cov : add Code Coverage and codecov.io integration (#2928)

* update .gitignore

* makefile: add coverage support (lcov, gcovr)

* add code-coverage workflow

* update code coverage workflow

* wun on ubuntu 20.04

* use gcc-8

* check why the job hang

* add env vars

* add LLAMA_CODE_COVERAGE=1 again

* - add CODECOV_TOKEN
- add missing make lcov-report

* install lcov

* update make file -pb flag

* remove unused  GGML_NITER from workflows

* wrap coverage output files in COV_TARGETS

4 files changed, 64 insertions, 2 deletions

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 20fd8c2b..9d0a6c22 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -18,7 +18,6 @@ on:
 env:
   BRANCH_NAME: ${{ github.head_ref || github.ref_name }}
   GGML_NLOOP: 3
-  GGML_NITER: 1
   GGML_N_THREADS: 1
 
 jobs:
diff --git a/.github/workflows/code-coverage.yml b/.github/workflows/code-coverage.yml
new file mode 100644
index 00000000..392db8a0
--- /dev/null
+++ b/.github/workflows/code-coverage.yml
@@ -0,0 +1,36 @@
+name: Code Coverage
+on: [push, pull_request]
+
+env:
+  GGML_NLOOP: 3
+  GGML_N_THREADS: 1
+
+jobs:
+  run:
+    runs-on: ubuntu-20.04
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential gcc-8 lcov
+
+      - name: Build
+        run: CC=gcc-8 make -j LLAMA_CODE_COVERAGE=1 tests
+
+      - name: Run tests
+        run: CC=gcc-8 make test
+
+      - name: Generate coverage report
+        run: |
+          make coverage
+          make lcov-report
+
+      - name: Upload coverage to Codecov
+        uses: codecov/codecov-action@v3
+        env:
+           CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
+        with:
+          files: lcov-report/coverage.info
diff --git a/.gitignore b/.gitignore
index 8b5f45a2..f9244fad 100644
--- a/.gitignore
+++ b/.gitignore
@@ -6,6 +6,10 @@
 *.exe
 *.dll
 *.log
+*.gcov
+*.gcno
+*.gcda
+*.dot
 .DS_Store
 .build/
 .cache/
@@ -17,6 +21,9 @@
 .vs/
 .vscode/
 
+lcov-report/
+gcovr-report/
+
 build*/
 out/
 tmp/
diff --git a/Makefile b/Makefile
index e214970f..c042bf0e 100644
--- a/Makefile
+++ b/Makefile
@@ -4,6 +4,9 @@ BUILD_TARGETS = main quantize quantize-stats perplexity embedding vdot train-tex
 # Binaries only useful for tests
 TEST_TARGETS = tests/test-llama-grammar tests/test-grammar-parser tests/test-double-float tests/test-grad0 tests/test-opt tests/test-quantize-fns tests/test-quantize-perf tests/test-sampling tests/test-tokenizer-0-llama tests/test-tokenizer-0-falcon tests/test-tokenizer-1
 
+# Code coverage output files
+COV_TARGETS = *.gcno tests/*.gcno *.gcda tests/*.gcda *.gcov tests/*.gcov lcov-report gcovr-report
+
 default: $(BUILD_TARGETS)
 
 test:
@@ -23,6 +26,18 @@ test:
 
 all: $(BUILD_TARGETS) $(TEST_TARGETS)
 
+coverage: ## Run code coverage
+	gcov -pb tests/*.cpp
+
+lcov-report: coverage ## Generate lcov report
+	mkdir -p lcov-report
+	lcov --capture --directory . --output-file lcov-report/coverage.info
+	genhtml lcov-report/coverage.info --output-directory lcov-report
+
+gcovr-report: coverage ## Generate gcovr report
+	mkdir -p gcovr-report
+	gcovr --root . --html --html-details --output gcovr-report/coverage.html
+
 ifndef UNAME_S
 UNAME_S := $(shell uname -s)
 endif
@@ -84,6 +99,11 @@ ifdef LLAMA_SERVER_VERBOSE
 	MK_CPPFLAGS += -DSERVER_VERBOSE=$(LLAMA_SERVER_VERBOSE)
 endif
 
+
+ifdef LLAMA_CODE_COVERAGE
+	CXXFLAGS += -fprofile-arcs -ftest-coverage -dumpbase ''
+endif
+
 ifdef LLAMA_DISABLE_LOGS
 	CFLAGS   += -DLOG_DISABLE_LOGS
 	CXXFLAGS += -DLOG_DISABLE_LOGS
@@ -399,7 +419,7 @@ libllama.so: llama.o ggml.o $(OBJS)
 	$(CXX) $(CXXFLAGS) -shared -fPIC -o $@ $^ $(LDFLAGS)
 
 clean:
-	rm -vf *.o tests/*.o *.so *.dll benchmark-matmult build-info.h $(BUILD_TARGETS) $(TEST_TARGETS)
+	rm -vrf *.o tests/*.o *.so *.dll benchmark-matmult build-info.h *.dot $(COV_TARGETS) $(BUILD_TARGETS) $(TEST_TARGETS)
 
 #
 # Examples