KhronosGroup / KhronosGroup/OpenCL-Docs
interaction between host process virtual address space and max mem alloc size
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 421
- Forks
- 132
- Avg merge
- 5d 13h
- Merged PRs (30d)
- 11
Description
See discussion in https://github.com/KhronosGroup/OpenCL-CTS/pull/1501.
Some OpenCL devices support more significantly more memory (CL_DEVICE_GLOBAL_MEM_SIZE) and larger memory allocations (CL_DEVICE_MAX_MEM_ALLOC_SIZE) than the host process virtual address space for 32-bit applications. Do we need to add any clarifications to the spec to document what should happen in this case? In particular, are there cases where it would make sense for CL_DEVICE_MAX_MEM_ALLOC_SIZE to exceed the size of the host process virtual address space?
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reading the discussion in OpenCL-CTS pull request 1501 and the linked OpenCL specification sections for CL_DEVICE_GLOBAL_MEM_SIZE and CL_DEVICE_MAX_MEM_ALLOC_SIZE. Determine whether the specification needs clarification for 32-bit applications whose host process virtual address space is smaller than the device limits, and document the expected behavior if a decision is reached.
Written by the indexing model from the issue text.
Assessment
- Domain
- documentation
- Issue type
- Documentation
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100