fix(stablediffusion-cpp): Switch back to upstream and update (#5880)

* sync(stablediffusion-cpp): Switch back to upstream and update Signed-off-by: Richard Palethorpe <io@richiejp.com> * fix(stablediffusion-ggml): NULL terminate options array to prevent segfault Signed-off-by: Richard Palethorpe <io@richiejp.com> * fix(build): Add BUILD_TYPE and BASE_IMAGE to all backends Signed-off-by: Richard Palethorpe <io@richiejp.com> --------- Signed-off-by: Richard Palethorpe <io@richiejp.com>
2025-12-30 22:20:20 -06:00 · 2025-07-24 15:03:18 +01:00
parent 4db1b80278
commit 8fe9fa98f2
5 changed files with 106 additions and 80 deletions
--- a/.github/workflows/bump_deps.yaml
+++ b/.github/workflows/bump_deps.yaml
@@ -21,7 +21,7 @@ jobs:
            variable: "BARKCPP_VERSION"
            branch: "main"
            file: "Makefile"
-          - repository: "richiejp/stable-diffusion.cpp"
+          - repository: "leejet/stable-diffusion.cpp"
            variable: "STABLEDIFFUSION_GGML_VERSION"
            branch: "master"
            file: "backend/go/stablediffusion-ggml/Makefile"
--- a/36
+++ b/36
@@ -145,7 +145,7 @@ backends/stablediffusion-ggml: docker-build-stablediffusion-ggml docker-save-sta

 backends/whisper: docker-build-whisper docker-save-whisper build
 	./local-ai backends install "ocifile://$(abspath ./backend-images/whisper.tar)"
-	
+
 backends/silero-vad: docker-build-silero-vad docker-save-silero-vad build
 	./local-ai backends install "ocifile://$(abspath ./backend-images/silero-vad.tar)"

@@ -359,19 +359,19 @@ backend-images:
 	mkdir -p backend-images

 docker-build-llama-cpp:
-	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg IMAGE_BASE=$(IMAGE_BASE) -t local-ai-backend:llama-cpp -f backend/Dockerfile.llama-cpp .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:llama-cpp -f backend/Dockerfile.llama-cpp .

 docker-build-bark-cpp:
-	docker build -t local-ai-backend:bark-cpp -f backend/Dockerfile.golang --build-arg BACKEND=bark-cpp .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:bark-cpp -f backend/Dockerfile.golang --build-arg BACKEND=bark-cpp .

 docker-build-piper:
-	docker build -t local-ai-backend:piper -f backend/Dockerfile.golang --build-arg BACKEND=piper .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:piper -f backend/Dockerfile.golang --build-arg BACKEND=piper .

 docker-build-local-store:
-	docker build -t local-ai-backend:local-store -f backend/Dockerfile.golang --build-arg BACKEND=local-store .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:local-store -f backend/Dockerfile.golang --build-arg BACKEND=local-store .

 docker-build-huggingface:
-	docker build -t local-ai-backend:huggingface -f backend/Dockerfile.golang --build-arg BACKEND=huggingface .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:huggingface -f backend/Dockerfile.golang --build-arg BACKEND=huggingface .

 docker-save-huggingface: backend-images
 	docker save local-ai-backend:huggingface -o backend-images/huggingface.tar
@@ -380,7 +380,7 @@ docker-save-local-store: backend-images
 	docker save local-ai-backend:local-store -o backend-images/local-store.tar

 docker-build-silero-vad:
-	docker build -t local-ai-backend:silero-vad -f backend/Dockerfile.golang --build-arg BACKEND=silero-vad .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:silero-vad -f backend/Dockerfile.golang --build-arg BACKEND=silero-vad .

 docker-save-silero-vad: backend-images
 	docker save local-ai-backend:silero-vad -o backend-images/silero-vad.tar
@@ -395,25 +395,25 @@ docker-save-bark-cpp: backend-images
 	docker save local-ai-backend:bark-cpp -o backend-images/bark-cpp.tar

 docker-build-stablediffusion-ggml:
-	docker build -t local-ai-backend:stablediffusion-ggml -f backend/Dockerfile.golang --build-arg BACKEND=stablediffusion-ggml .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:stablediffusion-ggml -f backend/Dockerfile.golang --build-arg BACKEND=stablediffusion-ggml .

 docker-save-stablediffusion-ggml: backend-images
 	docker save local-ai-backend:stablediffusion-ggml -o backend-images/stablediffusion-ggml.tar

 docker-build-rerankers:
-	docker build -t local-ai-backend:rerankers -f backend/Dockerfile.python --build-arg BACKEND=rerankers .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:rerankers -f backend/Dockerfile.python --build-arg BACKEND=rerankers .

 docker-build-vllm:
-	docker build -t local-ai-backend:vllm -f backend/Dockerfile.python --build-arg BACKEND=vllm .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:vllm -f backend/Dockerfile.python --build-arg BACKEND=vllm .

 docker-build-transformers:
-	docker build -t local-ai-backend:transformers -f backend/Dockerfile.python --build-arg BACKEND=transformers .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:transformers -f backend/Dockerfile.python --build-arg BACKEND=transformers .

 docker-build-diffusers:
-	docker build -t local-ai-backend:diffusers -f backend/Dockerfile.python --build-arg BACKEND=diffusers .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:diffusers -f backend/Dockerfile.python --build-arg BACKEND=diffusers .

 docker-build-kokoro:
-	docker build -t local-ai-backend:kokoro -f backend/Dockerfile.python --build-arg BACKEND=kokoro .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:kokoro -f backend/Dockerfile.python --build-arg BACKEND=kokoro .

 docker-build-whisper:
 	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:whisper -f backend/Dockerfile.golang --build-arg BACKEND=whisper  .
@@ -422,19 +422,19 @@ docker-save-whisper: backend-images
 	docker save local-ai-backend:whisper -o backend-images/whisper.tar

 docker-build-faster-whisper:
-	docker build -t local-ai-backend:faster-whisper -f backend/Dockerfile.python --build-arg BACKEND=faster-whisper .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:faster-whisper -f backend/Dockerfile.python --build-arg BACKEND=faster-whisper .

 docker-build-coqui:
-	docker build -t local-ai-backend:coqui -f backend/Dockerfile.python --build-arg BACKEND=coqui .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:coqui -f backend/Dockerfile.python --build-arg BACKEND=coqui .

 docker-build-bark:
-	docker build -t local-ai-backend:bark -f backend/Dockerfile.python --build-arg BACKEND=bark .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:bark -f backend/Dockerfile.python --build-arg BACKEND=bark .

 docker-build-chatterbox:
-	docker build -t local-ai-backend:chatterbox -f backend/Dockerfile.python --build-arg BACKEND=chatterbox .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:chatterbox -f backend/Dockerfile.python --build-arg BACKEND=chatterbox .

 docker-build-exllama2:
-	docker build -t local-ai-backend:exllama2 -f backend/Dockerfile.python --build-arg BACKEND=exllama2 .
+	docker build --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg BASE_IMAGE=$(BASE_IMAGE) -t local-ai-backend:exllama2 -f backend/Dockerfile.python --build-arg BACKEND=exllama2 .

 docker-build-backends: docker-build-llama-cpp docker-build-rerankers docker-build-vllm docker-build-transformers docker-build-diffusers docker-build-kokoro docker-build-faster-whisper docker-build-coqui docker-build-bark docker-build-chatterbox docker-build-exllama2

--- a/backend/go/stablediffusion-ggml/Makefile
+++ b/backend/go/stablediffusion-ggml/Makefile
@@ -18,8 +18,8 @@ GO_TAGS?=
 LD_FLAGS?=

 # stablediffusion.cpp (ggml)
-STABLEDIFFUSION_GGML_REPO?=https://github.com/richiejp/stable-diffusion.cpp
-STABLEDIFFUSION_GGML_VERSION?=10c6501bd05a697e014f1bee3a84e5664290c489
+STABLEDIFFUSION_GGML_REPO?=https://github.com/leejet/stable-diffusion.cpp
+STABLEDIFFUSION_GGML_VERSION?=1896b28ef2fd5b3643120e66979bea487385439f

 # Disable Shared libs as we are linking on static gRPC and we can't mix shared and static
 CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF
@@ -91,23 +91,18 @@ endif
 # (ggml can have different backends cpu, cuda, etc., each backend generates a .a archive)
 GGML_ARCHIVE_DIR := build/ggml/src/
 ALL_ARCHIVES := $(shell find $(GGML_ARCHIVE_DIR) -type f -name '*.a')
+ALL_OBJS := $(shell find $(GGML_ARCHIVE_DIR) -type f -name '*.o')

 # Name of the single merged library
 COMBINED_LIB := libggmlall.a

-# Rule to merge all the .a files into one
+# Instead of using the archives generated by GGML, use the object files directly to avoid overwriting objects with the same base name
 $(COMBINED_LIB): $(ALL_ARCHIVES)
-	@echo "Merging all .a into $(COMBINED_LIB)"
+	@echo "Merging all .o into $(COMBINED_LIB): $(ALL_OBJS)"
 	rm -f $@
-	mkdir -p merge-tmp
-	for a in $(ALL_ARCHIVES); do \
-		( cd merge-tmp && ar x ../$$a ); \
-	done
-	( cd merge-tmp && ar rcs ../$@ *.o )
+	ar -qc $@ $(ALL_OBJS)
 	# Ensure we have a proper index
 	ranlib $@
-	# Clean up
-	rm -rf merge-tmp

 build/libstable-diffusion.a:
 	@echo "Building SD with $(BUILD_TYPE) build type and $(CMAKE_ARGS)"
--- a/backend/go/stablediffusion-ggml/gosd.cpp
+++ b/backend/go/stablediffusion-ggml/gosd.cpp
@@ -53,9 +53,43 @@ sd_ctx_t* sd_c;

 sample_method_t sample_method;

+// Copied from the upstream CLI
+void sd_log_cb(enum sd_log_level_t level, const char* log, void* data) {
+    //SDParams* params = (SDParams*)data;
+    const char* level_str;
+
+    if (!log /*|| (!params->verbose && level <= SD_LOG_DEBUG)*/) {
+        return;
+    }
+
+    switch (level) {
+        case SD_LOG_DEBUG:
+            level_str = "DEBUG";
+            break;
+        case SD_LOG_INFO:
+            level_str = "INFO";
+            break;
+        case SD_LOG_WARN:
+            level_str = "WARN";
+            break;
+        case SD_LOG_ERROR:
+            level_str = "ERROR";
+            break;
+        default: /* Potential future-proofing */
+            level_str = "?????";
+            break;
+    }
+
+    fprintf(stderr, "[%-5s] ", level_str);
+    fputs(log, stderr);
+    fflush(stderr);
+}
+
 int load_model(char *model, char* options[], int threads, int diff) {
    fprintf (stderr, "Loading model!\n");

+    sd_set_log_callback(sd_log_cb, NULL);
+
    char *stableDiffusionModel = "";
    if (diff == 1 ) {
        stableDiffusionModel = model;
@@ -70,6 +104,8 @@ int load_model(char *model, char* options[], int threads, int diff) {
    char *scheduler = "";
    char *sampler = "";

+    fprintf(stderr, "parsing options\n");
+
    // If options is not NULL, parse options
    for (int i = 0; options[i] != NULL; i++) {
        char *optname = strtok(options[i], ":");
@@ -98,10 +134,13 @@ int load_model(char *model, char* options[], int threads, int diff) {
        }
    }

+    fprintf(stderr, "parsed options\n");
+
    int sample_method_found = -1;
-    for (int m = 0; m < N_SAMPLE_METHODS; m++) {
+    for (int m = 0; m < SAMPLE_METHOD_COUNT; m++) {
        if (!strcmp(sampler, sample_method_str[m])) {
            sample_method_found = m;
+            fprintf(stderr, "Found sampler: %s\n", sampler);
        }
    }
    if (sample_method_found == -1) {
@@ -111,7 +150,7 @@ int load_model(char *model, char* options[], int threads, int diff) {
    sample_method = (sample_method_t)sample_method_found;

    int schedule_found            = -1;
-    for (int d = 0; d < N_SCHEDULES; d++) {
+    for (int d = 0; d < SCHEDULE_COUNT; d++) {
        if (!strcmp(scheduler, schedule_str[d])) {
            schedule_found = d;
                fprintf (stderr, "Found scheduler: %s\n", scheduler);
@@ -125,30 +164,28 @@ int load_model(char *model, char* options[], int threads, int diff) {
    }

    schedule_t schedule = (schedule_t)schedule_found;
-    
+
    fprintf (stderr, "Creating context\n");
-    sd_ctx_t* sd_ctx = new_sd_ctx(model,
-                                  clip_l_path,
-                                  clip_g_path,
-                                  t5xxl_path,
-                                  stableDiffusionModel,
-                                  vae_path,
-                                  "",
-                                  "",
-                                  "",
-                                  "",
-                                  "",
-                                  false,
-                                  false,
-                                  false,
-                                  threads,
-                                  SD_TYPE_COUNT,
-                                  STD_DEFAULT_RNG,
-                                  schedule,
-                                  false,
-                                  false,
-                                  false,
-                                  false);
+    sd_ctx_params_t ctx_params;
+    sd_ctx_params_init(&ctx_params);
+    ctx_params.model_path = model;
+    ctx_params.clip_l_path = clip_l_path;
+    ctx_params.clip_g_path = clip_g_path;
+    ctx_params.t5xxl_path = t5xxl_path;
+    ctx_params.diffusion_model_path = stableDiffusionModel;
+    ctx_params.vae_path = vae_path;
+    ctx_params.taesd_path = "";
+    ctx_params.control_net_path = "";
+    ctx_params.lora_model_dir = "";
+    ctx_params.embedding_dir = "";
+    ctx_params.stacked_id_embed_dir = "";
+    ctx_params.vae_decode_only = false;
+    ctx_params.vae_tiling = false;
+    ctx_params.free_params_immediately = false;
+    ctx_params.n_threads = threads;
+    ctx_params.rng_type = STD_DEFAULT_RNG;
+    ctx_params.schedule = schedule;
+    sd_ctx_t* sd_ctx = new_sd_ctx(&ctx_params);

    if (sd_ctx == NULL) {
        fprintf (stderr, "failed loading model (generic error)\n");
@@ -169,29 +206,22 @@ int gen_image(char *text, char *negativeText, int width, int height, int steps,

    fprintf (stderr, "Generating image\n");

-    results = txt2img(sd_c,
-                            text,
-                            negativeText,
-                            -1, //clip_skip
-                            cfg_scale, // sfg_scale
-                            3.5f,
-			    0, // eta
-                            width,
-                            height,
-                            sample_method, 
-                            steps,
-                            seed,
-                            1,
-                            NULL,
-                            0.9f,
-                            20.f,
-                            false,
-                            "",
-                            skip_layers.data(),
-                            skip_layers.size(),
-                            0,
-                            0.01,
-                            0.2);
+    sd_img_gen_params_t p;
+    sd_img_gen_params_init(&p);
+
+    p.prompt = text;
+    p.negative_prompt = negativeText;
+    p.guidance.txt_cfg = cfg_scale;
+    p.guidance.slg.layers = skip_layers.data();
+    p.guidance.slg.layer_count = skip_layers.size();
+    p.width = width;
+    p.height = height;
+    p.sample_method = sample_method;
+    p.sample_steps = steps;
+    p.seed = seed;
+    p.input_id_images_path = "";
+
+    results = generate_image(sd_c, &p);

    if (results == NULL) {
        fprintf (stderr, "NO results\n");
--- a/backend/go/stablediffusion-ggml/gosd.go
+++ b/backend/go/stablediffusion-ggml/gosd.go
@@ -37,8 +37,8 @@ func (sd *SDGGML) Load(opts *pb.ModelOptions) error {

 	size := C.size_t(unsafe.Sizeof((*C.char)(nil)))
 	length := C.size_t(len(opts.Options))
-	options = (**C.char)(C.malloc(length * size))
-	view := (*[1 << 30]*C.char)(unsafe.Pointer(options))[0:len(opts.Options):len(opts.Options)]
+	options = (**C.char)(C.malloc((length + 1) * size))
+	view := (*[1 << 30]*C.char)(unsafe.Pointer(options))[0:len(opts.Options) + 1:len(opts.Options) + 1]

 	var diffusionModel int

@@ -66,6 +66,7 @@ func (sd *SDGGML) Load(opts *pb.ModelOptions) error {
 	for i, x := range oo {
 		view[i] = C.CString(x)
 	}
+	view[len(oo)] = nil

 	sd.cfgScale = opts.CFGScale