[Intel-xe] [PATCH 21/41] fixup! drm/xe/display: Implement display support

[Lucas De Marchi]

On Wed, Apr 12, 2023 at 10:25:16PM +0300, Jani Nikula wrote:
>On Wed, 12 Apr 2023, Lucas De Marchi <lucas.demarchi at intel.com> wrote:
>> On Wed, Apr 12, 2023 at 05:19:46PM +0300, Jani Nikula wrote:
>>>It's okay to include uapi/drm/i915_drm.h. It provides us with structs to
>>>make stuff compile, even if we don't use them in xe.
>>>
>>>The placement is perhaps a bit random, though.
>>>
>>>Signed-off-by: Jani Nikula <jani.nikula at intel.com>
>>>---
>>> drivers/gpu/drm/xe/compat-i915-headers/i915_vma.h | 1 +
>>> 1 file changed, 1 insertion(+)
>>>
>>>diff --git a/drivers/gpu/drm/xe/compat-i915-headers/i915_vma.h b/drivers/gpu/drm/xe/compat-i915-headers/i915_vma.h
>>>index 133c60cb3a09..c40bb27ac72a 100644
>>>--- a/drivers/gpu/drm/xe/compat-i915-headers/i915_vma.h
>>>+++ b/drivers/gpu/drm/xe/compat-i915-headers/i915_vma.h
>>>@@ -1,6 +1,7 @@
>>> #ifndef I915_VMA_H
>>> #define I915_VMA_H
>>>
>>>+#include <uapi/drm/i915_drm.h>
>>
>> my worry is that this creates an implicit dependency that ends up using
>> where it shouldn't, particularly when included in a header. Where
>> exactly this is needed? Can the include be moved to a .c?
>
>struct drm_intel_sprite_colorkey ckey member in struct
>intel_plane_state, in intel_display_types.h, and elsewhere. Can't be
>moved to a .c file.
>
>Without this, the member needs to be conditionally compiled, as well as
>all the code referencing it all over the place. Though in most places
>the code will just go away because it'll be unreferenced. But it needs
>to compile before being optimized away.

ok, it seems the alternative is worse.  LGTM

thanks
Lucas De Marchi

>
>BR,
>Jani.
>
>
>
>>
>> Lucas De Marchi
>>
>>> #include <drm/drm_mm.h>
>>>
>>> struct xe_bo;
>>>--
>>>2.39.2
>>>
>
>-- 
>Jani Nikula, Intel Open Source Graphics Center