[Intel-gfx] [PATCH] drm/i915/ggtt: Move ballooning support to i915_ggtt

[Chris Wilson]

Quoting Michal Wajdeczko (2019-11-21 17:58:50)
> Today VGT implements GGTT ballooning on its own, using some private
> static structure to hold info about reserved GGTT nodes and then
> manually update vm.reserved size owned by i915_ggtt.
> 
> As i915_ggtt already manages some custom reserved nodes (like uc_fw)
> move VGT ballooning support to i915_ggtt and make it more generic
> for possible future reuse in other scenarios.
> 
> As a bonus we can drop another place in driver that uses static data.
> 
> Signed-off-by: Michal Wajdeczko <michal.wajdeczko at intel.com>
> Cc: Xiong Zhang <xiong.y.zhang at intel.com>
> Cc: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Jani Nikula <jani.nikula at intel.com>
> ---
>  drivers/gpu/drm/i915/i915_gem_gtt.c | 47 +++++++++++++++++++
>  drivers/gpu/drm/i915/i915_gem_gtt.h | 18 ++++++++
>  drivers/gpu/drm/i915/i915_vgpu.c    | 72 ++++++-----------------------
>  3 files changed, 78 insertions(+), 59 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
> index 6239a9adbf14..d8b2084dab7e 100644
> --- a/drivers/gpu/drm/i915/i915_gem_gtt.c
> +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
> @@ -3827,6 +3827,53 @@ int i915_gem_gtt_insert(struct i915_address_space *vm,
>                                            start, end, DRM_MM_INSERT_EVICT);
>  }
>  
> +int i915_ggtt_balloon_space(struct i915_ggtt *ggtt,
> +                           enum i915_ggtt_balloon_id id,

Do we want to use a fixed number of slots?

So what exactly is wrong with the clients reserving nodes themselves?

I'm not seeing why exactly we want to move to a much more centralised
system and argue over ownership of i915_ggtt_balloon_id.
-Chris