Sohaib Nadeem

Hi @gcongiu, I have not been able to reproduce this on ROCm 6.2 using 2 MI210s and the reproducer code you provided. Can you try upgrading to ROCm 6.2 and...

Hi @mxz297, thanks for your patience. I was able to reproduce your issue on latest ROCm using the reproducer code provided. You are correct that the `hsa_api_` table of functions...

@mxz297 I'm going to close this ticket due to inactivity. If the above fix does not work, feel free to re-open the ticket and we can look into it further....

@erayinanc Thanks for confirming, I will close this issue now.

Hi @drtpotter, I have used the mat_mult_bugreport.cpp.txt you provided on a Radeon Pro VII (gfx906) but was unable to reproduce the issue. Can you upgrade to ROCm 6.2 and check...

Is the issue still present on the gfx1035 card as well?

Got it, thanks for your help in investigating this issue. While it is possible this may be a compiler bug specific to gfx906, support for gfx906 is now [deprecated](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/system-requirements.html#supported-gpus). I...

Hi @MrBurmark, sorry for the late follow-up. @feizheng10 has put in a change (https://github.com/ROCm/rocprofiler-compute/pull/555) to make the default normalization to be per kernel. Let me know if you have any...

Closing this issue as it is resolved. @MrBurmark Feel free to re-open the issue if you have any follow-up questions/concerns.

Hi @NeoChen1024, as @cgmb mentioned, there isn't any plan to officially support gfx1103/Radeon 780M. Let me know if you have any other questions, otherwise I can close out this issue.