[Bug]: CPU build failed to run with OneDNN linear for fp16 model #27524
Description
Your current environment
The output of python collect_env.py
==============================
System Info
==============================
OS : Ubuntu 22.04.4 LTS (x86_64)
GCC version : (Ubuntu 12.3.0-1ubuntu1~22.04.2) 12.3.0
Clang version : 14.0.0-1ubuntu1.1
CMake version : version 4.1.2
Libc version : glibc-2.35
==============================
PyTorch Info
==============================
PyTorch version : 2.8.0+cpu
Is debug build : False
CUDA used to build PyTorch : Could not collect
ROCM used to build PyTorch : N/A
==============================
Python Environment
==============================
Python version : 3.12.8 (main, Jan 14 2025, 22:49:14) [Clang 19.1.6 ] (64-bit runtime)
Python platform : Linux-6.8.0-79-generic-x86_64-with-glibc2.35
==============================
CUDA / GPU Info
==============================
Is CUDA available : False
CUDA runtime version : Could not collect
CUDA_MODULE_LOADING set to : N/A
GPU models and configuration : GPU 0: Quadro P600
Nvidia driver version : 550.144.03
cuDNN version : Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.9.1.1
/usr/lib/x86_64-linux-gnu/libcudnn_adv.so.9.1.1
/usr/lib/x86_64-linux-gnu/libcudnn_cnn.so.9.1.1
/usr/lib/x86_64-linux-gnu/libcudnn_engines_precompiled.so.9.1.1
/usr/lib/x86_64-linux-gnu/libcudnn_engines_runtime_compiled.so.9.1.1
/usr/lib/x86_64-linux-gnu/libcudnn_graph.so.9.1.1
/usr/lib/x86_64-linux-gnu/libcudnn_heuristic.so.9.1.1
/usr/lib/x86_64-linux-gnu/libcudnn_ops.so.9.1.1
HIP runtime version : N/A
MIOpen runtime version : N/A
Is XNNPACK available : True
==============================
CPU Info
==============================
架构： x86_64
CPU 运行模式： 32-bit, 64-bit
Address sizes: 46 bits physical, 48 bits virtual
字节序： Little Endian
CPU: 24
在线 CPU 列表： 0-23
厂商 ID： GenuineIntel
型号名称： Intel(R) Xeon(R) Silver 4116 CPU @ 2.10GHz
CPU 系列： 6
型号： 85
每个核的线程数： 1
每个座的核数： 12
座： 2
步进： 4
CPU 最大 MHz： 3000.0000
CPU 最小 MHz： 800.0000
BogoMIPS： 4200.00
标记： fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cdp_l3 pti intel_ppin ssbd mba ibrs ibpb stibp tpr_shadow flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm cqm mpx rdt_a avx512f avx512dq rdseed adx smap clflushopt clwb intel_pt avx512cd avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local dtherm ida arat pln pts vnmi pku ospke md_clear flush_l1d arch_capabilities
虚拟化： VT-x
L1d 缓存： 768 KiB (24 instances)
L1i 缓存： 768 KiB (24 instances)
L2 缓存： 24 MiB (24 instances)
L3 缓存： 33 MiB (2 instances)
NUMA 节点： 2
NUMA 节点0 CPU： 0-11
NUMA 节点1 CPU： 12-23
Vulnerability Gather data sampling: Mitigation; Microcode
Vulnerability Itlb multihit: KVM: Mitigation: VMX disabled
Vulnerability L1tf: Mitigation; PTE Inversion; VMX conditional cache flushes, SMT disabled
Vulnerability Mds: Mitigation; Clear CPU buffers; SMT disabled
Vulnerability Meltdown: Mitigation; PTI
Vulnerability Mmio stale data: Mitigation; Clear CPU buffers; SMT disabled
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Mitigation; IBRS
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Mitigation; Clear CPU buffers; SMT disabled
==============================
Versions of relevant libraries
==============================
[pip3] intel_extension_for_pytorch==2.8.0
[pip3] numpy==2.2.6
[pip3] pyzmq==27.1.0
[pip3] torch==2.8.0+cpu
[pip3] torchaudio==2.8.0+cpu
[pip3] torchvision==0.23.0+cpu
[pip3] transformers==4.57.1
[conda] numpy 2.3.4 pypi_0 pypi
[conda] nvidia-cublas-cu12 12.4.5.8 pypi_0 pypi
[conda] nvidia-cuda-cupti-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-cuda-nvrtc-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-cuda-runtime-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-cudnn-cu12 9.1.0.70 pypi_0 pypi
[conda] nvidia-cufft-cu12 11.2.1.3 pypi_0 pypi
[conda] nvidia-curand-cu12 10.3.5.147 pypi_0 pypi
[conda] nvidia-cusolver-cu12 11.6.1.9 pypi_0 pypi
[conda] nvidia-cusparse-cu12 12.3.1.170 pypi_0 pypi
[conda] nvidia-cusparselt-cu12 0.6.2 pypi_0 pypi
[conda] nvidia-ml-py 12.535.161 pypi_0 pypi
[conda] nvidia-nccl-cu12 2.21.5 pypi_0 pypi
[conda] nvidia-nvjitlink-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-nvtx-cu12 12.4.127 pypi_0 pypi
[conda] torch 2.9.0+cpu pypi_0 pypi
[conda] torchvision 0.24.0+cpu pypi_0 pypi
[conda] transformers 5.0.0.dev0 pypi_0 pypi
[conda] triton 3.2.0 pypi_0 pypi
==============================
vLLM Info
==============================
ROCM Version : Could not collect
vLLM Version : 0.1.dev10740+ga99564ac5 (git sha: a99564ac5)
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled
GPU Topology:
GPU0 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X 12-23 1 N/A
Legend:
X = Self
SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)
NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node
PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)
PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)
PIX = Connection traversing at most a single PCIe bridge
NV# = Connection traversing a bonded set of # NVLinks
==============================
Environment Variables
==============================
MAX_JOBS=24
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
🐛 Describe the bug
To reproduce, just run:
python examples/offline_inference/basic/basic.py
Error logs
[W1026 13:00:06.076972184 OperatorEntry.cpp:218] Warning: Warning only once for all operators, other operators may also be overridden.
Overriding a previously registered kernel for the same operator and the same dispatch key
operator: aten::_addmm_activation(Tensor self, Tensor mat1, Tensor mat2, *, Scalar beta=1, Scalar alpha=1, bool use_gelu=False) -> Tensor
registered at /pytorch/build/aten/src/ATen/RegisterSchema.cpp:6
dispatch key: AutocastCPU
previous kernel: registered at /pytorch/aten/src/ATen/autocast_mode.cpp:327
new kernel: registered at /opt/workspace/ipex-cpu-dev/csrc/cpu/autocast/autocast_mode.cpp:112 (function operator())
INFO 10-26 13:00:10 [importing.py:68] Triton not installed or not compatible; certain GPU-related functions will not be available.
INFO 10-26 13:00:12 [utils.py:253] non-default args: {'num_redundant_experts': None, 'eplb_window_size': None, 'eplb_step_interval': None, 'eplb_log_balancedness': None, 'enable_lora': None, 'model': '/data/LLM-model/opt-125m/'}
INFO 10-26 13:00:12 [model.py:667] Resolved architecture: OPTForCausalLM
INFO 10-26 13:00:12 [model.py:1751] Using max model len 2048
WARNING 10-26 13:00:12 [logger.py:133] Environment variable VLLM_CPU_KVCACHE_SPACE (GiB) for CPU backend is not set, using 4 by default.
INFO 10-26 13:00:12 [scheduler.py:211] Chunked prefill is enabled with max_num_batched_tokens=4096.
[W1026 13:00:15.405517015 OperatorEntry.cpp:218] Warning: Warning only once for all operators, other operators may also be overridden.
Overriding a previously registered kernel for the same operator and the same dispatch key
operator: aten::_addmm_activation(Tensor self, Tensor mat1, Tensor mat2, *, Scalar beta=1, Scalar alpha=1, bool use_gelu=False) -> Tensor
registered at /pytorch/build/aten/src/ATen/RegisterSchema.cpp:6
dispatch key: AutocastCPU
previous kernel: registered at /pytorch/aten/src/ATen/autocast_mode.cpp:327
new kernel: registered at /opt/workspace/ipex-cpu-dev/csrc/cpu/autocast/autocast_mode.cpp:112 (function operator())
INFO 10-26 13:00:19 [importing.py:68] Triton not installed or not compatible; certain GPU-related functions will not be available.
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [core.py:93] Initializing a V1 LLM engine (v0.1.dev10740+ga99564ac5) with config: model='/data/LLM-model/opt-125m/', speculative_config=None, tokenizer='/data/LLM-model/opt-125m/', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.float16, max_seq_len=2048, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=True, quantization=None, enforce_eager=False, kv_cache_dtype=auto, device_config=cpu, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None), seed=0, served_model_name=/data/LLM-model/opt-125m/, enable_prefix_caching=True, chunked_prefill_enabled=True, pooler_config=None, compilation_config={'level': None, 'mode': 2, 'debug_dump_path': None, 'cache_dir': '', 'backend': 'inductor', 'custom_ops': ['none'], 'splitting_ops': None, 'use_inductor': None, 'compile_sizes': None, 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'dce': True, 'size_asserts': False, 'nan_asserts': False, 'epilogue_fusion': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'use_cudagraph': True, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': [], 'cudagraph_copy_inputs': False, 'full_cuda_graph': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {}, 'max_cudagraph_capture_size': None, 'local_cache_dir': None}
(EngineCore_DP0 pid=599414) WARNING 10-26 13:00:21 [_logger.py:72] Pin memory is not supported on CPU.
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:164] auto thread-binding list (id, physical core): [(0, 0), (1, 1), (2, 2), (3, 3), (4, 4), (5, 5), (6, 6), (7, 7), (8, 8), (9, 9), (10, 10), (11, 11)]
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP threads binding of Process 599414:
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599414, core 0
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599643, core 1
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599644, core 2
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599645, core 3
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599646, core 4
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599647, core 5
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599648, core 6
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599649, core 7
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599650, core 8
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599651, core 9
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599652, core 10
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70] OMP tid: 599653, core 11
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_worker.py:70]
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [parallel_state.py:1325] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, TP rank 0, EP rank 0
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:21 [cpu_model_runner.py:67] Starting to load model /data/LLM-model/opt-125m/...
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:22 [cpu.py:146] Using Torch SDPA backend.
Loading pt checkpoint shards: 0% Completed | 0/1 [00:00<?, ?it/s]
Loading pt checkpoint shards: 100% Completed | 1/1 [00:00<00:00, 4.19it/s]
Loading pt checkpoint shards: 100% Completed | 1/1 [00:00<00:00, 4.18it/s]
(EngineCore_DP0 pid=599414)
(EngineCore_DP0 pid=599414) INFO 10-26 13:00:22 [default_loader.py:314] Loading weights took 0.25 seconds
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] EngineCore failed to start.
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] Traceback (most recent call last):
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/v1/engine/core.py", line 770, in run_engine_core
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] engine_core = EngineCoreProc(*args, **kwargs)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/v1/engine/core.py", line 538, in __init__
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] super().__init__(
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/v1/engine/core.py", line 102, in __init__
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] self.model_executor = executor_class(vllm_config)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] ^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/v1/executor/abstract.py", line 98, in __init__
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] self._init_executor()
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/v1/executor/uniproc_executor.py", line 47, in _init_executor
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] self.driver_worker.load_model()
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/v1/worker/gpu_worker.py", line 233, in load_model
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] self.model_runner.load_model(eep_scale_up=eep_scale_up)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/v1/worker/cpu_model_runner.py", line 68, in load_model
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] self.model = get_model(vllm_config=self.vllm_config)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/model_executor/model_loader/__init__.py", line 130, in get_model
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] return loader.load_model(vllm_config=vllm_config, model_config=model_config)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/model_executor/model_loader/base_loader.py", line 56, in load_model
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] process_weights_after_loading(model, model_config, target_device)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/model_executor/model_loader/utils.py", line 107, in process_weights_after_loading
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] quant_method.process_weights_after_loading(module)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/model_executor/layers/vocab_parallel_embedding.py", line 61, in process_weights_after_loading
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] dispatch_cpu_unquantized_gemm(layer, remove_weight=False)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/model_executor/layers/utils.py", line 188, in dispatch_cpu_unquantized_gemm
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] handler = ops.create_onednn_mm(origin_weight.t(), 32)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/vllm/_custom_ops.py", line 2437, in create_onednn_mm
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] handler.handler = torch.ops._C.create_onednn_mm_handler(
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] File "/data/develop-projects/github-repos/vllm-cpu/.venv/lib/python3.12/site-packages/torch/_ops.py", line 1243, in __call__
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] return self._op(*args, **kwargs)
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] ^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=599414) ERROR 10-26 13:00:22 [core.py:779] RuntimeError: could not create a primitive descriptor for the matmul primitive. Run workload with environment variable ONEDNN_VERBOSE=all to get additional diagnostic information.
However, models with bf16/fp32 dtype can work with OneDNN linear, and pytest -s -v tests/kernels/test_onednn.py can pass with fp16 on same build as well.
cc @bigPYJ1151 Any idea?
Before submitting a new issue...
- Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.