AI/rocm-systems
2
0
Fork 0
程式碼 問題 合併請求 Actions 套件 專案 版本發布 Wiki 動態

Merge remote-tracking branch 'nccl/master' into develop

瀏覽代碼 
[ROCm/rccl commit: 858b4e76eb]
此提交包含在：
Marzieh Berenjkoub
2026-01-20 13:01:49 -06:00
父節點 f336ad5133 9bc0e13ef0
當前提交 d7293281f3
共有 240 個檔案被更改，包括 16266 行新增 和 3578 行删除
下載 Patch 檔 下載差異檔 展開所有檔案 折疊所有檔案
projects/rccl/ext-profiler/example/CMakeLists.txt
+34
查看文件
@@ -0,0 +1,34 @@
# Find all C source files in current directory
set(SRC_FILES
${CMAKE_CURRENT_SOURCE_DIR}/plugin.cc
${CMAKE_CURRENT_SOURCE_DIR}/print_event.cc
)
# Create shared library
add_library(nccl-profiler-example SHARED ${SRC_FILES})
# Set include directories
target_include_directories(nccl-profiler-example PRIVATE
${CMAKE_CURRENT_SOURCE_DIR}/nccl
${CUDAToolkit_INCLUDE_DIRS}
)
# Set output name to match Makefile
set_target_properties(nccl-profiler-example PROPERTIES
OUTPUT_NAME "nccl-profiler-example"
PREFIX "lib"
POSITION_INDEPENDENT_CODE ON
LIBRARY_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/lib
)
add_custom_command(TARGET nccl-profiler-example POST_BUILD
COMMAND ${CMAKE_COMMAND} -E make_directory ${CMAKE_BINARY_DIR}/test/unit/plugins
COMMAND ${CMAKE_COMMAND} -E copy ${CMAKE_BINARY_DIR}/lib/libnccl-profiler-example.so ${CMAKE_BINARY_DIR}/test/unit/plugins
)
# Add custom target for clean (equivalent to Makefile clean target)
add_custom_target(clean-profiler-lib
COMMAND ${CMAKE_COMMAND} -E remove -f ${CMAKE_BINARY_DIR}/lib/libnccl-profiler-example.so
COMMAND ${CMAKE_COMMAND} -E remove -f ${CMAKE_BINARY_DIR}/test/unit/plugins/libnccl-profiler-example.so
COMMENT "Cleaning libnccl-profiler-example.so"
)
projects/rccl/ext-profiler/example/Makefile
+15 -8
查看文件
@@ -4,19 +4,26 @@
# See LICENSE.txt for license information
#
.DEFAULT_GOAL: build
include ../../makefiles/common.mk
SRCDIR ?= $(abspath ../..)
ROCM_PATH ?= $(wildcard /opt/rocm)
CXX = $(ROCM_PATH)/lib/llvm/bin/amdclang++
BUILDDIR ?= .
NCCLDIR := $(BUILDDIR)
HIPIFY_DIR := hipify-profiler
SRC_FILES := $(wildcard *.c)
SRC_FILES := $(wildcard *.cc)
HIPIFY_SRC := $(addprefix $(HIPIFY_DIR)/,$(SRC_FILES))
build: ${BUILDDIR}/librccl-profiler.so
build: ${BUILDDIR}/librccl-profiler-example.so
${BUILDDIR}/librccl-profiler.so: ${SRC_FILES}
${BUILDDIR}/librccl-profiler-example.so: $(HIPIFY_SRC)
@printf "Compiling %-35s > %s\n" $< $@
@mkdir -p ${BUILDDIR}
$(CC) -Inccl -fPIC -shared -o $@ $^
$(CXX) -D__HIP_PLATFORM_AMD__ -I$(HIPIFY_DIR) -I$(HIPIFY_DIR)/nccl -I$(ROCM_PATH)/include -fPIC -shared -o $@ $^
$(HIPIFY_DIR)/%.cc: %.cc
@mkdir -p $(HIPIFY_DIR)/nccl
@cp *.cc *.h $(HIPIFY_DIR)/
@cp nccl/*.h $(HIPIFY_DIR)/nccl/
@hipify-perl -inplace -quiet-warnings $(HIPIFY_DIR)/*.cc $(HIPIFY_DIR)/*.h
clean:
rm -f ${BUILDDIR}/librccl-profiler.so
rm -rf ${BUILDDIR}/librccl-profiler-example.so $(HIPIFY_DIR)
projects/rccl/ext-profiler/example/README.md
+59 -121
查看文件
@@ -13,8 +13,7 @@ change the size of the event window the profiler keeps track of.
## Building the profiler plugin
To use the example plugin, just type `make`. You will need a NCCL build's include directory present.
You can override `NCCL_HOME` to where the NCCL installation is on your system.
To build the example plugin shipped as part of NCCL, just type `make`.
## Using the profiler plugin
@@ -27,13 +26,13 @@ You can override `NCCL_HOME` to where the NCCL installation is on your system.
As an example, setting:
`NCCL_PROFILE_EVENT_MASK` to 1 (`ncclProfileGroup`) | 2 (`ncclProfileColl`) | 8 (`ncclProfileProxyOp`)
`NCCL_PROFILE_EVENT_MASK` to 256 (`ncclProfileGroupApi`) | 2 (`ncclProfileColl`) | 8 (`ncclProfileProxyOp`)
enables the profiling of the group, the collective and the proxy op events. The same events can be
enables the profiling of the group API, the collective and the proxy op events. The same events can be
expressed more concisely by setting `NCCL_PROFILE_EVENT_MASK` to 8 (`ncclProfileProxyOp`). Indeed,
in NCCL all the events above (in the event hierarchy) the one requested are also captured. The advantage
is that the profiler can easily correlate events that belong to the same NCCL operation and present
them accordingly.
them accordingly. Setting `NCCL_PROFILE_EVENT_MASK` to 4095 enables all events supported by the v5 profiler.
3. Set `NCCL_PROFILE_DUMP_FILE` to the name of the dump file for the collected traces. A file named
${NCCL_PROFILE_DUMP_FILE}-hostname-tid.txt is created. Profiler traces are saved using the chrome
@@ -57,11 +56,14 @@ The group, collective and p2p pools contain objects for the corresponding events
contains objects for `ProxyCtrl` events and the `ProxyDetach` pool contains objects for `ProxyOp` events
generated by remote proxies. A list of pools and their size is reported below:
- `NCCL_PROFILE_GROUP_POOL_SIZE` (16)
- `NCCL_PROFILE_COLL_POOL_SIZE` (16)
- `NCCL_PROFILE_P2P_POOL_SIZE` (1024)
- `NCCL_PROFILE_GROUP_API_POOL_SIZE` (256)
- `NCCL_PROFILE_COLL_API_POOL_SIZE` (256)
- `NCCL_PROFILE_P2P_API_POOL_SIZE` (256)
- `NCCL_PROFILE_KERNEL_LAUNCH_POOL_SIZE` (256)
- `NCCL_PROFILE_COLL_POOL_SIZE` (256)
- `NCCL_PROFILE_P2P_POOL_SIZE` (256)
- `NCCL_PROFILE_PROXY_CTRL_POOL_SIZE` (16)
- `NCCL_PROFILE_PROXY_DETACH_POOL_SIZE` (128)
- `NCCL_PROFILE_PROXY_DETACH_POOL_SIZE` (256)
Remote proxy operations are generated when PXN is in use. Refer to this article for more information
about PXN and how it works:
@@ -73,76 +75,58 @@ The example profiler generates traces using the json format. An example of trace
```
[
{"name": "Group", "cat": "GROUP", "ph": "b", "id": 0, "pid": 4157654, "tid": 1, "ts": 764234.611328, "args": {"groupId": 0}},
{"name": "AllReduce", "cat": "COLL", "ph": "b", "id": 0, "pid": 4157654, "tid": 1, "ts": 764237.294922, "args": {"SeqNum": 0, "CommHash": 673864846479792718, "Rank": 1, "Count": 32768, "Datatype": "ncclFloat32", "Algorithm": "RING", "Protocol": "LL", "nMaxChannels": 2}},
{"name": "Recv", "cat": "PROXY", "ph": "b", "id": 0, "pid": 4157654, "tid": 1, "ts": 768464.936523, "args": {"Channel": 0, "Peer": 0, "Steps": 14, "ChunkSize": 32768, "transSize": 229376, "POSTED": {"step": 14, "ts": 772020.300781}, "RECEIVED": {"step": 14, "ts": 772196.049805}, "TRANSMITTED": {"step": 14, "ts": 772197.326172}, "DONE": {"step": 14, "ts": 772201.538086}}},
{"name": "RecvBufferWait", "cat": "NET", "ph": "b", "id": 0, "pid": 4157654, "tid": 1, "ts": 768465.158203, "args": {"Step": 0}},
{"name": "RecvBufferWait", "cat": "NET", "ph": "e", "id": 0, "pid": 4157654, "tid": 1, "ts": 768477.924805},
{"name": "RecvWait", "cat": "NET", "ph": "b", "id": 0, "pid": 4157654, "tid": 1, "ts": 768477.924805, "args": {"Step": 0}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 0, "pid": 4157654, "tid": 1, "ts": 768547.197266},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 0, "pid": 4157654, "tid": 1, "ts": 768547.197266, "args": {"Step": 0}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 0, "pid": 4157654, "tid": 1, "ts": 768564.174805},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 0, "pid": 4157654, "tid": 1, "ts": 768564.174805, "args": {"Step": 0}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 0, "pid": 4157654, "tid": 1, "ts": 768568.276367},
{"name": "RecvBufferWait", "cat": "NET", "ph": "b", "id": 1, "pid": 4157654, "tid": 1, "ts": 768503.604492, "args": {"Step": 1}},
{"name": "RecvBufferWait", "cat": "NET", "ph": "e", "id": 1, "pid": 4157654, "tid": 1, "ts": 768504.549805},
{"name": "RecvWait", "cat": "NET", "ph": "b", "id": 1, "pid": 4157654, "tid": 1, "ts": 768504.549805, "args": {"Step": 1}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 1, "pid": 4157654, "tid": 1, "ts": 769994.490234},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 1, "pid": 4157654, "tid": 1, "ts": 769994.490234, "args": {"Step": 1}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 1, "pid": 4157654, "tid": 1, "ts": 769995.012695},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 1, "pid": 4157654, "tid": 1, "ts": 769995.012695, "args": {"Step": 1}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 1, "pid": 4157654, "tid": 1, "ts": 770006.914062},
{"name": "RecvBufferWait", "cat": "NET", "ph": "b", "id": 2, "pid": 4157654, "tid": 1, "ts": 768506.941406, "args": {"Step": 2}},
{"name": "RecvBufferWait", "cat": "NET", "ph": "e", "id": 2, "pid": 4157654, "tid": 1, "ts": 768507.435547},
{"name": "RecvWait", "cat": "NET", "ph": "b", "id": 2, "pid": 4157654, "tid": 1, "ts": 768507.435547, "args": {"Step": 2}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 2, "pid": 4157654, "tid": 1, "ts": 771452.536133},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 2, "pid": 4157654, "tid": 1, "ts": 771452.536133, "args": {"Step": 2}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 2, "pid": 4157654, "tid": 1, "ts": 771453.060547},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 2, "pid": 4157654, "tid": 1, "ts": 771453.060547, "args": {"Step": 2}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 2, "pid": 4157654, "tid": 1, "ts": 771468.458008},
{"name": "RecvBufferWait", "cat": "NET", "ph": "b", "id": 3, "pid": 4157654, "tid": 1, "ts": 768509.484375, "args": {"Step": 3}},
{"name": "RecvBufferWait", "cat": "NET", "ph": "e", "id": 3, "pid": 4157654, "tid": 1, "ts": 768510.250000},
{"name": "RecvWait", "cat": "NET", "ph": "b", "id": 3, "pid": 4157654, "tid": 1, "ts": 768510.250000, "args": {"Step": 3}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 3, "pid": 4157654, "tid": 1, "ts": 771904.499023},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 3, "pid": 4157654, "tid": 1, "ts": 771904.499023, "args": {"Step": 3}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 3, "pid": 4157654, "tid": 1, "ts": 771904.991211},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 3, "pid": 4157654, "tid": 1, "ts": 771904.991211, "args": {"Step": 3}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 3, "pid": 4157654, "tid": 1, "ts": 771910.500000},
{"name": "Send", "cat": "PROXY", "ph": "b", "id": 1, "pid": 4157654, "tid": 1, "ts": 768482.878906, "args": {"Channel": 0, "Peer": 2, "Steps": 14, "ChunkSize": 32768, "transSize": 229376, "POSTED": {"step": 14, "ts": 771995.675781}, "REM_FIFO_WAIT": {"step": 14, "ts": 772190.692383}, "TRANSMITTED": {"step": 14, "ts": 772191.516602}, "DONE": {"step": 14, "ts": 772208.473633}}},
{"name": "SendBufferWait", "cat": "NET", "ph": "b", "id": 14, "pid": 4157654, "tid": 1, "ts": 768483.019531, "args": {"Step": 0}},
{"name": "SendBufferWait", "cat": "NET", "ph": "e", "id": 14, "pid": 4157654, "tid": 1, "ts": 768483.300781},
{"name": "SendGpuWait", "cat": "NET", "ph": "b", "id": 14, "pid": 4157654, "tid": 1, "ts": 768483.300781, "args": {"Step": 0}},
{"name": "SendGpuWait", "cat": "NET", "ph": "e", "id": 14, "pid": 4157654, "tid": 1, "ts": 769594.615234},
{"name": "SendWait", "cat": "NET", "ph": "b", "id": 14, "pid": 4157654, "tid": 1, "ts": 769594.615234, "args": {"Step": 0}},
{"name": "SendWait", "cat": "NET", "ph": "e", "id": 14, "pid": 4157654, "tid": 1, "ts": 769618.889648},
{"name": "SendBufferWait", "cat": "NET", "ph": "b", "id": 15, "pid": 4157654, "tid": 1, "ts": 768505.083008, "args": {"Step": 1}},
{"name": "SendBufferWait", "cat": "NET", "ph": "e", "id": 15, "pid": 4157654, "tid": 1, "ts": 768505.163086},
{"name": "SendGpuWait", "cat": "NET", "ph": "b", "id": 15, "pid": 4157654, "tid": 1, "ts": 768505.163086, "args": {"Step": 1}},
{"name": "SendGpuWait", "cat": "NET", "ph": "e", "id": 15, "pid": 4157654, "tid": 1, "ts": 769610.555664},
{"name": "SendWait", "cat": "NET", "ph": "b", "id": 15, "pid": 4157654, "tid": 1, "ts": 769610.555664, "args": {"Step": 1}},
{"name": "SendWait", "cat": "NET", "ph": "e", "id": 15, "pid": 4157654, "tid": 1, "ts": 769622.517578},
{"name": "SendBufferWait", "cat": "NET", "ph": "b", "id": 16, "pid": 4157654, "tid": 1, "ts": 768507.937500, "args": {"Step": 2}},
{"name": "SendBufferWait", "cat": "NET", "ph": "e", "id": 16, "pid": 4157654, "tid": 1, "ts": 768508.017578},
{"name": "SendGpuWait", "cat": "NET", "ph": "b", "id": 16, "pid": 4157654, "tid": 1, "ts": 768508.017578, "args": {"Step": 2}},
{"name": "SendGpuWait", "cat": "NET", "ph": "e", "id": 16, "pid": 4157654, "tid": 1, "ts": 770002.129883},
{"name": "SendWait", "cat": "NET", "ph": "b", "id": 16, "pid": 4157654, "tid": 1, "ts": 770002.129883, "args": {"Step": 2}},
{"name": "SendWait", "cat": "NET", "ph": "e", "id": 16, "pid": 4157654, "tid": 1, "ts": 770013.848633},
{"name": "SendBufferWait", "cat": "NET", "ph": "b", "id": 17, "pid": 4157654, "tid": 1, "ts": 768510.742188, "args": {"Step": 3}},
{"name": "SendBufferWait", "cat": "NET", "ph": "e", "id": 17, "pid": 4157654, "tid": 1, "ts": 768510.822266},
{"name": "SendGpuWait", "cat": "NET", "ph": "b", "id": 17, "pid": 4157654, "tid": 1, "ts": 768510.822266, "args": {"Step": 3}},
{"name": "SendGpuWait", "cat": "NET", "ph": "e", "id": 17, "pid": 4157654, "tid": 1, "ts": 771461.563477},
{"name": "SendWait", "cat": "NET", "ph": "b", "id": 17, "pid": 4157654, "tid": 1, "ts": 771461.563477, "args": {"Step": 3}},
{"name": "SendWait", "cat": "NET", "ph": "e", "id": 17, "pid": 4157654, "tid": 1, "ts": 771469.171875},
{"name": "Group API", "cat": "GROUP_API", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 3433.595001, "args": {"groupApiId": 0, "groupDepth":1}},
{"name": "KernelLaunch", "cat": "KERNEL_LAUNCH", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 0.000000, "args": {"groupId": 0, "Stream": 0x5020000567d0}},
{"name": "KernelLaunch", "cat": "KERNEL_LAUNCH", "ph": "e", "id": 0, "pid": 225798, "tid": 1, "ts": 111991.558990},
{"name": "AllReduce", "cat": "COLL_API", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 0.000000, "args": {"count": 262144, "datatype": ncclFloat32, "root": 0, "GraphCaptured":0, "Stream": 0x5020000567d0}},
{"name": "AllReduce", "cat": "COLL", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 111994.477997, "args": {"SeqNum": 0, "CommHash": 1493613951195738943, "Rank": 0, "Count": 262144, "Datatype": "ncclFloat32", "Algorithm": "RING", "Protocol": "SIMPLE", "nChannels": 2}},
{"name": "KernelCh", "cat": "GPU", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 119711.888000, "args": {"Channel": 0, "StartGpuClk": 1756135989724672000, "StopGpuClk": 1756135989732831232}},
{"name": "ScheduleRecv", "cat": "PROXY", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 119652.709991, "args": {"Channel": 0, "Peer": 1, "Steps": 4, "ChunkSize": 4194304, "transSize": 524288}},
{"name": "ScheduleRecv", "cat": "PROXY", "ph": "e", "id": 0, "pid": 225798, "tid": 1, "ts": 119686.300995},
{"name": "ProgressRecv", "cat": "PROXY", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 119686.300995, "args": {"Channel": 0, "Peer": 1, "Steps": 4, "ChunkSize": 4194304, "transSize": 524288}},
{“name": "RecvWait", "cat": "NET", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 119707.677979, "args": {"Step": 0}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 0, "pid": 225798, "tid": 1, "ts": 119807.691986},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 119807.691986, "args": {"Step": 0}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 0, "pid": 225798, "tid": 1, "ts": 119867.338989},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 0, "pid": 225798, "tid": 1, "ts": 119867.338989, "args": {"Step": 0}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 0, "pid": 225798, "tid": 1, "ts": 120120.983002},
{"name": "RecvWait", "cat": "NET", "ph": "b", "id": 1, "pid": 225798, "tid": 1, "ts": 119733.647980, "args": {"Step": 1}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 1, "pid": 225798, "tid": 1, "ts": 119844.401001},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 1, "pid": 225798, "tid": 1, "ts": 119844.401001, "args": {"Step": 1}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 1, "pid": 225798, "tid": 1, "ts": 119890.567993},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 1, "pid": 225798, "tid": 1, "ts": 119890.567993, "args": {"Step": 1}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 1, "pid": 225798, "tid": 1, "ts": 120121.129974},
{"name": "RecvWait", "cat": "NET", "ph": "b", "id": 2, "pid": 225798, "tid": 1, "ts": 119753.023987, "args": {"Step": 2}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 2, "pid": 225798, "tid": 1, "ts": 120038.847992},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 2, "pid": 225798, "tid": 1, "ts": 120038.847992, "args": {"Step": 2}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 2, "pid": 225798, "tid": 1, "ts": 120085.685974},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 2, "pid": 225798, "tid": 1, "ts": 120085.685974, "args": {"Step": 2}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 2, "pid": 225798, "tid": 1, "ts": 120121.244995},
{"name": "RecvWait", "cat": "NET", "ph": "b", "id": 3, "pid": 225798, "tid": 1, "ts": 119772.510986, "args": {"Step": 3}},
{"name": "RecvWait", "cat": "NET", "ph": "e", "id": 3, "pid": 225798, "tid": 1, "ts": 120062.944977},
{"name": "RecvFlushWait", "cat": "NET", "ph": "b", "id": 3, "pid": 225798, "tid": 1, "ts": 120062.944977, "args": {"Step": 3}},
{"name": "RecvFlushWait", "cat": "NET", "ph": "e", "id": 3, "pid": 225798, "tid": 1, "ts": 120101.089996},
{"name": "RecvGpuWait", "cat": "NET", "ph": "b", "id": 3, "pid": 225798, "tid": 1, "ts": 120101.089996, "args": {"Step": 3}},
{"name": "RecvGpuWait", "cat": "NET", "ph": "e", "id": 3, "pid": 225798, "tid": 1, "ts": 120165.115997},
{"name": "ProgressRecv", "cat": "PROXY", "ph": "e", "id": 0, "pid": 225798, "tid": 1, "ts": 120165.356995},
{"name": "ScheduleSend", "cat": "PROXY", "ph": "b", "id": 1, "pid": 225798, "tid": 1, "ts": 119656.950989, "args": {"Channel": 0, "Peer": 1, "Steps": 4, "ChunkSize": 4194304, "transSize": 524288}},
{"name": "ScheduleSend", "cat": "PROXY", "ph": "e", "id": 1, "pid": 225798, "tid": 1, "ts": 119709.078979},
{"name": "ProgressSend", "cat": "PROXY", "ph": "b", "id": 1, "pid": 225798, "tid": 1, "ts": 119709.078979, "args": {"Channel": 0, "Peer": 1, "Steps": 4, "ChunkSize": 4194304, "transSize": 524288}},
{"name": "SendGpuWait", "cat": "NET", "ph": "b", "id": 4, "pid": 225798, "tid": 1, "ts": 119710.632996, "args": {"Step": 0}},
{"name": "SendGpuWait", "cat": "NET", "ph": "e", "id": 4, "pid": 225798, "tid": 1, "ts": 119808.636993},
{"name": "SendPeerWait", "cat": "NET", "ph": "b", "id": 4, "pid": 225798, "tid": 1, "ts": 119808.636993, "args": {"Step": 0}},
{"name": "SendPeerWait", "cat": "NET", "ph": "e", "id": 4, "pid": 225798, "tid": 1, "ts": 119818.972992},
... [ trace truncated for brevity ]
{"name": "AllReduce", "cat": "COLL", "ph": "e", "id": 0, "pid": 4157654, "tid": 1, "ts": 772209.317383},
{"name": "Group", "cat": "GROUP", "ph": "e", "id": 0, "pid": 4157654, "tid": 1, "ts": 772209.418945},
{"name": "AllReduce", "cat": "COLL", "ph": "e", "id": 17, "pid": 225798, "tid": 1, "ts": 170633.535980},
{"name": "AllReduce", "cat": "COLL_API", "ph": "e", "id": 17, "pid": 225798, "tid": 1, "ts": 170582.923981},
{"name": "Group API", "cat": "GROUP_API", "ph": "e", "id": 17, "pid": 225798, "tid": 1, "ts": 170637.582001},
{}]
```
Details about the fields used in the trace can be found at this link:
https://docs.google.com/document/d/1CvAClvFfyA5R-PhYUmn5OOQtYMH4h6I0nSsKchNAySU/preview?tab=t.0#heading=h.yr4qxyxotyw
The trace above is obtained by running a `ncclAllReduce` operation on 8 GPUs, communicating with each other through
The trace above is obtained by running a `ncclAllReduce` operation on 2 GPUs, communicating with each other through
the network interface. The `Group` event encloses all traces that are related to the single `ncclAllReduce` call.
(Note that for single collective invocations, where there are no explicit group calls, NCCL creates a group with only
one collective and this is what is presented in the traces above).
@@ -161,38 +145,17 @@ The `AllReduce` entry presents information about the `ncclAllReduce` operation.
- datatype : NCCL datatype
- algorithm : algorithm used to process the ncclAllReduce
- protocol : protocol used to process the ncclAllReduce
- nMaxChannels: max number of channels used to process the ncclAllReduce
- nChannels : Number of channels used to process the ncclAllReduce
If the proxy events are not active (e.g., the `ncclAllReduce` is intranode) the end timestamp will match the time
consumed by the CPU to launch the collective. For more details refer to `ext-profiler/README.md`, section `Profiling
of collective and p2p operations`.
### Proxy Send
The `Send` entry presents information about the `ProxyOp` processing in the progress thread. It contains the following
info in the args field:
- Channel : id of the channel used by this proxy operation to send data to the peer
- Peer : peer rank
- Steps : number of network steps required to transfer transSize bytes to the peer
- ChunkSize : chunk size used by NCCL to pipeline data through the proxy thread
- transSize : bytes transferred across the channel by this proxy operation
- POSTED : struct containing the number of buffer posts to the GPU and the time stamp for the last post
- REM_FIFO_WAIT: struct containing the number of remote buffer waits and the time stamp for the last wait
- TRANSMITTED : struct containing the number of network sends and the time stamp of the last send
- DONE : struct containing the number of network sends completed and the time stamp of the last send completed
In case of a network problem the POSTED, REM_FIFO_WAIT, TRANSMITTED and DONE might all have partially updated steps,
which could help identify at which point the network problem occurred.
The Proxy send trace gives a summary of the proxy progress thread activity for the channel. If more details are
needed, these can be obtained by enabling the proxy step event (`ncclProfileProxyStep`). In which case the trace
entries below are also reported by the profiler.
#### Proxy SendBufferWait
Presents, for every network step, the time the CPU proxy spends waiting for the channel staging buffer to become available.
#### Proxy SendGPUWait
#### Proxy SendGpuWait
Presents, for every network step, the time the CPU proxy spends waiting for the GPU to provide the data in the staging
buffer.
@@ -201,31 +164,6 @@ buffer.
Presents, for every network step, the time the CPU proxy spends waiting for the `isend` to complete
### Proxy Recv
The `Recv` entry presents information about the `ProxyOp` processing in the progress thread. It contains the following
info in the args field:
- Channel : id of the channel used by this proxy operation to recv data from the peer
- Peer : peer rank
- Steps : number of network steps required to transfer transSize bytes from the peer
- ChunkSize : chunk size used by NCCL to pipeline data through the proxy thread
- transSize : bytes transferred across the channel by this proxy operation
- POSTED : struct containing the number of recvs posted and the time stamp for the last recv posted
- RECEIVED : struct containing the number of recvs completed and the time stamp for the last recv completed
- TRANSMITTED: struct containing the number of recvs flushed to the GPU memory and the time stamp for the last recv flushed
- DONE : struct containing the number of flush completed and the time stamp for the last flush completed
The Proxy Recv trace gives a summary of the proxy progress thread activity for the channel. If more details are
needed, these can be obtained by enabling the proxy step event (`ncclProfileProxyStep`). In which case the trace
entries below are also reported by the profiler.
#### Proxy RecvBufferWait
Presents, for every network step, the time the CPU proxy spends waiting for the staging buffer for the channel to
become available.
#### Proxy RecvWait
Presents, for every network step, the time the CPU proxy spends waiting for a posted `irecv` to complete
@@ -234,6 +172,6 @@ Presents, for every network step, the time the CPU proxy spends waiting for a po
Presents, for every network step, the time the CPU proxy spends waitng for the recv data to be flushed to the GPU
#### Proxy RecvGPUWait
#### Proxy RecvGpuWait
Presents, for every network step, the time the CPU proxy spends waiting for the GPU to consume the recv data
projects/rccl/ext-profiler/example/event.c
-30
查看文件
@@ -1,30 +0,0 @@
/*************************************************************************
* Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
*
* See LICENSE.txt for license information
************************************************************************/
#include <stdio.h>
#include "event.h"
int taskEventQueueEmpty(struct group* g) {
return g->eventHead == NULL;
}
void taskEventQueueEnqueue(struct group* g, struct taskEventBase* event) {
event->next = NULL;
if (g->eventHead) g->eventTail->next = event;
else g->eventHead = event;
g->eventTail = event;
}
struct taskEventBase* taskEventQueueHead(struct group* g) {
return g->eventHead;
}
struct taskEventBase* taskEventQueueDequeue(struct group* g) {
struct taskEventBase* tmp = g->eventHead;
g->eventHead = g->eventHead->next;
if (g->eventHead == NULL) g->eventTail = NULL;
return tmp;
}
projects/rccl/ext-profiler/example/event.h
+142 -13
查看文件
@@ -10,10 +10,14 @@
#include <sys/types.h>
#include <stdint.h>
#include <unistd.h>
#include <cstring>
#include "err.h"
#include "profiler.h"
#include "queue.h"
#include <cuda_runtime.h>
#define MAX_CHANNELS 128 // Match RCCL's MAXCHANNELS
#define MAX_STEPS 16
#define MAX_STEPS 1024
#define MAX_OPS 16 // Up to 64K ranks for PAT
#define MAX_EVENTS_PER_REQ (8)
@@ -21,7 +25,7 @@ struct proxyOp;
struct proxyStep;
struct netPlugin {
uint8_t type;
uint64_t type;
int pluginType;
int pluginVer;
uint8_t pluginEvent;
@@ -63,7 +67,7 @@ struct kernelCh {
#define PROXY_STEP_MAX_STATES 3
struct proxyStep {
uint8_t type; // type of event: network transfer
uint64_t type; // type of event: network transfer
int state;
int step; // network transfer id in given channel
int isSend; // send/recv channel operation
@@ -76,7 +80,7 @@ struct proxyStep {
};
struct proxyOp {
uint8_t type; // type of event: proxy operation
uint64_t type; // type of event: proxy operation
uint8_t channelId; // channel id for this proxy operation
pid_t pid;
int rank;
@@ -97,7 +101,7 @@ struct group;
struct context;
struct proxyCtrl {
uint8_t type;
uint64_t type;
struct context* ctx; // profiler context
double startTs;
double stopTs;
@@ -107,12 +111,12 @@ struct proxyCtrl {
// task level event base structure
struct taskEventBase {
uint8_t type; // event type: collective/p2p
uint64_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
void* parent; // parent API event
struct taskEventBase* next; // next top level event
double startTs;
double stopTs;
};
@@ -147,7 +151,7 @@ struct p2p {
};
struct group {
uint8_t type;
uint64_t type;
struct context* ctx; // profiler context
int groupId;
int refCount;
@@ -158,6 +162,70 @@ struct group {
struct group* next; // next group event in queue
};
struct collApi {
uint64_t type;
struct groupApi* parent;
struct context* ctx; // profiler context
int collApiId;
int refCount;
cudaStream_t stream;
const char* func;
size_t count;
const char* datatype;
int root;
bool graphCaptured;
struct taskEventBase* eventHead; // queue head for task events
struct taskEventBase* eventTail; // queue tail for task events
double startTs;
double stopTs;
struct collApi* next;
};
struct p2pApi {
uint64_t type;
struct groupApi* parent;
struct context* ctx; // profiler context
int p2pApiId;
int refCount;
const char* func;
cudaStream_t stream;
size_t count;
const char* datatype;
bool graphCaptured;
struct taskEventBase* eventHead; // queue head for task events
struct taskEventBase* eventTail; // queue tail for task events
double startTs;
double stopTs;
struct p2pApi* next;
};
struct kernelLaunch {
uint64_t type;
struct groupApi* parent;
cudaStream_t stream;
int kernelLaunchId;
double startTs;
double stopTs;
struct kernelLaunch* next;
};
struct groupApi {
uint64_t type;
struct context* ctx;
int groupApiId;
int refCount;
bool graphCaptured;
int groupDepth;
struct profilerQueue<struct p2pApi, &p2pApi::next> p2pApiEvents;
struct profilerQueue<struct collApi, &collApi::next> collApiEvents;
struct profilerQueue<struct kernelLaunch, &kernelLaunch::next> kernelLaunchEvents;
double endOfncclGroupStartTs;
double startOfncclGroupEndTs;
double startTs;
double stopTs;
struct groupApi* next;
};
// arrays for different event objects
struct context {
const char* commName;
@@ -165,6 +233,26 @@ struct context {
int nranks;
int rank;
int groupApiPoolSize;
int groupApiPoolBase;
int groupApiPoolIndex;
struct groupApi* groupApiPool;
int collApiPoolSize;
int collApiPoolBase;
int collApiPoolIndex;
struct collApi* collApiPool;
int p2pApiPoolSize;
int p2pApiPoolBase;
int p2pApiPoolIndex;
struct p2pApi* p2pApiPool;
int kernelLaunchPoolSize;
int kernelLaunchPoolBase;
int kernelLaunchPoolIndex;
struct kernelLaunch* kernelLaunchPool;
int groupPoolSize;
int groupPoolBase;
int groupPoolIndex;
@@ -186,9 +274,50 @@ struct context {
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);
template <typename T>
inline int taskEventQueueEmpty(T *obj) {
return obj->eventHead == NULL;
}
template <typename T>
inline void taskEventQueueEnqueue(T* obj, struct taskEventBase* event) {
event->next = NULL;
if (obj->eventHead) obj->eventTail->next = event;
else obj->eventHead = event;
obj->eventTail = event;
}
template <typename T>
inline struct taskEventBase* taskEventQueueHead(T *obj) {
return obj->eventHead;
}
template <typename T>
inline struct taskEventBase* taskEventQueueDequeue(T* obj) {
struct taskEventBase* tmp = obj->eventHead;
obj->eventHead = obj->eventHead->next;
if (obj->eventHead == NULL) obj->eventTail = NULL;
return tmp;
}
template <typename T>
inline void resetTaskEvents(T *obj, struct context* ctx) {
while (!taskEventQueueEmpty(obj)) {
struct taskEventBase* base = taskEventQueueDequeue(obj);
if (base->type == ncclProfileColl) {
struct collective* c = (struct collective *)base;
// reset event proxyOps & proxySteps
memset(c->nProxyOps, 0, sizeof(int)*MAX_CHANNELS);
// release collective events in the group and return them to the collective pool
__atomic_fetch_add(&ctx->collPoolBase, 1, __ATOMIC_RELAXED);
} else if (base->type == ncclProfileP2p) {
struct p2p* p = (struct p2p *)base;
// reset event proxyOp and proxySteps
memset(&p->op, 0, sizeof(struct proxyOp)*MAX_CHANNELS);
// release p2p events in the group and return them to the p2p pool
__atomic_fetch_add(&ctx->p2pPoolBase, 1, __ATOMIC_RELAXED);
}
}
}
#endif
projects/rccl/ext-profiler/example/nccl/profiler.h
+21 -12
查看文件
@@ -11,17 +11,20 @@
#include <stdlib.h>
#include "common.h"
#include "err.h"
enum {
ncclProfileGroup = (1 << 0), // group event type
ncclProfileColl = (1 << 1), // host collective call event type
ncclProfileP2p = (1 << 2), // host point-to-point call event type
ncclProfileProxyOp = (1 << 3), // proxy operation event type
ncclProfileProxyStep = (1 << 4), // proxy step event type
ncclProfileProxyCtrl = (1 << 5), // proxy control event type
ncclProfileKernelCh = (1 << 6), // kernel channel event type
ncclProfileNetPlugin = (1 << 7), // network plugin-defined, events
ncclProfileGroup = (1 << 0), // group event type
ncclProfileColl = (1 << 1), // host collective call event type
ncclProfileP2p = (1 << 2), // host point-to-point call event type
ncclProfileProxyOp = (1 << 3), // proxy operation event type
ncclProfileProxyStep = (1 << 4), // proxy step event type
ncclProfileProxyCtrl = (1 << 5), // proxy control event type
ncclProfileKernelCh = (1 << 6), // kernel channel event type
ncclProfileNetPlugin = (1 << 7), // network plugin-defined, events
ncclProfileGroupApi = (1 << 8), // Group API events
ncclProfileCollApi = (1 << 9), // Collective API events
ncclProfileP2pApi = (1 << 10), // Point-to-Point API events
ncclProfileKernelLaunch = (1 << 11), // Kernel launch events
};
typedef enum {
@@ -56,21 +59,27 @@ typedef enum {
/* Kernel event states */
ncclProfilerKernelChStop = 22,
/* Group API States */
ncclProfilerEndGroupApiStart = 23,
ncclProfilerBeginGroupApiEnd = 24
} ncclProfilerEventState_t;
typedef ncclProfilerEventState_t ncclProfilerEventState_v1_t;
typedef ncclProfilerEventState_t ncclProfilerEventState_v2_t;
typedef ncclProfilerEventState_t ncclProfilerEventState_v3_t;
typedef ncclProfilerEventState_t ncclProfilerEventState_v4_t;
typedef ncclProfilerEventState_t ncclProfilerEventState_v5_t;
#include "profiler_v5.h"
#include "profiler_v4.h"
#include "profiler_v3.h"
#include "profiler_v2.h"
#include "profiler_v1.h"
#include "profiler_net.h"
typedef ncclProfiler_v4_t ncclProfiler_t;
typedef ncclProfilerEventDescr_v4_t ncclProfilerEventDescr_t;
typedef ncclProfilerEventStateArgs_v4_t ncclProfilerEventStateArgs_t;
typedef ncclProfiler_v5_t ncclProfiler_t;
typedef ncclProfilerEventDescr_v5_t ncclProfilerEventDescr_t;
typedef ncclProfilerEventStateArgs_v5_t ncclProfilerEventStateArgs_t;
#endif // end include guard
projects/rccl/ext-profiler/example/nccl/profiler_v5.h
+152
查看文件
@@ -0,0 +1,152 @@
/*************************************************************************
* Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
*
* See LICENSE.txt for license information
************************************************************************/
#ifndef PROFILER_V5_H_
#define PROFILER_V5_H_
#include <stdbool.h>
typedef struct {
uint64_t type; // event type descriptor: ncclProfileGroupApi, ...
void* parentObj; // pointer to the profiler parent object
int rank; // originating rank
union {
struct {
int graphCaptured;
int groupDepth;
} groupApi;
struct {
const char* func;
size_t count;
const char* datatype;
int root;
void* stream;
bool graphCaptured;
} collApi;
struct {
const char* func;
size_t count;
const char* datatype;
void* stream;
bool graphCaptured;
} p2pApi;
struct {
void* stream;
} kernelLaunch;
struct {
uint64_t seqNumber;
const char* func;
void const* sendBuff;
void* recvBuff;
size_t count;
int root;
const char* datatype;
uint8_t nChannels;
uint8_t nWarps;
const char* algo;
const char* proto;
void* parentGroup; // for backward compatibility with v4
} coll;
struct {
const char* func;
void* buff;
const char* datatype;
size_t count;
int peer;
uint8_t nChannels;
void* parentGroup; // for backward compatibility with v4
} p2p;
struct {
pid_t pid; // pid of the originating process
uint8_t channelId; // channel id for this proxy operation
int peer; // remote rank for send/recv
int nSteps; // number of steps for this proxy operation
int chunkSize; // amount of data transferred by this proxy operation
int isSend;
} proxyOp;
struct {
int step;
} proxyStep;
struct {
uint8_t channelId;
uint64_t pTimer; // start timestamp from GPU globaltimer
} kernelCh;
struct {
int64_t id;
void* data;
} netPlugin;
};
} ncclProfilerEventDescr_v5_t;
typedef union {
struct {
size_t transSize;
} proxyStep;
struct {
int appendedProxyOps;
} proxyCtrl;
struct {
void* data;
} netPlugin;
struct {
uint64_t pTimer;
} kernelCh;
} ncclProfilerEventStateArgs_v5_t;
typedef struct {
const char* name;
// init - initialize the profiler plugin
// Input
// - context : opaque profiler context object for separating profiler behavior across comms
// - commId : communicator id
// - commName : user assigned communicator name
// - nNodes : number of nodes in communicator
// - nranks : number of ranks in communicator
// - rank : rank identifier in communicator
// - logfn : logger function
// Output
// - eActivationMask: bitmask of active events set by the plugin
ncclResult_t (*init)(void** context, uint64_t commId, int* eActivationMask, const char* commName, int nNodes, int nranks, int rank, ncclDebugLogger_t logfn);
// startEvent - initialize and start a new event for the supplied event descriptor inside the eventset
// Input
// - context: opaque profiler context object
// - eDescr : pointer to ncclProfilerEventDescr_t object
// Output
// - eHandle: return event handle for supplied event descriptor object
ncclResult_t (*startEvent)(void* context, void** eHandle, ncclProfilerEventDescr_v5_t* eDescr);
// stopEvent - stop/finalize an event inside and event set
// Input
// - eHandle: handle to event object
ncclResult_t (*stopEvent)(void* eHandle);
// recordEventState - record event state transitions and event attribute updates
// Input
// - eHandle : handle to event object created through startEvent
// - eStateArgs: optional argument used to capture event attribute updates associated with the state transition
// - eState : event state transition
ncclResult_t (*recordEventState)(void* eHandle, ncclProfilerEventState_v5_t eState, ncclProfilerEventStateArgs_v5_t* eStateArgs);
// finalize - finalize the profiler plugin
// Input
// - context: opaque profiler context object
ncclResult_t (*finalize)(void* context);
} ncclProfiler_v5_t;
#endif
projects/rccl/ext-profiler/example/plugin.c → projects/rccl/ext-profiler/example/plugin.cc
+230 -34
查看文件
@@ -6,7 +6,7 @@
#include <stdio.h>
#include <pthread.h>
#include <string.h>
#include <cstring>
#include <linux/limits.h>
#include <sys/time.h>
#include <sys/types.h>
@@ -22,12 +22,20 @@ static int initialized; // initialization counter for profiler
static double startTime; // profiler start time
static const int defaultEActivationMask = ncclProfileColl | ncclProfileP2p;
static const int defaultGroupPoolSize = 16;
static const int defaultCollPoolSize = 16;
static const int defaultP2pPoolSize = 1024;
static const int defaultGroupApiPoolSize = 256;
static const int defaultCollApiPoolSize = 256;
static const int defaultP2pApiPoolSize = 256;
static const int defaultKernelLaunchPoolSize = 256;
static const int defaultGroupPoolSize = 256;
static const int defaultCollPoolSize = 256;
static const int defaultP2pPoolSize = 256;
static const int defaultProxyCtrlPoolSize = 16;
static const int defaultDetachPoolSize = 128;
static const int defaultDetachPoolSize = 256;
static int groupApiPoolSize;
static int collApiPoolSize;
static int p2pApiPoolSize;
static int kernelLaunchPoolSize;
static int groupPoolSize;
static int collPoolSize;
static int p2pPoolSize;
@@ -51,7 +59,7 @@ static pthread_mutex_t lock = PTHREAD_MUTEX_INITIALIZER;
static pid_t pid;
static int* eActivationMaskPtr;
__hidden ncclResult_t exampleProfilerInit(void** context, int* eActivationMask, const char* commName, uint64_t commHash, int nNodes, int nranks, int rank, ncclDebugLogger_t logfn) {
__hidden ncclResult_t exampleProfilerInit(void** context, uint64_t commId, int* eActivationMask, const char* commName, int nNodes, int nranks, int rank, ncclDebugLogger_t logfn) {
pthread_mutex_lock(&lock);
if (__atomic_fetch_add(&initialized, 1, __ATOMIC_RELAXED) == 0) {
// first thread initializes event mask, environment and detach pool
@@ -59,6 +67,18 @@ __hidden ncclResult_t exampleProfilerInit(void** context, int* eActivationMask,
str = getenv("NCCL_PROFILE_EVENT_MASK");
__atomic_store_n(eActivationMask, str ? atoi(str) : 0, __ATOMIC_RELAXED);
str = getenv("NCCL_PROFILE_GROUP_API_POOL_SIZE");
groupApiPoolSize = str ? atoi(str) : defaultGroupApiPoolSize;
str = getenv("NCCL_PROFILE_COLL_API_POOL_SIZE");
collApiPoolSize = str ? atoi(str) : defaultCollApiPoolSize;
str = getenv("NCCL_PROFILE_P2P_API_POOL_SIZE");
p2pApiPoolSize = str ? atoi(str) : defaultP2pApiPoolSize;
str = getenv("NCCL_PROFILE_KERNEL_LAUNCH_POOL_SIZE");
kernelLaunchPoolSize = str ? atoi(str) : defaultKernelLaunchPoolSize;
str = getenv("NCCL_PROFILE_GROUP_POOL_SIZE");
groupPoolSize = str ? atoi(str) : defaultGroupPoolSize;
@@ -96,11 +116,23 @@ __hidden ncclResult_t exampleProfilerInit(void** context, int* eActivationMask,
// pre-allocate memory for event object pools in dedicated profiler context
struct context* ctx = (struct context *)calloc(1, sizeof(*ctx));
ctx->commName = commName;
ctx->commHash = commHash;
ctx->commHash = commId;
ctx->nranks = nranks;
ctx->rank = rank;
logFn = logfn;
INFO(NCCL_INIT, "PROFILER/Plugin: init commName: %s commHash: %lu nranks: %d rank: %d", commName ? commName : "", commHash, nranks, rank);
INFO(NCCL_INIT, "PROFILER/Plugin: init commName: %s commHash: %lu nranks: %d rank: %d", commName ? commName : "", commId, nranks, rank);
ctx->groupApiPool = (struct groupApi *)calloc(groupApiPoolSize, sizeof(*ctx->groupApiPool));
if (ctx->groupApiPool == NULL) goto fail;
ctx->collApiPool = (struct collApi *)calloc(collApiPoolSize, sizeof(*ctx->collApiPool));
if (ctx->collApiPool == NULL) goto fail;
ctx->p2pApiPool = (struct p2pApi *)calloc(p2pApiPoolSize, sizeof(*ctx->p2pApiPool));
if (ctx->p2pApiPool == NULL) goto fail;
ctx->kernelLaunchPool = (struct kernelLaunch *)calloc(kernelLaunchPoolSize, sizeof(*ctx->kernelLaunchPool));
if (ctx->kernelLaunchPool == NULL) goto fail;
ctx->groupPool = (struct group *)calloc(groupPoolSize, sizeof(*ctx->groupPool));
if (ctx->groupPool == NULL) goto fail;
@@ -130,16 +162,22 @@ fail:
if (ctx->p2pPool) free(ctx->p2pPool);
if (ctx->collPool) free(ctx->collPool);
if (ctx->groupPool) free(ctx->groupPool);
if (ctx->collApiPool) free(ctx->collApiPool);
if (ctx->p2pApiPool) free(ctx->p2pApiPool);
if (ctx->kernelLaunchPool) free(ctx->kernelLaunchPool);
if (ctx->groupApiPool) free(ctx->groupApiPool);
free(ctx);
if (detachPool) free(detachPool);
return ncclSystemError;
}
static const char* profilerDumpFile;
__hidden ncclResult_t exampleProfilerFinalize(void* context) {
FILE* fh = NULL;
char filename[PATH_MAX] = { 0 };
struct context* ctx = (struct context *)context;
const char* dump = getenv("NCCL_PROFILE_DUMP_FILE");
const char* dump = profilerDumpFile ? profilerDumpFile : getenv("NCCL_PROFILE_DUMP_FILE");
if (dump) {
sprintf(filename, "%s_%lu_%d.json", dump, ctx->commHash, ctx->rank);
fh = fopen(filename, "w");
@@ -148,10 +186,12 @@ __hidden ncclResult_t exampleProfilerFinalize(void* context) {
INFO(NCCL_INIT, "PROFILER/Plugin: finalize commName: %s commHash: %lu nranks: %d rank: %d", ctx->commName ? ctx->commName : "", ctx->commHash, ctx->nranks, ctx->rank);
// print last N groups/collectives/p2ps
int start = (ctx->groupPoolIndex - groupPoolSize >= 0) ? ctx->groupPoolIndex - groupPoolSize : 0;
int end = ctx->groupPoolIndex;
// Note that since the v5 version of the profiler, group API events are now at the top of the hierarchy.
// Legacy Group events from v4 are still emitted for compatibility purposes when using the v4 profiler but excluded from this example.
int start = (ctx->groupApiPoolIndex - groupApiPoolSize >= 0) ? ctx->groupApiPoolIndex - groupApiPoolSize : 0;
int end = ctx->groupApiPoolIndex;
for (int i = start; i < end; i++) {
printEvent(fh, &ctx->groupPool[i%groupPoolSize]);
printEvent(fh, &ctx->groupApiPool[i%groupApiPoolSize]);
}
start = (ctx->proxyCtrlPoolIndex - proxyCtrlPoolSize >= 0) ? ctx->proxyCtrlPoolIndex - proxyCtrlPoolSize : 0;
@@ -161,6 +201,10 @@ __hidden ncclResult_t exampleProfilerFinalize(void* context) {
}
free(ctx->groupPool);
free(ctx->collApiPool);
free(ctx->p2pApiPool);
free(ctx->kernelLaunchPool);
free(ctx->groupApiPool);
free(ctx->collPool);
free(ctx->p2pPool);
free(ctx->proxyCtrlPool);
@@ -187,7 +231,113 @@ __hidden void updateEvent(void* handle);
__hidden ncclResult_t exampleProfilerStartEvent(void* context, void** eHandle, ncclProfilerEventDescr_t* eDescr) {
*eHandle = NULL;
struct context* ctx = (struct context *)context;
if (eDescr->type == ncclProfileGroup) {
if (eDescr->type == ncclProfileGroupApi) {
struct groupApi* event;
int groupApiId = __atomic_fetch_add(&ctx->groupApiPoolIndex, 1, __ATOMIC_RELAXED);
if ((groupApiId - __atomic_load_n(&ctx->groupApiPoolBase, __ATOMIC_RELAXED)) < groupApiPoolSize) {
// if there are available group API events grab one
event = &ctx->groupApiPool[groupApiId%groupApiPoolSize];
// Make sure all child events of the picked group API event are cleared
while (!profilerQueueEmpty(&event->collApiEvents)) {
struct collApi *collApiEvent = profilerQueueDequeue(&event->collApiEvents);
resetTaskEvents(collApiEvent, ctx);
__atomic_fetch_add(&ctx->collApiPoolBase, 1, __ATOMIC_RELAXED);
}
while (!profilerQueueEmpty(&event->p2pApiEvents)) {
struct p2pApi *p2pApiEvent = profilerQueueDequeue(&event->p2pApiEvents);
resetTaskEvents(p2pApiEvent, ctx);
__atomic_fetch_add(&ctx->p2pApiPoolBase, 1, __ATOMIC_RELAXED);
}
while (!profilerQueueEmpty(&event->kernelLaunchEvents)) {
profilerQueueDequeue(&event->kernelLaunchEvents);
__atomic_fetch_add(&ctx->kernelLaunchPoolBase, 1, __ATOMIC_RELAXED);
}
} else {
// else drop this event
__atomic_fetch_sub(&ctx->groupApiPoolIndex, 1, __ATOMIC_RELAXED);
return ncclSuccess;
}
event->type = ncclProfileGroupApi;
event->ctx = ctx;
event->groupApiId = groupApiId;
event->graphCaptured = eDescr->groupApi.graphCaptured;
event->groupDepth = eDescr->groupApi.groupDepth;
event->startTs = gettime() - startTime;
*eHandle = event;
} else if (eDescr->type == ncclProfileCollApi) {
if (eDescr->parentObj == NULL) return ncclSuccess;
struct collApi* event;
int collApiId = __atomic_fetch_add(&ctx->collApiPoolIndex, 1, __ATOMIC_RELAXED);
if ((collApiId - __atomic_load_n(&ctx->collApiPoolBase, __ATOMIC_RELAXED)) < collApiPoolSize) {
// if there are available Coll API events grab one
event = &ctx->collApiPool[collApiId%collApiPoolSize];
resetTaskEvents(event, ctx);
} else {
// else drop this event
__atomic_fetch_sub(&ctx->collApiPoolIndex, 1, __ATOMIC_RELAXED);
return ncclSuccess;
}
event->type = ncclProfileCollApi;
event->collApiId = collApiId;
event->ctx = ctx;
event->func = eDescr->collApi.func;
event->stream = (cudaStream_t) eDescr->collApi.stream;
event->count = eDescr->collApi.count;
event->datatype = eDescr->collApi.datatype;
event->root = eDescr->collApi.root;
event->graphCaptured = eDescr->collApi.graphCaptured;
struct groupApi* parent = (struct groupApi *) eDescr->parentObj;
event->parent = parent;
profilerQueueEnqueue(&parent->collApiEvents, event);
__atomic_fetch_add(&parent->refCount, 1, __ATOMIC_RELAXED);
*eHandle = event;
} else if (eDescr->type == ncclProfileP2pApi) {
if (eDescr->parentObj == NULL) return ncclSuccess;
struct p2pApi* event;
int p2pApiId = __atomic_fetch_add(&ctx->p2pApiPoolIndex, 1, __ATOMIC_RELAXED);
if ((p2pApiId - __atomic_load_n(&ctx->p2pApiPoolBase, __ATOMIC_RELAXED)) < p2pApiPoolSize) {
// if there are available p2p API events grab one
event = &ctx->p2pApiPool[p2pApiId%p2pApiPoolSize];
resetTaskEvents(event, ctx);
} else {
// else drop this event
__atomic_fetch_sub(&ctx->p2pApiPoolIndex, 1, __ATOMIC_RELAXED);
return ncclSuccess;
}
event->type = ncclProfileP2pApi;
event->p2pApiId = p2pApiId;
event->ctx = ctx;
event->func = eDescr->p2pApi.func;
event->stream = (cudaStream_t) eDescr->p2pApi.stream;
event->count = eDescr->p2pApi.count;
event->datatype = eDescr->p2pApi.datatype;
event->graphCaptured = eDescr->p2pApi.graphCaptured;
struct groupApi* parent = (struct groupApi *) eDescr->parentObj;
event->parent = parent;
profilerQueueEnqueue(&parent->p2pApiEvents, event);
__atomic_fetch_add(&parent->refCount, 1, __ATOMIC_RELAXED);
*eHandle = event;
} else if (eDescr->type == ncclProfileKernelLaunch) {
if (eDescr->parentObj == NULL) return ncclSuccess;
struct kernelLaunch* event;
int kernelLaunchId = __atomic_fetch_add(&ctx->kernelLaunchPoolIndex, 1, __ATOMIC_RELAXED);
if ((kernelLaunchId - __atomic_load_n(&ctx->kernelLaunchPoolBase, __ATOMIC_RELAXED)) < kernelLaunchPoolSize) {
// if there are available kernel API events grab one
event = &ctx->kernelLaunchPool[kernelLaunchId%kernelLaunchPoolSize];
} else {
// else drop this event
__atomic_fetch_sub(&ctx->kernelLaunchPoolIndex, 1, __ATOMIC_RELAXED);
return ncclSuccess;
}
event->type = ncclProfileKernelLaunch;
event->stream = (cudaStream_t) eDescr->kernelLaunch.stream;
struct groupApi* parent = (struct groupApi *) eDescr->parentObj;
event->parent = parent;
profilerQueueEnqueue(&parent->kernelLaunchEvents, event);
__atomic_fetch_add(&parent->refCount, 1, __ATOMIC_RELAXED);
*eHandle = event;
} else if (eDescr->type == ncclProfileGroup) {
if (eDescr->parentObj == NULL) return ncclSuccess;
struct group* event;
int groupId = __atomic_fetch_add(&ctx->groupPoolIndex, 1, __ATOMIC_RELAXED);
if ((groupId - __atomic_load_n(&ctx->groupPoolBase, __ATOMIC_RELAXED)) < groupPoolSize) {
@@ -222,7 +372,7 @@ __hidden ncclResult_t exampleProfilerStartEvent(void* context, void** eHandle, n
debugEvent(event, "GroupStart");
} else if (eDescr->type == ncclProfileColl) {
// the parent might be null if we run out of events
struct group* parent = (struct group *)eDescr->parentObj;
struct collApi* parent = (struct collApi *)eDescr->parentObj;
if (parent == NULL) return ncclSuccess;
struct collective* event;
@@ -253,12 +403,12 @@ __hidden ncclResult_t exampleProfilerStartEvent(void* context, void** eHandle, n
event->proto = eDescr->coll.proto;
*eHandle = event;
taskEventQueueEnqueue(parent, (struct taskEventBase *)event);
// increment the group ref counter so the event will staty open
// increment the group ref counter so the event will stay open
__atomic_fetch_add(&parent->refCount, 1, __ATOMIC_RELAXED);
debugEvent(event, "CollStart");
} else if (eDescr->type == ncclProfileP2p) {
// the parent might be null if we run out of events
struct group* parent = (struct group *)eDescr->parentObj;
struct p2pApi* parent = (struct p2pApi*) eDescr->parentObj;
if (parent == NULL) return ncclSuccess;
struct p2p* event;
@@ -458,8 +608,34 @@ __hidden ncclResult_t exampleProfilerStartEvent(void* context, void** eHandle, n
}
void updateEvent(void* handle) {
uint8_t type = *(uint8_t *)handle;
if (type == ncclProfileGroup) {
uint64_t type = *(uint64_t *)handle;
if (type == ncclProfileGroupApi) {
struct groupApi* event = (struct groupApi*) handle;
if (__atomic_sub_fetch(&event->refCount, 1, __ATOMIC_RELAXED) == 0) {
event->stopTs = gettime() - startTime;
__atomic_fetch_add(&event->ctx->groupApiPoolBase, 1, __ATOMIC_RELAXED);
}
} else if (type == ncclProfileCollApi) {
struct collApi* event = (struct collApi*) handle;
if (__atomic_sub_fetch(&event->refCount, 1, __ATOMIC_RELAXED) == 0) {
event->stopTs = gettime() - startTime;
__atomic_fetch_add(&event->ctx->collApiPoolBase, 1, __ATOMIC_RELAXED);
}
updateEvent(event->parent);
return;
} else if (type == ncclProfileP2pApi) {
struct p2pApi* event = (struct p2pApi*) handle;
if (__atomic_sub_fetch(&event->refCount, 1, __ATOMIC_RELAXED) == 0) {
event->stopTs = gettime() - startTime;
__atomic_fetch_add(&event->ctx->p2pApiPoolBase, 1, __ATOMIC_RELAXED);
}
updateEvent(event->parent);
event->stopTs = gettime() - startTime;
} else if (type == ncclProfileKernelLaunch) {
struct kernelLaunch* event = (struct kernelLaunch*) handle;
event->stopTs = gettime() - startTime;
updateEvent(event->parent);
} else if (type == ncclProfileGroup) {
struct group* event = (struct group *)handle;
if (__atomic_sub_fetch(&event->refCount, 1, __ATOMIC_RELAXED) == 0) {
event->stopTs = gettime() - startTime;
@@ -527,25 +703,35 @@ __hidden ncclResult_t exampleProfilerStopEvent(void* eHandle) {
// the event handle might be null if we run out of events
if (eHandle == NULL) return ncclSuccess;
uint8_t type = *(uint8_t *)eHandle;
if (type == ncclProfileGroup) {
// stopping the group event in NCCL core does not
// mean the group has completed. It means the group
// was submitted/enqueued so we need to keep the event open
uint64_t type = *(uint64_t *)eHandle;
// Stopping API events, Kernel Launch events, collective/p2p task events
// in NCCL core do not mean that they are complete. It means that the
// operation was enqueued so we need to keep the events open
if (type == ncclProfileGroupApi) {
struct groupApi* event = (struct groupApi*) eHandle;
event->stopTs = gettime() - startTime;
return ncclSuccess;
} else if (type == ncclProfileCollApi) {
struct collApi* event = (struct collApi*) eHandle;
event->stopTs = gettime() - startTime;
return ncclSuccess;
} else if (type == ncclProfileP2pApi) {
struct p2pApi* event = (struct p2pApi*) eHandle;
event->stopTs = gettime() - startTime;
return ncclSuccess;
} else if (type == ncclProfileKernelLaunch) {
struct kernelLaunch* event = (struct kernelLaunch*) eHandle;
event->stopTs = gettime() - startTime;
return ncclSuccess;
} else if (type == ncclProfileGroup) {
struct group* event = (struct group *)eHandle;
event->stopTs = gettime() - startTime;
return ncclSuccess;
} else if (type == ncclProfileColl) {
// stopping the collective event in NCCL core does not
// mean the collective has completed. It means the collective
// was submitted/enqueued so we need to keep the event open
struct collective* event = (struct collective *)eHandle;
event->base.stopTs = gettime() - startTime;
return ncclSuccess;
} else if (type == ncclProfileP2p) {
// stopping the p2p event in NCCL core does not
// mean the p2p has completed. It means the p2p
// was submitted/enqueued so we need to keep the event open
struct p2p* event = (struct p2p *)eHandle;
event->base.stopTs = gettime() - startTime;
return ncclSuccess;
@@ -559,8 +745,15 @@ __hidden ncclResult_t exampleProfilerRecordEventState(void* eHandle, ncclProfile
// the event handle might be null if we run out of events
if (eHandle == NULL) return ncclSuccess;
uint8_t type = *(uint8_t *)eHandle;
if (type == ncclProfileProxyOp) {
uint64_t type = *(uint64_t *)eHandle;
if (type == ncclProfileGroupApi) {
struct groupApi* event = (struct groupApi*) eHandle;
if (eState == ncclProfilerEndGroupApiStart) {
event->endOfncclGroupStartTs = gettime() - startTime;
} else if (eState == ncclProfilerBeginGroupApiEnd) {
event->startOfncclGroupEndTs = gettime() - startTime;
}
} else if (type == ncclProfileProxyOp) {
struct proxyOp* event = (struct proxyOp *)eHandle;
if (eState == ncclProfilerProxyOpInProgress_v4) {
event->progrTs = gettime() - startTime;
@@ -592,6 +785,8 @@ __hidden ncclResult_t exampleProfilerRecordEventState(void* eHandle, ncclProfile
case ncclProfilerProxyStepRecvGPUWait:
event->timestamp[PROXY_STEP_RECV_GPU_WAIT] = gettime() - startTime;
break;
default:
break;
}
} else if (type == ncclProfileProxyCtrl) {
struct proxyCtrl* event = (struct proxyCtrl *)eHandle;
@@ -609,7 +804,7 @@ __hidden ncclResult_t exampleProfilerRecordEventState(void* eHandle, ncclProfile
return ncclSuccess;
}
ncclProfiler_t ncclProfiler_v4 = {
ncclProfiler_t ncclProfiler_v5 = {
"Example-profiler",
exampleProfilerInit,
exampleProfilerStartEvent,
@@ -618,14 +813,15 @@ ncclProfiler_t ncclProfiler_v4 = {
exampleProfilerFinalize,
};
int exampleProfilerStart(int eActivationMask) {
__attribute__((visibility("default"))) int exampleProfilerStart(int eActivationMask, const char* name) {
profilerDumpFile = name;
if (__atomic_load_n(&initialized, __ATOMIC_RELAXED)) {
__atomic_store_n(eActivationMaskPtr, eActivationMask, __ATOMIC_RELAXED);
}
return ncclSuccess;
}
int exampleProfilerStop(void) {
__attribute__((visibility("default"))) int exampleProfilerStop(void) {
if (__atomic_load_n(&initialized, __ATOMIC_RELAXED)) {
__atomic_store_n(eActivationMaskPtr, 0, __ATOMIC_RELAXED);
}
projects/rccl/ext-profiler/example/plugin.h
+3 -2
查看文件
@@ -7,7 +7,8 @@
#ifndef PLUGIN_H_
#define PLUGIN_H_
int exampleProfilerStart(int eActivationMask);
int exampleProfilerStop(void);
__attribute__((visibility("default"))) int exampleProfilerStart(int eActivationMask, const char* name);
__attribute__((visibility("default"))) int exampleProfilerStop(void);
#endif
projects/rccl/ext-profiler/example/print_event.c → projects/rccl/ext-profiler/example/print_event.cc
+93 -6
查看文件
@@ -5,15 +5,59 @@
************************************************************************/
#include <stdio.h>
#include "err.h"
#include "profiler.h"
#include "event.h"
#include "print_event.h"
#include <cuda_runtime.h>
#define __hidden __attribute__ ((visibility("hidden")))
// FIXME: chrome tracing asynchronous events (following used) allow event nesting for events that have same id and category
// It appears that nesting more than three events causes issues. Therefore, every event is given an increasing id and a
// category that matches the type of event (GROUP, COLL, P2P, PROXY, NET)
// category that matches the type of event (GROUP API, COLL API, P2P API, GROUP, COLL, P2P, PROXY, NET)
static __thread int groupApiId;
__hidden void printGroupApiEventHeader(FILE* fh, struct groupApi* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"GROUP_API\", \"ph\": \"b\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f, \"args\": {\"groupApiId\": %d, \"groupDepth\":%d}},\n",
"Group API", groupApiId, getpid(), 1, event->startTs, event->groupApiId, event->groupDepth);
}
__hidden void printGroupApiEventTrailer(FILE* fh, struct groupApi* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"GROUP_API\", \"ph\": \"e\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f},\n",
"Group API", groupApiId++, getpid(), 1, event->stopTs);
}
static __thread int p2pApiId;
__hidden void printP2pApiEventHeader(FILE* fh, struct p2pApi* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"P2P_API\", \"ph\": \"b\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f, \"args\": {\"count\": %lu, \"datatype\": %s, \"GraphCaptured\":%d, \"Stream\": %p}},\n",
event->func, p2pApiId, getpid(), 1, event->startTs, event->count, event->datatype, event->graphCaptured, event->stream);
}
__hidden void printP2pApiEventTrailer(FILE* fh, struct p2pApi* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"P2P_API\", \"ph\": \"e\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f},\n",
event->func, p2pApiId++, getpid(), 1, event->stopTs);
}
static __thread int collApiId;
__hidden void printCollApiEventHeader(FILE* fh, struct collApi* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"COLL_API\", \"ph\": \"b\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f, \"args\": {\"count\": %lu, \"datatype\": %s, \"root\": %d, \"GraphCaptured\":%d, \"Stream\": %p}},\n",
event->func, collApiId, getpid(), 1, event->startTs, event->count, event->datatype, event->root, event->graphCaptured, event->stream);
}
__hidden void printCollApiEventTrailer(FILE* fh, struct collApi* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"COLL_API\", \"ph\": \"e\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f},\n",
event->func, collApiId++, getpid(), 1, event->stopTs);
}
static __thread int kernelLaunchId;
__hidden void printKernelLaunchEventHeader(FILE* fh, struct kernelLaunch* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"KERNEL_LAUNCH\", \"ph\": \"b\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f, \"args\": {\"groupId\": %d, \"Stream\": %p}},\n", "KernelLaunch", kernelLaunchId, getpid(), 1, event->startTs, event->kernelLaunchId, event->stream);
}
__hidden void printKernelLaunchEventTrailer(FILE* fh, struct kernelLaunch* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"KERNEL_LAUNCH\", \"ph\": \"e\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f},\n", "KernelLaunch", kernelLaunchId++, getpid(), 1, event->stopTs);
}
static __thread int groupId;
__hidden void printGroupEventHeader(FILE* fh, struct group* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"GROUP\", \"ph\": \"b\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f, \"args\": {\"groupId\": %d}},\n",
@@ -28,7 +72,7 @@ __hidden void printGroupEventTrailer(FILE* fh, struct group* event) {
static __thread int collId;
__hidden void printCollEventHeader(FILE* fh, struct collective* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"COLL\", \"ph\": \"b\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f, \"args\": {\"SeqNum\": %lu, \"CommHash\": %lu, \"Rank\": %d, \"Count\": %lu, \"Datatype\": \"%s\", \"Algorithm\": \"%s\", \"Protocol\": \"%s\", \"nChannels\": %d}},\n",
event->base.func, collId, getpid(), 1, event->base.startTs, event->seqNumber, event->base.parent->ctx->commHash, event->base.rank, event->count, event->datatype, event->algo, event->proto, event->nChannels);
event->base.func, collId, getpid(), 1, event->base.startTs, event->seqNumber, ((struct collApi*)event->base.parent)->ctx->commHash, event->base.rank, event->count, event->datatype, event->algo, event->proto, event->nChannels);
}
__hidden void printCollEventTrailer(FILE* fh, struct collective* event) {
@@ -39,7 +83,7 @@ __hidden void printCollEventTrailer(FILE* fh, struct collective* event) {
static __thread int p2pId;
__hidden void printP2pEventHeader(FILE* fh, struct p2p* event) {
fprintf(fh, "{\"name\": \"%s\", \"cat\": \"P2P\", \"ph\": \"b\", \"id\": %d, \"pid\": %d, \"tid\": %d, \"ts\": %f, \"args\": {\"CommHash\": %lu, \"Rank\": %d, \"Peer\": %d, \"Count\": %lu, \"Datatype\": \"%s\", \"nChannels\": %d}},\n",
event->base.func, p2pId, getpid(), 1, event->base.startTs, event->base.parent->ctx->commHash, event->base.rank, event->peer, event->count, event->datatype, event->nChannels);
event->base.func, p2pId, getpid(), 1, event->base.startTs, ((struct p2pApi*)event->base.parent)->ctx->commHash, event->base.rank, event->peer, event->count, event->datatype, event->nChannels);
}
__hidden void printP2pEventTrailer(FILE* fh, struct p2p* event) {
@@ -173,7 +217,7 @@ void debugEvent(void* eHandle, const char* tag) {
char filename[64] = { 0 };
sprintf(filename, "EventDebug-%d", getpid());
FILE* fh = fopen(filename, "a+");
uint8_t type = *(uint8_t *)eHandle;
uint64_t type = *(uint64_t *)eHandle;
if (type == ncclProfileGroup) {
struct group* event = (struct group *)eHandle;
fprintf(fh, "Group event %p tag = %s {\n", event, tag);
@@ -241,8 +285,51 @@ void debugEvent(void* eHandle, const char* tag) {
void printEvent(FILE* fh, void* handle) {
if (handle == NULL || fh == NULL) return;
uint8_t type = *(uint8_t *)handle;
if (type == ncclProfileGroup) {
uint64_t type = *(uint64_t *)handle;
if (type == ncclProfileGroupApi) {
struct groupApi* g = (struct groupApi*) handle;
printGroupApiEventHeader(fh, g);
struct kernelLaunch* kernelLaunchHead = profilerQueueHead(&g->kernelLaunchEvents);
while (kernelLaunchHead != NULL) {
printEvent(fh, kernelLaunchHead);
kernelLaunchHead = kernelLaunchHead->next;
}
struct collApi* collApiHead = profilerQueueHead(&g->collApiEvents);
while (collApiHead != NULL) {
printEvent(fh, collApiHead);
collApiHead = collApiHead->next;
}
struct p2pApi* p2pApiHead = profilerQueueHead(&g->p2pApiEvents);
while (p2pApiHead != NULL) {
printEvent(fh, p2pApiHead);
p2pApiHead = p2pApiHead->next;
}
printGroupApiEventTrailer(fh, g);
} else if (type == ncclProfileCollApi) {
struct collApi* collApiEvent = (struct collApi *) handle;
printCollApiEventHeader(fh, collApiEvent);
struct taskEventBase* base = taskEventQueueHead(collApiEvent);
while (base) {
struct taskEventBase* next = base->next;
printEvent(fh, base);
base = next;
}
printCollApiEventTrailer(fh, collApiEvent);
} else if (type == ncclProfileP2pApi) {
struct p2pApi* p2pApiEvent = (struct p2pApi *) handle;
printP2pApiEventHeader(fh, p2pApiEvent);
struct taskEventBase* base = taskEventQueueHead(p2pApiEvent);
while (base) {
struct taskEventBase* next = base->next;
printEvent(fh, base);
base = next;
}
printP2pApiEventTrailer(fh, p2pApiEvent);
} else if (type == ncclProfileKernelLaunch) {
struct kernelLaunch* kernelLaunchEvent = (struct kernelLaunch *) handle;
printKernelLaunchEventHeader(fh, kernelLaunchEvent);
printKernelLaunchEventTrailer(fh, kernelLaunchEvent);
} else if (type == ncclProfileGroup) {
struct group* g = (struct group *)handle;
printGroupEventHeader(fh, g);
struct taskEventBase* base = taskEventQueueHead(g);
projects/rccl/ext-profiler/example/queue.h
+50
查看文件
@@ -0,0 +1,50 @@
/*************************************************************************
* Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved.
*
* See LICENSE.txt for license information
************************************************************************/
#ifndef QUEUE_H
#define QUEUE_H
template<typename T, T *T::*next>
struct profilerQueue {
T *head, *tail;
};
template<typename T, T *T::*next>
inline void profilerQueueConstruct(profilerQueue<T,next> *me) {
me->head = nullptr;
me->tail = nullptr;
}
template<typename T, T *T::*next>
inline bool profilerQueueEmpty(profilerQueue<T,next> *me) {
return me->head == nullptr;
}
template<typename T, T *T::*next>
inline T* profilerQueueHead(profilerQueue<T,next> *me) {
return me->head;
}
template<typename T, T *T::*next>
inline T* profilerQueueTail(profilerQueue<T,next> *me) {
return me->tail;
}
template<typename T, T *T::*next>
inline void profilerQueueEnqueue(profilerQueue<T,next> *me, T *x) {
x->*next = nullptr;
(me->head ? me->tail->*next : me->head) = x;
me->tail = x;
}
template<typename T, T *T::*next>
inline T* profilerQueueDequeue(profilerQueue<T,next> *me) {
T *ans = me->head;
me->head = ans->*next;
if (me->head == nullptr) me->tail = nullptr;
return ans;
}
#endif