advisor failed with: GTPin ERROR: KernelCreate failed - Device does not exists

[nahso]

Hello, I am using advisor to profile the performance. The program finished perfectly when I ran it directly, but when I used the following command:

advisor --collect=roofline --profile-gpu --project-dir=./advisor/slm --search-dir src:r=../src/thundersvm -- ./run-demo.sh

advisor failed with the following message:

2022-09-26 08:44:49,838 INFO [default] before parallel for
2022-09-26 08:44:49,838 INFO [default] Device : Intel(R) Graphics [0x020a]
WARNING: SYCL_CACHE_DISABLE_PERSISTENT environment variable is deprecated and has no effect. By default, persistent device code caching is disabled
. Use SYCL_CACHE_PERSISTENT=1/0 to enable/disable.
GTPin ERROR: KernelCreate failed - Device does not exists
at: KernelCreate : 222
advisor: Warning: Cannot stop collection of GPU events
advisor: Collection stopped.
advisor: Warning: The application returned a non-zero exit value.
advisor: Opening result 19 % Loading 'userapicollector-2297991-61e4c904.trace'
advisor: Opening result 29 % Resolving information for `dash'
advisor: Warning: Cannot locate debugging information for file `/usr/bin/dash'.
advisor: Opening result 39 % Processing profile metrics and debug information
advisor: Error: Error 0x40000024 (No data) -- No data is collected. Possible reasons:
- Workload is too small. No samples are collected.
- The application environment is not specified correctly.
- The executable file has been stripped so cannot be profiled with algorithm analysis types.
See the Troubleshooting help topic for more details.
Also consider checking the collection log for additional information.
advisor: Opening result 100 % done
advisor: Preparing frequently used data 0 % done
advisor: Preparing frequently used data 100 % done
advisor: Warning: No GPU kernels detected

Program Elapsed Time: 9.63s

The beginning of the log was printed by:

void get_working_set_ins_sycl(const kernel_type *val, const int *col_ind, const int *row_ptr, const int *data_row_idx,
kernel_type *data_rows,int m, int n){
LOG(INFO) << "before parallel for";
LOG(INFO) << "Device : " << q.get_device().get_info<info::device::name>();
q.parallel_for(sycl::range<1>(m), [=](sycl::id<1> i) {
int row = data_row_idx[i];
for (int j = row_ptr[row]; j < row_ptr[row + 1]; ++j) {
int col = col_ind[j];
data_rows[i*n + col] = val[j]; //row majorx
// data_rows[i + col*m] = val[j]; //col major
}
}).wait();
LOG(INFO) << "after parallel for";
}

Can you please tell me what's the reason?

[AlekhyaV_Intel]

Hi,

Thank you for posting in Intel Communities. We would like to have the following information to debug your issue further:

1) Exact OS details

2) Processor details

3) Intel Advisor version

4) Sample Reproducer (sample code, steps to reproduce, commands you've used, etc.)

5) Self Checker logs: For getting self checker logs, please run the below commands.

Linux: 

python3 /opt/intel/oneapi/advisor/latest/bin64/adv_self_check.py (need to run as administarator or super user)

Windows: 

Go to this path- C:\Program Files (x86)\Intel\oneAPI\advisor\latest\sys_check

and run python advisor_sys_check (run as administrator)

Regards,

Alekhya

[AlekhyaV_Intel]

Hi,

We have not heard back from you. Could you please give us an update regarding this issue?

Regards,

Alekhya

[AlekhyaV_Intel]

Hi,

We have not heard back from you. So, we will close this thread now. If you need any further assistance, please post a new question as this thread will no longer be monitored by Intel.

Regards,

Alekhya