Gen12LP device selection timeout

[REGULY__ISTVAN]

Hello,

I am trying to run SYCL code on the Gen12LP devices in the devcloud. However, queue creation takes a very long time, or even gets stuck. Even running "clinfo" will get stuck when querying the Gen12LP devices. This happens on several different nodes of devcloud, including s011-n001, s012-n001. On s011-n002 clinfo simply segfaults. On s011-n006 clinfo does not report any GPUs available

I do not have anything in my shell environment that should affect this.

To reproduce: run clinfo, or the simple-sycl-app.cpp from https://intel.github.io/llvm-docs/GetStartedGuide.html modified to select the GPU.

Thanks,

Istvan

[Gopika_Intel]

Hi Istvan,

Thank you for using the Intel Devcloud forum. Sorry for the inconvenience caused. We’ve connected with the internal team. We’ll get back to you.

Regards

Gopika

[Gopika_Intel · Accepted solution]

Hi,

The issue has been fixed from our end. Please check if your issue has been resolved and also give us a confirmation about the same.

Regards

Gopika

[REGULY__ISTVAN]

Resolved, thank you!

[Gopika_Intel]

Hi Istvan,

Thank you for the confirmation. We would no longer be monitoring this thread. For further queries and issues please raise a new thread, we’d be happy to help you.

Regards

Gopika