Skip to content

Commit 88e3df2

Browse files
aryanputtaaryanputta
authored
docs(core): drop incorrect handle_type requirement for cuMemRetainAllocationHandle (#2418)
Defect 4 of #2388. Signed-off-by: Aryan <[email protected]>
1 parent 608cc5f commit 88e3df2

1 file changed

Lines changed: 3 additions & 4 deletions

File tree

‎cuda_core/cuda/core/_memory/_virtual_memory_resource.py‎

Lines changed: 3 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -46,10 +46,9 @@ class VirtualMemoryResourceOptions:
4646
location_type: :obj:`~_memory.VirtualMemoryLocationType` | str
4747
Controls the location of the allocation.
4848
handle_type: :obj:`~_memory.VirtualMemoryHandleType` | str
49-
Export handle type for the physical allocation. Use
50-
``"posix_fd"`` on Linux if you plan to
51-
import/export the allocation (required for cuMemRetainAllocationHandle).
52-
Use `None` if you don't need an exportable handle.
49+
Export handle type for the physical allocation. Use ``"posix_fd"`` on
50+
Linux if you plan to import/export the allocation. Use `None` if you
51+
don't need an exportable handle.
5352
gpu_direct_rdma: bool
5453
Hint that the allocation should be GDR-capable (if supported).
5554
granularity: :obj:`~_memory.VirtualMemoryGranularityType` | str

0 commit comments

Comments
 (0)