Skip to content

Fix unforked llamacpp #546

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
doringeman merged 6 commits into from
Jan 8, 2026
Merged

Fix unforked llamacpp #546

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
6c3db6b
fix(llamacpp): add missing cmake flags
doringeman Jan 8, 2026
a8b3c91
style(llamacpp): use spaces for consistent help output alignment
doringeman Jan 8, 2026
aa3c859
fix(llamacpp): install renamed binary correctly
doringeman Jan 8, 2026
591986a
fix(llamacpp): fix rpath for macOS binary
doringeman Jan 8, 2026
927a488
fix(scheduling): use /health endpoint for backend readiness checks
doringeman Jan 8, 2026
9bf9df1
chore(llamacpp): use $(CURDIR) instead of $(shell pwd)
doringeman Jan 8, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 12 additions & 6 deletions llamacpp/Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,9 @@ ifeq ($(DETECTED_OS),macOS)
-DGGML_NATIVE=OFF \
-DGGML_OPENMP=OFF \
-DLLAMA_CURL=OFF \
-DLLAMA_BUILD_COMMON=ON \
-DLLAMA_BUILD_SERVER=ON \
-DLLAMA_BUILD_TOOLS=ON \
-GNinja \
-S $(NATIVE_DIR)
@echo "Building..."
Expand All @@ -43,6 +46,9 @@ ifeq ($(DETECTED_OS),macOS)
rm -rf $(INSTALL_DIR)/lib/cmake
rm -rf $(INSTALL_DIR)/lib/pkgconfig
rm -rf $(INSTALL_DIR)/include
@echo "Fixing rpath..."
install_name_tool -delete_rpath "$(shell pwd)/$(BUILD_DIR)/bin" $(INSTALL_DIR)/bin/com.docker.llama-server
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot Jan 8, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

Using the built-in $(CURDIR) variable is preferred over $(shell pwd). It's more efficient as it avoids forking a shell process and is a standard practice in GNU Makefiles.

	install_name_tool -delete_rpath "$(CURDIR)/$(BUILD_DIR)/bin" $(INSTALL_DIR)/bin/com.docker.llama-server

Copy link
Contributor Author

@doringeman doringeman Jan 8, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done in chore(llamacpp): use $(CURDIR) instead of $(shell pwd).

install_name_tool -add_rpath "@executable_path/../lib" $(INSTALL_DIR)/bin/com.docker.llama-server
@echo "Build complete! Binaries are in $(INSTALL_DIR)"
else ifeq ($(DETECTED_OS),Linux)
@echo "Linux build not implemented yet"
Expand Down Expand Up @@ -87,9 +93,9 @@ install-dir:

help:
@echo "Available targets:"
@echo " build - Build llama.cpp (macOS only for now)"
@echo " install-deps - Install build dependencies"
@echo " build-dir - Print build directory path"
@echo " install-dir - Print install directory path"
@echo " clean - Clean build artifacts"
@echo " help - Show this help"
@echo " build - Build llama.cpp (macOS only for now)"
@echo " install-deps - Install build dependencies"
@echo " build-dir - Print build directory path"
@echo " install-dir - Print install directory path"
@echo " clean - Clean build artifacts"
@echo " help - Show this help"
6 changes: 2 additions & 4 deletions llamacpp/native/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,10 +41,8 @@ if (DDLLAMA_BUILD_SERVER)

add_custom_target(com.docker.llama-server ALL DEPENDS "${LLAMA_SERVER_DST}")

# Install the renamed binary using TARGETS instead of PROGRAMS for better cross-platform support
install(TARGETS llama-server
RUNTIME DESTINATION bin
RENAME "com.docker.llama-server${CMAKE_EXECUTABLE_SUFFIX}")
# Install the renamed binary
install(PROGRAMS "${LLAMA_SERVER_DST}" DESTINATION bin)
endif()

if (WIN32 AND DDLLAMA_BUILD_UTILS)
Expand Down
5 changes: 3 additions & 2 deletions pkg/inference/scheduling/runner.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -222,8 +222,9 @@ func (r *runner) wait(ctx context.Context) error {
return r.err
default:
}
// Create and execute a request targeting a known-valid endpoint.
readyRequest, err := http.NewRequestWithContext(ctx, http.MethodGet, "http://localhost/v1/models", http.NoBody)
// Create and execute a request targeting the health endpoint.
// Note: /health returns 503 during model loading, 200 when ready.
readyRequest, err := http.NewRequestWithContext(ctx, http.MethodGet, "http://localhost/health", http.NoBody)
if err != nil {
return fmt.Errorf("readiness request creation failed: %w", err)
}
Expand Down