[Intel-gfx] [i-g-t PATCH v1 0/2] Avoid calling DRM_IOCTL_I915_GET_PIPE_FROM_CRTC_ID on !i915

Daniel Vetter daniel at ffwll.ch
Wed Apr 20 14:14:48 UTC 2016


On Tue, Apr 19, 2016 at 01:40:26PM +0200, Tomeu Vizoso wrote:
> Hi,
> 
> following Daniel's comments on a previous series that completely removed
> usage of DRM_IOCTL_I915_GET_PIPE_FROM_CRTC_ID, I'm sending this one that
> removes it only on codepaths that can be run on drivers other than i915.
> 
> It does so by using the already-available crtc indexes instead of
> querying for the pipe number (which should return the same value).

Problem is a bit that KMS ids in general are really hard to tell apart. In
the kernel we started to have proper names for all of them, and let
drivers set those on a per-chip basis even, to make sure they match the hw
specs.

I do think keeping the pipe id around is somewhat useful, since for most
hw it's a valid concept to enumerate crtcs like that.
-Daniel

> 
> Thanks,
> 
> Tomeu
> 
> 
> Tomeu Vizoso (2):
>   lib: Remove superfluous kmstest_connector_config.pipe
>   tests/kms_setmode: Remove superfluous call to
>     kmstest_get_pipe_from_crtc_id
> 
>  lib/igt_kms.c                     | 24 +++++++++++-------------
>  lib/igt_kms.h                     |  1 -
>  tests/kms_crtc_background_color.c |  2 +-
>  tests/kms_flip.c                  | 10 +++++-----
>  tests/kms_flip_tiling.c           |  2 +-
>  tests/kms_panel_fitting.c         |  2 +-
>  tests/kms_plane_scaling.c         |  2 +-
>  tests/kms_render.c                |  2 +-
>  tests/kms_setmode.c               |  5 +----
>  tests/testdisplay.c               |  1 -
>  10 files changed, 22 insertions(+), 29 deletions(-)
> 
> -- 
> 2.5.5
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch


More information about the Intel-gfx mailing list