[Intel-gfx] [PATCH v2 4/8] drm/i915: introduce mem->reserved

[Chris Wilson]

Quoting Matthew Auld (2021-01-26 15:12:55)
> From: Abdiel Janulgue <abdiel.janulgue at linux.intel.com>
> 
> In the following patch we need to reserve regions unaccessible to the
> driver during initialization, so add mem->reserved for collecting such
> regions.
> 
> Cc: Imre Deak <imre.deak at intel.com>
> Signed-off-by: Abdiel Janulgue <abdiel.janulgue at linux.intel.com>
> Signed-off-by: Matthew Auld <matthew.auld at intel.com>
> ---
>  drivers/gpu/drm/i915/intel_memory_region.c    |  2 +
>  drivers/gpu/drm/i915/intel_memory_region.h    |  2 +
>  .../drm/i915/selftests/intel_memory_region.c  | 89 +++++++++++++++++++
>  3 files changed, 93 insertions(+)
> 
> diff --git a/drivers/gpu/drm/i915/intel_memory_region.c b/drivers/gpu/drm/i915/intel_memory_region.c
> index b1b610bfff09..803a414c9523 100644
> --- a/drivers/gpu/drm/i915/intel_memory_region.c
> +++ b/drivers/gpu/drm/i915/intel_memory_region.c
> @@ -156,6 +156,7 @@ int intel_memory_region_init_buddy(struct intel_memory_region *mem)
>  
>  void intel_memory_region_release_buddy(struct intel_memory_region *mem)
>  {
> +       i915_buddy_free_list(&mem->mm, &mem->reserved);
>         i915_buddy_fini(&mem->mm);
>  }
>  
> @@ -185,6 +186,7 @@ intel_memory_region_create(struct drm_i915_private *i915,
>         mutex_init(&mem->objects.lock);
>         INIT_LIST_HEAD(&mem->objects.list);
>         INIT_LIST_HEAD(&mem->objects.purgeable);
> +       INIT_LIST_HEAD(&mem->reserved);
>  
>         mutex_init(&mem->mm_lock);
>  
> diff --git a/drivers/gpu/drm/i915/intel_memory_region.h b/drivers/gpu/drm/i915/intel_memory_region.h
> index 6ffc0673f005..8c9947bba3e8 100644
> --- a/drivers/gpu/drm/i915/intel_memory_region.h
> +++ b/drivers/gpu/drm/i915/intel_memory_region.h
> @@ -89,6 +89,8 @@ struct intel_memory_region {
>         unsigned int id;
>         char name[8];
>  
> +       struct list_head reserved;

So you added this, but this is not exercised by the selftest either.

I think you want to complete the API with
intel_memory_region_reserve(mem, start, size/end)

And then inspect the mem->reserved rather than the local list.

Presumably check the mem->reserved is empty on entry, and then it's
safe to call i915_buddy_free_list(&mem->mm, &mem->reserved) directly.
-Chris