Project 0: Thomas Shaw

README.md

@@ -3,11 +3,101 @@ Project 0 Getting Started
 
 **University of Pennsylvania, CIS 5650: GPU Programming and Architecture, Project 0**
 
-* (TODO) YOUR NAME HERE
-  * (TODO) [LinkedIn](), [personal website](), [twitter](), etc.
-* Tested on: (TODO) Windows 22, i7-2222 @ 2.22GHz 22GB, GTX 222 222MB (Moore 2222 Lab)
+* Thomas Shaw
+  * [LinkedIn](https://www.linkedin.com/in/thomas-shaw-54468b222), [personal website](https://tlshaw.me), [GitHub](https://github.com/printer83mph), etc.
+* Tested on: Fedora 42, Ryzen 7 5700x @ 4.67GHz, 32GB, RTX 2070 8GB
 
-### (TODO: Your README)
+### Results
 
-Include screenshots, analysis, etc. (Remember, this is public, so don't put
-anything here that you don't want to share with the world.)
+Configuring my machine to work with CUDA was a bit frustrating, as I previously had an installation of Fedora 40 with proprietary non-CUDA drivers. The process to both A) upgrade to 42, which I had wanted to do for a while, and B) install CUDA toolkit + drivers, was a bit overcomplicated, so after fiddling for a couple days, I went with a clean slate (fresh install) and things worked out much nicer.
+
+I used NVIDIA's "Package Manager Install" guide, which required some extra steps since I had Secure Boot enabled, but otherwise was straightforward.
+
+I'm primarily using Visual Studio Code as my code editing environment, with `clangd` as a language server + linter + formatter. With some minimal configuration, this works nicely.
+
+```
+# .clangd
+
+CompileFlags:
+  Add:
+    - -std=c++11
+    - --cuda-path=/usr/local/cuda
+    - --cuda-gpu-arch=sm_75
+    - -L/usr/local/cuda/lib64
+    - -I/usr/local/cuda/include
+  Remove:
+    - "-forward-unknown-to-host-compiler"
+    - "-arch=native"
+    - "--expt-*"
+    - "--options-file"
+    - "-G"
+```
+
+I combined this with the following build configuration:
+
+```sh
+CC=gcc CXX=g++ CUDACXX=nvcc cmake .. -G "Ninja" \
+  -DCMAKE_BUILD_TYPE=Debug \
+  -DCMAKE_CUDA_FLAGS="-g -G" \
+  -DCMAKE_EXPORT_COMPILE_COMMANDS=ON
+```
+
+One of the main issues I ran into after a fresh install was that Wayland (compositor) didn't like the built binary. I wasn't sure exactly why without digging further into GTK, but the following error would appear:
+
+```
+$ cuda-gl-check/build/bin/cuda-gl-check
+libdecor-gtk-WARNING: Failed to initialize GTK
+Failed to load plugin 'libdecor-gtk.so': failed to init
+```
+
+This was resolved by setting the following environment variables when running the binary, whether debugging or just running raw:
+
+```sh
+export WAYLAND_DISPLAY=""
+export XDG_SESSION_TYPE=x11
+```
+
+I also plugged these into `launch.json`:
+
+```json
+{
+  "$schema": "vscode://schemas/launch",
+  // Use IntelliSense to learn about possible attributes.
+  // Hover to view descriptions of existing attributes.
+  // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+  "version": "0.2.0",
+  "configurations": [
+    {
+      "name": "CUDA C++: Launch",
+      "type": "cuda-gdb",
+      "request": "launch",
+      "environment": [
+        {"name": "WAYLAND_DISPLAY", "value": ""},
+        {"name": "XDG_SESSION_TYPE", "value": "x11"}
+      ],
+      "program": "${workspaceFolder}/build/bin/cuda-gl-check"
+    },
+    {
+      "name": "CUDA C++: Attach",
+      "type": "cuda-gdb",
+      "request": "attach"
+    }
+  ]
+}
+```
+
+### Screencaps
+
+![gl check window](images/gl-check-screencap.png)
+
+![nsight gui](images/nsight-gui.png)
+
+![nsight summary](images/nsight-summary.png)
+
+![nsight analysis](images/nsight-analysis.png)
+
+![webgl report](images/webgl-report.png)
+
+For WebGPU, I had to enable the "Unsafe WebGPU Support" flag in Chrome.
+
+![webGPU report](images/webgpu-report.png)

cuda-gl-check/.clangd

@@ -0,0 +1,13 @@
+CompileFlags:
+  Add:
+    - -std=c++11
+    - --cuda-path=/usr/local/cuda
+    - --cuda-gpu-arch=sm_75
+    - -L/usr/local/cuda/lib64
+    - -I/usr/local/cuda/include
+  Remove:
+    - "-forward-unknown-to-host-compiler"
+    - "-arch=native"
+    - "--expt-*"
+    - "--options-file"
+    - "-G"

cuda-gl-check/.gitignore

@@ -215,6 +215,10 @@ install_manifest.txt
 *.app
 
 
+### Clangd ###
+.cache/clangd
+
+
 ### CUDA ###
 *.i
 *.ii

cuda-gl-check/.vscode/launch.json

@@ -0,0 +1,24 @@
+{
+  "$schema": "vscode://schemas/launch",
+  // Use IntelliSense to learn about possible attributes.
+  // Hover to view descriptions of existing attributes.
+  // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+  "version": "0.2.0",
+  "configurations": [
+    {
+      "name": "CUDA C++: Launch",
+      "type": "cuda-gdb",
+      "request": "launch",
+      "environment": [
+        {"name": "WAYLAND_DISPLAY", "value": ""},
+        {"name": "XDG_SESSION_TYPE", "value": "x11"}
+      ],
+      "program": "${workspaceFolder}/build/bin/cuda-gl-check"
+    },
+    {
+      "name": "CUDA C++: Attach",
+      "type": "cuda-gdb",
+      "request": "attach"
+    }
+  ]
+}

cuda-gl-check/.vscode/settings.json

@@ -0,0 +1,11 @@
+{
+  "files.associations": {
+    "*.cu": "cuda-cpp"
+  },
+  "[cpp]": {
+    "editor.defaultFormatter": "llvm-vs-code-extensions.vscode-clangd"
+  },
+  "[cuda-cpp]": {
+    "editor.defaultFormatter": "llvm-vs-code-extensions.vscode-clangd"
+  }
+}