|
XenevaOS
|
#include <amdgpu_drm.h>
| __u32 drm_amdgpu_gem_va::_pad |
| __u32 drm_amdgpu_gem_va::flags |
AMDGPU_VM_PAGE_*
| __u32 drm_amdgpu_gem_va::handle |
GEM object handle
| __u64 drm_amdgpu_gem_va::input_fence_syncobj_handles |
Array of sync object handle to wait for given input fences
| __u64 drm_amdgpu_gem_va::map_size |
Specify mapping size. Must be correctly aligned.
| __u32 drm_amdgpu_gem_va::num_syncobj_handles |
the number of syncobj handles in @input_fence_syncobj_handles
| __u64 drm_amdgpu_gem_va::offset_in_bo |
Specify offset inside of BO to assign. Must be correctly aligned.
| __u32 drm_amdgpu_gem_va::operation |
AMDGPU_VA_OP_*
| __u64 drm_amdgpu_gem_va::va_address |
va address to assign . Must be correctly aligned.
| __u64 drm_amdgpu_gem_va::vm_timeline_point |
vm_timeline_point is a sequence number used to add new timeline point.
| __u32 drm_amdgpu_gem_va::vm_timeline_syncobj_out |
The vm page table update fence is installed in given vm_timeline_syncobj_out at vm_timeline_point.