requiredAllocSize in min_max_mem_alloc_size does not match current CL_DEVICE_MAX_MEM_ALLOC_SIZE minimum for either profile
Maintainers usually reply within 1 day
Nobody has claimed this yet.
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 48/100
- Issue type
- Bug
- Clarity
- Mostly clear
- Activity status
- Active
- Tech stack
- cpp
- Domain
- testing-qa
Research direction
Start in test_conformance/api/test_api_min_max.cpp at min_max_mem_alloc_size, then compare its profile-specific values with the CL_DEVICE_MAX_MEM_ALLOC_SIZE specification and embedded_profile.asciidoc. Resolve whether the CTS expectation or specification is authoritative, update the affected documentation or test behavior accordingly, and run the relevant API conformance test.
Written by the indexing model from the issue text.
Description
Related to #2795
min_max_mem_alloc_size in test_conformance/api/test_api_min_max.cpp hardcodes:
if (gIsEmbedded)
requiredAllocSize = 1 * 1024 * 1024;
else
requiredAllocSize = 128 * 1024 * 1024;
Current spec text for CL_DEVICE_MAX_MEM_ALLOC_SIZE:
Max size of memory object allocation in bytes. The minimum value is max(min(1024 x 1024 x 1024, 1/4th of CL_DEVICE_GLOBAL_MEM_SIZE), 32 x 1024 x 1024).
That puts the actual floor at 32 MB for full profile, not 128 MB.
CL_DEVICE_MAX_MEM_ALLOC_SIZE has no entry in the embedded profile relaxation table (embedded-device-queries-table in embedded_profile.asciidoc), so the same 32 MB floor applies to embedded devices too. The 1 MB used here is roughly 32x below that.
Questions:
- Should requiredAllocSize be lowered to 32 MB for both profiles to match the documented minimum, or is there a reason CTS intentionally holds a stricter bar for full profile and a laxer one for embedded?
- If the latter, should the spec wording be revisited instead?
- Dominant language
- C++
- Stars
- 234
- Forks
- 240
- Avg merge
- 6d 12h
- Merged PRs (30d)
- 15
Getting set up
- No Dockerfile or Docker Compose file
- No pull request template
- Read the contributing 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.
More from KhronosGroup/OpenCL-CTS
-
cognizant-backlog
Difficulty 2/5 1-3 hours Newbie friendliness 61/100
KhronosGroup/OpenCL-CTS#2882 · 1 comment ·
Maintainers usually reply within 1 day
-
cl_khr_command_buffer: negative_enqueue_inconsistent_device expects CL_INCOMPATIBLE_COMMAND_QUEUE_KHR instead of CL_INVALID_DEVICEPossibly taken @kpet claimed this 11 days ago. Open
Difficulty 1/5 Under an hour Newbie friendliness 82/100
KhronosGroup/OpenCL-CTS#2872 · 1 comment · 1 assignee ·
Maintainers usually reply within 1 day
-
test_compiler: feature macros for cl_ext_float_atomics are not testedPossibly taken @shajder claimed this 10 days ago. Opencognizant-triage
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
KhronosGroup/OpenCL-CTS#2818 ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 70/100
KhronosGroup/OpenCL-CTS#2699 · 1 comment ·
Maintainers usually reply within 1 day
-
cleanup/refactoring
Difficulty 2/5 1-3 hours Newbie friendliness 62/100
KhronosGroup/OpenCL-CTS#2057 ·
Maintainers usually reply within 1 day
All issues in KhronosGroup/OpenCL-CTS
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Maintainers usually reply within 1 day
-
useless statementOpen
Difficulty 1/5 Under an hour Newbie friendliness 86/100
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
-
Difficulty 2/5 1-3 hours Newbie friendliness 65/100
grizzlypeak3d/DJV#962 ·
-
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
Maintainers usually reply within 1 day