feat(primbench): make AMD SMI and NVML soft dependencies by MyNameIsTrez · Pull Request #5637 · ROCm/rocm-libraries

MyNameIsTrez · 2026-03-19T23:01:41Z

Motivation

The PR [rocPRIM] Warn/Prevent rocPRIM benchmarks from being built on Windows was merged as a hotfix, since rocPRIM's benchmarks were not buildable on Windows anymore.

This PR its commit fix(rocprim): revert Windows hotfix PR #5494 reverts that hotfix, and provides a proper fix, by turning AMD SMI and NVML from hard dependencies into soft dependencies.

This PR will be undrafted once the PR feat(primbench): support an extra backend on which it is based has been merged.

Technical Details

Here is the updated Dependencies section of the readme:

Test Plan

VS Code dev containers

CUDA Dockerfile + devcontainer.json

Dockerfile:

FROM nvidia/cuda:12.9.1-devel-ubuntu24.04

RUN apt update && apt install -y git cmake ninja-build wget

RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg

RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list

RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF'
Package: *
Pin: origin repo.radeon.com
Pin-Priority: 600
EOF

RUN apt update && apt install -y hip-base

ENV HIP_PLATFORM=nvidia

devcontainer.json:

{
    "build": {
        "dockerfile": "Dockerfile"
    },
    "name": "cuda-minimal",
    "privileged": true,
    "runArgs": [
        "--gpus=all"
    ]
}

HIP Dockerfile + devcontainer.json

Dockerfile:

FROM rocm/rocm-terminal:latest

devcontainer.json:

{
    "build": {
        "dockerfile": "Dockerfile"
    },
    "name": "hip-minimal",
    "runArgs": [
        "--device=/dev/kfd",
        "--device=/dev/dri"
    ]
}

Test Result

I have confirmed that Stream HPC's CI is now able to build rocPRIM's benchmarks on Windows. It is crucial however for AMD's CI, or one of AMD's reviewers, to confirm that rocPRIM's benchmarks now build and run successfully on Windows for them too.

Monitoring still works

hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv

nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv

Monitoring can now be turned off

hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv

nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv

When monitoring is disabled, the GPU °C column is excluded from the printed output:

Status of copy             Noise  Bytes/sec  Specialization   Index/2
─────────────────────────────────────────────────────────────────────
Success after 1s            0.1%   4.08e+11  type: char             0
Success after 1s            0.0%   5.33e+11  type: long long        1

The monitoring, temperature_type and gpu_temp_celcius keys are also excluded from the JSON output:

23,27d22
<             "monitoring": {
<                 "name": "amdsmi",
<                 "version": "25.3.0"
<             },
<             "temperature_type": "edge",
73,76d67
<             },
<             "gpu_temp_celsius": {
<                 "start": 50,
<                 "end": 48

Forgetting to pass -DPRIMBENCH_NO_MONITORING throws documented compilation errors

hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. && ./copy_benchmark

nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. && ./copy_benchmark

ld.lld: error: undefined symbol: amdsmi_get_lib_version
>>> referenced by copy_benchmark.cpp

/usr/bin/ld: /tmp/tmpxft_00007baa_00000000-11_copy_benchmark.o: in function `primbench::detail::monitor::get_temp() const':
tmpxft_00007baa_00000000-6_copy_benchmark.cudafe1.cpp:(.text._ZNK9primbench6detail7monitor8get_tempEv[_ZNK9primbench6detail7monitor8get_tempEv]+0x35): undefined reference to `nvmlDeviceGetTemperature'

Submission Checklist

Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

codecov-commenter · 2026-03-20T01:23:15Z

Codecov Report

✅ All modified and coverable lines are covered by tests.

❌ Your project status has failed because the head coverage (77.72%) is below the target coverage (80.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #5637   +/-   ##
========================================
  Coverage    65.53%   65.53%           
========================================
  Files         2045     2045           
  Lines       318558   318558           
  Branches     41872    41872           
========================================
  Hits        208741   208741           
  Misses       92486    92486           
  Partials     17331    17331

Flag	Coverage Δ	*Carryforward flag
hipBLAS	`90.67% <ø> (ø)`	Carriedforward from 70cdadb
hipBLASLt	`43.42% <ø> (ø)`	Carriedforward from 70cdadb
hipCUB	`82.21% <ø> (ø)`	Carriedforward from 70cdadb
hipDNN	`86.23% <ø> (ø)`	Carriedforward from 70cdadb
hipFFT	`58.16% <ø> (ø)`	Carriedforward from 70cdadb
hipRAND	`76.12% <ø> (ø)`	Carriedforward from 70cdadb
hipSOLVER	`68.73% <ø> (ø)`	Carriedforward from 70cdadb
hipSPARSE	`84.70% <ø> (ø)`	Carriedforward from 70cdadb
rocBLAS	`48.12% <ø> (ø)`	Carriedforward from 70cdadb
rocFFT	`52.39% <ø> (ø)`	Carriedforward from 70cdadb
rocPRIM	`38.96% <ø> (ø)`
rocRAND	`57.09% <ø> (ø)`	Carriedforward from 70cdadb
rocSOLVER	`77.72% <ø> (ø)`	Carriedforward from 70cdadb
rocSPARSE	`71.33% <ø> (ø)`	Carriedforward from 70cdadb

*This pull request uses carry forward flags. Click here to find out more.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

MyNameIsTrez · 2026-04-14T15:43:06Z

Rebased.

MyNameIsTrez · 2026-04-15T18:05:18Z

I added the commit feat(primbench): dynamic kernel warming in order to unhardcode the magic values threads_per_block = 256 and num_items = 1 << 20. Both the HIP and CUDA primbench examples still take roughly the same amount of time to warm up the GPU, which I expected.

## Motivation The PR [feat(rocrand): use primbench](#4454) requires [primbench](https://github.com/ROCm/rocm-libraries/blob/develop/shared/primbench/README.md) to support CUDA, since rocRAND contains cuRAND benchmarks. The PR [feat(primbench): make AMD SMI and NVML soft dependencies](#5637) also depends on this `feat(primbench): support an extra backend` PR, as it depends on the refactored code. ## Technical Details I've spent a few hours on cleaning up the commit history. I recommend reviewing this PR by clicking on the first commit in the [Commits tab](https://github.com/ROCm/rocm-libraries/pull/5512/commits), and then clicking the `Next` button to scroll through the commits chronologically. The first commit is `feat(primbench): support CUDA` and is by far the largest and most complex, so I recommend reviewing the other commits first since they're nice and small. The `feat(primbench): support vector CLI args` commit was also required for rocRAND, since its benchmarks accept an array of values: `--lambda 1 2 3`. ### Two flags were removed Even though this PR adds CUDA support to `primbench.hpp`, it went *down* from 4075 lines to 3593 lines. This is mostly the result of me deciding to remove the flags `--output-hip-device-properties-context` and `--output-amdsmi-context`. These flags were used to dump an extreme amount of extra context to the JSON output, but I have not actually found the extra context useful *even once* while updating all 53 rocPRIM benchmarks to use primbench. Porting these flags to CUDA would have been a nightmare, and removing these two flags has made the library significantly more maintainable. In the future we might decide to output a few of the most useful context fields to the JSON output again. ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ### Setup ```sh git checkout users/mynameistrez/add-cuda-support-to-primbench && \ cd shared/primbench ``` ### Running the CUDA example benchmark ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark ``` ### Running the HIP example benchmark ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark ``` ## Test Result The above commands work for me, and the JSON and CSV output looks correct. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

feat(primbench): support an extra backend ## Motivation The PR [feat(rocrand): use primbench](ROCm/rocm-libraries#4454) requires [primbench](https://github.com/ROCm/rocm-libraries/blob/develop/shared/primbench/README.md) to support CUDA, since rocRAND contains cuRAND benchmarks. The PR [feat(primbench): make AMD SMI and NVML soft dependencies](ROCm/rocm-libraries#5637) also depends on this `feat(primbench): support an extra backend` PR, as it depends on the refactored code. ## Technical Details I've spent a few hours on cleaning up the commit history. I recommend reviewing this PR by clicking on the first commit in the [Commits tab](https://github.com/ROCm/rocm-libraries/pull/5512/commits), and then clicking the `Next` button to scroll through the commits chronologically. The first commit is `feat(primbench): support CUDA` and is by far the largest and most complex, so I recommend reviewing the other commits first since they're nice and small. The `feat(primbench): support vector CLI args` commit was also required for rocRAND, since its benchmarks accept an array of values: `--lambda 1 2 3`. ### Two flags were removed Even though this PR adds CUDA support to `primbench.hpp`, it went *down* from 4075 lines to 3593 lines. This is mostly the result of me deciding to remove the flags `--output-hip-device-properties-context` and `--output-amdsmi-context`. These flags were used to dump an extreme amount of extra context to the JSON output, but I have not actually found the extra context useful *even once* while updating all 53 rocPRIM benchmarks to use primbench. Porting these flags to CUDA would have been a nightmare, and removing these two flags has made the library significantly more maintainable. In the future we might decide to output a few of the most useful context fields to the JSON output again. ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ### Setup ```sh git checkout users/mynameistrez/add-cuda-support-to-primbench && \ cd shared/primbench ``` ### Running the CUDA example benchmark ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark ``` ### Running the HIP example benchmark ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark ``` ## Test Result The above commands work for me, and the JSON and CSV output looks correct. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

stanleytsang-amd · 2026-04-16T21:53:01Z

@MyNameIsTrez Please resolve the merge conflicts, thanks!

feat(primbench): make AMD SMI and NVML soft dependencies (#5637) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ## Motivation The PR [[rocPRIM] Warn/Prevent rocPRIM benchmarks from being built on Windows](ROCm/rocm-libraries#5494) was merged as a hotfix, since rocPRIM's benchmarks were not buildable on Windows anymore. This PR its commit `fix(rocprim): revert Windows hotfix PR #5494` reverts that hotfix, and provides a *proper* fix, by turning AMD SMI and NVML from hard dependencies into soft dependencies. This PR will be undrafted once the PR [feat(primbench): support an extra backend](ROCm/rocm-libraries#5512) on which it is based has been merged. ## Technical Details [Here](https://github.com/ROCm/rocm-libraries/tree/users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies/shared/primbench#dependencies) is the updated Dependencies section of the readme: <img width="1353" height="755" alt="image" src="https://github.com/user-attachments/assets/ff8b98b5-1d1c-4fae-a6be-fd2ee7e7415c" /> ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ## Test Result I have confirmed that Stream HPC's CI is now able to build rocPRIM's benchmarks on Windows. It is crucial however for AMD's CI, or one of AMD's reviewers, to confirm that rocPRIM's benchmarks now build and run successfully on Windows for them too. <details> <summary>Monitoring still works</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` </details> <details> <summary>Monitoring can now be turned off</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` When monitoring is disabled, the `GPU °C` column is excluded from the printed output: ``` Status of copy Noise Bytes/sec Specialization Index/2 ───────────────────────────────────────────────────────────────────── Success after 1s 0.1% 4.08e+11 type: char 0 Success after 1s 0.0% 5.33e+11 type: long long 1 ``` The `monitoring`, `temperature_type` and `gpu_temp_celcius` keys are also excluded from the JSON output: ```diff 23,27d22 < "monitoring": { < "name": "amdsmi", < "version": "25.3.0" < }, < "temperature_type": "edge", 73,76d67 < }, < "gpu_temp_celsius": { < "start": 50, < "end": 48 ``` </details> <details> <summary>Forgetting to pass <code>-DPRIMBENCH_NO_MONITORING</code> throws documented compilation errors</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. && ./copy_benchmark ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. && ./copy_benchmark ``` ``` ld.lld: error: undefined symbol: amdsmi_get_lib_version >>> referenced by copy_benchmark.cpp ``` ``` /usr/bin/ld: /tmp/tmpxft_00007baa_00000000-11_copy_benchmark.o: in function `primbench::detail::monitor::get_temp() const': tmpxft_00007baa_00000000-6_copy_benchmark.cudafe1.cpp:(.text._ZNK9primbench6detail7monitor8get_tempEv[_ZNK9primbench6detail7monitor8get_tempEv]+0x35): undefined reference to `nvmlDeviceGetTemperature' ``` </details> ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

## Motivation The PR [feat(rocrand): use primbench](#4454) requires [primbench](https://github.com/ROCm/rocm-libraries/blob/develop/shared/primbench/README.md) to support CUDA, since rocRAND contains cuRAND benchmarks. The PR [feat(primbench): make AMD SMI and NVML soft dependencies](#5637) also depends on this `feat(primbench): support an extra backend` PR, as it depends on the refactored code. ## Technical Details I've spent a few hours on cleaning up the commit history. I recommend reviewing this PR by clicking on the first commit in the [Commits tab](https://github.com/ROCm/rocm-libraries/pull/5512/commits), and then clicking the `Next` button to scroll through the commits chronologically. The first commit is `feat(primbench): support CUDA` and is by far the largest and most complex, so I recommend reviewing the other commits first since they're nice and small. The `feat(primbench): support vector CLI args` commit was also required for rocRAND, since its benchmarks accept an array of values: `--lambda 1 2 3`. ### Two flags were removed Even though this PR adds CUDA support to `primbench.hpp`, it went *down* from 4075 lines to 3593 lines. This is mostly the result of me deciding to remove the flags `--output-hip-device-properties-context` and `--output-amdsmi-context`. These flags were used to dump an extreme amount of extra context to the JSON output, but I have not actually found the extra context useful *even once* while updating all 53 rocPRIM benchmarks to use primbench. Porting these flags to CUDA would have been a nightmare, and removing these two flags has made the library significantly more maintainable. In the future we might decide to output a few of the most useful context fields to the JSON output again. ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ### Setup ```sh git checkout users/mynameistrez/add-cuda-support-to-primbench && \ cd shared/primbench ``` ### Running the CUDA example benchmark ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark ``` ### Running the HIP example benchmark ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark ``` ## Test Result The above commands work for me, and the JSON and CSV output looks correct. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

## Motivation The PR [[rocPRIM] Warn/Prevent rocPRIM benchmarks from being built on Windows](#5494) was merged as a hotfix, since rocPRIM's benchmarks were not buildable on Windows anymore. This PR its commit `fix(rocprim): revert Windows hotfix PR #5494` reverts that hotfix, and provides a *proper* fix, by turning AMD SMI and NVML from hard dependencies into soft dependencies. This PR will be undrafted once the PR [feat(primbench): support an extra backend](#5512) on which it is based has been merged. ## Technical Details [Here](https://github.com/ROCm/rocm-libraries/tree/users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies/shared/primbench#dependencies) is the updated Dependencies section of the readme: <img width="1353" height="755" alt="image" src="https://github.com/user-attachments/assets/ff8b98b5-1d1c-4fae-a6be-fd2ee7e7415c" /> ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ## Test Result I have confirmed that Stream HPC's CI is now able to build rocPRIM's benchmarks on Windows. It is crucial however for AMD's CI, or one of AMD's reviewers, to confirm that rocPRIM's benchmarks now build and run successfully on Windows for them too. <details> <summary>Monitoring still works</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` </details> <details> <summary>Monitoring can now be turned off</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` When monitoring is disabled, the `GPU °C` column is excluded from the printed output: ``` Status of copy Noise Bytes/sec Specialization Index/2 ───────────────────────────────────────────────────────────────────── Success after 1s 0.1% 4.08e+11 type: char 0 Success after 1s 0.0% 5.33e+11 type: long long 1 ``` The `monitoring`, `temperature_type` and `gpu_temp_celcius` keys are also excluded from the JSON output: ```diff 23,27d22 < "monitoring": { < "name": "amdsmi", < "version": "25.3.0" < }, < "temperature_type": "edge", 73,76d67 < }, < "gpu_temp_celsius": { < "start": 50, < "end": 48 ``` </details> <details> <summary>Forgetting to pass <code>-DPRIMBENCH_NO_MONITORING</code> throws documented compilation errors</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. && ./copy_benchmark ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. && ./copy_benchmark ``` ``` ld.lld: error: undefined symbol: amdsmi_get_lib_version >>> referenced by copy_benchmark.cpp ``` ``` /usr/bin/ld: /tmp/tmpxft_00007baa_00000000-11_copy_benchmark.o: in function `primbench::detail::monitor::get_temp() const': tmpxft_00007baa_00000000-6_copy_benchmark.cudafe1.cpp:(.text._ZNK9primbench6detail7monitor8get_tempEv[_ZNK9primbench6detail7monitor8get_tempEv]+0x35): undefined reference to `nvmlDeviceGetTemperature' ``` </details> ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

## Motivation The PR [feat(rocrand): use primbench](ROCm#4454) requires [primbench](https://github.com/ROCm/rocm-libraries/blob/develop/shared/primbench/README.md) to support CUDA, since rocRAND contains cuRAND benchmarks. The PR [feat(primbench): make AMD SMI and NVML soft dependencies](ROCm#5637) also depends on this `feat(primbench): support an extra backend` PR, as it depends on the refactored code. ## Technical Details I've spent a few hours on cleaning up the commit history. I recommend reviewing this PR by clicking on the first commit in the [Commits tab](https://github.com/ROCm/rocm-libraries/pull/5512/commits), and then clicking the `Next` button to scroll through the commits chronologically. The first commit is `feat(primbench): support CUDA` and is by far the largest and most complex, so I recommend reviewing the other commits first since they're nice and small. The `feat(primbench): support vector CLI args` commit was also required for rocRAND, since its benchmarks accept an array of values: `--lambda 1 2 3`. ### Two flags were removed Even though this PR adds CUDA support to `primbench.hpp`, it went *down* from 4075 lines to 3593 lines. This is mostly the result of me deciding to remove the flags `--output-hip-device-properties-context` and `--output-amdsmi-context`. These flags were used to dump an extreme amount of extra context to the JSON output, but I have not actually found the extra context useful *even once* while updating all 53 rocPRIM benchmarks to use primbench. Porting these flags to CUDA would have been a nightmare, and removing these two flags has made the library significantly more maintainable. In the future we might decide to output a few of the most useful context fields to the JSON output again. ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ### Setup ```sh git checkout users/mynameistrez/add-cuda-support-to-primbench && \ cd shared/primbench ``` ### Running the CUDA example benchmark ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark ``` ### Running the HIP example benchmark ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark ``` ## Test Result The above commands work for me, and the JSON and CSV output looks correct. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

## Motivation The PR [feat(rocrand): use primbench](#4454) requires [primbench](https://github.com/ROCm/rocm-libraries/blob/develop/shared/primbench/README.md) to support CUDA, since rocRAND contains cuRAND benchmarks. The PR [feat(primbench): make AMD SMI and NVML soft dependencies](#5637) also depends on this `feat(primbench): support an extra backend` PR, as it depends on the refactored code. ## Technical Details I've spent a few hours on cleaning up the commit history. I recommend reviewing this PR by clicking on the first commit in the [Commits tab](https://github.com/ROCm/rocm-libraries/pull/5512/commits), and then clicking the `Next` button to scroll through the commits chronologically. The first commit is `feat(primbench): support CUDA` and is by far the largest and most complex, so I recommend reviewing the other commits first since they're nice and small. The `feat(primbench): support vector CLI args` commit was also required for rocRAND, since its benchmarks accept an array of values: `--lambda 1 2 3`. ### Two flags were removed Even though this PR adds CUDA support to `primbench.hpp`, it went *down* from 4075 lines to 3593 lines. This is mostly the result of me deciding to remove the flags `--output-hip-device-properties-context` and `--output-amdsmi-context`. These flags were used to dump an extreme amount of extra context to the JSON output, but I have not actually found the extra context useful *even once* while updating all 53 rocPRIM benchmarks to use primbench. Porting these flags to CUDA would have been a nightmare, and removing these two flags has made the library significantly more maintainable. In the future we might decide to output a few of the most useful context fields to the JSON output again. ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ### Setup ```sh git checkout users/mynameistrez/add-cuda-support-to-primbench && \ cd shared/primbench ``` ### Running the CUDA example benchmark ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark ``` ### Running the HIP example benchmark ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark ``` ## Test Result The above commands work for me, and the JSON and CSV output looks correct. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

## Motivation The PR [[rocPRIM] Warn/Prevent rocPRIM benchmarks from being built on Windows](#5494) was merged as a hotfix, since rocPRIM's benchmarks were not buildable on Windows anymore. This PR its commit `fix(rocprim): revert Windows hotfix PR #5494` reverts that hotfix, and provides a *proper* fix, by turning AMD SMI and NVML from hard dependencies into soft dependencies. This PR will be undrafted once the PR [feat(primbench): support an extra backend](#5512) on which it is based has been merged. ## Technical Details [Here](https://github.com/ROCm/rocm-libraries/tree/users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies/shared/primbench#dependencies) is the updated Dependencies section of the readme: <img width="1353" height="755" alt="image" src="https://github.com/user-attachments/assets/ff8b98b5-1d1c-4fae-a6be-fd2ee7e7415c" /> ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ## Test Result I have confirmed that Stream HPC's CI is now able to build rocPRIM's benchmarks on Windows. It is crucial however for AMD's CI, or one of AMD's reviewers, to confirm that rocPRIM's benchmarks now build and run successfully on Windows for them too. <details> <summary>Monitoring still works</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` </details> <details> <summary>Monitoring can now be turned off</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` When monitoring is disabled, the `GPU °C` column is excluded from the printed output: ``` Status of copy Noise Bytes/sec Specialization Index/2 ───────────────────────────────────────────────────────────────────── Success after 1s 0.1% 4.08e+11 type: char 0 Success after 1s 0.0% 5.33e+11 type: long long 1 ``` The `monitoring`, `temperature_type` and `gpu_temp_celcius` keys are also excluded from the JSON output: ```diff 23,27d22 < "monitoring": { < "name": "amdsmi", < "version": "25.3.0" < }, < "temperature_type": "edge", 73,76d67 < }, < "gpu_temp_celsius": { < "start": 50, < "end": 48 ``` </details> <details> <summary>Forgetting to pass <code>-DPRIMBENCH_NO_MONITORING</code> throws documented compilation errors</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. && ./copy_benchmark ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. && ./copy_benchmark ``` ``` ld.lld: error: undefined symbol: amdsmi_get_lib_version >>> referenced by copy_benchmark.cpp ``` ``` /usr/bin/ld: /tmp/tmpxft_00007baa_00000000-11_copy_benchmark.o: in function `primbench::detail::monitor::get_temp() const': tmpxft_00007baa_00000000-6_copy_benchmark.cudafe1.cpp:(.text._ZNK9primbench6detail7monitor8get_tempEv[_ZNK9primbench6detail7monitor8get_tempEv]+0x35): undefined reference to `nvmlDeviceGetTemperature' ``` </details> ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

## Motivation The PR [feat(rocrand): use primbench](ROCm#4454) requires [primbench](https://github.com/ROCm/rocm-libraries/blob/develop/shared/primbench/README.md) to support CUDA, since rocRAND contains cuRAND benchmarks. The PR [feat(primbench): make AMD SMI and NVML soft dependencies](ROCm#5637) also depends on this `feat(primbench): support an extra backend` PR, as it depends on the refactored code. ## Technical Details I've spent a few hours on cleaning up the commit history. I recommend reviewing this PR by clicking on the first commit in the [Commits tab](https://github.com/ROCm/rocm-libraries/pull/5512/commits), and then clicking the `Next` button to scroll through the commits chronologically. The first commit is `feat(primbench): support CUDA` and is by far the largest and most complex, so I recommend reviewing the other commits first since they're nice and small. The `feat(primbench): support vector CLI args` commit was also required for rocRAND, since its benchmarks accept an array of values: `--lambda 1 2 3`. ### Two flags were removed Even though this PR adds CUDA support to `primbench.hpp`, it went *down* from 4075 lines to 3593 lines. This is mostly the result of me deciding to remove the flags `--output-hip-device-properties-context` and `--output-amdsmi-context`. These flags were used to dump an extreme amount of extra context to the JSON output, but I have not actually found the extra context useful *even once* while updating all 53 rocPRIM benchmarks to use primbench. Porting these flags to CUDA would have been a nightmare, and removing these two flags has made the library significantly more maintainable. In the future we might decide to output a few of the most useful context fields to the JSON output again. ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ### Setup ```sh git checkout users/mynameistrez/add-cuda-support-to-primbench && \ cd shared/primbench ``` ### Running the CUDA example benchmark ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark ``` ### Running the HIP example benchmark ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark ``` ## Test Result The above commands work for me, and the JSON and CSV output looks correct. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

## Motivation The PR [[rocPRIM] Warn/Prevent rocPRIM benchmarks from being built on Windows](ROCm#5494) was merged as a hotfix, since rocPRIM's benchmarks were not buildable on Windows anymore. This PR its commit `fix(rocprim): revert Windows hotfix PR ROCm#5494` reverts that hotfix, and provides a *proper* fix, by turning AMD SMI and NVML from hard dependencies into soft dependencies. This PR will be undrafted once the PR [feat(primbench): support an extra backend](ROCm#5512) on which it is based has been merged. ## Technical Details [Here](https://github.com/ROCm/rocm-libraries/tree/users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies/shared/primbench#dependencies) is the updated Dependencies section of the readme: <img width="1353" height="755" alt="image" src="https://github.com/user-attachments/assets/ff8b98b5-1d1c-4fae-a6be-fd2ee7e7415c" /> ## Test Plan ### VS Code dev containers <details> <summary>CUDA <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM nvidia/cuda:12.9.1-devel-ubuntu24.04 RUN apt update && apt install -y git cmake ninja-build wget RUN wget -qO- https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/keyrings/rocm.gpg RUN echo "deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/7.2 noble main" > /etc/apt/sources.list.d/rocm.list RUN tee /etc/apt/preferences.d/rocm-pin-600 <<'EOF' Package: * Pin: origin repo.radeon.com Pin-Priority: 600 EOF RUN apt update && apt install -y hip-base ENV HIP_PLATFORM=nvidia ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "cuda-minimal", "privileged": true, "runArgs": [ "--gpus=all" ] } ``` </details> <details> <summary>HIP <code>Dockerfile</code> + <code>devcontainer.json</code></summary> `Dockerfile`: ```Dockerfile FROM rocm/rocm-terminal:latest ``` `devcontainer.json`: ```json { "build": { "dockerfile": "Dockerfile" }, "name": "hip-minimal", "runArgs": [ "--device=/dev/kfd", "--device=/dev/dri" ] } ``` </details> ## Test Result I have confirmed that Stream HPC's CI is now able to build rocPRIM's benchmarks on Windows. It is crucial however for AMD's CI, or one of AMD's reviewers, to confirm that rocPRIM's benchmarks now build and run successfully on Windows for them too. <details> <summary>Monitoring still works</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -lamd_smi && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -lnvidia-ml && ./copy_benchmark --json-out monitoring.json --csv-out monitoring.csv ``` </details> <details> <summary>Monitoring can now be turned off</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. -DPRIMBENCH_NO_MONITORING && ./copy_benchmark --json-out no_monitoring.json --csv-out no_monitoring.csv ``` When monitoring is disabled, the `GPU °C` column is excluded from the printed output: ``` Status of copy Noise Bytes/sec Specialization Index/2 ───────────────────────────────────────────────────────────────────── Success after 1s 0.1% 4.08e+11 type: char 0 Success after 1s 0.0% 5.33e+11 type: long long 1 ``` The `monitoring`, `temperature_type` and `gpu_temp_celcius` keys are also excluded from the JSON output: ```diff 23,27d22 < "monitoring": { < "name": "amdsmi", < "version": "25.3.0" < }, < "temperature_type": "edge", 73,76d67 < }, < "gpu_temp_celsius": { < "start": 50, < "end": 48 ``` </details> <details> <summary>Forgetting to pass <code>-DPRIMBENCH_NO_MONITORING</code> throws documented compilation errors</summary> ```sh hipcc -o copy_benchmark examples/hip/copy_benchmark.cpp -I. && ./copy_benchmark ``` ```sh nvcc -o copy_benchmark examples/cuda/copy_benchmark.cu -I. && ./copy_benchmark ``` ``` ld.lld: error: undefined symbol: amdsmi_get_lib_version >>> referenced by copy_benchmark.cpp ``` ``` /usr/bin/ld: /tmp/tmpxft_00007baa_00000000-11_copy_benchmark.o: in function `primbench::detail::monitor::get_temp() const': tmpxft_00007baa_00000000-6_copy_benchmark.cudafe1.cpp:(.text._ZNK9primbench6detail7monitor8get_tempEv[_ZNK9primbench6detail7monitor8get_tempEv]+0x35): undefined reference to `nvmlDeviceGetTemperature' ``` </details> ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

github-actions Bot added documentation project: rocprim shared: primbench labels Mar 19, 2026

assistant-librarian Bot added the organization: ROCm label Mar 19, 2026

MyNameIsTrez force-pushed the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch 2 times, most recently from c32e0fe to a9fbbe1 Compare March 20, 2026 12:57

This was referenced Mar 20, 2026

feat(primbench): support an extra backend #5512

Merged

[rocPRIM] Warn/Prevent rocPRIM benchmarks from being built on Windows… #5521

Closed

MyNameIsTrez force-pushed the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch 3 times, most recently from f5b4ffa to bc28665 Compare April 1, 2026 21:10

MyNameIsTrez mentioned this pull request Apr 8, 2026

feat(rocrand): use primbench #4454

Merged

1 task done

MyNameIsTrez force-pushed the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch from 28984c0 to 0bb409a Compare April 10, 2026 17:07

MyNameIsTrez marked this pull request as ready for review April 10, 2026 17:07

MyNameIsTrez requested review from a team as code owners April 10, 2026 17:07

MyNameIsTrez force-pushed the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch from 0bb409a to f93a41c Compare April 10, 2026 17:22

spolifroni-amd approved these changes Apr 13, 2026

View reviewed changes

MyNameIsTrez force-pushed the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch from f93a41c to 5a84121 Compare April 14, 2026 15:24

MyNameIsTrez force-pushed the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch from 2140e55 to 669c836 Compare April 15, 2026 18:02

MyNameIsTrez added 4 commits April 16, 2026 22:24

feat(primbench): make AMD SMI and NVML soft dependencies

b833284

docs(primbench): document soft dependencies

d60662d

fix(rocprim): revert Windows hotfix PR #5494

208e2bc

fix(rocprim): on Windows pass -DPRIMBENCH_NO_MONITORING

8605bc5

MyNameIsTrez added 7 commits April 16, 2026 22:24

fix(primbench): fix macro conditions being evaluated more than once

c80d358

fix(primbench): fix Windows compilation errors

5597104

fix(primbench): include cstring

330ccbc

fix(primbench): store string pointers returned by meta() correctly

b7c1404

docs(primbench): document Windows usage

f602a30

fix(primbench): stop flooding logs

2d3ac9c

feat(primbench): dynamic kernel warming

d98dc96

MyNameIsTrez force-pushed the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch from 669c836 to d98dc96 Compare April 16, 2026 22:25

stanleytsang-amd approved these changes Apr 22, 2026

View reviewed changes

stanleytsang-amd merged commit dd6042e into develop Apr 22, 2026
23 of 24 checks passed

stanleytsang-amd deleted the users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies branch April 22, 2026 14:33

stanleytsang-amd mentioned this pull request Apr 22, 2026

rocPRIM benchmark build fails: missing #include <cassert> in primbench.hpp ROCm/TheRock#4739

Closed

lajagapp mentioned this pull request Apr 22, 2026

[primbench] Add missing '#include <cassert>' in primbench.hpp #6662

Closed

Navigation Menu

Repository navigation

feat(primbench): make AMD SMI and NVML soft dependencies - #5637

feat(primbench): make AMD SMI and NVML soft dependencies#5637
stanleytsang-amd merged 11 commits into
developfrom
users/mynameistrez/make-amdsmi-and-nvml-soft-dependencies

MyNameIsTrez commented Mar 19, 2026 •

edited

Loading

Uh oh!

codecov-commenter commented Mar 20, 2026 •

edited

Loading

Uh oh!

MyNameIsTrez commented Apr 14, 2026

Uh oh!

MyNameIsTrez commented Apr 15, 2026 •

edited

Loading

Uh oh!

stanleytsang-amd commented Apr 16, 2026

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

MyNameIsTrez commented Mar 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Motivation

Technical Details

Test Plan

VS Code dev containers

Test Result

Submission Checklist

Uh oh!

codecov-commenter commented Mar 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

MyNameIsTrez commented Apr 14, 2026

Uh oh!

MyNameIsTrez commented Apr 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

stanleytsang-amd commented Apr 16, 2026

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

MyNameIsTrez commented Mar 19, 2026 •

edited

Loading

codecov-commenter commented Mar 20, 2026 •

edited

Loading

MyNameIsTrez commented Apr 15, 2026 •

edited

Loading