Files
derek.holloway 7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
Flatten subprojects into a single base monorepo
2026-09-02 16:34:47 -07:00

257 lines
8.1 KiB
C++

/*************************************************************************
* Copyright (c) 2025 Advanced Micro Devices, Inc. All rights reserved.
*
* See LICENSE.txt for license information
************************************************************************/
#include <gtest/gtest.h>
#include "common/ErrCode.hpp"
#include "common/ProcessIsolatedTestRunner.hpp"
#include "enqueue.h"
namespace RcclUnitTesting
{
class EnqueueCountTests : public ::testing::Test
{
};
TEST_F(EnqueueCountTests, ncclFuncSendCount_AllTests)
{
ProcessIsolatedTestRunner::ExecutionOptions options;
options.stopOnFirstFailure = false;
options.verboseLogging = true;
RUN_ISOLATED_TESTS_WITH_OPTIONS(
options,
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncSendCount_AllReduce",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncSendCount(ncclFuncAllReduce, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncSendCount_Broadcast",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncSendCount(ncclFuncBroadcast, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncSendCount_Reduce",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncSendCount(ncclFuncReduce, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncSendCount_AllGather",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncSendCount(ncclFuncAllGather, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncSendCount_ReduceScatter",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncSendCount(ncclFuncReduceScatter, nRanks, count);
EXPECT_EQ(result, count * nRanks);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncSendCount_ZeroCount",
[]()
{
size_t result = ncclFuncSendCount(ncclFuncAllReduce, 4, 0);
EXPECT_EQ(result, 0);
}
).withNumGpus(0)
);
}
TEST_F(EnqueueCountTests, ncclFuncRecvCount_AllTests)
{
ProcessIsolatedTestRunner::ExecutionOptions options;
options.stopOnFirstFailure = false;
options.verboseLogging = true;
RUN_ISOLATED_TESTS_WITH_OPTIONS(
options,
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncRecvCount_AllReduce",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncRecvCount(ncclFuncAllReduce, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncRecvCount_Broadcast",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncRecvCount(ncclFuncBroadcast, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncRecvCount_Reduce",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncRecvCount(ncclFuncReduce, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncRecvCount_AllGather",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncRecvCount(ncclFuncAllGather, nRanks, count);
EXPECT_EQ(result, count * nRanks);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncRecvCount_ReduceScatter",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncRecvCount(ncclFuncReduceScatter, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncRecvCount_ZeroCount",
[]()
{
size_t result = ncclFuncRecvCount(ncclFuncAllReduce, 4, 0);
EXPECT_EQ(result, 0);
}
).withNumGpus(0)
);
}
TEST_F(EnqueueCountTests, ncclFuncMaxSendRecvCount_AllTests)
{
ProcessIsolatedTestRunner::ExecutionOptions options;
options.stopOnFirstFailure = false;
options.verboseLogging = true;
RUN_ISOLATED_TESTS_WITH_OPTIONS(
options,
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncMaxSendRecvCount_AllReduce",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncMaxSendRecvCount(ncclFuncAllReduce, nRanks, count);
EXPECT_EQ(result, count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncMaxSendRecvCount_AllGather",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncMaxSendRecvCount(ncclFuncAllGather, nRanks, count);
EXPECT_EQ(result, count * nRanks);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncMaxSendRecvCount_ReduceScatter",
[]()
{
size_t count = 1000;
int nRanks = 4;
size_t result = ncclFuncMaxSendRecvCount(ncclFuncReduceScatter, nRanks, count);
EXPECT_EQ(result, count * nRanks);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncMaxSendRecvCount_ZeroCount",
[]()
{
size_t result = ncclFuncMaxSendRecvCount(ncclFuncAllReduce, 4, 0);
EXPECT_EQ(result, 0);
}
).withNumGpus(0)
);
}
TEST_F(EnqueueCountTests, ncclFuncCounts_EdgeCases)
{
ProcessIsolatedTestRunner::ExecutionOptions options;
options.stopOnFirstFailure = false;
options.verboseLogging = true;
RUN_ISOLATED_TESTS_WITH_OPTIONS(
options,
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncCounts_SingleRank",
[]()
{
size_t count = 1000;
int nRanks = 1;
EXPECT_EQ(ncclFuncSendCount(ncclFuncAllReduce, nRanks, count), count);
EXPECT_EQ(ncclFuncRecvCount(ncclFuncAllReduce, nRanks, count), count);
EXPECT_EQ(ncclFuncMaxSendRecvCount(ncclFuncAllReduce, nRanks, count), count);
}
).withNumGpus(0),
ProcessIsolatedTestRunner::TestConfig(
"ncclFuncCounts_LargeRankCount",
[]()
{
size_t count = 1000;
int nRanks = 1024;
EXPECT_EQ(ncclFuncSendCount(ncclFuncAllGather, nRanks, count), count);
EXPECT_EQ(ncclFuncRecvCount(ncclFuncAllGather, nRanks, count), count * nRanks);
EXPECT_EQ(
ncclFuncMaxSendRecvCount(ncclFuncAllGather, nRanks, count),
count * nRanks
);
}
).withNumGpus(0)
);
}
} // namespace RcclUnitTesting