[WebGPU] Enable Cast to int64 by default

[fanchenkong1]

Description

Support casting to int64 from float32 via IEEE-754 bit decomposition.

• Introduce a new float_to_int64 helper that emits the truncated-toward-zero value in full int64 range.
• to_ type now always allows int64, regardless of enable_int64; casting from int64 stays gated by enable_int64.
• Adds cast_op_test.cc coverage for the newly introduced conversions.

Motivation and Context

While running the mask-generation vision encoder (Xenova/sam-vit-base) on the WebGPU EP via Transformers.js, float32-to-int64 cast nodes fall back to the CPU provider under the default session configuration, because casting to int64 was previously gated behind enable_int64 flag, introducing host memcpy and synchronization overhead.

Making cast-to-int64 correct across the full int64 range lets it run on the
WebGPU EP by default, keeping these nodes on-device and eliminating the stalls.

Performance Impact

Measured on the vision_encoder.onnx of Xenova/sam-vit-base (mask-generation, SAM ViT-base vision encoder) on the
WebGPU EP.

Platform	Latency reduction	Speedup
Intel Wildcat Lake	−22.8%	1.30×
Intel Panther Lake	−17.1%	1.21×

This change yields a 1.2–1.3× speedup on the SAM ViT-base vision encoder under default configuration.

[daijh]

@qjia7 @guschmue PTAL.

[qjia7]

nit: Please use output.SetByOffset instead of indirect accessing y.

[Copilot]

Pull request overview

This PR updates the WebGPU EP’s Cast kernel to allow casting to int64 by default (even when enable_int64 is false), and implements float32/float16 → int64 conversion in WGSL via IEEE-754 bit decomposition to avoid CPU fallback and associated device/host sync overhead.

Changes:

• Add an IEEE-754 bit-decomposition path for float → int64 in the WebGPU Cast shader, including lane-safe stores for int64 outputs.
• Adjust WebGPU Cast kernel type constraints so T2 (output) always allows int64, while T1 (input) still gates int64 on enable_int64.
• Add CPU-side Cast tests covering large float→int64 values and several int32/uint32/bool→int64 and size%4 regression cases.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
onnxruntime/core/providers/webgpu/tensor/cast.h	Extends CastProgram parameters and adds output_size uniform for int64 tail handling.
onnxruntime/core/providers/webgpu/tensor/cast.cc	Implements float→int64 WGSL helper and relaxes output type constraints for int64; updates shader codegen paths.
onnxruntime/test/providers/cpu/tensor/cast_op_test.cc	Adds test coverage for float32/float16/int32/uint32/bool to int64 conversions and non-multiple-of-4 sizes.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.