llama/pom.xml: write JNI headers relative to the module, not the cwd - #452
Merged
Merged
Conversation
The 16 classifier-profile compile executions (cuda, opencl-android,
windows-msvc, vulkan-*, rocm-*, sycl-*, openvino-*, ...) passed `-h src/main/cpp`,
which javac resolves against the process working directory. Building the core
from the repository root (`mvn -pl llama -am ... -P vulkan-windows`) therefore
wrote net_ladenthin_llama_*.h into ./src/main/cpp at the root -- untracked files
outside the module -- instead of llama/src/main/cpp, where .gitignore expects
them. They now use ${project.basedir}/src/main/cpp.
Verified: `mvn -pl llama -am clean compile -P vulkan-windows` from the root puts
all five headers into llama/src/main/cpp and creates no ./src.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01E2h8gXyyE5UeimkL9vQv9G
bernardladenthin
had a problem deploying
to
maven-central
September 21, 2026 22:02 — with 
GitHub Actions
Failure
GitHub Actions
Failure
bernardladenthin
had a problem deploying
to
startgate
September 21, 2026 22:02 — with
GitHub Actions
Error
GitHub Actions
Error
bernardladenthin
had a problem deploying
to
maven-central
September 21, 2026 22:02 — with
GitHub Actions
Failure
GitHub Actions
Failure
This branch had an error being deployed
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
The 16 classifier-profile compile executions in
llama/pom.xmlpass-h src/main/cpp. javac resolves that against the process working directory, not the module. Building the core from the repository root — e.g.mvn -pl llama -am install -P vulkan-windows, the local recipe for installing a snapshot with GPU natives — writesnet_ladenthin_llama_*.hinto./src/main/cppat the root: untracked files outside the module, instead ofllama/src/main/cpp, where.gitignore(llama/src/main/cpp/net_ladenthin_llama_*.h) expects them. Building from insidellama/happened to work, which is why CI never noticed.Fix
<arg>src/main/cpp</arg>→<arg>${project.basedir}/src/main/cpp</arg>in all 16 executions. No other change.Verified
From the repository root on Windows 11:
mvn -pl llama -am clean compile -P vulkan-windowsputs all five headers intollama/src/main/cpp/and no./srcis created;git statusshows onlyllama/pom.xml.🤖 Generated with Claude Code
https://claude.ai/code/session_01E2h8gXyyE5UeimkL9vQv9G