[PATCH v7 03/24] drm/gpuvm: Introduce drm_gpuvm_madvise_ops_create

Danilo Krummrich dakr at kernel.org
Mon Aug 18 17:32:25 UTC 2025


On Thu Aug 14, 2025 at 12:59 PM CEST, Himal Prasad Ghimiray wrote:
> This ops is used to iterate over GPUVA's in the user-provided range
> and split the existing sparse VMA's if the start or end of the input
> range lies within it. The operations can create up to 2 REMAPS and 2 MAPs.
>
> The primary use case is for drivers to assign attributes to GPU VAs in
> the specified range without performing unmaps or merging mappings,
> supporting fine-grained control over sparse va's.
>
> v2
> - use drm_gpuvm_sm_map_ops_create with flags instead of defining new
>   ops_create (Danilo)
> - Add doc (Danilo)
>
> v3
> - Fix doc
> - Fix unmapping check
>
> v4
> - Fix mapping for non madvise ops
>
> v5
> - Fix mapping (Matthew Brost)
> - Rebase on top of struct changes
>
> v6
> - flag moved to map_req
>
> v7 (Danilo)
> - Use different functions
> - Add kernel-doc
> - Modify op_unmap_cb and op_map_cb to handle madvise and NULL ptr
> - use gem_obj check in single place

(Same comment as in patch 1.)

> Cc: Danilo Krummrich <dakr at kernel.org>
> Cc: Matthew Brost <matthew.brost at intel.com>
> Cc: Boris Brezillon <bbrezillon at kernel.org>
> Cc: <dri-devel at lists.freedesktop.org>
> Signed-off-by: Himal Prasad Ghimiray<himal.prasad.ghimiray at intel.com>
> Reviewed-by: Matthew Brost <matthew.brost at intel.com> #v6

This looks pretty clean now, and thanks for adding the documentation.

Acked-by: Danilo Krummrich <dakr at kernel.org>

Feel free to take this through the Xe tree if necessary.


More information about the Intel-xe mailing list