aoti-debug
Diagnose and fix AOTInductor compilation and runtime errors in PyTorch, focusing on device and shape consistency.
Install
mkdir -p .claude/skills/aoti-debug && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/4024" && unzip -o skill.zip -d .claude/skills/aoti-debug && rm skill.zipInstalls to .claude/skills/aoti-debug
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Debug AOTInductor (AOTI) errors and crashes. Use when encountering AOTI segfaults, device mismatch errors, constant loading failures, or runtime errors from aot_compile, aot_load, aoti_compile_and_package, or aoti_load_package.Key capabilities
- →Validate device and shape matching for AOTI
- →Identify problematic kernels using intermediate value debuggers
- →Force deterministic CUDA behavior for debugging
- →Check input properties with runtime flags
How it works
It routes errors to specific debugging guides and provides environment variables to validate input guards, force synchronous execution, and print intermediate kernel values.
Inputs & outputs
When to use aoti-debug
- →Fix AOTI segfaults
- →Resolve device mismatch errors
- →Debug compile failures
About this skill
AOTI Debugging Guide
This skill helps diagnose and fix common AOTInductor issues.
Error Pattern Routing
Check the error message and route to the appropriate sub-guide:
Triton Index Out of Bounds
If the error matches this pattern:
Assertion `index out of bounds: 0 <= tmpN < ksM` failed
→ Follow the guide in triton-index-out-of-bounds.md
All Other Errors
Continue with the sections below.
First Step: Always Check Device and Shape Matching
For ANY AOTI error (segfault, exception, crash, wrong output), ALWAYS check these first:
- Compile device == Load device: The model must be loaded on the same device type it was compiled on
- Input devices match: Runtime inputs must be on the same device as the compiled model
- Input shapes match: Runtime input shapes must match the shapes used during compilation (or satisfy dynamic shape constraints)
# During compilation - note the device and shapes
model = MyModel().eval() # What device? CPU or .cuda()?
inp = torch.randn(2, 10) # What device? What shape?
compiled_so = torch._inductor.aot_compile(model, (inp,))
# During loading - device type MUST match compilation
loaded = torch._export.aot_load(compiled_so, "???") # Must match model/input device above
# During inference - device and shapes MUST match
out = loaded(inp.to("???")) # Must match compile device, shape must match
If any of these don't match, you will get errors ranging from segfaults to exceptions to wrong outputs.
Key Constraint: Device Type Matching
AOTI requires compile and load to use the same device type.
- If you compile on CUDA, you must load on CUDA (device index can differ)
- If you compile on CPU, you must load on CPU
- Cross-device loading (e.g., compile on GPU, load on CPU) is NOT supported
Common Error Patterns
1. Device Mismatch Segfault
Symptom: Segfault, exception, or crash during aot_load() or model execution.
Example error messages:
The specified pointer resides on host memory and is not registered with any CUDA device- Crash during constant loading in AOTInductorModelBase
Expected out tensor to have device cuda:0, but got cpu instead
Cause: Compile and load device types don't match (see "First Step" above).
Solution: Ensure compile and load use the same device type. If compiled on CPU, load on CPU. If compiled on CUDA, load on CUDA.
2. Input Device Mismatch at Runtime
Symptom: RuntimeError during model execution.
Cause: Input device doesn't match compile device (see "First Step" above).
Better Debugging: Run with AOTI_RUNTIME_CHECK_INPUTS=1 for clearer errors. This flag validates all input properties including device type, dtype, sizes, and strides:
AOTI_RUNTIME_CHECK_INPUTS=1 python your_script.py
This produces actionable error messages like:
Error: input_handles[0]: unmatched device type, expected: 0(cpu), but got: 1(cuda)
Debugging CUDA Illegal Memory Access (IMA) Errors
If you encounter CUDA illegal memory access errors, follow this systematic approach:
Step 1: Sanity Checks
Before diving deep, try these debugging flags:
AOTI_RUNTIME_CHECK_INPUTS=1
TORCHINDUCTOR_NAN_ASSERTS=1
These flags take effect at compilation time (at codegen time):
AOTI_RUNTIME_CHECK_INPUTS=1checks if inputs satisfy the same guards used during compilationTORCHINDUCTOR_NAN_ASSERTS=1adds codegen before and after each kernel to check for NaN
Step 2: Pinpoint the CUDA IMA
CUDA IMA errors can be non-deterministic. Use these flags to trigger the error deterministically:
PYTORCH_NO_CUDA_MEMORY_CACHING=1
CUDA_LAUNCH_BLOCKING=1
These flags take effect at runtime:
PYTORCH_NO_CUDA_MEMORY_CACHING=1disables PyTorch's Caching Allocator, which allocates bigger buffers than needed immediately. This is usually why CUDA IMA errors are non-deterministic.CUDA_LAUNCH_BLOCKING=1forces kernels to launch one at a time. Without this, you get "CUDA kernel errors might be asynchronously reported" warnings since kernels launch asynchronously.
Step 3: Identify Problematic Kernels with Intermediate Value Debugger
Use the AOTI Intermediate Value Debugger to pinpoint the problematic kernel:
AOT_INDUCTOR_DEBUG_INTERMEDIATE_VALUE_PRINTER=3
This prints kernels one by one at runtime. Together with previous flags, this shows which kernel was launched right before the error.
To inspect inputs to a specific kernel:
AOT_INDUCTOR_FILTERED_KERNELS_TO_PRINT="triton_poi_fused_add_ge_logical_and_logical_or_lt_231,_add_position_embeddings_kernel_5" AOT_INDUCTOR_DEBUG_INTERMEDIATE_VALUE_PRINTER=2
If inputs to the kernel are unexpected, inspect the kernel that produces the bad input.
Additional Debugging Tools
Logging and Tracing
- tlparse / TORCH_TRACE: Provides complete output codes and records guards used
- TORCH_LOGS: Use
TORCH_LOGS="+inductor,output_code"to see more PT2 internal logs - TORCH_SHOW_CPP_STACKTRACES: Set to
1to see more stack traces
Common Sources of Issues
- Dynamic shapes: Historically a source of many IMAs. Pay special attention when debugging dynamic shape scenarios.
- Custom ops: Especially when implemented in C++ with dynamic shapes. The meta function may need to be Symint'ified.
API Notes
Deprecated API
torch._export.aot_compile() # Deprecated
torch._export.aot_load() # Deprecated
Current API
torch._inductor.aoti_compile_and_package()
torch._inductor.aoti_load_package()
The new API stores device metadata in the package, so aoti_load_package() automatically uses the correct device type. You can only change the device index (e.g., cuda:0 vs cuda:1), not the device type.
Environment Variables Summary
| Variable | When | Purpose |
|---|---|---|
AOTI_RUNTIME_CHECK_INPUTS=1 | Compile time | Validate inputs match compilation guards |
TORCHINDUCTOR_NAN_ASSERTS=1 | Compile time | Check for NaN before/after kernels |
PYTORCH_NO_CUDA_MEMORY_CACHING=1 | Runtime | Make IMA errors deterministic |
CUDA_LAUNCH_BLOCKING=1 | Runtime | Force synchronous kernel launches |
AOT_INDUCTOR_DEBUG_INTERMEDIATE_VALUE_PRINTER=3 | Compile time | Print kernels at runtime |
TORCH_LOGS="+inductor,output_code" | Runtime | See PT2 internal logs |
TORCH_SHOW_CPP_STACKTRACES=1 | Runtime | Show C++ stack traces |
When not to use it
- →When the error is unrelated to AOTInductor compilation or loading
- →When cross-device loading is required
Prerequisites
Limitations
- →Cross-device loading is not supported
- →CUDA illegal memory access errors can be non-deterministic
How it compares
It provides a systematic routing approach for AOTI-specific crashes rather than generic PyTorch debugging.
Compared to similar skills
aoti-debug side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| aoti-debug (this skill) | 1 | 5mo | Review | Advanced |
| langsmith-observability | 4 | 7mo | Review | Intermediate |
| mflux-debugging | 3 | 2mo | No flags | Advanced |
| debug-distributed | 1 | 6mo | Review | Advanced |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by pytorch
View all by pytorch →You might also like
langsmith-observability
davila7
LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating model outputs against datasets, monitoring production systems, or building systematic testing pipelines for AI applications.
mflux-debugging
filipstrand
Debug MLX ports by comparing against a PyTorch/diffusers reference via exported tensors/images (export-then-compare).
debug-distributed
inclusionAI
Guide for debugging distributed training issues in AReaL. Use when user encounters hangs, wrong results, OOM, or communication errors.
benchmark-kernel
flashinfer-ai
Guide for benchmarking FlashInfer kernels with CUPTI timing
debug-model
vladmandic
Debug a broken SD.Next or Diffusers model integration. Use when a newly added or ported model fails to load, misdetects, crashes during prompt encoding or sampling, or produces incorrect outputs.
tokenizer-checker
openvinotoolkit
Validate a HuggingFace tokenizer with OpenVINO Tokenizers and OpenVINO GenAI. Use when: checking if a tokenizer converts and works correctly, verifying tokenizer/detokenizer accuracy, testing normalization steps, checking GenAI Tokenizer compatibility.