Profiler improvements
* Add events for CUDA kernel start and end.
* Allow network plugins to generate profiling events
* Enable profiling on a per-operation basis, rather than per-communicator.
* Add support for graph capturing.
Add implicit launch order
* Allow to prevent deadlocks when using multiple NCCL communicators per
device by implicitly ordering NCCL operations using the host program
order. Disabled by default, set NCCL_LAUNCH_ORDER_IMPLICIT=1 to enable.
* Add a complementary mechanism to detect host threads racing to launch
to the same device. Enabled by default, set NCCL_LAUNCH_RACE_FATAL=0 to
disable.
Optimize the PAT algorithm
* Separate the computation and execution of PAT steps on different warps,
allowing to run up to 16 PAT steps in parallel to significantly
accelerate PAT and reduce its linear part.
Add support for setting QoS per communicator
* Add a new trafficClass field to the communicator configuration, to
allow the application to select a particular traffic class for a
given communicator. The meaning of the traffic class is
network-specific and should be set in accordance with the network
configuration.
* For the IB/RoCE plugin, existing config variables such as NCCL_IB_SL
and NCCL_IB_TC take precedence.
Allow to enable GPU Direct RDMA specifically on C2C platforms
* Disabled by default, set NCCL_NET_GDR_C2C=1 to enable.
Do not disable user buffer registration unless PXN is really used
* Only disable UB when a communicator has more than one rank per
node on any node.
RAS subsystem improvements
* Report operation counts separately for each collective operation type.
* Provide details about missing communicator ranks and reliably
distinguish ranks that are no longer a given communicator's members
(now reported as NOCOMM) from those that failed to respond.
Add support for timestamps to NCCL diagnostic messages
* On by default for WARN messages; NCCL_DEBUG_TIMESTAMP_LEVELS can be
used to enable them for other debug levels as well.
* The format can be changed using the NCCL_DEBUG_TIMESTAMP_FORMAT config
variable.
Reduce the memory usage with NVLink SHARP (NVLS)
* Potentially save hundreds of MBs of device memory, considering the
multicast buffer size granularity separately from the address alignment.
Update performance tuning for recent Intel CPUs
* Improve algorithm/protocol selection on recent CPUs such as Emerald
Rapids and Sapphire Rapids.
Improve channel scheduling when mixing LL and Simple operations.
* Make LL operations account for 4x more traffic to ensure LL and simple
operations complete at the same time.
Refactor the plugin code
* Clean up and harmonize the support code across the network, tuner,
and profiler plugins.
Add support for comment lines (starting with #) in the nccl.conf file
* Issue #1540.
Make user buffer registration problems print an INFO instead of a WARN.
Drop support for network plugin interface version 5.
Fix a race condition with split-shared communicators
* NCCL could hang during connection setup if multiple communicators
were grouped together that share resources.
Fix a performance regression when using NCCL_CROSS_NIC=1
* NCCL would unnecessarily alternate rings, breaking the GPU-NIC
associations.
Make GID index detection code more resilient
* Dynamic GID detection code was giving up too soon if the
detected index was not available (e.g., wasn't mapped to the
container's sysfs).
* Issues #1538, #1573.
Fix a race condition with non-blocking operation
* Fix issue when creating a non-blocking communicator after a non-
blocking collective operation on another communicator.
Fix shared memory usage on recent Blackwell GPUs.
* Issues NVIDIA/nccl-tests#287, NVIDIA/nccl-tests#291, #1637.
Fix an error with NIC fusion and IB SHARP when recreating communicators
* Disable the unloading of network plugins
Make the auto-merge failures in the NIC fusion non-fatal
* This could happen when trying to merge IB and RoCE devices.
Fixes to ncclCommAbort
* Fix hangs due to the progress thread spinning indefinitely on the
network progress.
* Reduce the abort time by up to two orders of magnitude.
Fix a crash when libnccl.so was dynamically unloaded
* The RAS subsystem was missing a clean-up handler.
Fix a hang if the network plugin's test() call returns an error.
Fix a hang on heterogeneous architectures
* Ensure we harmonize the tuning to avoid different tuning choices,
causing a hang.
Fix double-free on failed ncclCommInitRank and ncclCommFinalize.
Fix a potential list traversal bug during a group launch of multiple
communicators
* Issue #1599.
Unify the handling of NCCL configuration variables
* Under rare circumstances, some variables specified in the config file
could be ignored.
[ROCm/rccl commit: f44ac759fe]
* Limit P2P channels per peer to not exceeding max channels
* [UT] test single GPU cases for all collectives
* [UT] fix out of range root value
[ROCm/rccl commit: 4237caad69]
* misc/msccl: read graphCaptureStatus for every collective call
* fix a bug in checking whether UBR is enabled in MSCCLPP
* cmake: Fix patch reversal order
* misc/msccl: add logging
[ROCm/rccl commit: 23c0b7bd84]
* Improve RDMA flushing by write dummy payload with RO=0
* Rename env var for disabling this change to RCCL_GDR_FLUSH_GPU_MEM_NO_RELAXED_ORDERING
[ROCm/rccl commit: 60c1264d27]
* Enable GDRCopy only on gfx94x
* Use cudaFree instead of hipFree
* Add warning if failed to get device property
* Remove extra return
[ROCm/rccl commit: 32dc7ef47c]
COMPILING_TARGETS is not actually used for --offload-arch option,
instead GPU_TARGETS is being used implicitly when we call
find_package(hip REQUIRED) (See hip-config-amd.cmake).
[ROCm/rccl commit: 2f1c0bb213]
* Initial Script ready for review
* Added RCCL-tests and RCCL versions
* Added output folder and README
* Base format built
* Added ROCm version
* Added function to center titles and Vram information
* Added HIP version
* Cleaned formatting
* UCX version and MPI version
* Added NUMA balancing
* Added rocminfo
* Removed notes
* Changed regex for broadcom Nic
* Removed note by the ACS info
* Added Hostname to summary and details
* Print summary to terminal
* Added argparse
* Added flags and readme
* Added GPU ID
* fixed spelling
* renamed script again
* Added file descriptor and locked mem checks
* Added file descriptor and locked mem checks
* Removed extra spaces from summary table
* printing output file location
* Removed sudo in code and ACS flag
[ROCm/rccl commit: 4ba94d6662]
* Print KL/CL/KE events for all warps
* Fix count off-by-one issue
* Fix opCount in KE and restore CPU thread option
* Simplify count calculation
[ROCm/rccl commit: ebf7e2305e]
* Accept an EXPLICIT_ROCM_VERSION and use that vs inspecting the environment if provided.
* Use CMake's built in file reading support vs execute_process (without error checking) to avoid silent but deadly later failures.
* Properly quote some comparisons to avoid syntax errors if they happen to have an empty string.
* Guard against ROCM_PATH being an empty string, avoiding stray path extensions to root directories, etc.
Co-authored-by: Stella Laurenzo <stellaraccident@gmail.com>
[ROCm/rccl commit: 92ac136db5]
* Switched cmake_host_system_information feature to a manual parse to remain cmake 3.5 compliant.
* Updating minimum cmake to 3.16 to conform with the rest of ROCm. This change still applies.
[ROCm/rccl commit: 42ab425037]
* working tests for a single message size
* move call_RCCL routine StandaloneUtils, create .cpp file for StandaloneUtils so that it can be included in several tests
* simplify test invocation
* remove unecessary logs and exit from ncclCommRegister
* set expected results for allGather
* skip test if nranks doesn't match number of gpus, call getAndDistributeNCCLid only from parent process
* fix improper size of expected-results vector
* Removing unused changes.
* Refactored to create a new file for the forked collectives call, as StandaloneUtils is for the Standalone tests. Renamed the functions to be slightly more accurate and follow existing naming conventions.
* Apply suggestions from code review
Co-authored-by: corey-derochie-amd <161367113+corey-derochie-amd@users.noreply.github.com>
---------
Co-authored-by: isaki001 <isakioti@banff-pla-r27-38.pla.dcgpu>
Co-authored-by: corey-derochie-amd <161367113+corey-derochie-amd@users.noreply.github.com>
Co-authored-by: Corey Derochie <corey.derochie@amd.com>
[ROCm/rccl commit: 3398fa78fe]
* erase handle key from mscclpp communicator during deregistration
* remove check on buffer size being a multiple of 32 from registration/deregistration routines since these checks are applied during enqueue
* add check for greater than zero buffer size in mscclpp registration
[ROCm/rccl commit: 19105206f6]
Wenkai Du
Nilesh M Negi
Kamil Iskra
Tim
isaki001
Nusrat Islam
Bertan Dogancay
dependabot[bot]
Pedram Alizadeh
gilbertlee-amd
akolliasAMD
Sohaib Nadeem
Nikhil-Nunna
corey-derochie-amd
mberenjk
rahulc1984
Dingming Wu
Vijay Srinivasan
Nikhil-Nunna
AbandiGa
Edgar Gabriel
Jeffrey E Erickson
Jeffrey Novotny