chore: ⬆️ Update antirez/ds4 to 110afdd8886586f18fc9b28bc5533152dd10e728

backend/cpp/ds4/CMakeLists.txt

@@ -85,8 +85,12 @@ elseif(DS4_GPU STREQUAL "cpu")
 endif()
 
 # Upstream splits distributed inference, tensor-parallel transport, the SSD
-# expert cache, and layer placement into GPU-agnostic translation units. Link
-# them regardless of DS4_GPU.
+# expert cache, layer placement, and image preprocessing into GPU-agnostic
+# translation units. Link them regardless of DS4_GPU. ds4_image.o holds the
+# vision preprocessing and DeepSeek4/GLM5.3 image layout helpers that ds4.c
+# calls unconditionally, so leaving it out breaks the link even for text-only
+# use. Keep this list in sync with DS4_OBJ_TARGET in the Makefile.
+list(APPEND DS4_OBJS "${DS4_DIR}/ds4_image.o")
 list(APPEND DS4_OBJS "${DS4_DIR}/ds4_distributed.o")
 list(APPEND DS4_OBJS "${DS4_DIR}/ds4_tp.o")
 list(APPEND DS4_OBJS "${DS4_DIR}/ds4_ssd.o")

backend/cpp/ds4/Makefile

@@ -1,10 +1,10 @@
 # ds4 backend Makefile.
 #
-# Upstream pin lives below as DS4_VERSION?=8db89fe083ae4d17c9a2428ccd29803d3ae8f577
+# Upstream pin lives below as DS4_VERSION?=110afdd8886586f18fc9b28bc5533152dd10e728
 # (.github/bump_deps.sh) can find and update it - matches the
 # llama-cpp / ik-llama-cpp / turboquant convention.
 
-DS4_VERSION?=8db89fe083ae4d17c9a2428ccd29803d3ae8f577
+DS4_VERSION?=110afdd8886586f18fc9b28bc5533152dd10e728
 DS4_REPO?=https://github.com/antirez/ds4
 
 CURRENT_MAKEFILE_DIR := $(dir $(abspath $(lastword $(MAKEFILE_LIST))))
@@ -80,20 +80,23 @@ endif
 endif
 
 # Upstream splits distributed inference, tensor-parallel transport, the SSD
-# expert cache, and layer placement into GPU-agnostic translation units. They
-# are shared by every GPU mode, so append them unconditionally below.
+# expert cache, layer placement, and image preprocessing into GPU-agnostic
+# translation units. They are shared by every GPU mode, so append them
+# unconditionally below. ds4_image.o carries the vision preprocessing and the
+# DeepSeek4/GLM5.3 image layout helpers that ds4.c calls, so it must be linked
+# even when no image is ever passed to the engine.
 ifeq ($(BUILD_TYPE),cublas)
     CMAKE_ARGS += -DDS4_GPU=cuda
-    DS4_OBJ_TARGET := ds4.o ds4_cuda.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o \
+    DS4_OBJ_TARGET := ds4.o ds4_image.o ds4_cuda.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o \
         cuda/mmq/ds4_ggml_stubs.o cuda/mmq/ds4_mmq.o cuda/mmq/ds4_mmq_d2r.o \
         cuda/mmq/quantize.o cuda/mmq/mmid.o cuda/mmq/mmvq.o cuda/mmq/ds4_repack.o
 else ifeq ($(UNAME_S),Darwin)
     CMAKE_ARGS += -DDS4_GPU=metal
-    DS4_OBJ_TARGET := ds4.o ds4_metal.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
+    DS4_OBJ_TARGET := ds4.o ds4_image.o ds4_metal.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 else
     # CPU reference path (Linux only - macOS CPU path is broken by VM bug per ds4 README).
     CMAKE_ARGS += -DDS4_GPU=cpu
-    DS4_OBJ_TARGET := ds4_cpu.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
+    DS4_OBJ_TARGET := ds4_cpu.o ds4_image.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 endif
 
 ifneq ($(NATIVE),true)
@@ -116,14 +119,12 @@ ds4:
 
 # Build ds4's engine object files via its own Makefile, which already encodes
 # the right per-platform compile flags (Objective-C/Metal on Darwin, nvcc on Linux+CUDA).
+# DS4_OBJ_TARGET is the single source of truth for the per-variant object list;
+# a second hand-kept copy here is what silently went stale on an upstream bump.
+# DS4_ARCH_MAKEVARS is only ever assigned inside the cublas branch above, so it
+# expands empty on the Darwin and CPU paths and needs no branch of its own here.
 ds4/ds4.o: ds4
-ifeq ($(BUILD_TYPE),cublas)
 	+$(MAKE) -C ds4 $(DS4_ARCH_MAKEVARS) $(DS4_OBJ_TARGET)
-else ifeq ($(UNAME_S),Darwin)
-	+$(MAKE) -C ds4 ds4.o ds4_metal.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
-else
-	+$(MAKE) -C ds4 ds4_cpu.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
-endif
 
 grpc-server: ds4/ds4.o
 	mkdir -p $(BUILD_DIR)