fix qlinear nan bug

[aarushjain29]

Motivation

ONNX Runtime's CPUExecutionProvider and MIGraphX disagreed on QuantizeLinear output when the input is NaN:

input: [nan]
scale: 0.5
onnxruntime out: [-128]
migraphx out: [ 127] ← MIGraphX

Technical Details

Changelog Category

Add a CHANGELOG.md entry for any option other than Not Applicable

• • Added: New functionality.
• • Changed: Changes to existing functionality.
• • Removed: Functionality or support that has been removed. (Compared to a previous release)
• • Optimized: Component performance that has been optimized or improved.
• • Resolved Issues: Known issues from a previous version that have been resolved.
• • Not Applicable: This PR is not to be included in the changelog.

[Copilot]

Pull request overview

This PR adjusts MIGraphX’s quantizelinear operator to handle NaN inputs deterministically during quantization, aligning behavior with common ONNX runtime expectations (notably ONNX Runtime CPU EP).

Changes:

• Add explicit NaN handling for integral quantization outputs: NaN inputs now saturate to the quantized type’s minimum value.
• Prevent NaN from flowing through the clamp logic (std::min/std::max) where it would otherwise silently select the max bound.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #4914      +/-   ##
===========================================
- Coverage    92.66%   92.62%   -0.04%     
===========================================
  Files          588      587       -1     
  Lines        30412    30416       +4     
===========================================
- Hits         28180    28171       -9     
- Misses        2232     2245      +13     

Files with missing lines	Coverage Δ
src/include/migraphx/op/quantizelinear.hpp	100.00% <100.00%> (+2.63%)	⬆️

... and 21 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[gh-app-migraphx-bot-pr-write]

Test	Batch	New Rate (22efc1)	Old Rate (283f84)	Diff	Status
torchvision-resnet50	64	1,490.56	3,154.40	-52.75%	🔴
torchvision-resnet50_fp16	64	3,380.59	6,643.79	-49.12%	🔴
torchvision-densenet121	32	1,442.53	2,697.51	-46.52%	🔴
torchvision-densenet121_fp16	32	2,567.72	4,525.49	-43.26%	🔴
torchvision-inceptionv3	32	901.87	1,795.56	-49.77%	🔴
torchvision-inceptionv3_fp16	32	2,365.74	2,809.78	-15.80%	🔴
cadene-inceptionv4	16	438.81	824.23	-46.76%	🔴
cadene-resnext64x4	16	450.42	783.73	-42.53%	🔴
slim-mobilenet	64	4,230.48	8,391.24	-49.58%	🔴
slim-nasnetalarge	64	nan	228.48	nan	❌
slim-resnet50v2	64	1,624.49	3,315.41	-51.00%	🔴
bert-mrpc-onnx	8	1,173.29	1,172.27	0.09%	✅
bert-mrpc-tf	1	190.90	489.81	-61.03%	🔴
pytorch-examples-wlang-gru	1	378.50	410.78	-7.86%	🔴
pytorch-examples-wlang-lstm	1	577.09	495.68	16.43%	🔆
torchvision-resnet50_1	1	748.78	786.81	-4.83%	✅
cadene-dpn92_1	1	452.49	443.27	2.08%	✅
cadene-resnext101_1	1	363.76	363.99	-0.06%	✅
onnx-taau-downsample	1	391.84	399.95	-2.03%	✅
dlrm-criteoterabyte	1	30.71	32.43	-5.31%	🔴
dlrm-criteoterabyte_fp16	1	50.49	51.89	-2.68%	✅
agentmodel	1	10,385.27	8,911.54	16.54%	🔆
unet_fp16	2	53.62	56.90	-5.76%	🔴
resnet50v1_fp16	1	1,008.95	945.91	6.66%	🔆
resnet50v1_int8	1	937.37	924.22	1.42%	✅
bert_base_cased_fp16	64	1,090.61	1,098.59	-0.73%	✅
bert_large_uncased_fp16	32	280.08	346.40	-19.15%	🔴
bert_large_fp16	1	204.48	204.56	-0.04%	✅
distilgpt2_fp16	16	2,086.86	2,096.23	-0.45%	✅
yolov5s	1	566.41	564.81	0.28%	✅
tinyllama	1	45.83	45.97	-0.31%	✅
vicuna-fastchat	1	43.94	44.04	-0.22%	✅
whisper-tiny-encoder	1	417.10	417.87	-0.19%	✅
whisper-tiny-decoder	1	138.32	413.65	-66.56%	🔴
llama2_7b	1	20.35	20.33	0.11%	✅
qwen1.5-7b	1	23.52	23.60	-0.34%	✅
phi3-3.8b	1	24.80	26.73	-7.20%	🔴
llama3-8b	1	21.62	21.73	-0.52%	✅
whisper-large-encoder	1	10.28	10.28	-0.08%	✅
whisper-large-decoder	1	109.37	105.60	3.57%	✅
mistral-7b	1	23.63	23.77	-0.60%	✅
FLUX.1-schnell	1	763.67	764.64	-0.13%	✅

Regressions detected 🔴

[gh-app-migraphx-bot-pr-write]

Test	Status	Result
bert-mrpc-onnx	✅	PASSED: MIGraphX meets tolerance
bert-mrpc-tf	❌	ERROR - check error output traceback Traceback (most recent call last): File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 377, in main() File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 313, in main import tensorflow as tf File "/usr/local/lib/python3.10/dist-packages/tensorflow/init.py", line 38, in from tensorflow.python.tools import module_util as _module_util File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/init.py", line 36, in from tensorflow.python import pywrap_tensorflow as _pywrap_tensorflow File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 26, in self_check.preload_check() File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/platform/self_check.py", line 63, in preload_check from tensorflow.python.platform import _pywrap_cpu_feature_guard ImportError: libamdhip64.so.6: cannot open shared object file: No such file or directory
pytorch-examples-wlang-gru	✅	PASSED: MIGraphX meets tolerance
pytorch-examples-wlang-lstm	✅	PASSED: MIGraphX meets tolerance
dlrm-criteoterabyte	✅	PASSED: MIGraphX meets tolerance
agentmodel	✅	PASSED: MIGraphX meets tolerance
unet	✅	PASSED: MIGraphX meets tolerance
resnet50v1	✅	PASSED: MIGraphX meets tolerance
bert_base_cased_fp16	✅	PASSED: MIGraphX meets tolerance
bert_large_uncased_fp16	🔴	FAILED: MIGraphX is not within tolerance - check verbose output
bert_large	✅	PASSED: MIGraphX meets tolerance
yolov5s	✅	PASSED: MIGraphX meets tolerance
tinyllama	✅	PASSED: MIGraphX meets tolerance
vicuna-fastchat	✅	PASSED: MIGraphX meets tolerance
whisper-tiny-encoder	✅	PASSED: MIGraphX meets tolerance
whisper-tiny-decoder	✅	PASSED: MIGraphX meets tolerance
distilgpt2_fp16	✅	PASSED: MIGraphX meets tolerance
llama2_7b	✅	PASSED: MIGraphX meets tolerance
qwen1.5-7b	✅	PASSED: MIGraphX meets tolerance
phi3-3.8b	✅	PASSED: MIGraphX meets tolerance
llama3-8b	✅	PASSED: MIGraphX meets tolerance
whisper-large-encoder	❌	ERROR - check error output traceback Traceback (most recent call last): File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 377, in main() File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 224, in main model = migraphx.parse_onnx(model_name, default_dim_value=batch) RuntimeError: /data/src/include/migraphx/op/convolution.hpp:102: normalize_compute_shape: CONVOLUTION: mismatched channel numbers
whisper-large-decoder	✅	PASSED: MIGraphX meets tolerance
mistral-7b	✅	PASSED: MIGraphX meets tolerance
FLUX.1-schnell	✅	PASSED: MIGraphX meets tolerance

[TedThemistokleous]

Fix tidy but otherwise lgtm

[kahmed10]

TODO: when testing is updated to check actual values on ref+gpu, add a test for this case