-
Notifications
You must be signed in to change notification settings - Fork 1.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Nginx does not run #1
Comments
Am I correct in assuming that this will potentially make nginx behave in unexpected ways and thus it is not something one should attempt for any somewhat serious use of nginx + gvisor? To quote the nginx docs at http://nginx.org/en/docs/ngx_core_module.html#master_process:
|
I'm not totally sure what the "master_process" directive does, but the docs do state that it should not be used in production. http://nginx.org/en/docs/faq/daemon_master_process_off.html We are still working on the |
Is there an example to repro this behaviour? I am not having an issue with nginx:latest. |
gVisor now supports FIOASYNC ioctl. I just verified that nginx runs without the "master_process: off" directive. |
Updates #1 PiperOrigin-RevId: 201760129
Updates google#1 PiperOrigin-RevId: 201760129 Change-Id: Ifd8ce9e0f93c6771083dc9bf8d35a2800c13481a
glibc's malloc also uses SYS_TIME. Permit it. #0 0x0000000000de6267 in time () #1 0x0000000000db19d8 in get_nprocs () #2 0x0000000000d8a31a in arena_get2.part () #3 0x0000000000d8ab4a in malloc () #4 0x0000000000d3c6b5 in __sanitizer::InternalAlloc(unsigned long, __sanitizer::SizeClassAllocatorLocalCache<__sanitizer::SizeClassAllocator32<0ul, 140737488355328ull, 0ul, __sanitizer::SizeClassMap<3ul, 4ul, 8ul, 17ul, 64ul, 14ul>, 20ul, __sanitizer::TwoLevelByteMap<32768ull, 4096ull, __sanitizer::NoOpMapUnmapCallback>, __sanitizer::NoOpMapUnmapCallback> >*, unsigned long) () #5 0x0000000000d4cd70 in __tsan_go_start () #6 0x00000000004617a3 in racecall () #7 0x00000000010f4ea0 in runtime.findfunctab () #8 0x000000000043f193 in runtime.racegostart () Signed-off-by: Dmitry Vyukov <dvyukov@google.com> [mpratt@google.com: updated comments and commit message] Signed-off-by: Michael Pratt <mpratt@google.com> Change-Id: Ibe2d0dc3035bf5052d5fb802cfaa37c5e0e7a09a PiperOrigin-RevId: 203042627
glibc's malloc also uses SYS_TIME. Permit it. #0 0x0000000000de6267 in time () google#1 0x0000000000db19d8 in get_nprocs () google#2 0x0000000000d8a31a in arena_get2.part () google#3 0x0000000000d8ab4a in malloc () google#4 0x0000000000d3c6b5 in __sanitizer::InternalAlloc(unsigned long, __sanitizer::SizeClassAllocatorLocalCache<__sanitizer::SizeClassAllocator32<0ul, 140737488355328ull, 0ul, __sanitizer::SizeClassMap<3ul, 4ul, 8ul, 17ul, 64ul, 14ul>, 20ul, __sanitizer::TwoLevelByteMap<32768ull, 4096ull, __sanitizer::NoOpMapUnmapCallback>, __sanitizer::NoOpMapUnmapCallback> >*, unsigned long) () google#5 0x0000000000d4cd70 in __tsan_go_start () google#6 0x00000000004617a3 in racecall () google#7 0x00000000010f4ea0 in runtime.findfunctab () google#8 0x000000000043f193 in runtime.racegostart () Signed-off-by: Dmitry Vyukov <dvyukov@google.com> [mpratt@google.com: updated comments and commit message] Signed-off-by: Michael Pratt <mpratt@google.com> Change-Id: Ibe2d0dc3035bf5052d5fb802cfaa37c5e0e7a09a PiperOrigin-RevId: 203042627
glibc's malloc also uses SYS_TIME. Permit it. #0 0x0000000000de6267 in time () #1 0x0000000000db19d8 in get_nprocs () #2 0x0000000000d8a31a in arena_get2.part () #3 0x0000000000d8ab4a in malloc () google#4 0x0000000000d3c6b5 in __sanitizer::InternalAlloc(unsigned long, __sanitizer::SizeClassAllocatorLocalCache<__sanitizer::SizeClassAllocator32<0ul, 140737488355328ull, 0ul, __sanitizer::SizeClassMap<3ul, 4ul, 8ul, 17ul, 64ul, 14ul>, 20ul, __sanitizer::TwoLevelByteMap<32768ull, 4096ull, __sanitizer::NoOpMapUnmapCallback>, __sanitizer::NoOpMapUnmapCallback> >*, unsigned long) () google#5 0x0000000000d4cd70 in __tsan_go_start () google#6 0x00000000004617a3 in racecall () google#7 0x00000000010f4ea0 in runtime.findfunctab () google#8 0x000000000043f193 in runtime.racegostart () Signed-off-by: Dmitry Vyukov <dvyukov@google.com> [mpratt@google.com: updated comments and commit message] Signed-off-by: Michael Pratt <mpratt@google.com> Change-Id: Ibe2d0dc3035bf5052d5fb802cfaa37c5e0e7a09a PiperOrigin-RevId: 203042627 Upstream-commit: 6144751
Below command under hostinet network will lead to panic: $ cat /proc/net/tcp It's caused by the wrong SizeOfTCPInfo. #0 runtime.panicindex() google#1 encoding/binary.littleEndian.Uint64 google#2 encoding/binary.(*littleEndian).Uint64 google#3 gvisor.dev/gvisor/pkg/binary.unmarshal google#4 gvisor.dev/gvisor/pkg/binary.unmarshal google#5 gvisor.dev/gvisor/pkg/binary.Unmarshal google#6 gvisor.dev/gvisor/pkg/sentry/socket/hostinet.(*socketOperations).State google#7 gvisor.dev/gvisor/pkg/sentry/fs/proc.(*netTCP).ReadSeqFileData Correct SizeOfTCPInfo from 104 to 192 to fix it. Fixes google#640 Signed-off-by: Jianfeng Tan <henry.tjf@antfin.com>
This change adds more information about what needs to be done to implement `/dev/fuse`
Distributed training isn't working with PyTorch on certain A100 nodes. Adds the missing ioctl `UVM_UNMAP_EXTERNAL` allowing for certain NCCL operations to succeed when using [`torch.distributed`](https://pytorch.org/docs/stable/distributed.html), fixing distributed training. ## Reproduction This affects numerous A100 40GB and 80GB instances in our fleet. This reproduction requires 4 A100 GPUs, either 40GB or 80GB. - **NVIDIA Driver Version**: 550.54.15 - **CUDA Version**: 12.4 - **NVIDIA device**: NVIDIA A100 80GB PCIe ### Steps 1. **Install gvisor** ```bash URL="https://storage.googleapis.com/gvisor/releases/master/latest/${ARCH}" wget -nc "${URL}/runsc" "${URL}/runsc.sha512" chmod +x runsc sudo cp runsc /usr/local/bin/runsc sudo /usr/local/bin/runsc install sudo systemctl reload docker ``` 2. **Add GPU enabling gvisor options** ```json { "runtimes": { "nvidia": { "path": "nvidia-container-runtime", "runtimeArgs": [] }, "runsc": { "path": "/usr/local/bin/runsc", "runtimeArgs": ["--nvproxy", "--nvproxy-docker", "-debug-log=/tmp/runsc/", "-debug", "-strace"] } } } ``` Reload configs with `sudo systemctl reload docker`. 3. **Run reproduction NCCL test** This test creates one main process and N peer processes. Each peer process sends a torch `Tensor` to the main process using NCCL. ```Dockerfile # Dockerfile FROM python:3.9.15-slim-bullseye RUN pip install torch numpy COPY <<EOF repro.py import argparse import datetime import os import torch import torch.distributed as dist import torch.multiprocessing as mp def setup(rank, world_size): os.environ["MASTER_ADDR"] = "localhost" os.environ["MASTER_PORT"] = "12355" dist.init_process_group("nccl", rank=rank, world_size=world_size, timeout=datetime.timedelta(seconds=600)) torch.cuda.set_device(rank) def cleanup(): dist.destroy_process_group() def send_tensor(rank, world_size): try: setup(rank, world_size) # rank receiving all tensors target_rank = world_size - 1 dist.barrier() tensor = torch.ones(5).cuda(rank) if rank < target_rank: print(f"[RANK {rank}] sending tensor: {tensor}") dist.send(tensor=tensor, dst=target_rank) elif rank == target_rank: for other_rank in range(target_rank): tensor = torch.zeros(5).cuda(target_rank) dist.recv(tensor=tensor, src=other_rank) print(f"[RANK {target_rank}] received tensor from rank={other_rank}: {tensor}") print("PASS: NCCL working.") except Exception as e: print(f"[RANK {rank}] error in send_tensor: {e}") raise finally: cleanup() def main(world_size: int = 2): mp.spawn(send_tensor, args=(world_size,), nprocs=world_size, join=True) if __name__ == "__main__": parser = argparse.ArgumentParser(description="Run torch-based NCCL tests") parser.add_argument("world_size", type=int, help="number of GPUs to run test on") args = parser.parse_args() if args.world_size < 2: raise RuntimeError(f"world_size needs to be larger than 1 {args.world_size}") main(args.world_size) EOF ENTRYPOINT ["python", "repro.py", "4"] ``` Build image with: ``` docker build -f Dockerfile . ``` Then run it with: ``` sudo docker run -it --shm-size=2.00gb --runtime=runsc --gpus='"device=GPU-742ea7fc-dd4f-612c-e860-499bf200a815,GPU-94a801d8-7713-acf6-337d-338b7cfdf19e,GPU-0d19cef2-10ce-e445-a0be-3d330e36c1fd,GPU-ac5046fb-020c-93e8-2784-f44aedbc5bbd"' 040a44863fb1 ``` #### Failure (truncated) ``` ... Exception raised from recvBytes at ../torch/csrc/distributed/c10d/Utils.hpp:672 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7edda14cf897 in /usr/local/lib/python3.11/site-packages/torch/lib/libc10.so) frame #1: <unknown function> + 0x5b3a23e (0x7edd8d73a23e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #2: c10d::TCPStore::doWait(c10::ArrayRef<std::string>, std::chrono::duration<long, std::ratio<1l, 1000l> >) + 0x2c7 (0x7edd8d734c87 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #3: c10d::TCPStore::doGet(std::string const&) + 0x32 (0x7edd8d734f82 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #4: c10d::TCPStore::get(std::string const&) + 0xa1 (0x7edd8d735fd1 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #5: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #6: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #7: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #8: c10d::ProcessGroupNCCL::broadcastUniqueNCCLID(ncclUniqueId*, bool, std::string const&, int) + 0xa9 (0x7edd54da9189 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #9: c10d::ProcessGroupNCCL::getNCCLComm(std::string const&, c10::Device&, c10d::OpType, int, bool) + 0xc50 (0x7edd54db0610 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #10: c10d::ProcessGroupNCCL::recv(std::vector<at::Tensor, std::allocator<at::Tensor> >&, int, int) + 0x5f8 (0x7edd54dcf978 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #11: <unknown function> + 0x5adc309 (0x7edd8d6dc309 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #12: <unknown function> + 0x5ae6f10 (0x7edd8d6e6f10 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #13: <unknown function> + 0x5ae6fa5 (0x7edd8d6e6fa5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #14: <unknown function> + 0x5124446 (0x7edd8cd24446 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #15: <unknown function> + 0x1acf4b8 (0x7edd896cf4b8 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #16: <unknown function> + 0x5aee004 (0x7edd8d6ee004 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #17: <unknown function> + 0x5af36b5 (0x7edd8d6f36b5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #18: <unknown function> + 0xd2fe8e (0x7edda032fe8e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) frame #19: <unknown function> + 0x47f074 (0x7edd9fa7f074 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) <omitting python frames> frame #35: <unknown function> + 0x29d90 (0x7edda2029d90 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #36: __libc_start_main + 0x80 (0x7edda2029e40 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #37: <unknown function> + 0x108e (0x55f950b0c08e in /usr/local/bin/python) . This may indicate a possible application crash on rank 0 or a network set up issue. ... ``` ### Fix gvisor debug logs show: ``` W0702 20:36:17.577055 445833 uvm.go:148] [ 22: 84] nvproxy: unknown uvm ioctl 66 = 0x42 ``` I've implemented that ioctl in this PR. This is the output after the fix. ``` [RANK 2] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:2') [RANK 0] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:0') [RANK 1] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:1') [RANK 3] received tensor from rank=0: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=1: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=2: tensor([1., 1., 1., 1., 1.], device='cuda:3') PASS: NCCL working. ``` FUTURE_COPYBARA_INTEGRATE_REVIEW=#10610 from luiscape:master ee88734 PiperOrigin-RevId: 649146570
Distributed training isn't working with PyTorch on certain A100 nodes. Adds the missing ioctl `UVM_UNMAP_EXTERNAL` allowing for certain NCCL operations to succeed when using [`torch.distributed`](https://pytorch.org/docs/stable/distributed.html), fixing distributed training. ## Reproduction This affects numerous A100 40GB and 80GB instances in our fleet. This reproduction requires 4 A100 GPUs, either 40GB or 80GB. - **NVIDIA Driver Version**: 550.54.15 - **CUDA Version**: 12.4 - **NVIDIA device**: NVIDIA A100 80GB PCIe ### Steps 1. **Install gvisor** ```bash URL="https://storage.googleapis.com/gvisor/releases/master/latest/${ARCH}" wget -nc "${URL}/runsc" "${URL}/runsc.sha512" chmod +x runsc sudo cp runsc /usr/local/bin/runsc sudo /usr/local/bin/runsc install sudo systemctl reload docker ``` 2. **Add GPU enabling gvisor options** ```json { "runtimes": { "nvidia": { "path": "nvidia-container-runtime", "runtimeArgs": [] }, "runsc": { "path": "/usr/local/bin/runsc", "runtimeArgs": ["--nvproxy", "--nvproxy-docker", "-debug-log=/tmp/runsc/", "-debug", "-strace"] } } } ``` Reload configs with `sudo systemctl reload docker`. 3. **Run reproduction NCCL test** This test creates one main process and N peer processes. Each peer process sends a torch `Tensor` to the main process using NCCL. ```Dockerfile # Dockerfile FROM python:3.9.15-slim-bullseye RUN pip install torch numpy COPY <<EOF repro.py import argparse import datetime import os import torch import torch.distributed as dist import torch.multiprocessing as mp def setup(rank, world_size): os.environ["MASTER_ADDR"] = "localhost" os.environ["MASTER_PORT"] = "12355" dist.init_process_group("nccl", rank=rank, world_size=world_size, timeout=datetime.timedelta(seconds=600)) torch.cuda.set_device(rank) def cleanup(): dist.destroy_process_group() def send_tensor(rank, world_size): try: setup(rank, world_size) # rank receiving all tensors target_rank = world_size - 1 dist.barrier() tensor = torch.ones(5).cuda(rank) if rank < target_rank: print(f"[RANK {rank}] sending tensor: {tensor}") dist.send(tensor=tensor, dst=target_rank) elif rank == target_rank: for other_rank in range(target_rank): tensor = torch.zeros(5).cuda(target_rank) dist.recv(tensor=tensor, src=other_rank) print(f"[RANK {target_rank}] received tensor from rank={other_rank}: {tensor}") print("PASS: NCCL working.") except Exception as e: print(f"[RANK {rank}] error in send_tensor: {e}") raise finally: cleanup() def main(world_size: int = 2): mp.spawn(send_tensor, args=(world_size,), nprocs=world_size, join=True) if __name__ == "__main__": parser = argparse.ArgumentParser(description="Run torch-based NCCL tests") parser.add_argument("world_size", type=int, help="number of GPUs to run test on") args = parser.parse_args() if args.world_size < 2: raise RuntimeError(f"world_size needs to be larger than 1 {args.world_size}") main(args.world_size) EOF ENTRYPOINT ["python", "repro.py", "4"] ``` Build image with: ``` docker build -f Dockerfile . ``` Then run it with: ``` sudo docker run -it --shm-size=2.00gb --runtime=runsc --gpus='"device=GPU-742ea7fc-dd4f-612c-e860-499bf200a815,GPU-94a801d8-7713-acf6-337d-338b7cfdf19e,GPU-0d19cef2-10ce-e445-a0be-3d330e36c1fd,GPU-ac5046fb-020c-93e8-2784-f44aedbc5bbd"' 040a44863fb1 ``` #### Failure (truncated) ``` ... Exception raised from recvBytes at ../torch/csrc/distributed/c10d/Utils.hpp:672 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7edda14cf897 in /usr/local/lib/python3.11/site-packages/torch/lib/libc10.so) frame #1: <unknown function> + 0x5b3a23e (0x7edd8d73a23e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #2: c10d::TCPStore::doWait(c10::ArrayRef<std::string>, std::chrono::duration<long, std::ratio<1l, 1000l> >) + 0x2c7 (0x7edd8d734c87 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #3: c10d::TCPStore::doGet(std::string const&) + 0x32 (0x7edd8d734f82 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #4: c10d::TCPStore::get(std::string const&) + 0xa1 (0x7edd8d735fd1 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #5: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #6: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #7: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #8: c10d::ProcessGroupNCCL::broadcastUniqueNCCLID(ncclUniqueId*, bool, std::string const&, int) + 0xa9 (0x7edd54da9189 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #9: c10d::ProcessGroupNCCL::getNCCLComm(std::string const&, c10::Device&, c10d::OpType, int, bool) + 0xc50 (0x7edd54db0610 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #10: c10d::ProcessGroupNCCL::recv(std::vector<at::Tensor, std::allocator<at::Tensor> >&, int, int) + 0x5f8 (0x7edd54dcf978 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #11: <unknown function> + 0x5adc309 (0x7edd8d6dc309 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #12: <unknown function> + 0x5ae6f10 (0x7edd8d6e6f10 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #13: <unknown function> + 0x5ae6fa5 (0x7edd8d6e6fa5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #14: <unknown function> + 0x5124446 (0x7edd8cd24446 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #15: <unknown function> + 0x1acf4b8 (0x7edd896cf4b8 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #16: <unknown function> + 0x5aee004 (0x7edd8d6ee004 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #17: <unknown function> + 0x5af36b5 (0x7edd8d6f36b5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #18: <unknown function> + 0xd2fe8e (0x7edda032fe8e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) frame #19: <unknown function> + 0x47f074 (0x7edd9fa7f074 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) <omitting python frames> frame #35: <unknown function> + 0x29d90 (0x7edda2029d90 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #36: __libc_start_main + 0x80 (0x7edda2029e40 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #37: <unknown function> + 0x108e (0x55f950b0c08e in /usr/local/bin/python) . This may indicate a possible application crash on rank 0 or a network set up issue. ... ``` ### Fix gvisor debug logs show: ``` W0702 20:36:17.577055 445833 uvm.go:148] [ 22: 84] nvproxy: unknown uvm ioctl 66 = 0x42 ``` I've implemented that ioctl in this PR. This is the output after the fix. ``` [RANK 2] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:2') [RANK 0] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:0') [RANK 1] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:1') [RANK 3] received tensor from rank=0: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=1: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=2: tensor([1., 1., 1., 1., 1.], device='cuda:3') PASS: NCCL working. ``` FUTURE_COPYBARA_INTEGRATE_REVIEW=#10610 from luiscape:master ee88734 PiperOrigin-RevId: 649146570
Distributed training isn't working with PyTorch on certain A100 nodes. Adds the missing ioctl `UVM_UNMAP_EXTERNAL` allowing for certain NCCL operations to succeed when using [`torch.distributed`](https://pytorch.org/docs/stable/distributed.html), fixing distributed training. ## Reproduction This affects numerous A100 40GB and 80GB instances in our fleet. This reproduction requires 4 A100 GPUs, either 40GB or 80GB. - **NVIDIA Driver Version**: 550.54.15 - **CUDA Version**: 12.4 - **NVIDIA device**: NVIDIA A100 80GB PCIe ### Steps 1. **Install gvisor** ```bash URL="https://storage.googleapis.com/gvisor/releases/master/latest/${ARCH}" wget -nc "${URL}/runsc" "${URL}/runsc.sha512" chmod +x runsc sudo cp runsc /usr/local/bin/runsc sudo /usr/local/bin/runsc install sudo systemctl reload docker ``` 2. **Add GPU enabling gvisor options** ```json { "runtimes": { "nvidia": { "path": "nvidia-container-runtime", "runtimeArgs": [] }, "runsc": { "path": "/usr/local/bin/runsc", "runtimeArgs": ["--nvproxy", "--nvproxy-docker", "-debug-log=/tmp/runsc/", "-debug", "-strace"] } } } ``` Reload configs with `sudo systemctl reload docker`. 3. **Run reproduction NCCL test** This test creates one main process and N peer processes. Each peer process sends a torch `Tensor` to the main process using NCCL. ```Dockerfile # Dockerfile FROM python:3.9.15-slim-bullseye RUN pip install torch numpy COPY <<EOF repro.py import argparse import datetime import os import torch import torch.distributed as dist import torch.multiprocessing as mp def setup(rank, world_size): os.environ["MASTER_ADDR"] = "localhost" os.environ["MASTER_PORT"] = "12355" dist.init_process_group("nccl", rank=rank, world_size=world_size, timeout=datetime.timedelta(seconds=600)) torch.cuda.set_device(rank) def cleanup(): dist.destroy_process_group() def send_tensor(rank, world_size): try: setup(rank, world_size) # rank receiving all tensors target_rank = world_size - 1 dist.barrier() tensor = torch.ones(5).cuda(rank) if rank < target_rank: print(f"[RANK {rank}] sending tensor: {tensor}") dist.send(tensor=tensor, dst=target_rank) elif rank == target_rank: for other_rank in range(target_rank): tensor = torch.zeros(5).cuda(target_rank) dist.recv(tensor=tensor, src=other_rank) print(f"[RANK {target_rank}] received tensor from rank={other_rank}: {tensor}") print("PASS: NCCL working.") except Exception as e: print(f"[RANK {rank}] error in send_tensor: {e}") raise finally: cleanup() def main(world_size: int = 2): mp.spawn(send_tensor, args=(world_size,), nprocs=world_size, join=True) if __name__ == "__main__": parser = argparse.ArgumentParser(description="Run torch-based NCCL tests") parser.add_argument("world_size", type=int, help="number of GPUs to run test on") args = parser.parse_args() if args.world_size < 2: raise RuntimeError(f"world_size needs to be larger than 1 {args.world_size}") main(args.world_size) EOF ENTRYPOINT ["python", "repro.py", "4"] ``` Build image with: ``` docker build -f Dockerfile . ``` Then run it with: ``` sudo docker run -it --shm-size=2.00gb --runtime=runsc --gpus='"device=GPU-742ea7fc-dd4f-612c-e860-499bf200a815,GPU-94a801d8-7713-acf6-337d-338b7cfdf19e,GPU-0d19cef2-10ce-e445-a0be-3d330e36c1fd,GPU-ac5046fb-020c-93e8-2784-f44aedbc5bbd"' 040a44863fb1 ``` #### Failure (truncated) ``` ... Exception raised from recvBytes at ../torch/csrc/distributed/c10d/Utils.hpp:672 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7edda14cf897 in /usr/local/lib/python3.11/site-packages/torch/lib/libc10.so) frame #1: <unknown function> + 0x5b3a23e (0x7edd8d73a23e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #2: c10d::TCPStore::doWait(c10::ArrayRef<std::string>, std::chrono::duration<long, std::ratio<1l, 1000l> >) + 0x2c7 (0x7edd8d734c87 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #3: c10d::TCPStore::doGet(std::string const&) + 0x32 (0x7edd8d734f82 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #4: c10d::TCPStore::get(std::string const&) + 0xa1 (0x7edd8d735fd1 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #5: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #6: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #7: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #8: c10d::ProcessGroupNCCL::broadcastUniqueNCCLID(ncclUniqueId*, bool, std::string const&, int) + 0xa9 (0x7edd54da9189 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #9: c10d::ProcessGroupNCCL::getNCCLComm(std::string const&, c10::Device&, c10d::OpType, int, bool) + 0xc50 (0x7edd54db0610 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #10: c10d::ProcessGroupNCCL::recv(std::vector<at::Tensor, std::allocator<at::Tensor> >&, int, int) + 0x5f8 (0x7edd54dcf978 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #11: <unknown function> + 0x5adc309 (0x7edd8d6dc309 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #12: <unknown function> + 0x5ae6f10 (0x7edd8d6e6f10 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #13: <unknown function> + 0x5ae6fa5 (0x7edd8d6e6fa5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #14: <unknown function> + 0x5124446 (0x7edd8cd24446 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #15: <unknown function> + 0x1acf4b8 (0x7edd896cf4b8 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #16: <unknown function> + 0x5aee004 (0x7edd8d6ee004 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #17: <unknown function> + 0x5af36b5 (0x7edd8d6f36b5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #18: <unknown function> + 0xd2fe8e (0x7edda032fe8e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) frame #19: <unknown function> + 0x47f074 (0x7edd9fa7f074 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) <omitting python frames> frame #35: <unknown function> + 0x29d90 (0x7edda2029d90 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #36: __libc_start_main + 0x80 (0x7edda2029e40 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #37: <unknown function> + 0x108e (0x55f950b0c08e in /usr/local/bin/python) . This may indicate a possible application crash on rank 0 or a network set up issue. ... ``` ### Fix gvisor debug logs show: ``` W0702 20:36:17.577055 445833 uvm.go:148] [ 22: 84] nvproxy: unknown uvm ioctl 66 = 0x42 ``` I've implemented that ioctl in this PR. This is the output after the fix. ``` [RANK 2] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:2') [RANK 0] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:0') [RANK 1] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:1') [RANK 3] received tensor from rank=0: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=1: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=2: tensor([1., 1., 1., 1., 1.], device='cuda:3') PASS: NCCL working. ``` FUTURE_COPYBARA_INTEGRATE_REVIEW=#10610 from luiscape:master ee88734 PiperOrigin-RevId: 649146570
Distributed training isn't working with PyTorch on certain A100 nodes. Adds the missing ioctl `UVM_UNMAP_EXTERNAL` allowing for certain NCCL operations to succeed when using [`torch.distributed`](https://pytorch.org/docs/stable/distributed.html), fixing distributed training. ## Reproduction This affects numerous A100 40GB and 80GB instances in our fleet. This reproduction requires 4 A100 GPUs, either 40GB or 80GB. - **NVIDIA Driver Version**: 550.54.15 - **CUDA Version**: 12.4 - **NVIDIA device**: NVIDIA A100 80GB PCIe ### Steps 1. **Install gvisor** ```bash URL="https://storage.googleapis.com/gvisor/releases/master/latest/${ARCH}" wget -nc "${URL}/runsc" "${URL}/runsc.sha512" chmod +x runsc sudo cp runsc /usr/local/bin/runsc sudo /usr/local/bin/runsc install sudo systemctl reload docker ``` 2. **Add GPU enabling gvisor options** ```json { "runtimes": { "nvidia": { "path": "nvidia-container-runtime", "runtimeArgs": [] }, "runsc": { "path": "/usr/local/bin/runsc", "runtimeArgs": ["--nvproxy", "--nvproxy-docker", "-debug-log=/tmp/runsc/", "-debug", "-strace"] } } } ``` Reload configs with `sudo systemctl reload docker`. 3. **Run reproduction NCCL test** This test creates one main process and N peer processes. Each peer process sends a torch `Tensor` to the main process using NCCL. ```Dockerfile # Dockerfile FROM python:3.9.15-slim-bullseye RUN pip install torch numpy COPY <<EOF repro.py import argparse import datetime import os import torch import torch.distributed as dist import torch.multiprocessing as mp def setup(rank, world_size): os.environ["MASTER_ADDR"] = "localhost" os.environ["MASTER_PORT"] = "12355" dist.init_process_group("nccl", rank=rank, world_size=world_size, timeout=datetime.timedelta(seconds=600)) torch.cuda.set_device(rank) def cleanup(): dist.destroy_process_group() def send_tensor(rank, world_size): try: setup(rank, world_size) # rank receiving all tensors target_rank = world_size - 1 dist.barrier() tensor = torch.ones(5).cuda(rank) if rank < target_rank: print(f"[RANK {rank}] sending tensor: {tensor}") dist.send(tensor=tensor, dst=target_rank) elif rank == target_rank: for other_rank in range(target_rank): tensor = torch.zeros(5).cuda(target_rank) dist.recv(tensor=tensor, src=other_rank) print(f"[RANK {target_rank}] received tensor from rank={other_rank}: {tensor}") print("PASS: NCCL working.") except Exception as e: print(f"[RANK {rank}] error in send_tensor: {e}") raise finally: cleanup() def main(world_size: int = 2): mp.spawn(send_tensor, args=(world_size,), nprocs=world_size, join=True) if __name__ == "__main__": parser = argparse.ArgumentParser(description="Run torch-based NCCL tests") parser.add_argument("world_size", type=int, help="number of GPUs to run test on") args = parser.parse_args() if args.world_size < 2: raise RuntimeError(f"world_size needs to be larger than 1 {args.world_size}") main(args.world_size) EOF ENTRYPOINT ["python", "repro.py", "4"] ``` Build image with: ``` docker build -f Dockerfile . ``` Then run it with: ``` sudo docker run -it --shm-size=2.00gb --runtime=runsc --gpus='"device=GPU-742ea7fc-dd4f-612c-e860-499bf200a815,GPU-94a801d8-7713-acf6-337d-338b7cfdf19e,GPU-0d19cef2-10ce-e445-a0be-3d330e36c1fd,GPU-ac5046fb-020c-93e8-2784-f44aedbc5bbd"' 040a44863fb1 ``` #### Failure (truncated) ``` ... Exception raised from recvBytes at ../torch/csrc/distributed/c10d/Utils.hpp:672 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7edda14cf897 in /usr/local/lib/python3.11/site-packages/torch/lib/libc10.so) frame #1: <unknown function> + 0x5b3a23e (0x7edd8d73a23e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #2: c10d::TCPStore::doWait(c10::ArrayRef<std::string>, std::chrono::duration<long, std::ratio<1l, 1000l> >) + 0x2c7 (0x7edd8d734c87 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #3: c10d::TCPStore::doGet(std::string const&) + 0x32 (0x7edd8d734f82 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #4: c10d::TCPStore::get(std::string const&) + 0xa1 (0x7edd8d735fd1 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #5: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #6: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #7: c10d::PrefixStore::get(std::string const&) + 0x31 (0x7edd8d6ea371 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #8: c10d::ProcessGroupNCCL::broadcastUniqueNCCLID(ncclUniqueId*, bool, std::string const&, int) + 0xa9 (0x7edd54da9189 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #9: c10d::ProcessGroupNCCL::getNCCLComm(std::string const&, c10::Device&, c10d::OpType, int, bool) + 0xc50 (0x7edd54db0610 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #10: c10d::ProcessGroupNCCL::recv(std::vector<at::Tensor, std::allocator<at::Tensor> >&, int, int) + 0x5f8 (0x7edd54dcf978 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #11: <unknown function> + 0x5adc309 (0x7edd8d6dc309 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #12: <unknown function> + 0x5ae6f10 (0x7edd8d6e6f10 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #13: <unknown function> + 0x5ae6fa5 (0x7edd8d6e6fa5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #14: <unknown function> + 0x5124446 (0x7edd8cd24446 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #15: <unknown function> + 0x1acf4b8 (0x7edd896cf4b8 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #16: <unknown function> + 0x5aee004 (0x7edd8d6ee004 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #17: <unknown function> + 0x5af36b5 (0x7edd8d6f36b5 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so) frame #18: <unknown function> + 0xd2fe8e (0x7edda032fe8e in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) frame #19: <unknown function> + 0x47f074 (0x7edd9fa7f074 in /usr/local/lib/python3.11/site-packages/torch/lib/libtorch_python.so) <omitting python frames> frame #35: <unknown function> + 0x29d90 (0x7edda2029d90 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #36: __libc_start_main + 0x80 (0x7edda2029e40 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #37: <unknown function> + 0x108e (0x55f950b0c08e in /usr/local/bin/python) . This may indicate a possible application crash on rank 0 or a network set up issue. ... ``` ### Fix gvisor debug logs show: ``` W0702 20:36:17.577055 445833 uvm.go:148] [ 22: 84] nvproxy: unknown uvm ioctl 66 = 0x42 ``` I've implemented that ioctl in this PR. This is the output after the fix. ``` [RANK 2] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:2') [RANK 0] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:0') [RANK 1] sending tensor: tensor([1., 1., 1., 1., 1.], device='cuda:1') [RANK 3] received tensor from rank=0: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=1: tensor([1., 1., 1., 1., 1.], device='cuda:3') [RANK 3] received tensor from rank=2: tensor([1., 1., 1., 1., 1.], device='cuda:3') PASS: NCCL working. ``` FUTURE_COPYBARA_INTEGRATE_REVIEW=#10610 from luiscape:master ee88734 PiperOrigin-RevId: 649146570
Depending on the configuration, you may see
nginx
fail with the error:Support for
FIOASYNC
is in progress, but it’s not available yet. For now, addthe line below to
/etc/nginx/nginx.conf
:The text was updated successfully, but these errors were encountered: