[Intel-gfx] [PATCH v2 4/4] drm/i915/uapi: document behaviour for DG2 64K support

[Jordan Justen]

Robert Beckett <bob.beckett at collabora.com> writes:

> From: Matthew Auld <matthew.auld at intel.com>
>
> On discrete platforms like DG2, we need to support a minimum page size
> of 64K when dealing with device local-memory. This is quite tricky for
> various reasons, so try to document the new implicit uapi for this.
>
> v2: Fixed suggestions on formatting [Daniel]
>
> Signed-off-by: Matthew Auld <matthew.auld at intel.com>
> Signed-off-by: Ramalingam C <ramalingam.c at intel.com>
> Signed-off-by: Robert Beckett <bob.beckett at collabora.com>
> cc: Simon Ser <contact at emersion.fr>
> cc: Pekka Paalanen <ppaalanen at gmail.com>
> Cc: Jordan Justen <jordan.l.justen at intel.com>
> Cc: Kenneth Graunke <kenneth at whitecape.org>
> Cc: mesa-dev at lists.freedesktop.org
> Cc: Tony Ye <tony.ye at intel.com>
> Cc: Slawomir Milczarek <slawomir.milczarek at intel.com>
> ---
>  include/uapi/drm/i915_drm.h | 44 ++++++++++++++++++++++++++++++++-----
>  1 file changed, 39 insertions(+), 5 deletions(-)
>
> diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h
> index 5e678917da70..486b7b96291e 100644
> --- a/include/uapi/drm/i915_drm.h
> +++ b/include/uapi/drm/i915_drm.h
> @@ -1118,10 +1118,16 @@ struct drm_i915_gem_exec_object2 {
>  	/**
>  	 * When the EXEC_OBJECT_PINNED flag is specified this is populated by
>  	 * the user with the GTT offset at which this object will be pinned.
> +	 *
>  	 * When the I915_EXEC_NO_RELOC flag is specified this must contain the
>  	 * presumed_offset of the object.
> +	 *
>  	 * During execbuffer2 the kernel populates it with the value of the
>  	 * current GTT offset of the object, for future presumed_offset writes.
> +	 *
> +	 * See struct drm_i915_gem_create_ext for the rules when dealing with
> +	 * alignment restrictions with I915_MEMORY_CLASS_DEVICE, on devices with
> +	 * minimum page sizes, like DG2.
>  	 */
>  	__u64 offset;
>  
> @@ -3145,11 +3151,39 @@ struct drm_i915_gem_create_ext {
>  	 *
>  	 * The (page-aligned) allocated size for the object will be returned.
>  	 *
> -	 * Note that for some devices we have might have further minimum
> -	 * page-size restrictions(larger than 4K), like for device local-memory.
> -	 * However in general the final size here should always reflect any
> -	 * rounding up, if for example using the I915_GEM_CREATE_EXT_MEMORY_REGIONS
> -	 * extension to place the object in device local-memory.
> +	 *
> +	 * **DG2 64K min page size implications:**

Long term, I'm not sure that the "**" (for emphasis) is needed here or
below. It's interesting at the moment, but will be just another thing
baked into the kernel/user code in a month from now. :)

> +	 *
> +	 * On discrete platforms, starting from DG2, we have to contend with GTT
> +	 * page size restrictions when dealing with I915_MEMORY_CLASS_DEVICE
> +	 * objects.  Specifically the hardware only supports 64K or larger GTT
> +	 * page sizes for such memory. The kernel will already ensure that all
> +	 * I915_MEMORY_CLASS_DEVICE memory is allocated using 64K or larger page
> +	 * sizes underneath.
> +	 *
> +	 * Note that the returned size here will always reflect any required
> +	 * rounding up done by the kernel, i.e 4K will now become 64K on devices
> +	 * such as DG2.
> +	 *
> +	 * **Special DG2 GTT address alignment requirement:**
> +	 *
> +	 * The GTT alignment will also need be at least 2M for  such objects.
> +	 *
> +	 * Note that due to how the hardware implements 64K GTT page support, we
> +	 * have some further complications:
> +	 *
> +	 *   1) The entire PDE(which covers a 2MB virtual address range), must
> +	 *   contain only 64K PTEs, i.e mixing 4K and 64K PTEs in the same
> +	 *   PDE is forbidden by the hardware.
> +	 *
> +	 *   2) We still need to support 4K PTEs for I915_MEMORY_CLASS_SYSTEM
> +	 *   objects.
> +	 *
> +	 * To keep things simple for userland, we mandate that any GTT mappings
> +	 * must be aligned to and rounded up to 2MB. As this only wastes virtual
> +	 * address space and avoids userland having to copy any needlessly
> +	 * complicated PDE sharing scheme (coloring) and only affects GD2, this
> +	 * id deemed to be a good compromise.

typos: GD2, id

Isn't much of this more relavent to the vma offset at exec time? Is
there actually any new restriction on the size field during buffer
creation?

I see Matthew references these notes from the offset comments, so if the
kernel devs prefer it here, then you can add my Acked-by on this patch.

-Jordan

>  	 */
>  	__u64 size;
>  	/**
> -- 
> 2.25.1