Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Bug Report] dprint_tensix_dest_reg Bug #17481

Open
hschoi4448 opened this issue Feb 3, 2025 · 1 comment
Open

[Bug Report] dprint_tensix_dest_reg Bug #17481

hschoi4448 opened this issue Feb 3, 2025 · 1 comment
Assignees
Labels
bug Something isn't working moreh moreh contribution P2

Comments

@hschoi4448
Copy link
Contributor

hschoi4448 commented Feb 3, 2025

Describe the bug
During recent debugging, I tried using dprint_tensix_dest_reg to check the value of the destination register before packing and discovered a bug.

My code performs a copy operation from an fp32 CB to another fp32 CB, with dprint_tensix_dest_reg inserted in between.

The values in the input and output CBs are identical, but the value printed by dprint_tensix_dest_reg differs from both the input and output.

After some debugging, I found that when constructing the float32 destination register value, it uses two 16-bit values. However, using the HEX() function to dprint the output revealed that the second 16-bit value is 0.

You can found test related kernel code here:

To Reproduce
Steps to reproduce the behavior:

  1. Go to branch: https://github.com/tenstorrent/tt-metal/tree/hyungsuk/bug_dprint_tensix_dest_reg
  2. Run cmd: pytest ./tests/ttnn/unit_tests/operations/test_moreh_softmax.py
  3. See DPRINT result

Image

Expected behavior
dprint_tensix_dest_reg should be able to correctly print fp32 values.

Screenshots

Please complete the following environment information:

  • OS: Ubuntu 22.04.4 LTS
  • Version of software (eg. commit dac9673)
  • Arch: wormhole_b0

Additional context
Add any other context about the problem here.

@hschoi4448 hschoi4448 added the bug Something isn't working label Feb 3, 2025
hschoi4448 added a commit that referenced this issue Feb 3, 2025
@hschoi4448 hschoi4448 added the moreh moreh contribution label Feb 3, 2025
@zzigler-tt zzigler-tt added P1 P2 and removed P1 labels Feb 3, 2025
@zzigler-tt
Copy link

Assigning to @ayerofieiev-tt to triage, designated as P2 for now. Please update as needed. TY

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working moreh moreh contribution P2
Projects
None yet
Development

No branches or pull requests

3 participants