Replies: 1 comment 1 reply
|
A quick way is to somehow modify your OpenCL code to prevent the compiler from using the flatscratch registers. A better way is to add support for the register. The related code is there https://github.com/sarchlab/mgpusim/blob/v4/amd/emu/wavefront.go#L104. Also, I noticed that the EnableSgptPrivateSegmentWaveByteOffset is used, which is not supported by MGPUSim. I am not sure how that field is used. Is it possible to search for related documentation and let me know how that works? |
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Hello Professor Sun, I've encountered some issues related to registers recently. It seems to be due to the fact that I am loading eight 16-byte plaintext blocks for encryption within one thread. How should I modify the system to support these register configurations?

All reactions