Symmetric memory API and symmetric kernels * Redesign from the ground up, enabling major latency and bandwidth improvements. * Add new API calls to register user-allocated memory among communicator ranks into a NCCL window: ncclCommWindowRegister() and ncclCommWindowDeregister(). The calls currently support symmetric registration for P2P and NVLS, and require VMM memory buffers (i.e., CUMEM must be operational). * Implement specialized kernels taking advantage of symmetrically registered memory, with performance gains expected particularly for small to medium message sizes. * The kernels support 32 bit floating point types and smaller, and sum as the reduction operator, with no more than one collective operation per group. * Floating point summation is always done in fp32 accumulators (with the exception of fp8 on NVLS, where it uses fp16 inside the switch). Thus, the accuracy with fp8 and fp16 data types should be much improved. * This initial implementation supports non-network communicators only (P2P and NVLS transports). * To explore this functionality users need to use the new memory registration API calls with the NCCL_WIN_COLL_SYMMETRIC flag and all ranks of a communicator must pass buffers at the same offset in the same registration when invoking a collective NCCL operation. Add support for DGX Spark. Add support for DirectNIC (CX8) to the internal IB plugin. Add a new ncclCommShrink() API call * It is a non-collective call similar to ncclCommSplit(), which makes it possible to exclude some (possibly unresponsive) ranks from the parent communicator. Add support for loading multiple network plugins * This enables the creation of generic containers that can work across a range of providers. * Allow NCCL_NET_PLUGIN to accept a comma-separated list of plugins to load. NVLink SHARP (NVLS) improvements * Implement NVLS+IB SHARP support for AllGather and ReduceScatter with user buffer registration. This improves performance and reduces the number of CTAs needed to achieve peak bandwidth. * Gracefully fall back by default to other transports if NVLS initialization fails (the old behavior of returning an error code from a NCCL call can be preserved by setting NCCL_NVLS_ENABLE=1). * Decrease the NVLS channel count to 24 on Blackwell systems with multiple NVLink domains per communicator. * Enable fine-tuning of NCCL behavior per communicator using new "ncclConfig_t" members "collnetEnable", "CTAPolicy", and "nvlsCTAs". Profiler improvements * Extend the init function by adding communicator name, comm id (hash), rank, number of ranks, number of nodes, and the NCCL log function to the argument list. This makes the name and the comm id available to all events in the communicator without explicitly passing them to each individual event. Add the communicator id and rank to the profiler trace filename. Now, the communicator name can be set via a new "ncclConfig_t" member "commName". * Improve the accuracy of the GPU kernel events by providing GPU-generated timestamps for the start and stop of every NCCL operation. * Harmonize proxy events, removing overlaps between ProxyOp and ProxyStep states. * Add support for network-defined event updates (through "recordEventState"). * Report the correct number of channels used by every collective/p2p operation (used to be set to nMaxChannels for collectives and absent for p2ps). * Fix the logic on proxyCtrl Idle/Active events (Issue #1162). * Fix an issue where the network proxy profiler could lose track of an event identifier (Issue #1682). * Improve the backward compatibility with plugins older than v4. * Ensure that the work counters are 0-initialized. * Fix a potential race condition in the network profiler that could result in an event being linked to a wrong parent. MNNVL improvements * Increase to 16 the number of NICs used to communicate between MNNVL domains on GB200 systems, to optimize the performance of collective operations. * Add support for more complex MNNVL topologies with up to 32 NICs per node. * If the MNNVL fabric initialization was unsuccessful, NCCL will now fail by default, so as to avoid inadvertently falling back to a potentially much slower network transport. Such failures are typically due to a misconfigured IMEX support on the system. To continue without MNNVL, restart the job with NCCL_MNNVL_ENABLE=0. * Fix a potential hang in alltoall-like communication patterns at a scale of over 80 ranks. * Make NCCL_P2P_DISABLE=1 imply NCCL_MNNVL_ENABLE=0 (so the latter no longer needs to be specified on MNNVL systems). * Fix an initialization failure when NCCL_TOPO_FILE is used on MNNVL systems. * Fix the graph search to exclude non-local NICs. * Fix the SHM transport to use fabric handles on MNNVL systems. NIC Fusion improvements * Disable the creation of fused NICs for physical devices that haven't been merged. * Flatten multiple ports to a single PCI device within the internal IB plugin and reparent dual-port NICs under the first PCI parent. If the parent is not a PCI switch, PCI devices for fused NICs won't be duplicated. * Route traffic on GB200-CX8 systems through DirectNIC, not the host interface. Improve support for platforms with C2C connectivity (e.g., GB200) * Enable GPUDirect RDMA for the NICs by default. * Add support for P2C (PXN over C2C) and the LL128 protocol. Extend NCCL fault tolerance in multithreaded scenarios * Support the creation of multiple nonblocking communicators within a single group and polling in parallel for the completion using multiple threads (one per communicator). Enable ncclImplicitOrderLaunch for CUDA 12.9+ * This can potentially speed up NCCL_IMPLICIT_LAUNCH_ORDER. Improve the netSocket transport latency and control * Provide finer control over the size of the socket send/receive buffers, the task size, and the number of sockets that a single peer can open. * Add support for the inlining of small messages behind the header when using multiple sockets per connection. Improve the readability of the CPU affinity in the debug output * Print it as a range string rather than a bitmask. Fix a potential race condition in graph execution * A contention could arise when mixing graph and non-graph execution. Improve PXN connection code * Avoid duplicate and unused connections. RAS fixes * Fix a memory corruption at job termination time in case of a previously failed initialization of a RAS socket connection. * Fix a race condition leading to a crash when generating a RAS report during communicator initialization (Issues #1669, #1718). * Fix a potential race condition when gathering data for a RAS status report. Fix a potential memory corruption in ncclCommSplit() * Memory could get corrupted when resource sharing was in use and the size of the NVLink domain in the new communicator was smaller than in the old one. Fix asynchronous graph upload * Fix a small memory leak. * Fix oversychronization. Add a check for out-of-memory conditions in ncclMemAlloc() Clean up the NCCL socket code * accept() will retry also if just reading the magic failed (Issue #1613). * connect() will retry also if poll() did not return a POLLOUT event (Issue #1618). * Add error checking in a few instances (Issue #1539). * Fix the loop condition in ncclFindInterfaceMatchSubnet() (Issue #1574). * Clean up the debug output, downgrading WARN messages to INFO in non-critical cases, and printing the peer's address where relevant. Switch NCCL_DEBUG_FILE to line buffering * This should help avoid mixed-up partial output lines in multithreaded cases. Other minor fixes * Improve the checks for buffer overflows in the graph code (Issue #1585). * Extend logging and state clearing to all four events in the internal IB plugin (Issue #1650). * Fix the error path in case IB communication is not ready (Issue #1489). * Add ECE logging for IB fabric. * Fix various minor issues in the graph module (Issue #1635). * Clean up the debug output in the graph code, downgrading WARN messages to INFO in non-critical cases. * Add a missing argument to a directSend() call (Issue #1628). * Remove duplicate code in sendProxySetup() (Issue #1420). * Fix the order of arguments of cudaDeviceCanAccessPeer() (Issue #1507). * Fix compiler warnings with GCC 14. * Fix a typo in a comment (Issue #1236).
195 lines
5.2 KiB
C
195 lines
5.2 KiB
C
/*************************************************************************
|
|
* Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
|
|
*
|
|
* See LICENSE.txt for license information
|
|
************************************************************************/
|
|
|
|
#ifndef EVENT_H_
|
|
#define EVENT_H_
|
|
|
|
#include <sys/types.h>
|
|
#include <stdint.h>
|
|
#include <unistd.h>
|
|
#include "profiler.h"
|
|
|
|
#define MAX_CHANNELS 32
|
|
#define MAX_STEPS 16
|
|
#define MAX_OPS 16 // Up to 64K ranks for PAT
|
|
#define MAX_EVENTS_PER_REQ (8)
|
|
|
|
struct proxyOp;
|
|
struct proxyStep;
|
|
|
|
struct netPlugin {
|
|
uint8_t type;
|
|
int pluginType;
|
|
int pluginVer;
|
|
uint8_t pluginEvent;
|
|
union {
|
|
struct {
|
|
int device;
|
|
int qpNum;
|
|
int opcode;
|
|
uint64_t wr_id;
|
|
size_t length;
|
|
} qp;
|
|
struct {
|
|
int fd;
|
|
int op;
|
|
size_t length;
|
|
} sock;
|
|
};
|
|
double startTs;
|
|
double stopTs;
|
|
struct proxyStep* parent;
|
|
};
|
|
|
|
struct kernelCh {
|
|
uint8_t type;
|
|
uint8_t channelId;
|
|
struct taskEventBase* parent;
|
|
double startTs;
|
|
double stopTs;
|
|
uint64_t startGpuClk;
|
|
uint64_t stopGpuClk;
|
|
};
|
|
|
|
#define PROXY_STEP_SEND_GPU_WAIT 0
|
|
#define PROXY_STEP_SEND_PEER_WAIT 1
|
|
#define PROXY_STEP_SEND_WAIT 2
|
|
#define PROXY_STEP_RECV_WAIT 0
|
|
#define PROXY_STEP_RECV_FLUSH_WAIT 1
|
|
#define PROXY_STEP_RECV_GPU_WAIT 2
|
|
#define PROXY_STEP_MAX_STATES 3
|
|
|
|
struct proxyStep {
|
|
uint8_t type; // type of event: network transfer
|
|
int state;
|
|
int step; // network transfer id in given channel
|
|
int isSend; // send/recv channel operation
|
|
double timestamp[PROXY_STEP_MAX_STATES];
|
|
double startTs;
|
|
double stopTs;
|
|
struct proxyOp* parent;
|
|
struct netPlugin net[MAX_EVENTS_PER_REQ];
|
|
int nNetEvents;
|
|
};
|
|
|
|
struct proxyOp {
|
|
uint8_t type; // type of event: proxy operation
|
|
uint8_t channelId; // channel id for this proxy operation
|
|
pid_t pid;
|
|
int rank;
|
|
int peer; // peer rank for this proxy operation
|
|
int nSteps; // total number of network transfers for this proxy operation
|
|
int chunkSize; // chunk size for this proxy operation
|
|
int isSend; // send/recv channel operation
|
|
size_t transSize; // transfer data size for this proxy operation
|
|
double startTs;
|
|
double progrTs; // In progress state transition
|
|
double stopTs;
|
|
int stepCount; // last processed network operation for this proxy operation
|
|
struct proxyStep step[MAX_STEPS]; // array of network transfer events
|
|
struct taskEventBase* parent; // parent event p2p/collective
|
|
};
|
|
|
|
struct group;
|
|
struct context;
|
|
|
|
struct proxyCtrl {
|
|
uint8_t type;
|
|
struct context* ctx; // profiler context
|
|
double startTs;
|
|
double stopTs;
|
|
int state;
|
|
int appended; // appended proxy operations
|
|
};
|
|
|
|
// task level event base structure
|
|
struct taskEventBase {
|
|
uint8_t type; // event type: collective/p2p
|
|
int rank; // rank of the operation in NCCL communicator
|
|
const char* func; // ncclFunc*
|
|
int refCount; // number of references for this operation
|
|
struct group* parent; // parent event group
|
|
struct taskEventBase* next; // next top level event in group
|
|
double startTs;
|
|
double stopTs;
|
|
};
|
|
|
|
struct collective {
|
|
struct taskEventBase base; // base structure for this event
|
|
uint64_t seqNumber; // sequence number for this collective in communicator
|
|
void const* sendBuff;
|
|
void* recvBuff;
|
|
size_t count;
|
|
int root;
|
|
const char* datatype;
|
|
uint8_t nChannels;
|
|
const char* algo;
|
|
const char* proto;
|
|
int nWarps;
|
|
struct proxyOp op[MAX_CHANNELS][2*MAX_OPS];
|
|
int nProxyOps[MAX_CHANNELS];
|
|
struct kernelCh kernel[MAX_CHANNELS];
|
|
};
|
|
|
|
struct p2p {
|
|
struct taskEventBase base; // base structure for this event
|
|
uint8_t func;
|
|
void const* buff;
|
|
size_t count;
|
|
const char* datatype;
|
|
int peer;
|
|
uint8_t nChannels;
|
|
struct proxyOp op[MAX_CHANNELS];
|
|
struct kernelCh kernel[MAX_CHANNELS];
|
|
};
|
|
|
|
struct group {
|
|
uint8_t type;
|
|
struct context* ctx; // profiler context
|
|
int groupId;
|
|
int refCount;
|
|
struct taskEventBase* eventHead; // queue head for task events
|
|
struct taskEventBase* eventTail; // queue tail for task events
|
|
double startTs;
|
|
double stopTs;
|
|
struct group* next; // next group event in queue
|
|
};
|
|
|
|
// arrays for different event objects
|
|
struct context {
|
|
const char* commName;
|
|
uint64_t commHash;
|
|
int nranks;
|
|
int rank;
|
|
|
|
int groupPoolSize;
|
|
int groupPoolBase;
|
|
int groupPoolIndex;
|
|
struct group* groupPool;
|
|
|
|
int collPoolSize;
|
|
int collPoolBase;
|
|
int collPoolIndex;
|
|
struct collective* collPool;
|
|
|
|
int p2pPoolSize;
|
|
int p2pPoolBase;
|
|
int p2pPoolIndex;
|
|
struct p2p* p2pPool;
|
|
|
|
int proxyCtrlPoolSize;
|
|
int proxyCtrlPoolBase;
|
|
int proxyCtrlPoolIndex;
|
|
struct proxyCtrl* proxyCtrlPool;
|
|
};
|
|
|
|
int taskEventQueueEmpty(struct group* g);
|
|
void taskEventQueueEnqueue(struct group* g, struct taskEventBase* event);
|
|
struct taskEventBase* taskEventQueueHead(struct group* g);
|
|
struct taskEventBase* taskEventQueueDequeue(struct group* g);
|
|
|
|
#endif
|