2.23.4-1

Add scalable init API
 * Add new ncclCommInitRankScalable to allow for passing multiple
   unique IDs to the init function.
 * Spreads the load onto multiple bootstrap roots, allowing for
   constant bootstrap time.
 * Requires multiple ranks to create a unique ID, and the CPU-side
   ID exchange code to call allgather[v] instead of broadcast.

Accelerate init bootstrap operations
 * Reduce the number of calls to allgather.
 * Allow roots to reply early to ranks when information is already
   available.
 * Add an option to use ncclNet instead of sockets to perform
   bootstrap allgather operations.

Add PAT algorithms for Allgather and ReduceScatter
 * Parallel Aggregated Trees, variation of Bruck algorithm.
 * Logarithmic number of network steps for small sizes at scale.
 * Only supports one rank per node at the moment.

Add support for registered buffers for intra-node communication.
 * Allow registered user buffers to be accessed directly intra-node
 * Avoids extra copies in algorithms which permit it, saving
   memory bandwidth and helping with compute overlap.

Add profiler plugin API
 * New plugin API for profiling
 * Supports various levels of profiling, with a hierarchy.

Asynchronous graph allocation
 * Make calls to cudaMalloc and cudaMemcpy during graph allocation
   asynchronous.
 * Significantly speeds up graph capture.

Use fatal IB asynchronous events to stop network operation
 * Avoids many other error messages
 * Only fatal errors are affected; potentially transient errors
   (e.g. port down) do not cause an immediate stop.

Set P2P level to PXB on AMD CPUs when using more than 2 GPUs per node
 * P2P would cause a significant performance degradation when using
   many GPUs, and therefore many interleaved data flows.
 * Disable P2P through the CPU when we have 3+ GPUs per node; keep it
   enabled when we only have 2 GPUs.

Improve the init logs to report the real NCCL function.
 * Make the log report ncclCommInitRank or ncclCommSplit, rather than
   the generic ncclCommInitRankFunc.

Add a parameter to set the location of the user configuration file.
 * Add NCCL_CONF_FILE environment variable to set where the user's
   configuration file resides.

Increase default IB timeout
 * Increase IB timeout value from 18 to 20.
 * Should help avoid fatal errors on large RoCE systems.

Add new check for nvidia peermem
 * On linux kernels 6.6+, /sys/kernel/mm/memory_peers is no longer
   present; check for /sys/module/nvidia_peermem/version instead.

Fix old performance regression when mixing small and large operations.
 * Improves distribution of work on channels.

Fix crash when NUMA IDs are equal to -1.
 * Can happen when a NIC is a virtual NIC, or when linux doesn't
   know which NUMA node a device is attached to
 * Issue NVIDIA/nccl-tests#233

Fix tree graph search when NCCL_CROSS_NIC is set to 1.
 * Would force NCCL to use the balanced_tree pattern, thereby
   disabling LL128 on platforms with 1 GPU+1 NIC per PCI switch.
 * Would also try to use alternate rings even though it was not
   needed.

Compiler tweaks and fixes
 * PR #1177
 * PR #1228

Fix stack smash
 * PR #1325

Fixes for multi-node NVLink + IB operation

Coverity fixes and comments.


[ROCm/rccl commit: 68b542363f]

projects/rccl/src/misc/ipcsocket.cc

@@ -41,6 +41,7 @@ ncclResult_t ncclIpcSocketInit(ncclIpcSocket *handle, int rank, uint64_t hash, v
   int len = snprintf(temp, NCCL_IPC_SOCKNAME_LEN, NCCL_IPC_SOCKNAME_STR, rank, hash);
   if (len > (sizeof(cliaddr.sun_path) - 1)) {
     WARN("UDS: Cannot bind provided name to socket. Name too large");
+    close(fd);
     return ncclInternalError;
   }
 #ifndef USE_ABSTRACT_SOCKET
@@ -66,7 +67,7 @@ ncclResult_t ncclIpcSocketInit(ncclIpcSocket *handle, int rank, uint64_t hash, v
   // Mark socket as non-blocking
   if (handle->abortFlag) {
     int flags;
-    EQCHECK(flags = fcntl(fd, F_GETFL), -1);
+    SYSCHECK(flags = fcntl(fd, F_GETFL), "fcntl");
     SYSCHECK(fcntl(fd, F_SETFL, flags | O_NONBLOCK), "fcntl");
   }
 
@@ -186,20 +187,16 @@ ncclResult_t ncclIpcSocketSendMsg(ncclIpcSocket *handle, void *hdr, int hdrLen,
   cliaddr.sun_path[0] = '\0'; // Linux abstract socket trick
 #endif
 
-  TRACE(NCCL_INIT, "UDS: Sending hdr %p len %d to UDS socket %s", hdr, hdrLen, temp);
+  TRACE(NCCL_INIT, "UDS: Sending hdr %p len %d fd %d to UDS socket %s", hdr, hdrLen, sendFd, temp);
 
-  if (sendFd != -1) {
-    TRACE(NCCL_INIT, "UDS: Sending fd %d to UDS socket %s", sendFd, temp);
+  msg.msg_control = control_un.control;
+  msg.msg_controllen = sizeof(control_un.control);
 
-    msg.msg_control = control_un.control;
-    msg.msg_controllen = sizeof(control_un.control);
-
-    cmptr = CMSG_FIRSTHDR(&msg);
-    cmptr->cmsg_len = CMSG_LEN(sizeof(int));
-    cmptr->cmsg_level = SOL_SOCKET;
-    cmptr->cmsg_type = SCM_RIGHTS;
-    memmove(CMSG_DATA(cmptr), &sendFd, sizeof(sendFd));
-  }
+  cmptr = CMSG_FIRSTHDR(&msg);
+  cmptr->cmsg_len = CMSG_LEN(sizeof(int));
+  cmptr->cmsg_level = SOL_SOCKET;
+  cmptr->cmsg_type = SCM_RIGHTS;
+  memmove(CMSG_DATA(cmptr), &sendFd, sizeof(sendFd));
 
   msg.msg_name = (void *)&cliaddr;
   msg.msg_namelen = sizeof(struct sockaddr_un);