From aee5e863ad1ad4c283f7f74d3da8e910c85c1250 Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Thu, 19 Mar 2026 11:21:10 +0900
Subject: [PATCH 01/15] Remove make dependency

---
 ci/run.sh | 28 ++++++++++++----------------
 1 file changed, 12 insertions(+), 16 deletions(-)

diff --git a/ci/run.sh b/ci/run.sh
index eaf6358c0d..5da6bfb8b8 100755
--- a/ci/run.sh
+++ b/ci/run.sh
@@ -242,13 +242,13 @@ function gg_run_ctest_debug {
 
     set -e
 
-    # Check cmake, make and ctest are installed
+    # Check cmake and ctest are installed
     gg_check_build_requirements
 
     (time cmake -DCMAKE_BUILD_TYPE=Debug ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time make -j$(nproc)                                  ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake --build . --config Debug -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
 
-    (time ctest --output-on-failure -L main -E "test-opt|test-backend-ops" ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
+    (time ctest -C Debug --output-on-failure -L main -E "test-opt|test-backend-ops" ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
 
     set +e
 }
@@ -273,16 +273,16 @@ function gg_run_ctest_release {
 
     set -e
 
-    # Check cmake, make and ctest are installed
+    # Check cmake and ctest are installed
     gg_check_build_requirements
 
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
 
     if [ -z ${GG_BUILD_LOW_PERF} ]; then
-        (time ctest --output-on-failure -L 'main|python' ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
+        (time ctest -C Release --output-on-failure -L 'main|python' ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
     else
-        (time ctest --output-on-failure -L main -E test-opt ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
+        (time ctest -C Release --output-on-failure -L main -E test-opt ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
     fi
 
     set +e
@@ -340,7 +340,7 @@ function gg_run_ctest_with_model_debug {
     cd build-ci-debug
     set -e
 
-    (LLAMACPP_TEST_MODELFILE="$model" time ctest --output-on-failure -L model) 2>&1 | tee -a $OUT/${ci}-ctest.log
+    (LLAMACPP_TEST_MODELFILE="$model" time ctest -C Debug --output-on-failure -L model) 2>&1 | tee -a $OUT/${ci}-ctest.log
 
     set +e
     cd ..
@@ -353,7 +353,7 @@ function gg_run_ctest_with_model_release {
     cd build-ci-release
     set -e
 
-    (LLAMACPP_TEST_MODELFILE="$model" time ctest --output-on-failure -L model) 2>&1 | tee -a $OUT/${ci}-ctest.log
+    (LLAMACPP_TEST_MODELFILE="$model" time ctest -C Release --output-on-failure -L model) 2>&1 | tee -a $OUT/${ci}-ctest.log
 
     # test memory leaks
     #if [[ ! -z ${GG_BUILD_METAL} ]]; then
@@ -408,7 +408,7 @@ function gg_run_qwen3_0_6b {
     set -e
 
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf  --outtype f16
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-bf16.gguf --outtype bf16
@@ -557,7 +557,7 @@ function gg_run_embd_bge_small {
     set -e
 
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
@@ -602,7 +602,7 @@ function gg_run_rerank_tiny {
     set -e
 
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
@@ -652,10 +652,6 @@ function gg_check_build_requirements {
         gg_printf 'cmake not found, please install'
     fi
 
-    if ! command -v make &> /dev/null; then
-        gg_printf 'make not found, please install'
-    fi
-
     if ! command -v ctest &> /dev/null; then
         gg_printf 'ctest not found, please install'
     fi

From 80a998594398ab13061d825c99358b922dca6228 Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Mon, 23 Mar 2026 11:42:49 +0900
Subject: [PATCH 02/15] Added option to specify Ninja generator

---
 ci/run.sh | 27 +++++++++++++++++----------
 1 file changed, 17 insertions(+), 10 deletions(-)

diff --git a/ci/run.sh b/ci/run.sh
index 5da6bfb8b8..2a98b941b6 100755
--- a/ci/run.sh
+++ b/ci/run.sh
@@ -57,6 +57,13 @@ SRC=`pwd`
 CMAKE_EXTRA="-DLLAMA_FATAL_WARNINGS=${LLAMA_FATAL_WARNINGS:-ON} -DLLAMA_OPENSSL=OFF -DGGML_SCHED_NO_REALLOC=ON"
 CTEST_EXTRA=""
 
+# Default to use make unless specified for compatibility
+CMAKE_GENERATOR=(-G "Unix Makefiles")
+
+if [ -n "${GG_BUILD_NINJA}" ]; then
+    CMAKE_GENERATOR=(-G Ninja)
+fi
+
 if [ ! -z ${GG_BUILD_METAL} ]; then
     CMAKE_EXTRA="${CMAKE_EXTRA} -DGGML_METAL=ON"
 fi
@@ -245,8 +252,8 @@ function gg_run_ctest_debug {
     # Check cmake and ctest are installed
     gg_check_build_requirements
 
-    (time cmake -DCMAKE_BUILD_TYPE=Debug ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time cmake --build . --config Debug -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Debug ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake --build . --config Debug -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     (time ctest -C Debug --output-on-failure -L main -E "test-opt|test-backend-ops" ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
 
@@ -276,8 +283,8 @@ function gg_run_ctest_release {
     # Check cmake and ctest are installed
     gg_check_build_requirements
 
-    (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     if [ -z ${GG_BUILD_LOW_PERF} ]; then
         (time ctest -C Release --output-on-failure -L 'main|python' ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
@@ -407,8 +414,8 @@ function gg_run_qwen3_0_6b {
 
     set -e
 
-    (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf  --outtype f16
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-bf16.gguf --outtype bf16
@@ -556,8 +563,8 @@ function gg_run_embd_bge_small {
 
     set -e
 
-    (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
@@ -601,8 +608,8 @@ function gg_run_rerank_tiny {
 
     set -e
 
-    (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
-    (time cmake --build . --config Release -j$(nproc)        ) 2>&1 | tee -a $OUT/${ci}-make.log
+    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 

From f552c4559b85e222aab37f654da764af4283fee7 Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Mon, 23 Mar 2026 12:15:45 +0900
Subject: [PATCH 03/15] use ninja-build as default for several CI

---
 .github/workflows/build.yml | 26 ++++++++++++++------------
 1 file changed, 14 insertions(+), 12 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 6d500d3098..847afcf47e 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -1079,12 +1079,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-low-perf:
     runs-on: ubuntu-22.04-arm
@@ -1105,12 +1105,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-x64-cpu-high-perf:
     runs-on: ubuntu-22.04
@@ -1131,12 +1131,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-high-perf:
     runs-on: ubuntu-22.04-arm
@@ -1157,12 +1157,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NO_SVE=1 GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NO_SVE=1 GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-high-perf-sve:
     runs-on: ubuntu-22.04-arm
@@ -1183,12 +1183,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-kleidiai:
      runs-on: ubuntu-22.04-arm
@@ -1209,12 +1209,12 @@ jobs:
          id: depends
          run: |
            sudo apt-get update
-           sudo apt-get install -y build-essential
+           sudo apt-get install -y build-essential ninja-build
 
        - name: Test
          id: ggml-ci
          run: |
-           GG_BUILD_KLEIDIAI=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+           GG_BUILD_KLEIDIAI=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-kleidiai-graviton4:
      runs-on: ah-ubuntu_22_04-c8g_8x
@@ -1232,6 +1232,7 @@ jobs:
            sudo DEBIAN_FRONTEND=noninteractive NEEDRESTART_MODE=a \
            apt-get install -y \
             build-essential \
+            ninja-build \
             python3-venv \
             gpg \
             wget \
@@ -1262,4 +1263,5 @@ jobs:
          run: |
            GG_BUILD_KLEIDIAI=1 \
            GG_BUILD_EXTRA_TESTS_0=1 \
+           GG_BUILD_NINJA=1 \
            bash ./ci/run.sh ./tmp/results ./tmp/mnt

From 60b8cb0a2f3dd10ce9fddd6b96cdd4c429e8118f Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Tue, 24 Mar 2026 12:39:26 +0900
Subject: [PATCH 04/15] Revert "use ninja-build as default for several CI"

This reverts commit f552c4559b85e222aab37f654da764af4283fee7.
---
 .github/workflows/build.yml | 26 ++++++++++++--------------
 1 file changed, 12 insertions(+), 14 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 847afcf47e..6d500d3098 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -1079,12 +1079,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ninja-build
+          sudo apt-get install build-essential
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-low-perf:
     runs-on: ubuntu-22.04-arm
@@ -1105,12 +1105,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ninja-build
+          sudo apt-get install build-essential
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-x64-cpu-high-perf:
     runs-on: ubuntu-22.04
@@ -1131,12 +1131,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ninja-build
+          sudo apt-get install build-essential
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-high-perf:
     runs-on: ubuntu-22.04-arm
@@ -1157,12 +1157,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ninja-build
+          sudo apt-get install build-essential
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NO_SVE=1 GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NO_SVE=1 GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-high-perf-sve:
     runs-on: ubuntu-22.04-arm
@@ -1183,12 +1183,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ninja-build
+          sudo apt-get install build-essential
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-kleidiai:
      runs-on: ubuntu-22.04-arm
@@ -1209,12 +1209,12 @@ jobs:
          id: depends
          run: |
            sudo apt-get update
-           sudo apt-get install -y build-essential ninja-build
+           sudo apt-get install -y build-essential
 
        - name: Test
          id: ggml-ci
          run: |
-           GG_BUILD_KLEIDIAI=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+           GG_BUILD_KLEIDIAI=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-kleidiai-graviton4:
      runs-on: ah-ubuntu_22_04-c8g_8x
@@ -1232,7 +1232,6 @@ jobs:
            sudo DEBIAN_FRONTEND=noninteractive NEEDRESTART_MODE=a \
            apt-get install -y \
             build-essential \
-            ninja-build \
             python3-venv \
             gpg \
             wget \
@@ -1263,5 +1262,4 @@ jobs:
          run: |
            GG_BUILD_KLEIDIAI=1 \
            GG_BUILD_EXTRA_TESTS_0=1 \
-           GG_BUILD_NINJA=1 \
            bash ./ci/run.sh ./tmp/results ./tmp/mnt

From 29487c45f0c7454aa140ee22bf4f311f2b9172e4 Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Tue, 24 Mar 2026 12:49:35 +0900
Subject: [PATCH 05/15] changed use plain string rather than arrays

---
 ci/run.sh | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/ci/run.sh b/ci/run.sh
index 2a98b941b6..a17d01e0c0 100755
--- a/ci/run.sh
+++ b/ci/run.sh
@@ -58,10 +58,10 @@ CMAKE_EXTRA="-DLLAMA_FATAL_WARNINGS=${LLAMA_FATAL_WARNINGS:-ON} -DLLAMA_OPENSSL=
 CTEST_EXTRA=""
 
 # Default to use make unless specified for compatibility
-CMAKE_GENERATOR=(-G "Unix Makefiles")
+CMAKE_GENERATOR="Unix Makefiles"
 
 if [ -n "${GG_BUILD_NINJA}" ]; then
-    CMAKE_GENERATOR=(-G Ninja)
+    CMAKE_GENERATOR="Ninja"
 fi
 
 if [ ! -z ${GG_BUILD_METAL} ]; then
@@ -252,7 +252,7 @@ function gg_run_ctest_debug {
     # Check cmake and ctest are installed
     gg_check_build_requirements
 
-    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Debug ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Debug ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time cmake --build . --config Debug -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     (time ctest -C Debug --output-on-failure -L main -E "test-opt|test-backend-ops" ${CTEST_EXTRA}) 2>&1 | tee -a $OUT/${ci}-ctest.log
@@ -283,7 +283,7 @@ function gg_run_ctest_release {
     # Check cmake and ctest are installed
     gg_check_build_requirements
 
-    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     if [ -z ${GG_BUILD_LOW_PERF} ]; then
@@ -414,7 +414,7 @@ function gg_run_qwen3_0_6b {
 
     set -e
 
-    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf  --outtype f16
@@ -563,7 +563,7 @@ function gg_run_embd_bge_small {
 
     set -e
 
-    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
@@ -608,7 +608,7 @@ function gg_run_rerank_tiny {
 
     set -e
 
-    (time cmake "${CMAKE_GENERATOR[@]}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
+    (time cmake -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time cmake --build . --config Release -j$(nproc)) 2>&1 | tee -a $OUT/${ci}-make.log
 
     python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf

From e7304e2f0e59a3db902fa1b93d922cdfc7511838 Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Tue, 24 Mar 2026 14:44:59 +0900
Subject: [PATCH 06/15] Enabled ninja build by default for experimentation

---
 .github/workflows/build.yml | 63 +++++++++++++++++++++++--------------
 1 file changed, 39 insertions(+), 24 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 6d500d3098..8fe42dbed0 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -80,6 +80,7 @@ jobs:
         run: |
           sysctl -a
           cmake -B build \
+            -G "Ninja Multi-Config" \
             -DCMAKE_BUILD_RPATH="@loader_path" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DLLAMA_BUILD_BORINGSSL=ON \
@@ -118,6 +119,7 @@ jobs:
           # Metal is disabled due to intermittent failures with Github runners not having a GPU:
           # https://github.com/ggml-org/llama.cpp/actions/runs/8635935781/job/23674807267#step:5:2313
           cmake -B build \
+            -G "Ninja Multi-Config" \
             -DCMAKE_BUILD_RPATH="@loader_path" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DLLAMA_BUILD_BORINGSSL=ON \
@@ -165,7 +167,7 @@ jobs:
         id: cmake_build
         run: |
           export CMAKE_PREFIX_PATH=dawn
-          cmake -B build -DGGML_WEBGPU=ON -DGGML_METAL=OFF -DGGML_BLAS=OFF
+          cmake -B build -G "Ninja Multi-Config" -DGGML_WEBGPU=ON -DGGML_METAL=OFF -DGGML_BLAS=OFF
           cmake --build build --config Release -j $(sysctl -n hw.logicalcpu)
 
       - name: Test
@@ -209,7 +211,7 @@ jobs:
           sudo apt-get install -y --no-install-recommends \
             python3 python3-pip python3-dev \
             libjpeg-dev build-essential libssl-dev \
-            git-lfs
+            git-lfs ninja-build
 
       - name: Python Dependencies
         id: python_depends
@@ -229,6 +231,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build \
+            -G "Ninja Multi-Config" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DGGML_RPC=ON
           cmake --build build --config Release -j $(nproc)
@@ -274,12 +277,13 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential libssl-dev
+          sudo apt-get install build-essential libssl-dev ninja-build
 
       - name: Build
         id: cmake_build
         run: |
           cmake -B build \
+            -G "Ninja Multi-Config" \
             -DGGML_RPC=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -300,12 +304,13 @@ jobs:
       - name: Dependencies
         id: depends
         run: |
-          sudo apt-get install -y glslc libvulkan-dev libssl-dev
+          sudo apt-get install -y glslc libvulkan-dev libssl-dev ninja-build
 
       - name: Configure
         id: cmake_configure
         run: |
           cmake -B build \
+            -G "Ninja" \
             -DCMAKE_BUILD_TYPE=RelWithDebInfo \
             -DGGML_BACKEND_DL=ON \
             -DGGML_CPU_ALL_VARIANTS=ON \
@@ -336,7 +341,8 @@ jobs:
         run: |
           sudo add-apt-repository -y ppa:kisak/kisak-mesa
           sudo apt-get update -y
-          sudo apt-get install -y build-essential mesa-vulkan-drivers libxcb-xinput0 libxcb-xinerama0 libxcb-cursor-dev libssl-dev
+          sudo apt-get install -y build-essential mesa-vulkan-drivers \
+            libxcb-xinput0 libxcb-xinerama0 libxcb-cursor-dev libssl-dev ninja-build
 
       - name: Get latest Vulkan SDK version
         id: vulkan_sdk_version
@@ -377,6 +383,7 @@ jobs:
         run: |
           export Dawn_DIR=dawn/lib64/cmake/Dawn
           cmake -B build \
+            -G "Ninja Multi-Config" \
             -DGGML_WEBGPU=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -415,11 +422,12 @@ jobs:
         run: |
           source emsdk/emsdk_env.sh
           emcmake cmake -B build-wasm \
+            -G "Ninja Multi-Config" \
             -DGGML_WEBGPU=ON \
             -DLLAMA_OPENSSL=OFF \
             -DEMDAWNWEBGPU_DIR=emdawnwebgpu_pkg
 
-          cmake --build build-wasm --target test-backend-ops -j $(nproc)
+          cmake --build build-wasm --config Release --target test-backend-ops -j $(nproc)
 
   ubuntu-22-hip:
     runs-on: ubuntu-22.04
@@ -434,7 +442,7 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install -y build-essential git cmake rocblas-dev hipblas-dev libssl-dev rocwmma-dev
+          sudo apt-get install -y build-essential git cmake rocblas-dev hipblas-dev libssl-dev rocwmma-dev ninja-build
 
       - name: ccache
         uses: ggml-org/ccache-action@v1.2.21
@@ -447,6 +455,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build -S . \
+            -G "Ninja Multi-Config" \
             -DCMAKE_HIP_COMPILER="$(hipconfig -l)/clang" \
             -DGGML_HIP_ROCWMMA_FATTN=ON \
             -DGGML_HIP=ON
@@ -465,7 +474,7 @@ jobs:
         id: depends
         run: |
           apt-get update
-          apt-get install -y build-essential git cmake libssl-dev
+          apt-get install -y build-essential git cmake libssl-dev ninja-build
 
       - name: ccache
         uses: ggml-org/ccache-action@v1.2.21
@@ -478,6 +487,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build -S . \
+            -G "Ninja Multi-Config" \
             -DGGML_MUSA=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -502,7 +512,7 @@ jobs:
         shell: bash
         run: |
           sudo apt update
-          sudo apt install intel-oneapi-compiler-dpcpp-cpp libssl-dev
+          sudo apt install intel-oneapi-compiler-dpcpp-cpp libssl-dev ninja-build
 
       - name: install oneAPI MKL library
         shell: bash
@@ -525,6 +535,7 @@ jobs:
         run: |
           source /opt/intel/oneapi/setvars.sh
           cmake -B build \
+            -G "Ninja Multi-Config" \
             -DGGML_SYCL=ON \
             -DCMAKE_C_COMPILER=icx \
             -DCMAKE_CXX_COMPILER=icpx
@@ -551,7 +562,7 @@ jobs:
         shell: bash
         run: |
           sudo apt update
-          sudo apt install intel-oneapi-compiler-dpcpp-cpp libssl-dev
+          sudo apt install intel-oneapi-compiler-dpcpp-cpp libssl-dev ninja-build
 
       - name: install oneAPI MKL library
         shell: bash
@@ -574,6 +585,7 @@ jobs:
         run: |
           source /opt/intel/oneapi/setvars.sh
           cmake -B build \
+            -G "Ninja Multi-Config" \
             -DGGML_SYCL=ON \
             -DCMAKE_C_COMPILER=icx \
             -DCMAKE_CXX_COMPILER=icpx \
@@ -679,7 +691,7 @@ jobs:
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_OPENMP=OFF -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
           - build: 'vulkan-x64'
             arch: 'x64'
-            defines: '-DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
+            defines: '-G "Ninja" -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
           - build: 'llvm-arm64'
             arch: 'arm64'
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
@@ -977,7 +989,7 @@ jobs:
           sudo apt-get update
 
           # Install necessary packages
-          sudo apt-get install -y libatomic1 libtsan2 gcc-14 g++-14 rustup cmake build-essential libssl-dev wget ccache git-lfs
+          sudo apt-get install -y libatomic1 libtsan2 gcc-14 g++-14 rustup cmake build-essential libssl-dev wget ccache git-lfs ninja-build
 
           # Set gcc-14 and g++-14 as the default compilers
           sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-14 100
@@ -1028,6 +1040,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build \
+            -G "Ninja" \
             -DCMAKE_BUILD_TYPE=Release \
             -DGGML_OPENMP=OFF \
             -DLLAMA_BUILD_EXAMPLES=ON \
@@ -1079,12 +1092,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-low-perf:
     runs-on: ubuntu-22.04-arm
@@ -1105,12 +1118,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-x64-cpu-high-perf:
     runs-on: ubuntu-22.04
@@ -1131,12 +1144,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-high-perf:
     runs-on: ubuntu-22.04-arm
@@ -1157,12 +1170,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NO_SVE=1 GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_HIGH_PERF=1 GG_BUILD_NO_SVE=1 GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-high-perf-sve:
     runs-on: ubuntu-22.04-arm
@@ -1183,12 +1196,12 @@ jobs:
         id: depends
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential
+          sudo apt-get install build-essential ninja-build
 
       - name: Test
         id: ggml-ci
         run: |
-          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          LLAMA_ARG_THREADS=$(nproc) GG_BUILD_NO_BF16=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-kleidiai:
      runs-on: ubuntu-22.04-arm
@@ -1209,12 +1222,12 @@ jobs:
          id: depends
          run: |
            sudo apt-get update
-           sudo apt-get install -y build-essential
+           sudo apt-get install -y build-essential ninja-build
 
        - name: Test
          id: ggml-ci
          run: |
-           GG_BUILD_KLEIDIAI=1 GG_BUILD_EXTRA_TESTS_0=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+           GG_BUILD_KLEIDIAI=1 GG_BUILD_EXTRA_TESTS_0=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
 
   ggml-ci-arm64-cpu-kleidiai-graviton4:
      runs-on: ah-ubuntu_22_04-c8g_8x
@@ -1232,6 +1245,7 @@ jobs:
            sudo DEBIAN_FRONTEND=noninteractive NEEDRESTART_MODE=a \
            apt-get install -y \
             build-essential \
+            ninja-build \
             python3-venv \
             gpg \
             wget \
@@ -1262,4 +1276,5 @@ jobs:
          run: |
            GG_BUILD_KLEIDIAI=1 \
            GG_BUILD_EXTRA_TESTS_0=1 \
+           GG_BUILD_NINJA=1 \
            bash ./ci/run.sh ./tmp/results ./tmp/mnt

From 5bc9a63ae306b6ae63ffb1b1db715f377fbcbd5d Mon Sep 17 00:00:00 2001
From: Aaron Teo <aaron.teo1@ibm.com>
Date: Tue, 24 Mar 2026 13:50:53 +0800
Subject: [PATCH 07/15] ci: add run.sh to test conditions to trigger GitHub CI
 and self-hosted runners

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
---
 .github/workflows/build-self-hosted.yml | 1 +
 .github/workflows/build.yml             | 1 +
 2 files changed, 2 insertions(+)

diff --git a/.github/workflows/build-self-hosted.yml b/.github/workflows/build-self-hosted.yml
index 2944cb8401..52cd850074 100644
--- a/.github/workflows/build-self-hosted.yml
+++ b/.github/workflows/build-self-hosted.yml
@@ -27,6 +27,7 @@ on:
     types: [opened, synchronize, reopened]
     paths: [
       '.github/workflows/build-self-hosted.yml',
+      'ci/run.sh',
       '**/CMakeLists.txt',
       '**/.cmake',
       '**/*.h',
diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 8fe42dbed0..8a78677985 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -29,6 +29,7 @@ on:
     paths: [
       '.github/workflows/build.yml',
       '.github/workflows/build-cmake-pkg.yml',
+      'ci/run.sh',
       '**/CMakeLists.txt',
       '**/.cmake',
       '**/*.h',

From a923ae3977edc8c584fbb4625001727b87992db0 Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Tue, 24 Mar 2026 15:11:11 +0900
Subject: [PATCH 08/15] Enabled ninja build by default on self-hosted envs for
 experimentation

---
 .github/workflows/build-self-hosted.yml | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/.github/workflows/build-self-hosted.yml b/.github/workflows/build-self-hosted.yml
index 52cd850074..8905825ba6 100644
--- a/.github/workflows/build-self-hosted.yml
+++ b/.github/workflows/build-self-hosted.yml
@@ -68,7 +68,7 @@ jobs:
         id: ggml-ci
         run: |
           nvidia-smi
-          GG_BUILD_CUDA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
+          GG_BUILD_CUDA=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
 
   ggml-ci-nvidia-vulkan-cm:
     runs-on: [self-hosted, Linux, NVIDIA]
@@ -82,7 +82,7 @@ jobs:
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 GGML_VK_DISABLE_COOPMAT2=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
+          GG_BUILD_VULKAN=1 GGML_VK_DISABLE_COOPMAT2=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
 
   ggml-ci-nvidia-vulkan-cm2:
     runs-on: [self-hosted, Linux, NVIDIA, COOPMAT2]
@@ -96,7 +96,7 @@ jobs:
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
+          GG_BUILD_VULKAN=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
 
   # TODO: provision AMX-compatible machine
   #ggml-ci-cpu-amx:
@@ -153,7 +153,7 @@ jobs:
       - name: Test
         id: ggml-ci
         run: |
-          GG_BUILD_METAL=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
+          GG_BUILD_METAL=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-mac-webgpu:
     runs-on: [self-hosted, macOS, ARM64]
@@ -180,7 +180,7 @@ jobs:
       - name: Test
         id: ggml-ci
         run: |
-          GG_BUILD_WEBGPU=1 GG_BUILD_WEBGPU_DAWN_PREFIX="$GITHUB_WORKSPACE/dawn" \
+          GG_BUILD_WEBGPU=1 GG_BUILD_WEBGPU_DAWN_PREFIX="$GITHUB_WORKSPACE/dawn" GG_BUILD_NINJA=1 \
             bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-mac-vulkan:
@@ -195,7 +195,7 @@ jobs:
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
+          GG_BUILD_VULKAN=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-linux-intel-vulkan:
     runs-on: [self-hosted, Linux, Intel]
@@ -211,7 +211,7 @@ jobs:
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
+          GG_BUILD_VULKAN=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-intel-openvino-gpu-low-perf:
     runs-on: [self-hosted, Linux, Intel, OpenVINO]
@@ -243,4 +243,4 @@ jobs:
         id: ggml-ci
         run: |
           source ./openvino_toolkit/setupvars.sh
-          GG_BUILD_OPENVINO=1 GGML_OPENVINO_DEVICE=GPU GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          GG_BUILD_OPENVINO=1 GGML_OPENVINO_DEVICE=GPU GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt

From 0a6263b1ee67e46023cfef33c81c213e5cb149c1 Mon Sep 17 00:00:00 2001
From: Aaron Teo <aaron.teo1@ibm.com>
Date: Tue, 24 Mar 2026 14:13:52 +0800
Subject: [PATCH 09/15] ci: revert generator to ninja instead of ninja
 multi-config

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
---
 .github/workflows/build.yml | 32 ++++++++++++++++----------------
 1 file changed, 16 insertions(+), 16 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 8a78677985..3bce418e36 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -81,7 +81,7 @@ jobs:
         run: |
           sysctl -a
           cmake -B build \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DCMAKE_BUILD_RPATH="@loader_path" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DLLAMA_BUILD_BORINGSSL=ON \
@@ -120,7 +120,7 @@ jobs:
           # Metal is disabled due to intermittent failures with Github runners not having a GPU:
           # https://github.com/ggml-org/llama.cpp/actions/runs/8635935781/job/23674807267#step:5:2313
           cmake -B build \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DCMAKE_BUILD_RPATH="@loader_path" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DLLAMA_BUILD_BORINGSSL=ON \
@@ -168,7 +168,7 @@ jobs:
         id: cmake_build
         run: |
           export CMAKE_PREFIX_PATH=dawn
-          cmake -B build -G "Ninja Multi-Config" -DGGML_WEBGPU=ON -DGGML_METAL=OFF -DGGML_BLAS=OFF
+          cmake -B build -G "Ninja" -DGGML_WEBGPU=ON -DGGML_METAL=OFF -DGGML_BLAS=OFF
           cmake --build build --config Release -j $(sysctl -n hw.logicalcpu)
 
       - name: Test
@@ -232,7 +232,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DGGML_RPC=ON
           cmake --build build --config Release -j $(nproc)
@@ -284,7 +284,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DGGML_RPC=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -384,7 +384,7 @@ jobs:
         run: |
           export Dawn_DIR=dawn/lib64/cmake/Dawn
           cmake -B build \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DGGML_WEBGPU=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -423,7 +423,7 @@ jobs:
         run: |
           source emsdk/emsdk_env.sh
           emcmake cmake -B build-wasm \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DGGML_WEBGPU=ON \
             -DLLAMA_OPENSSL=OFF \
             -DEMDAWNWEBGPU_DIR=emdawnwebgpu_pkg
@@ -456,7 +456,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build -S . \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DCMAKE_HIP_COMPILER="$(hipconfig -l)/clang" \
             -DGGML_HIP_ROCWMMA_FATTN=ON \
             -DGGML_HIP=ON
@@ -488,7 +488,7 @@ jobs:
         id: cmake_build
         run: |
           cmake -B build -S . \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DGGML_MUSA=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -536,7 +536,7 @@ jobs:
         run: |
           source /opt/intel/oneapi/setvars.sh
           cmake -B build \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DGGML_SYCL=ON \
             -DCMAKE_C_COMPILER=icx \
             -DCMAKE_CXX_COMPILER=icpx
@@ -586,7 +586,7 @@ jobs:
         run: |
           source /opt/intel/oneapi/setvars.sh
           cmake -B build \
-            -G "Ninja Multi-Config" \
+            -G "Ninja" \
             -DGGML_SYCL=ON \
             -DCMAKE_C_COMPILER=icx \
             -DCMAKE_CXX_COMPILER=icpx \
@@ -686,19 +686,19 @@ jobs:
         include:
           - build: 'cpu-x64 (static)'
             arch: 'x64'
-            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DBUILD_SHARED_LIBS=OFF'
+            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DBUILD_SHARED_LIBS=OFF'
           - build: 'openblas-x64'
             arch: 'x64'
-            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_OPENMP=OFF -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
+            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_OPENMP=OFF -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
           - build: 'vulkan-x64'
             arch: 'x64'
             defines: '-G "Ninja" -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
           - build: 'llvm-arm64'
             arch: 'arm64'
-            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
+            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
           - build: 'llvm-arm64-opencl-adreno'
             arch: 'arm64'
-            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/opencl-arm64-release" -DGGML_OPENCL=ON -DGGML_OPENCL_USE_ADRENO_KERNELS=ON'
+            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/opencl-arm64-release" -DGGML_OPENCL=ON -DGGML_OPENCL_USE_ADRENO_KERNELS=ON'
 
     steps:
       - name: Clone
@@ -867,7 +867,7 @@ jobs:
         # TODO: Remove GGML_CUDA_CUB_3DOT2 flag once CCCL 3.2 is bundled within CTK and that CTK version is used in this project
         run: |
           call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvarsall.bat" x64
-          cmake -S . -B build -G "Ninja Multi-Config" ^
+          cmake -S . -B build -G "Ninja" ^
             -DLLAMA_BUILD_SERVER=ON ^
             -DLLAMA_BUILD_BORINGSSL=ON ^
             -DGGML_NATIVE=OFF ^

From 3908b9675b23084370d842e1703751e518011c13 Mon Sep 17 00:00:00 2001
From: Aaron Teo <aaron.teo1@ibm.com>
Date: Tue, 24 Mar 2026 14:18:31 +0800
Subject: [PATCH 10/15] ci: install ninja-build for self-hosted workflows

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
---
 .github/workflows/build-self-hosted.yml | 40 +++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/.github/workflows/build-self-hosted.yml b/.github/workflows/build-self-hosted.yml
index 8905825ba6..f70c19e52a 100644
--- a/.github/workflows/build-self-hosted.yml
+++ b/.github/workflows/build-self-hosted.yml
@@ -64,6 +64,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Test
         id: ggml-ci
         run: |
@@ -78,6 +83,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Test
         id: ggml-ci
         run: |
@@ -92,6 +102,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Test
         id: ggml-ci
         run: |
@@ -150,6 +165,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Test
         id: ggml-ci
         run: |
@@ -163,6 +183,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Dawn Dependency
         id: dawn-depends
         run: |
@@ -191,6 +216,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Test
         id: ggml-ci
         run: |
@@ -207,6 +237,11 @@ jobs:
         with:
           persist-credentials: false
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Test
         id: ggml-ci
         run: |
@@ -226,6 +261,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
+      - name: Install Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ninja-build
+
       - name: Setup OpenVINO Toolkit
         uses: ./.github/actions/linux-setup-openvino
         with:

From c06031138fb866aaf0e69f434b5f59f61515e400 Mon Sep 17 00:00:00 2001
From: Aaron Teo <aaron.teo1@ibm.com>
Date: Tue, 24 Mar 2026 14:23:47 +0800
Subject: [PATCH 11/15] ci: revert ninja from self-hosted runners

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
---
 .github/workflows/build-self-hosted.yml | 52 ++++---------------------
 1 file changed, 8 insertions(+), 44 deletions(-)

diff --git a/.github/workflows/build-self-hosted.yml b/.github/workflows/build-self-hosted.yml
index f70c19e52a..2510db7bdd 100644
--- a/.github/workflows/build-self-hosted.yml
+++ b/.github/workflows/build-self-hosted.yml
@@ -27,7 +27,6 @@ on:
     types: [opened, synchronize, reopened]
     paths: [
       '.github/workflows/build-self-hosted.yml',
-      'ci/run.sh',
       '**/CMakeLists.txt',
       '**/.cmake',
       '**/*.h',
@@ -64,16 +63,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Test
         id: ggml-ci
         run: |
           nvidia-smi
-          GG_BUILD_CUDA=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
+          GG_BUILD_CUDA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
 
   ggml-ci-nvidia-vulkan-cm:
     runs-on: [self-hosted, Linux, NVIDIA]
@@ -83,16 +77,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Test
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 GGML_VK_DISABLE_COOPMAT2=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
+          GG_BUILD_VULKAN=1 GGML_VK_DISABLE_COOPMAT2=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
 
   ggml-ci-nvidia-vulkan-cm2:
     runs-on: [self-hosted, Linux, NVIDIA, COOPMAT2]
@@ -111,7 +100,7 @@ jobs:
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
+          GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
 
   # TODO: provision AMX-compatible machine
   #ggml-ci-cpu-amx:
@@ -165,15 +154,10 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Test
         id: ggml-ci
         run: |
-          GG_BUILD_METAL=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
+          GG_BUILD_METAL=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-mac-webgpu:
     runs-on: [self-hosted, macOS, ARM64]
@@ -183,11 +167,6 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Dawn Dependency
         id: dawn-depends
         run: |
@@ -205,7 +184,7 @@ jobs:
       - name: Test
         id: ggml-ci
         run: |
-          GG_BUILD_WEBGPU=1 GG_BUILD_WEBGPU_DAWN_PREFIX="$GITHUB_WORKSPACE/dawn" GG_BUILD_NINJA=1 \
+          GG_BUILD_WEBGPU=1 GG_BUILD_WEBGPU_DAWN_PREFIX="$GITHUB_WORKSPACE/dawn" \
             bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-mac-vulkan:
@@ -216,16 +195,11 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Test
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
+          GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-linux-intel-vulkan:
     runs-on: [self-hosted, Linux, Intel]
@@ -237,16 +211,11 @@ jobs:
         with:
           persist-credentials: false
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Test
         id: ggml-ci
         run: |
           vulkaninfo --summary
-          GG_BUILD_VULKAN=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
+          GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
 
   ggml-ci-intel-openvino-gpu-low-perf:
     runs-on: [self-hosted, Linux, Intel, OpenVINO]
@@ -261,11 +230,6 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Setup OpenVINO Toolkit
         uses: ./.github/actions/linux-setup-openvino
         with:
@@ -283,4 +247,4 @@ jobs:
         id: ggml-ci
         run: |
           source ./openvino_toolkit/setupvars.sh
-          GG_BUILD_OPENVINO=1 GGML_OPENVINO_DEVICE=GPU GG_BUILD_LOW_PERF=1 GG_BUILD_NINJA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
+          GG_BUILD_OPENVINO=1 GGML_OPENVINO_DEVICE=GPU GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt

From 231d441a4af34dbb024fd6bc47bcaae43b135a57 Mon Sep 17 00:00:00 2001
From: Aaron Teo <aaron.teo1@ibm.com>
Date: Tue, 24 Mar 2026 14:25:10 +0800
Subject: [PATCH 12/15] ci: missed one self-hosted step

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
---
 .github/workflows/build-self-hosted.yml | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/.github/workflows/build-self-hosted.yml b/.github/workflows/build-self-hosted.yml
index 2510db7bdd..2944cb8401 100644
--- a/.github/workflows/build-self-hosted.yml
+++ b/.github/workflows/build-self-hosted.yml
@@ -91,11 +91,6 @@ jobs:
         id: checkout
         uses: actions/checkout@v6
 
-      - name: Install Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y ninja-build
-
       - name: Test
         id: ggml-ci
         run: |

From 7842cf622c34e6962b540d82f03ae5adf59a4524 Mon Sep 17 00:00:00 2001
From: Aaron Teo <aaron.teo1@ibm.com>
Date: Tue, 24 Mar 2026 14:33:58 +0800
Subject: [PATCH 13/15] ci: fix windows ci errors from an errenous revert

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
---
 .github/workflows/build.yml | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 3bce418e36..8161d08d53 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -686,19 +686,19 @@ jobs:
         include:
           - build: 'cpu-x64 (static)'
             arch: 'x64'
-            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DBUILD_SHARED_LIBS=OFF'
+            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DBUILD_SHARED_LIBS=OFF'
           - build: 'openblas-x64'
             arch: 'x64'
-            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_OPENMP=OFF -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
+            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_OPENMP=OFF -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
           - build: 'vulkan-x64'
             arch: 'x64'
-            defines: '-G "Ninja" -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
+            defines: '-G "Ninja Multi-Config" -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
           - build: 'llvm-arm64'
             arch: 'arm64'
-            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
+            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
           - build: 'llvm-arm64-opencl-adreno'
             arch: 'arm64'
-            defines: '-G "Ninja" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/opencl-arm64-release" -DGGML_OPENCL=ON -DGGML_OPENCL_USE_ADRENO_KERNELS=ON'
+            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/opencl-arm64-release" -DGGML_OPENCL=ON -DGGML_OPENCL_USE_ADRENO_KERNELS=ON'
 
     steps:
       - name: Clone

From a987c02a567e735f7964df54457fc73a24dd220f Mon Sep 17 00:00:00 2001
From: "Nakasaka, Masato" <masato.nakasaka@intel.com>
Date: Tue, 24 Mar 2026 16:50:38 +0900
Subject: [PATCH 14/15] Added explicit build types for Ninja

Also reverted some needless change
---
 .github/workflows/build.yml | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 8161d08d53..fc19bb9c84 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -82,6 +82,7 @@ jobs:
           sysctl -a
           cmake -B build \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DCMAKE_BUILD_RPATH="@loader_path" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DLLAMA_BUILD_BORINGSSL=ON \
@@ -121,6 +122,7 @@ jobs:
           # https://github.com/ggml-org/llama.cpp/actions/runs/8635935781/job/23674807267#step:5:2313
           cmake -B build \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DCMAKE_BUILD_RPATH="@loader_path" \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DLLAMA_BUILD_BORINGSSL=ON \
@@ -168,7 +170,7 @@ jobs:
         id: cmake_build
         run: |
           export CMAKE_PREFIX_PATH=dawn
-          cmake -B build -G "Ninja" -DGGML_WEBGPU=ON -DGGML_METAL=OFF -DGGML_BLAS=OFF
+          cmake -B build -G "Ninja" -DCMAKE_BUILD_TYPE=Release -DGGML_WEBGPU=ON -DGGML_METAL=OFF -DGGML_BLAS=OFF
           cmake --build build --config Release -j $(sysctl -n hw.logicalcpu)
 
       - name: Test
@@ -233,6 +235,7 @@ jobs:
         run: |
           cmake -B build \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DGGML_RPC=ON
           cmake --build build --config Release -j $(nproc)
@@ -285,6 +288,7 @@ jobs:
         run: |
           cmake -B build \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DGGML_RPC=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -385,6 +389,7 @@ jobs:
           export Dawn_DIR=dawn/lib64/cmake/Dawn
           cmake -B build \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DGGML_WEBGPU=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -424,6 +429,7 @@ jobs:
           source emsdk/emsdk_env.sh
           emcmake cmake -B build-wasm \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DGGML_WEBGPU=ON \
             -DLLAMA_OPENSSL=OFF \
             -DEMDAWNWEBGPU_DIR=emdawnwebgpu_pkg
@@ -457,6 +463,7 @@ jobs:
         run: |
           cmake -B build -S . \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DCMAKE_HIP_COMPILER="$(hipconfig -l)/clang" \
             -DGGML_HIP_ROCWMMA_FATTN=ON \
             -DGGML_HIP=ON
@@ -489,6 +496,7 @@ jobs:
         run: |
           cmake -B build -S . \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DGGML_MUSA=ON
           cmake --build build --config Release -j $(nproc)
 
@@ -537,6 +545,7 @@ jobs:
           source /opt/intel/oneapi/setvars.sh
           cmake -B build \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DGGML_SYCL=ON \
             -DCMAKE_C_COMPILER=icx \
             -DCMAKE_CXX_COMPILER=icpx
@@ -587,6 +596,7 @@ jobs:
           source /opt/intel/oneapi/setvars.sh
           cmake -B build \
             -G "Ninja" \
+            -DCMAKE_BUILD_TYPE=Release \
             -DGGML_SYCL=ON \
             -DCMAKE_C_COMPILER=icx \
             -DCMAKE_CXX_COMPILER=icpx \
@@ -692,7 +702,7 @@ jobs:
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_OPENMP=OFF -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
           - build: 'vulkan-x64'
             arch: 'x64'
-            defines: '-G "Ninja Multi-Config" -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
+            defines: '-G "Ninja" -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
           - build: 'llvm-arm64'
             arch: 'arm64'
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
@@ -867,7 +877,7 @@ jobs:
         # TODO: Remove GGML_CUDA_CUB_3DOT2 flag once CCCL 3.2 is bundled within CTK and that CTK version is used in this project
         run: |
           call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvarsall.bat" x64
-          cmake -S . -B build -G "Ninja" ^
+          cmake -S . -B build -G "Ninja Multi-Config" ^
             -DLLAMA_BUILD_SERVER=ON ^
             -DLLAMA_BUILD_BORINGSSL=ON ^
             -DGGML_NATIVE=OFF ^

From b9cb6b651b8b0ac1ecc66faa2b8b9758c5a2c964 Mon Sep 17 00:00:00 2001
From: Aaron Teo <aaron.teo1@ibm.com>
Date: Tue, 24 Mar 2026 16:49:43 +0800
Subject: [PATCH 15/15] ci: use ninja multi-config for vulkan-x64 build

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
---
 .github/workflows/build.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index fc19bb9c84..ea0b6e15cc 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -702,7 +702,7 @@ jobs:
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_OPENMP=OFF -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
           - build: 'vulkan-x64'
             arch: 'x64'
-            defines: '-G "Ninja" -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
+            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/x64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DGGML_RPC=ON -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DGGML_VULKAN=ON'
           - build: 'llvm-arm64'
             arch: 'arm64'
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'