Collecting environment information...
==============================
        System Info
==============================
OS                           : Ubuntu 24.04.4 LTS (x86_64)
GCC version                  : (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0
Clang version                : Could not collect
CMake version                : Could not collect
Libc version                 : glibc-2.39

==============================
       PyTorch Info
==============================
PyTorch version              : 2.11.0+cu130
Is debug build               : False
CUDA used to build PyTorch   : 13.0
ROCM used to build PyTorch   : N/A
XPU used to build PyTorch    : N/A

==============================
      Python Environment
==============================
Python version               : 3.12.13 (main, Apr 14 2026, 14:29:00) [Clang 22.1.3 ] (64-bit runtime)
Python platform              : Linux-6.17.0-19-generic-x86_64-with-glibc2.39
    
==============================
       CUDA / GPU Info
==============================
Is CUDA available            : True
CUDA runtime version         : Could not collect
CUDA_MODULE_LOADING set to   : 
GPU models and configuration : 
GPU 0: NVIDIA A100-SXM4-80GB
GPU 1: NVIDIA A100-SXM4-80GB
GPU 2: NVIDIA A100-SXM4-80GB
GPU 3: NVIDIA A100-SXM4-80GB
GPU 4: NVIDIA A100-SXM4-80GB
GPU 5: NVIDIA A100-SXM4-80GB
GPU 6: NVIDIA A100-SXM4-80GB
GPU 7: NVIDIA A100-SXM4-80GB

Nvidia driver version        : 595.58.03
cuDNN version                : Could not collect
HIP runtime version          : N/A
MIOpen runtime version       : N/A
Is XNNPACK available         : True

==============================
          CPU Info
==============================
Architecture:                            x86_64
CPU op-mode(s):                          32-bit, 64-bit
Address sizes:                           43 bits physical, 48 bits virtual
Byte Order:                              Little Endian
CPU(s):                                  256
On-line CPU(s) list:                     0-255
Vendor ID:                               AuthenticAMD
Model name:                              AMD EPYC 7742 64-Core Processor
CPU family:                              23
Model:                                   49
Thread(s) per core:                      2
Core(s) per socket:                      64
Socket(s):                               2
Stepping:                                0
Frequency boost:                         enabled
CPU(s) scaling MHz:                      45%
CPU max MHz:                             3416.0681
CPU min MHz:                             1500.0000
BogoMIPS:                                4491.54
Flags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip rdpid overflow_recov succor smca sev sev_es
Virtualization:                          AMD-V
L1d cache:                               4 MiB (128 instances)
L1i cache:                               4 MiB (128 instances)
L2 cache:                                64 MiB (128 instances)
L3 cache:                                512 MiB (32 instances)
NUMA node(s):                            8
NUMA node0 CPU(s):                       0-15,128-143
NUMA node1 CPU(s):                       16-31,144-159
NUMA node2 CPU(s):                       32-47,160-175
NUMA node3 CPU(s):                       48-63,176-191
NUMA node4 CPU(s):                       64-79,192-207
NUMA node5 CPU(s):                       80-95,208-223
NUMA node6 CPU(s):                       96-111,224-239
NUMA node7 CPU(s):                       112-127,240-255
Vulnerability Gather data sampling:      Not affected
Vulnerability Ghostwrite:                Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Old microcode:             Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Mitigation; untrained return thunk; SMT enabled with STIBP protection
Vulnerability Spec rstack overflow:      Mitigation; Safe RET
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; Retpolines; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Mitigation; IBPB before exit to userspace

==============================
Versions of relevant libraries
==============================
[pip3] flashinfer-python==0.6.7
[pip3] numpy==2.2.6
[pip3] nvidia-cublas==13.1.0.3
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cupti==13.0.85
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvrtc==13.0.88
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime==13.0.96
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-cu13==9.19.0.56
[pip3] nvidia-cudnn-frontend==1.18.0
[pip3] nvidia-cufft==12.0.0.61
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-cufile==1.15.1.6
[pip3] nvidia-cufile-cu12==1.13.1.3
[pip3] nvidia-curand==10.4.0.35
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver==12.0.4.66
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse==12.6.3.3
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cusparselt-cu13==0.8.0
[pip3] nvidia-cutlass-dsl==4.4.2
[pip3] nvidia-cutlass-dsl-libs-base==4.4.2
[pip3] nvidia-ml-py==13.595.45
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nccl-cu13==2.28.9
[pip3] nvidia-nvjitlink==13.0.88
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvshmem-cu12==3.4.5
[pip3] nvidia-nvshmem-cu13==3.4.5
[pip3] nvidia-nvtx==13.0.85
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] pyzmq==27.1.0
[pip3] torch==2.11.0
[pip3] torch_c_dlpack_ext==0.1.5
[pip3] torchaudio==2.11.0
[pip3] torchvision==0.26.0
[pip3] transformers==4.57.6
[pip3] triton==3.6.0
[conda] Could not collect

==============================
         vLLM Info
==============================
ROCM Version                 : Could not collect
vLLM Version                 : 0.19.1rc1.dev269+gf7e62e3d6 (git sha: f7e62e3d6)
vLLM Build Flags:
  CUDA Archs: Not Set; ROCm: Disabled; XPU: Disabled
GPU Topology:
        GPU0    GPU1    GPU2    GPU3    GPU4    GPU5    GPU6    GPU7    NIC0    NIC1    NIC2    NIC3    CPU Affinity    NUMA Affinity   GPU NUMA ID
GPU0     X      NV12    NV12    NV12    NV12    NV12    NV12    NV12    SYS     SYS     SYS     SYS     48-63,176-191   3               N/A
GPU1    NV12     X      NV12    NV12    NV12    NV12    NV12    NV12    SYS     SYS     SYS     SYS     48-63,176-191   3               N/A
GPU2    NV12    NV12     X      NV12    NV12    NV12    NV12    NV12    SYS     SYS     SYS     SYS     16-31,144-159   1               N/A
GPU3    NV12    NV12    NV12     X      NV12    NV12    NV12    NV12    SYS     SYS     SYS     SYS     16-31,144-159   1               N/A
GPU4    NV12    NV12    NV12    NV12     X      NV12    NV12    NV12    SYS     SYS     SYS     SYS     112-127,240-255 7               N/A
GPU5    NV12    NV12    NV12    NV12    NV12     X      NV12    NV12    SYS     SYS     SYS     SYS     112-127,240-255 7               N/A
GPU6    NV12    NV12    NV12    NV12    NV12    NV12     X      NV12    SYS     SYS     SYS     SYS     80-95,208-223   5               N/A
GPU7    NV12    NV12    NV12    NV12    NV12    NV12    NV12     X      SYS     SYS     SYS     SYS     80-95,208-223   5               N/A
NIC0    SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS      X      PIX     SYS     SYS
NIC1    SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS     PIX      X      SYS     SYS
NIC2    SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS      X      PIX
NIC3    SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS     SYS     PIX      X 

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

NIC Legend:

  NIC0: mlx5_4
  NIC1: mlx5_5
  NIC2: mlx5_10
  NIC3: mlx5_11

==============================
     Environment Variables
==============================
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_local

---

sorted_ids = torch.empty((max_loras * max_num_tokens_padded,), dtype=torch.int32, device=topk_ids.device)
expert_ids = torch.empty((max_loras * max_num_m_blocks,), dtype=torch.int32, device=topk_ids.device)
num_tokens_post_pad = torch.empty((max_loras,), dtype=torch.int32, device=topk_ids.device)

---

File "vllm/compilation/cuda_graph.py", line 355, in __call__
    entry.cudagraph.replay()
File "torch/cuda/graphs.py", line 143, in replay
    super().replay()
torch.AcceleratorError: CUDA error: an illegal memory access was encountered

---

#!/usr/bin/env python3
"""
Offline reproducer for the concurrent base + LoRA MoE segfault.

Uses vLLM's LLM API directly — no server needed. Sends interleaved
base + LoRA requests in a single batch to trigger the
moe_lora_align_block_size stale-buffer crash during CUDA graph replay.

Usage:
    python repro_offline.py
    python repro_offline.py --model nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16
    python repro_offline.py --model nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 --adapter ace-nguyen/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16-dummy-moe-lora --tp 2
    python repro_offline.py --cudagraph-mode full
    python repro_offline.py --cudagraph-mode none   # eager mode, no graphs
"""

import argparse

from vllm import LLM, SamplingParams
from vllm.lora.request import LoRARequest

# Valid CUDAGraphMode values (from pydantic validation):
#   0 = NONE, 1 = PIECEWISE, 2 = FULL, (2,0) = FULL, (2,1) = FULL+PIECEWISE
CUDAGRAPH_MODES = {
    "full": 2,                # full only, no piecewise
    "piecewise": 1,           # piecewise only, no full
    "both": (2, 1),           # full + piecewise (vLLM default)
    "none": None,             # eager mode, no graphs at all
}


def main():
    p = argparse.ArgumentParser(description="Offline base+LoRA crash reproducer")
    p.add_argument(
        "--model",
        default="nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16",
        help="Base model path",
    )
    p.add_argument(
        "--adapter",
        default="ace-nguyen/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16-dummy-moe-lora",
        help="LoRA adapter path",
    )
    p.add_argument("--adapter-name", default="test-lora", help="LoRA adapter name")
    p.add_argument("--tp", type=int, default=1, help="Tensor parallel size")
    p.add_argument(
        "--cudagraph-mode",
        choices=list(CUDAGRAPH_MODES.keys()),
        default=None,
        help="CUDA graph mode: full, piecewise, full_and_piecewise, none (default: vLLM default)",
    )
    p.add_argument("-n", type=int, default=10, help="Number of request pairs (base+lora)")
    args = p.parse_args()

    print(f"Model:          {args.model}")
    print(f"Adapter:        {args.adapter}")
    print(f"TP:             {args.tp}")
    print(f"CUDAGraph mode: {args.cudagraph_mode or 'default'}")
    print(f"Pairs:          {args.n}")
    print()

    llm_kwargs = dict(
        model=args.model,
        tensor_parallel_size=args.tp,
        enable_lora=True,
        trust_remote_code=True,
    )

    if args.cudagraph_mode is not None:
        mode = CUDAGRAPH_MODES[args.cudagraph_mode]
        if mode is None:
            llm_kwargs["enforce_eager"] = True
        else:
            llm_kwargs["compilation_config"] = {"cudagraph_mode": mode}

    llm = LLM(**llm_kwargs)

    sampling = SamplingParams(
        max_tokens=32,
        min_tokens=32,
        ignore_eos=True,
    )

    lora_req = LoRARequest(args.adapter_name, 1, args.adapter)

    # Build interleaved batch: base, lora, base, lora, ...
    prompts = []
    lora_requests = []
    for i in range(args.n):
        # Base model request
        prompts.append(f"Base request {i}: Count from 1 to 50:")
        lora_requests.append(None)
        # LoRA request
        prompts.append(f"LoRA request {i}: Count from 1 to 50:")
        lora_requests.append(lora_req)

    print(f"Generating {len(prompts)} requests (interleaved base+LoRA)...")

    # This triggers the crash if the bug is present — the first generate()
    # replays CUDA graphs captured during warmup, and stale buffer addresses
    # in moe_lora_align_block_size cause a segfault.
    outputs = llm.generate(
        prompts,
        sampling_params=sampling,
        lora_request=lora_requests,
    )

    print(f"\nCompleted {len(outputs)} requests successfully.")
    for i, out in enumerate(outputs[:4]):
        tag = "BASE" if lora_requests[i] is None else "LORA"
        text = out.outputs[0].text[:80].replace("\n", " ")
        print(f"  [{tag}] {text}...")

    print("\nPASSED")

if __name__ == "__main__":
    main()