Skip to content

Qualcomm AI Engine Direct - Adding QNN backend support for div.Tensor_mode core ATen op#19785

Merged
psiddh merged 1 commit into
pytorch:mainfrom
CodeLinaro:div_tensor
Jun 15, 2026
Merged

Qualcomm AI Engine Direct - Adding QNN backend support for div.Tensor_mode core ATen op#19785
psiddh merged 1 commit into
pytorch:mainfrom
CodeLinaro:div_tensor

Conversation

@qti-horodnic

@qti-horodnic qti-horodnic commented May 26, 2026

Copy link
Copy Markdown
Contributor

Summary

Added support for the core ATen op div.Tensor_mode using a decomposition pass and the div, trunc, floor ops, based on the selected mode:

    div(x, y, rounding_mode=None)    -> div(x, y)
    div(x, y, rounding_mode="trunc") -> trunc(div(x, y))
    div(x, y, rounding_mode="floor") -> floor(div(x, y))

Test plan

python backends/qualcomm/tests/test_qnn_delegate.py -k TestQNNQuantizedOperator.test_qnn_backend_div_mode --model SM8750 --host aisw-vm15-labsd --device 545ee4aa --build_folder build-android
python backends/qualcomm/tests/test_qnn_delegate.py -k TestQNNFloatingPointOperator.test_qnn_backend_div_mode --model SM8750 --host aisw-vm15-labsd --device 545ee4aa --build_folder build-android

python backends/qualcomm/tests/test_qnn_delegate.py -k TestQNNQuantizedOperator.test_qnn_backend_div_scalar_mode --model SM8750 --host aisw-vm15-labsd --device 545ee4aa --build_folder build-android
python backends/qualcomm/tests/test_qnn_delegate.py -k TestQNNFloatingPointOperator.test_qnn_backend_div_scalar_mode --model SM8750 --host aisw-vm15-labsd --device 545ee4aa --build_folder build-android

cc @cccclai @cbilgin @abhinaykukkadapu

@pytorch-bot

pytorch-bot Bot commented May 26, 2026

Copy link
Copy Markdown

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19785

Note: Links to docs will display an error until the docs builds have been completed.

❌ 3 New Failures

As of commit 694b064 with merge base 0378fc4 (image):

NEW FAILURES - The following jobs have failed:

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@meta-cla meta-cla Bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label May 26, 2026
@linux-foundation-easycla

linux-foundation-easycla Bot commented May 26, 2026

Copy link
Copy Markdown

CLA Signed
The committers listed above are authorized under a signed CLA.

  • ✅ login: qti-horodnic / name: Arik Horodniceanu (37652bf)

@qti-horodnic

Copy link
Copy Markdown
Contributor Author

@pytorchbot label "release notes: qualcomm"

@pytorch-bot pytorch-bot Bot added the release notes: qualcomm Changes to the Qualcomm backend delegate label May 26, 2026
Comment thread backends/qualcomm/_passes/qnn_pass_manager.py
@digantdesai digantdesai added the module: qnn Issues related to Qualcomm's QNN delegate and code under backends/qualcomm/ label May 27, 2026
@psiddh

psiddh commented Jun 4, 2026

Copy link
Copy Markdown
Contributor

Need to resolve merge conflict

@qti-horodnic

Copy link
Copy Markdown
Contributor Author

Need to resolve merge conflict

@psiddh
Thanks, resolved the conflicts now

Comment thread backends/qualcomm/_passes/qnn_pass_manager.py Outdated
@qti-horodnic

Copy link
Copy Markdown
Contributor Author

@psiddh

Hi, can we merge this pr or did you want any other changes?

@qti-horodnic

Copy link
Copy Markdown
Contributor Author

@psiddh

Hi, can we merge this pr or did you want any other changes?

Hi @psiddh, can we merge this PR? Not sure why the linter is stuck, I ran it locally and it looks fine

@psiddh psiddh merged commit 9c4471c into pytorch:main Jun 15, 2026
178 of 181 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: qnn Issues related to Qualcomm's QNN delegate and code under backends/qualcomm/ release notes: qualcomm Changes to the Qualcomm backend delegate

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants