[FEATURE] Add SYCL feature flags to rllm-llamacpp build (To add support for Intel GPUs) #96
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@microsoft-github-policy-service agree |
TheMrCodes
commented
Apr 18, 2024
rllm/llama-cpp-low/build.rs
Outdated
Comment on lines
49
to
89
|
|
||
| let dirs = [ | ||
| "/opt/intel/oneapi/compiler/latest/lib", | ||
| "/opt/intel/oneapi/mkl/latest/lib", | ||
| //"/opt/intel/oneapi/dnnl/latest/lib", | ||
| ]; | ||
|
|
||
| // *.a => static | ||
| // *.so => dynamic | ||
| for dir in dirs.iter() { | ||
| println!("cargo:rustc-link-search={}", dir); | ||
| for file in std::fs::read_dir(dir).unwrap() { | ||
| let file = file.unwrap(); | ||
| let file_name = file.file_name(); | ||
| let file_name = file_name.to_str().unwrap(); | ||
| if !file_name.starts_with("lib") { continue; } | ||
| if file_name.contains("lp64") && !file_name.contains("ilp64") { continue; } | ||
| if file_name.contains("seq") { continue; } | ||
| if file_name == "libmkl_gnu_thread.so" { continue; } | ||
| let file_name = file_name.trim_start_matches("lib"); | ||
|
|
||
| if file_name.ends_with(".so") { | ||
| let file_name = &file_name[..file_name.len()-3]; | ||
| println!("cargo:rustc-link-lib=dylib={}", file_name); | ||
| } else if file_name.ends_with(".a") { | ||
| let file_name = &file_name[..file_name.len()-2]; | ||
| println!("cargo:rustc-link-lib=static={}", file_name); | ||
| } | ||
| } | ||
| } | ||
| //panic!("stop here"); | ||
|
|
||
| //println!("cargo:rustc-link-search=native=/opt/intel/oneapi/compiler/latest/lib"); | ||
| //println!("cargo:rustc-link-lib=intlc"); | ||
| //println!("cargo:rustc-link-lib=svml"); | ||
| //println!("cargo:rustc-link-lib=sycl"); | ||
| //println!("cargo:rustc-link-search=native=/opt/intel/oneapi/mkl/latest/lib"); | ||
| //println!("cargo:rustc-link-lib=mkl_core"); | ||
| //println!("cargo:rustc-link-lib=mkl_sycl_blas"); | ||
| //println!("cargo:rustc-link-lib=mkl_sycl"); | ||
| } |
There was a problem hiding this comment.
Currently struggling with dynamically linken the libraries but also resolving only needed without duplicates and with resolved dependencies. Don't really know whats a good way here...
|
The basic build works now. 🚀 Known Issues:
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The goal of this pull request is to add support for Intel GPU to the build script of rllm llama.cpp
This is done by
The code will be tested on a machine with Intel 13Gen Processor & Intel Arc A770 GPU