Skip to content

Latest commit

 

History

History
76 lines (46 loc) · 3.13 KB

BUILDING.md

File metadata and controls

76 lines (46 loc) · 3.13 KB
Raw

How to Compile

FileOnQ.Imaging.Raw uses a mix of .NET Framework, .NET, native C/C++ projects, and CUDA kernels for GPU processing. You will need to have additional build tools installed to properly build this locally. This readme serves as a guide for new developers to contribute.

Visual Studio and .NET

This project uses a combination of Visual Studio and .NET CLI tools for building and developing.

  • Visual Studio 2019 (16.10.3) or newer
    • Workload: Desktop development with C++
    • Component: Windows 10 SDK (10.0.18)
  • .NET CLI 5.0.300 or newer

If you are using an older version of Visual Studio or .NET CLI you won't have access to the _OR_GREATER preprocessor directives. This will result in runtime issues when running the library locally.

Build Dependencies

  • LibRaw: add these dependencies by running the below command in the repo's root folder.

    git submodule update --init --recursive

    or clone them with the repo using

    git clone --recurse-submodules https://github.com/FileOnQ.Imaging.Raw.git

  • CUDA: You will need to download and install CUDA SDK 11.4

Compile

First time compilation takes a while. Future builds won't take as long even if you do a clean and rebuild.

By compiling the main project FileOnQ.Imaging.Raw you will be compiling all native dependencies. No need to run any additional compilation instructions.

Visual Studio

Right click on FileOnQ.Imaging.Raw and select build

CLI

dotnet build src/FileOnQ.Imaging.Raw/FileOnQ.Imaging.Raw.csproj

Troubleshooting

If after doing all of the above and compiling is causing errors. Try opening a developer command prompt as admin, navigate to the repo's root directory and run one of the following

msbuild /p:Platform="x64"
msbuild /p:Platform="x86"

Build Scripts and Structure

The entire build process is orchestrated from FileOnQ.Imaging.Raw.csproj which includes downloading and compiling third party libraries. During the build you'll need to clone several git repositories and compile them. This leverages a combination of dotnet build targets and bat files. The build targets can be found in the src/FileOnQ.Imaging.Raw/Build directory. The build scripts or batch files can be found in the build directory.

The build targets and scripts should work out of the box if the required dependencies above have been installed.

Build Targets

The build targets located in the main project FileOnQ.Imaging.Raw are used to orchestrate builds for all supported platform architectures. They all follow a similar process

  1. Clone repository for each supported platform architecture
  2. Run build script (bat file)
  3. Copy generated assemblies over to src/FileOnQ.Imaging.Raw/runtimes/platform/native

Build Scripts (bat)

The build scripts load the Visual Studio Developer Command Prompt for the desired platform architecture and then run the build commands. This is really important when compiling native libraries for x86, x64, ARM, etc. Once the tools are loaded it compiles the code using the native toolchains.