added some manual calls to unzip necessary files

Author: gregbolet

CMakeLists.txt

@@ -401,11 +401,16 @@ IF(BUILD_CUDA)
     UNZIPFILE("${HECBENCH_SRC_DIR}/slu-cuda/src/nicslu.tar")
     UNZIPFILE("${HECBENCH_SRC_DIR}/axhelm-cuda/BlasLapack.tar.gz")
     UNZIPFILE("${HECBENCH_SRC_DIR}/testSNAP-omp/refdata.tar.gz")
+    UNZIPFILE("${HECBENCH_SRC_DIR}/lsqt-cuda/examples.tar.gz")
+    UNZIPFILE("${HECBENCH_SRC_DIR}/meanshift-cuda/dataset.tar.gz")
+    UNZIPFILE("${HECBENCH_SRC_DIR}/ced-cuda/frames.tar.gz")
 ENDIF()
 
 IF(BUILD_OMP)
     UNZIPFILE("${HECBENCH_SRC_DIR}/axhelm-omp/BlasLapack.tar.gz")
     UNZIPFILE("${HECBENCH_SRC_DIR}/testSNAP-omp/refdata.tar.gz")
+    UNZIPFILE("${HECBENCH_SRC_DIR}/lsqt-cuda/examples.tar.gz")
+    UNZIPFILE("${HECBENCH_SRC_DIR}/meanshift-cuda/dataset.tar.gz")
 ENDIF()
 
 set(TARGETS "")

Dockerfile

@@ -1,5 +1,7 @@
 FROM nvidia/cuda:12.6.0-devel-ubuntu24.04
 
+ARG HOSTOS=linux 
+
 # Set non-interactive frontend for apt-get to avoid prompts
 ENV DEBIAN_FRONTEND=noninteractive
 
@@ -10,6 +12,7 @@ SHELL ["/bin/bash", "-c"]
 RUN apt-get update && \
     apt-get upgrade -y && \
     apt-get install -y wget make git gfortran libomp-18-dev libboost-all-dev clang-18 clang-tools-18 unzip && \
+    apt-get install -y imagemagick && \
     wget https://github.com/Kitware/CMake/releases/download/v3.28.0/cmake-3.28.0-linux-x86_64.sh && \
     chmod +x cmake-3.28.0-linux-x86_64.sh && \
     ./cmake-3.28.0-linux-x86_64.sh --skip-license --prefix=/usr/local && \
@@ -57,15 +60,20 @@ COPY . .
 
 # if we are on a windows host, we need to remove the CRLF characters from all the files
 #RUN find . -type f -exec sed -i 's/\r$//' {} +
-RUN find . -type f \
-    -not -name "*.gz" \
-    -not -name "*.zip" \
-    -not -name "*.rar" \
-    -not -name "*.7z" \
-    -not -name "*.tar" \
-    -not -name "*.bz2" \
-    -not -name "*.tar.*" \
-    -exec sed -i 's/\r$//' {} +
+RUN if [ "$HOSTOS" = "windows" ]; then \
+    echo "Removing CRLF characters from files..."; \
+    find . -type f \
+        -not -name "*.gz" \
+        -not -name "*.zip" \
+        -not -name "*.rar" \
+        -not -name "*.7z" \
+        -not -name "*.tar" \
+        -not -name "*.bz2" \
+        -not -name "*.tar.*" \
+        -exec sed -i 's/\r$//' {} +; \
+    else \
+    echo "HOSTOS is not windows; skipping CRLF removal."; \
+    fi 
 
 # write out to the bashrc to source conda and activate the environment on container startup
 RUN echo 'conda activate gpu-flopbench' >> ~/.bashrc

README.md

@@ -18,9 +18,9 @@ Target codes thus-far:
 This work was presented at the [HPDC 2025](https://hpdc.sci.utah.edu/2025/) [AI4Sys Workshop](https://ai4sys.github.io/).
 Upon having our paper accepted, we pre-published on arXiv in case people wanted to cite us.
 
-[paper link here](https://dl.acm.org/doi/10.1145/3731545.3743645) 
+📃📃 [paper link here](https://dl.acm.org/doi/10.1145/3731545.3743645) 🔗🔗
 
-BibTeX reference below.
+📜 BibTeX reference below.
 ```
 @inproceedings{10.1145/3731545.3743645,
 author = {Bolet, Gregory and Georgakoudis, Giorgis and Menon, Harshitha and Parasyris, Konstantinos and Hasabnis, Niranjan and Estes, Hayden and Cameron, Kirk and Oren, Gal},
@@ -54,12 +54,12 @@ For ease-of-reproducibility, we supply a `Dockerfile` with the necessary steps t
 The following is a list of steps to help you get set up and into the main bash shell of the container.
 
 ‼️‼️
-We note that the base container image will take up about 40 GB of storage space; once we start building codes and gathering profiling data, the disk usage will jump up to about 50 GB.
+We note that the base container image will take up about 15 GB of storage space, which then jumps to 40 GB when we build the container; once we start building codes and gathering profiling data, the disk usage will jump up to about 50 GB.
 Please ensure your system has enough storage space before continuing.
 ‼️‼️
 
 ```
-git clone git@github.com:gregbolet/gpu-FLOPBench.git ./gpu-flopbench
+git clone git@github.com:gregbolet/gpu-flopbench.git ./gpu-flopbench
 cd ./gpu-flopbench
 docker build --progress=plain -t 'gpu-flopbench' .
 docker run -ti --gpus all --name gpu-flopbench-container --runtime=nvidia -e NVIDIA_DRIVER_CAPABILITIES=compute,utility -e NVIDIA_VISIBLE_DEVICES=all gpu-flopbench
@@ -104,7 +104,7 @@ In future work we would like to make this process of building the codes agnostic
 
 Start by simply cloning our repo.
 ```
-git clone git@github.com:gregbolet/gpu-FLOPBench.git ./gpu-flopbench
+git clone git@github.com:gregbolet/gpu-flopbench.git ./gpu-flopbench
 cd ./gpu-flopbench
 ```
 
@@ -116,6 +116,7 @@ By default, we have everything building with `clang++` and `clang`, this should
 source ./runBuild.sh
 ```
 NOTE: If you're running this from a Docker container generated from our Dockerfile, it should work out-of-the-box.
+
 We originally had the CUDA codes building with `nvcc`, but to be able to also build SYCL and OMP codes, we switched to just LLVM. You may still be able to build the codes with `nvcc`, but it may take some modifications to the build pipeline.
 We have future plans to sample SYCL and OMP codes, but for now, this work focuses on CUDA codes.
 
@@ -202,7 +203,7 @@ python3 simpleScrapeKernels.py
 The scraped output will be a file called `simple-scraped-kernels-with-sass.json` in JSON format. 
 We particularly do this simple form of scraping because we're struggling to have a proper AST traversal script that can properly extract CUDA kernels from source. 
 This is a future step we're working on. For now, this file contains all the source files from each executable that was built in the `build` directory.
-Because this is an update version, we include SASS code in the scrape, but these are not used in the final results of this paper.
+Because this is an updated version, we include SASS code in the scrape, but these are not used in the final results of this paper.
 
 ## Pruning the Scraped Kernels