Module debug 

Debugging utilities for execution tracing and tensor inspection.

This module provides comprehensive debugging tools for TensorLogic execution:

• ExecutionTracer: Record execution flow through computation graphs
• TensorInspector: Examine intermediate tensor values and statistics
• BreakpointManager: Pause execution at specific nodes for inspection
• ExecutionRecorder: Record full execution history for replay and analysis

Example

use tensorlogic_infer::debug::{ExecutionTracer, TensorInspector, BreakpointManager};

// Set up tracing
let mut tracer = ExecutionTracer::new();
tracer.enable();

// Add breakpoints
let mut breakpoints = BreakpointManager::new();
breakpoints.add_node_breakpoint(5);

// Execute with debugging
// ... execution code ...

// Analyze trace
let trace = tracer.get_trace();
for entry in trace.entries() {
    println!("Node {}: {}ms", entry.node_id, entry.duration_ms());
}

Structs

BreakpointHit

Breakpoint hit information.

BreakpointManager

Manager for execution breakpoints.

ExecutionRecorder

Full execution recorder for replay and analysis.

ExecutionReport

Comprehensive execution report.

ExecutionTrace

Execution trace containing recorded operations.

ExecutionTracer

Execution tracer for recording operation flow.

OperationHandle

Handle for an in-progress operation recording.

TensorInspector

Tensor inspector for examining intermediate values.

TensorStats

Tensor statistics for inspection.

TraceEntry

Execution trace entry recording a single operation.

TraceSummary

Summary statistics for an execution trace.

Enums

Breakpoint

Breakpoint type for execution control.