Error messages from QNN are turned into verbose level messages

Describe the issue

error1002.onnx.zip

To reproduce

Run the attached model with the QNN EP. With normal log handling, you get just a single un-helpful error message:

Failed to finalize QNN graph. Error code: 1002 at location qnn_model.cc:167 FinalizeGraphs

However, if you turn on verbose logging, you get to see the real errors:

tcm_migration.cc:2088:ERROR:Operator named q::*InputSlicePad (0x1654900000002) not sufficiently tiled to fit in TCM. Requires 12441600 bytes

graph_prepare.cc:2808:ERROR:Graph prepare TCM Migration action failed

graph_prepare.cc:2868:ERROR:Graph prepare failed during optimization with err: 17, Fatal Optimize

These errors should not be labeled verbose. They need to be in the error category.

Urgency

No response

Platform

Windows

OS Version

Android 14

ONNX Runtime Installation

Built from Source

ONNX Runtime Version or Commit ID

1.21

ONNX Runtime API

Python

Architecture

ARM64

Execution Provider

Other / Unknown

Execution Provider Library Version

QNN