- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hang in clFinish on gfx902 (Vega M GH)
Steps to reproduce:
1. Prepare a binary for gfx902 using the CL_CONTEXT_OFFLINE_DEVICES_AMD approach (offline compilation)
2. Find the AMD platform by name ('Advanced Micro Devices, Inc')
3. Create an OpenCL context with CL_DEVICE_TYPE_GPU
4. Load and build the program (clCreateProgramWithBinary, clBuildProgram)
5. Create 4 queues
6. Enqueue several kernel calls in one of the queues
7. Call clFinish on the queues
Expected outcome:
clFinish returns
Observed outcome:
clFinish hangs, call stack:
ntdll.dll!NtWaitForSingleObject() Unknown
KERNELBASE.dll!WaitForSingleObjectEx() Unknown
amdocl64.dll!00007ff806264e49() Unknown
amdocl64.dll!00007ff80625df88() Unknown
amdocl64.dll!00007ff80625fd5a() Unknown
amdocl64.dll!00007ff80624a51c() Unknown
Unfortunately I do not have direct access to the device to prepare a mimimalistic test case, but can provide an EXE for further investigation on request.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you for reporting it.
Please provide the executable and the offline binary that can be used to reproduce the issue.
Thanks.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you dipak.
Please check your inbox.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you. I'll check and reply via PM.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I've created a ticket against the issue. Once I get any update on this, I'll let you know.
Thanks.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Just for reference, another user ( samsam ) has reported a similar issue here: RX Vega M GH is detected incorrectly using AMD driver .
Thanks.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you Dipak. Another similar thread: https://community.amd.com/thread/251040
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Yes, the thread was posted by the same user in our support forum.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
timchist‌ feel free to use to reproduce my minimalistic project... sorry that headers in main.c are C files, but I have no time to improve... here it is at GitHub - sowson/gpucomp: Hello GPU Compute World!
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you, sowson. As I said earlier, unfortunately I have no direct access to the GPU. Perhaps dipak‌ may find it useful.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi dipak,
are there any updates on this case?
Thanks
