DistributedDataParallel with compile(..., mode="max-autotune") hangs in 2.5+

🐛 Describe the bug

I have a DCNN-like network that I'm training with DDP and 2 GPUs. This always got a speedup from compile(net, mode="max-autotune"). However, since PyTorch 2.5+ and still in current Nightly, the compile step hangs forever. In nvtop I can see memory usage go up, slight burst of GPU activity, then memory usage resets back to baseline. This repeats forever. Compilation works correct if I disable "max-autotune".

Patching to make def is_big_gpu(index) always return True - a bugfix which is needed to get both GPUs in the machine to tune, does not change the outcome or behavior, i.e. both the original version and the patched one will hang.

I need to patch PyTorch with #136332 to compile the network.

I tried env TORCH_COMPILE_DEBUG=1 but most of the output that this produced seems to be empty files.

Please let me know what kind of debug output would be needed to investigate this. This is a regression compared to PyTorch 2.3 and 2.4.

Versions

Collecting environment information... PyTorch version: 2.5.1+cu124 Is debug build: False CUDA used to build PyTorch: 12.4 ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.5 LTS (x86_64) GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 Clang version: 14.0.0-1ubuntu1.1 CMake version: version 3.22.1 Libc version: glibc-2.35

Python version: 3.10.12 (main, Sep 11 2024, 15:47:36) [GCC 11.4.0] (64-bit runtime) Python platform: Linux-6.5.0-27-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: Could not collect CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA GeForce RTX 4070 SUPER GPU 1: NVIDIA GeForce RTX 3080 Ti

Nvidia driver version: 560.35.03 cuDNN version: Probably one of the following: /usr/lib/x86_64-linux-gnu/libcudnn.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.9.7 HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True

CPU: 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): 24 On-line CPU(s) list: 0-23 Vendor ID: AuthenticAMD Model name: AMD Ryzen 9 3900X 12-Core Processor CPU family: 23 Model: 113 Thread(s) per core: 2 Core(s) per socket: 12 Socket(s): 1 Stepping: 0 Frequency boost: enabled CPU max MHz: 4672.0698 CPU min MHz: 2200.0000 BogoMIPS: 7585.24 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 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 ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 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: 384 KiB (12 instances) L1i cache: 384 KiB (12 instances) L2 cache: 6 MiB (12 instances) L3 cache: 64 MiB (4 instances) NUMA node(s): 1 NUMA node0 CPU(s): 0-23 Vulnerability Gather data sampling: 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 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 Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected

Versions of relevant libraries: [pip3] numpy==2.1.3 [pip3] nvidia-cublas-cu12==12.4.5.8 [pip3] nvidia-cuda-cupti-cu12==12.4.127 [pip3] nvidia-cuda-nvrtc-cu12==12.4.127 [pip3] nvidia-cuda-runtime-cu12==12.4.127 [pip3] nvidia-cudnn-cu12==9.1.0.70 [pip3] nvidia-cufft-cu12==11.2.1.3 [pip3] nvidia-curand-cu12==10.3.5.147 [pip3] nvidia-cusolver-cu12==11.6.1.9 [pip3] nvidia-cusparse-cu12==12.3.1.170 [pip3] nvidia-nccl-cu12==2.21.5 [pip3] nvidia-nvjitlink-cu12==12.4.127 [pip3] nvidia-nvtx-cu12==12.4.127 [pip3] pytorch_optimizer==3.2.0 [pip3] torch==2.5.1 [pip3] torchaudio==2.5.1 [pip3] torchvision==0.20.1 [pip3] triton==3.1.0 [conda] Could not collect

cc @ezyang @gchanan @zou3519 @kadeng @msaroufim @H-Huang @awgu @kwen2501 @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @c-p-i-o @chauhang @penguinwu

pytorch / pytorch

DistributedDataParallel with compile(..., mode="max-autotune") hangs in 2.5+ #140395

🐛 Describe the bug

Versions