OpenCL* for CPU
Ask questions and share information on Intel® SDK for OpenCL™ Applications and OpenCL™ implementations for Intel® CPU
Announcements
This forum covers OpenCL* for CPU only. OpenCL* for GPU questions can be asked in the GPU Compute Software forum. Intel® FPGA SDK for OpenCL™ questions can be ask in the FPGA Intel® High Level Design forum.
1663 Discussions

clGetMemObjectInfo bug in SDK 2012

jbnote
Beginner
132 Views
Hi,

I'm encountering the following problem when calling clGetMemObjectInfo.

According to spec from http://www.khronos.org/registry/cl/sdk/1.0/docs/man/xhtml/clGetMemObjectInfo.html I should be able to call the function like this:
size_t size;
cl_int err = clGetMemObjectInfo(memobj, param_name, 0, NULL, &size);
assert(err == CL_SUCCESS);
and get back the size of the return parameter in size [granted this does not make sense for this function where all objects returned have a size which is compile-time known, but this is for uniformity's sake].
Intel's implementation always returns CL_INVALID_VALUE, contrary to the spec (param_value is NULL). Further calls (i'm doing this first call as a safeguard) with effective size and destination pointers actually set do not exhibit the problem, so i'm pretty sure it's not an issue of param_name or memobj being incorrect. Other OpenCL implementations (AMD, Nvidia) are working as expected.
Kind regards, JB
0 Kudos
1 Reply
Raghupathi_M_Intel
132 Views
Hi JB,

Thanks for posting the issue. Indeed there seems to be a difference in Intel's implementation compared to (at least) Nvidia. It looks like a bug and Iwill get back to you when a fix will be available.

Thanks,
Raghu
Reply