how that goes.> diff --git a/drivers/gpu/drm/nouveau/Kconfig b/drivers/gpu/drm/nouveau/Kconfig > index 9c2108b48524..f2585416507e 100644 > --- a/drivers/gpu/drm/nouveau/Kconfig > +++ b/drivers/gpu/drm/nouveau/Kconfig > @@ -3,21 +3,14 @@ config DRM_NOUVEAU > tristate "Nouveau (NVIDIA) cards" > depends on DRM && PCI && MMU > depends on AGP || !AGP > + depends on ACPI_VIDEO || !ACPI > + depends on BACKLIGHT_CLASS_DEVICE > + depends on MXM_WMI || !X86 || !ACPI > select IOMMU_API > select FW_LOADER > select DRM_KMS_HELPER > select DRM_TTM > select DRM_TTM_HELPER > - select BACKLIGHT_CLASS_DEVICE > - select ACPI_VIDEO if ACPI && X86 && INPUT > - select X86_PLATFORM_DEVICES if ACPI && X86 > - select ACPI_WMI if ACPI && X86 > - select MXM_WMI if ACPI && X86 > - select POWER_SUPPLY > - # Similar to i915, we need to select ACPI_VIDEO and it's dependencies > - select INPUT if ACPI && X86 > - select THERMAL if ACPI && X86 > - select ACPI_VIDEO if ACPI && X86 > select SND_HDA_COMPONENT if SND_HDA_CORE > help > Choose this option for open-source NVIDIA support. >
Arnd Bergmann
2021-Jul-24 14:58 UTC
[Nouveau] [PATCH] nouveau: make backlight support non optional
On Sat, Jul 24, 2021 at 4:14 PM Karol Herbst <kherbst at redhat.com> wrote:> > we use the MXM_WMI in code. We also have to keep arm in mind and not > break stuff there. So I will try to play around with your changes and > see how that goes.Ok, should find any randconfig build failures for arm, arm64 or x86 over the weekend. I also this on linux-next today ld: drivers/gpu/drm/i915/display/intel_panel.o: in function `intel_backlight_device_register': intel_panel.c:(.text+0x2804): undefined reference to `backlight_device_register' ld: intel_panel.c:(.text+0x284e): undefined reference to `backlight_device_register' ld: drivers/gpu/drm/i915/display/intel_panel.o: in function `intel_backlight_device_unregister': intel_panel.c:(.text+0x28b1): undefined reference to `backlight_device_unregister' and I added this same thing there to see how it goes: diff --git a/drivers/gpu/drm/i915/Kconfig b/drivers/gpu/drm/i915/Kconfig index 87825d36335b..69c6b7aec49e 100644 --- a/drivers/gpu/drm/i915/Kconfig +++ b/drivers/gpu/drm/i915/Kconfig @@ -3,6 +3,8 @@ config DRM_I915 tristate "Intel 8xx/9xx/G3x/G4x/HD Graphics" depends on DRM depends on X86 && PCI + depends on ACPI_VIDEO || !ACPI + depends on BACKLIGHT_CLASS_DEVICE select INTEL_GTT select INTERVAL_TREE # we need shmfs for the swappable backing store, and in particular @@ -16,10 +18,6 @@ config DRM_I915 select IRQ_WORK # i915 depends on ACPI_VIDEO when ACPI is enabled # but for select to work, need to select ACPI_VIDEO's dependencies, ick - select DRM_I915_BACKLIGHT if ACPI - select INPUT if ACPI - select ACPI_VIDEO if ACPI - select ACPI_BUTTON if ACPI select SYNC_FILE select IOSF_MBI select CRC32 @@ -64,13 +62,7 @@ config DRM_I915_FORCE_PROBE Use "*" to force probe the driver for all known devices. config DRM_I915_BACKLIGHT - tristate "Control backlight support" - depends on DRM_I915 - default DRM_I915 - select BACKLIGHT_CLASS_DEVICE - help - Say Y here if you want to control the backlight of your display - (e.g. a laptop panel). + def_tristate DRM_I915 config DRM_I915_CAPTURE_ERROR bool "Enable capturing GPU state following a hang"