sycl-toolchain-hip maxworkgroup return negative number?
- Dominant language
- LLVM
- Stars
- 1.5k
- Forks
- 854
- Avg merge
- 3d 17h
- Merged PRs (30d)
- 137
Description
**Describe the bug**
sycl/plugin/hip/pi_hip.cpp in version sycl-2022-12
how does the function `get_max_work_group_size` to get the number?
```
size_t *reqdThreadsPerBlock = kernel->reqdThreadsPerBlock_;
maxWorkGroupSize = command_queue->device_->get_max_work_group_size();
```
**To Reproduce**
I add the printf in source file at sycl/plugin/hip/pi_hip.cpp
The output is
```
---> piKernelSetArg(
: 0x1f8dbf0
: 4
: 8
: 0x1a5eee0
) ---> pi_result : PI_SUCCESS
[out]void * : 0x1a5eee0
---> piEnqueueKernelLaunch(
: 0x1f72640
: 0x1f8dbf0
: 2
: 0x2012ff0
: 0x2012fc0
: 0x2012fd8
: 1
pi_event * : 0x195ddb0[ 0x195dbb0 ... ]
pi_event * : 0x19ca3c8[ 0 ... ]
reqdThreadsPerBlock is [0,0,0]
maxWorkGroupSize is : -1694498771 , maxThreadsPerBlock is : 140727211951808
maxWorkGroupSize is 140727211951808 , threadsPerBlockdim is [16,16,1]
Segmentation fault (core dumped)
```
1. Include code snippet as short as possible
2. Specify the command which should be used to compile the program
3. Specify the comment which should be used to launch the program
4. Indicate what is wrong and what was expected
**Environment (please complete the following information):**
- Ubuntu 20.04
- Target device and vendor: AMD MI100
- DPC++ version: clang-16 sycl-2022-12
- Dependencies version: rocm
**Additional context**
Add any other context about the problem here.
Contributor guide
Assessment
This issue has not been assessed yet.