[PATCH v7 00/24] MADVISE FOR XE
Rodrigo Vivi
rodrigo.vivi at intel.com
Mon Aug 18 20:28:09 UTC 2025
On Thu, Aug 14, 2025 at 04:29:26PM +0530, Himal Prasad Ghimiray wrote:
> -v8
> Define new function in gpuvm for madvise
> Fix tile check in ops_create
Please re-send this series again. But this time, ensure that dri-devel
is in the --to= of the series!
In 8 revisions, it looks like not a single time it was sent completely
to dri-devel. Big changes like this needs to be pretty visible.
Ensure to get all the appropriate acks in all the relevant patches as
well after resending to dri-devel.
Thanks,
Rodrigo.
>
> -v7
> Change gpuvm layering on gpuvm_map_req struct
> Fix EAGAIN return on garbage collector splitting vma
>
> -v6
> Rebase on gpuvm patches
> Address review comments
>
> -v5
> Restore attributes to default after free from userspace
> Add defragment worker to merge cpu mirror vma with default attributes
> Avoid using VMA in uapi
> address review comments
>
> -v4:
> fix atomic policies
> fix attribute copy
> address review comments
>
> Provides a user API to assign attributes like pat_index, atomic
> operation type, and preferred location for SVM ranges.
> The Kernel Mode Driver (KMD) may split existing VMAs to cover input
> ranges, assign user-provided attributes, and invalidate existing PTEs so
> that the next page fault/prefetch can use the new attributes.
>
> Boris Brezillon (2):
> drm/gpuvm: Pass map arguments through a struct
> drm/gpuvm: Kill drm_gpuva_init()
>
> Himal Prasad Ghimiray (22):
> drm/gpuvm: Introduce drm_gpuvm_madvise_ops_create
> drm/xe/uapi: Add madvise interface
> drm/xe/vm: Add attributes struct as member of vma
> drm/xe/vma: Move pat_index to vma attributes
> drm/xe/vma: Modify new_vma to accept struct xe_vma_mem_attr as
> parameter
> drm/gpusvm: Make drm_gpusvm_for_each_* macros public
> drm/xe/svm: Split system allocator vma incase of madvise call
> drm/xe: Allow CPU address mirror VMA unbind with gpu bindings for
> madvise
> drm/xe/svm: Add xe_svm_ranges_zap_ptes_in_range() for PTE zapping
> drm/xe: Implement madvise ioctl for xe
> drm/xe/svm: Add svm ranges migration policy on atomic access
> drm/xe/madvise: Update migration policy based on preferred location
> drm/xe/svm: Support DRM_XE_SVM_MEM_RANGE_ATTR_PAT memory attribute
> drm/xe/uapi: Add flag for consulting madvise hints on svm prefetch
> drm/xe/svm: Consult madvise preferred location in prefetch
> drm/xe/bo: Add attributes field to xe_bo
> drm/xe/bo: Update atomic_access attribute on madvise
> drm/xe/madvise: Skip vma invalidation if mem attr are unchanged
> drm/xe/vm: Add helper to check for default VMA memory attributes
> drm/xe: Reset VMA attributes to default in SVM garbage collector
> drm/xe: Enable madvise ioctl for xe
> drm/xe/uapi: Add UAPI for querying VMA count and memory attributes
>
> drivers/gpu/drm/drm_gpusvm.c | 122 ++-----
> drivers/gpu/drm/drm_gpuvm.c | 287 ++++++++++++----
> drivers/gpu/drm/imagination/pvr_vm.c | 15 +-
> drivers/gpu/drm/msm/msm_gem_vma.c | 33 +-
> drivers/gpu/drm/nouveau/nouveau_uvmm.c | 11 +-
> drivers/gpu/drm/panthor/panthor_mmu.c | 13 +-
> drivers/gpu/drm/xe/Makefile | 1 +
> drivers/gpu/drm/xe/xe_bo.c | 29 +-
> drivers/gpu/drm/xe/xe_bo_types.h | 8 +
> drivers/gpu/drm/xe/xe_device.c | 4 +
> drivers/gpu/drm/xe/xe_gt_pagefault.c | 35 +-
> drivers/gpu/drm/xe/xe_pt.c | 39 ++-
> drivers/gpu/drm/xe/xe_svm.c | 254 ++++++++++++--
> drivers/gpu/drm/xe/xe_svm.h | 23 ++
> drivers/gpu/drm/xe/xe_vm.c | 438 ++++++++++++++++++++++--
> drivers/gpu/drm/xe/xe_vm.h | 10 +-
> drivers/gpu/drm/xe/xe_vm_madvise.c | 445 +++++++++++++++++++++++++
> drivers/gpu/drm/xe/xe_vm_madvise.h | 15 +
> drivers/gpu/drm/xe/xe_vm_types.h | 57 +++-
> include/drm/drm_gpusvm.h | 70 ++++
> include/drm/drm_gpuvm.h | 38 ++-
> include/uapi/drm/xe_drm.h | 274 +++++++++++++++
> 22 files changed, 1922 insertions(+), 299 deletions(-)
> create mode 100644 drivers/gpu/drm/xe/xe_vm_madvise.c
> create mode 100644 drivers/gpu/drm/xe/xe_vm_madvise.h
>
> --
> 2.34.1
>
More information about the Intel-xe
mailing list