bump NCCL floor to 2.18.1.1, include nccl.h where it's needed (#4661)

Contributes to rapidsai/build-planning#102 Some RAPIDS libraries are using `ncclCommSplit()`, which was introduced in `nccl==2.18.1.1`. This is part of a series of PRs across RAPIDS updating libraries' pins to `nccl>=2.18.1.1` to ensure they get a new-enough version that supports that. Authors: - James Lamb (https://github.com/jameslamb) Approvers: - Chuck Hastings (https://github.com/ChuckHastings) - Vyas Ramasubramani (https://github.com/vyasr) - https://github.com/jakirkham URL: #4661
rapidsai · Sep 25, 2024 · 27f2256 · 27f2256
1 parent 59217ad
commit 27f2256
Show file tree

Hide file tree

Showing 7 changed files with 9 additions and 4 deletions.
diff --git a/conda/environments/all_cuda-118_arch-x86_64.yaml b/conda/environments/all_cuda-118_arch-x86_64.yaml
@@ -33,7 +33,7 @@ dependencies:
 - libraft==24.10.*,>=0.0.0a0
 - librmm==24.10.*,>=0.0.0a0
 - nbsphinx
-- nccl>=2.9.9
+- nccl>=2.18.1.1
 - networkx>=2.5.1
 - networkx>=3.0
 - ninja

diff --git a/conda/environments/all_cuda-125_arch-x86_64.yaml b/conda/environments/all_cuda-125_arch-x86_64.yaml
@@ -39,7 +39,7 @@ dependencies:
 - libraft==24.10.*,>=0.0.0a0
 - librmm==24.10.*,>=0.0.0a0
 - nbsphinx
-- nccl>=2.9.9
+- nccl>=2.18.1.1
 - networkx>=2.5.1
 - networkx>=3.0
 - ninja

diff --git a/conda/recipes/libcugraph/conda_build_config.yaml b/conda/recipes/libcugraph/conda_build_config.yaml
@@ -17,7 +17,7 @@ doxygen_version:
   - ">=1.8.11"
 
 nccl_version:
-  - ">=2.9.9"
+  - ">=2.18.1.1"
 
 c_stdlib:
   - sysroot

diff --git a/cpp/include/cugraph/mtmg/instance_manager.hpp b/cpp/include/cugraph/mtmg/instance_manager.hpp
@@ -20,6 +20,8 @@
 
 #include <raft/comms/std_comms.hpp>
 
+#include <nccl.h>
+
 #include <vector>
 
 namespace cugraph {

diff --git a/cpp/include/cugraph/mtmg/resource_manager.hpp b/cpp/include/cugraph/mtmg/resource_manager.hpp
@@ -27,6 +27,8 @@
 #include <rmm/mr/device/owning_wrapper.hpp>
 #include <rmm/mr/device/pool_memory_resource.hpp>
 
+#include <nccl.h>
+
 #include <execution>
 
 namespace cugraph {

diff --git a/cpp/tests/mtmg/multi_node_threaded_test.cu b/cpp/tests/mtmg/multi_node_threaded_test.cu
@@ -39,6 +39,7 @@
 #include <thrust/unique.h>
 
 #include <gtest/gtest.h>
+#include <nccl.h>
 
 #include <filesystem>
 #include <fstream>

diff --git a/dependencies.yaml b/dependencies.yaml
@@ -530,7 +530,7 @@ dependencies:
           - aiohttp
           - fsspec>=0.6.0
           - requests
-          - nccl>=2.9.9
+          - nccl>=2.18.1.1
           - ucx-proc=*=gpu
           - &ucx_py_unsuffixed ucx-py==0.40.*,>=0.0.0a0
       - output_types: pyproject