[igt-dev] [PATCH 1/3] drm header update: temp commit

[Zbigniew Kempczyński]

On Fri, Jul 28, 2023 at 09:57:02AM +0200, Albert Esteve wrote:
>    On Fri, Jul 28, 2023 at 9:42 AM Zbigniew Kempczyński
>    <zbigniew.kempczynski at intel.com> wrote:
> 
>      On Fri, Jul 21, 2023 at 03:04:56PM +0200, Albert Esteve wrote:
>      > Update the drm included header to obtain
>      > VIRTUAL_CURSOR_PLANE capability and be
>      > able to test cursor hotspot setting.
>      >
>      > This commit should be removed once the kernel
>      > patch is applied and update the headers
>      > as appropiate.
>      >
>      > Signed-off-by: Albert Esteve <aesteve at redhat.com>
>      > ---
>      >  include/drm-uapi/drm.h | 54
>      +++++++++++++++++++++++++++++++++++++++++-
>      >  1 file changed, 53 insertions(+), 1 deletion(-)
>      >
> 
>      I'm sorry Albert but we don't allow to add not accepted changes
>      upstream regarding uapi headers. Only verbatim copy of uapi
>      headers may land in igt.
> 
>      In patch [PATCH 2/3] igt_kms: add hotspot plane property you're
>      unconditionally using definition which doesn't exists upstream
>      yet. Likely it will just return from ioctl but I got doubts it
>      should be written this way.
> 
>      I would suggest to introduce:
> 
>      #define LOCAL_DRM_CLIENT_CAP_CURSOR_PLANE_HOTSPOT▸ 6
> 
>      in igt_kms.h instead drm.h. And when drm.h will land in igt
>      remove this localized definition replacing with official one.
> 
>    Hi Zbigniew,
>    Sure, that is why I mention in the commit that it should be removed
>    once the kernel patch lands.
>    There is a dependency, and I was not expecting this patch to be
>    upstreamed before the kernel patch lands.
>    But I'm ok with your suggestion. It will result in a cleaner patch to
>    review :)
>    Thanks!

In the meantime I've just verified I'm not talking nonsense and it
seems I'm talking. At the moment doing verbatim copy of kernel
uapi headers conflicts due to missing __user definition and
unresolved __packed in single place in i915_drm.h. I'm thinking how
to solve this with minimal effort as I would like to avoid
changes in uapi headers or have as minimum changes as possible.

--
Zbigniew