Cmake option to use conv2d direct

Author: daniandtheweb

CMakeLists.txt

@@ -31,6 +31,7 @@ option(SD_VULKAN                     "sd: vulkan backend" OFF)
 option(SD_OPENCL                     "sd: opencl backend" OFF)
 option(SD_SYCL                       "sd: sycl backend" OFF)
 option(SD_MUSA                       "sd: musa backend" OFF)
+option(SD_CONV2D_DIRECT              "sd: enable conv2d direct support" OFF)
 option(SD_FAST_SOFTMAX               "sd: x1.5 faster softmax, indeterministic (sometimes, same seed don't generate same image), cuda only" OFF)
 option(SD_BUILD_SHARED_LIBS          "sd: build shared libs" OFF)
 #option(SD_BUILD_SERVER               "sd: build server example"                           ON)
@@ -77,6 +78,11 @@ if(SD_MUSA)
     endif()
 endif()
 
+if(SD_CONV2D_DIRECT)
+    message("-- Use CONV2D Direct for VAE")
+    add_definitions(-DSD_USE_CONV2D_DIRECT)
+endif ()
+
 set(SD_LIB stable-diffusion)
 
 file(GLOB SD_LIB_SOURCES

ggml_extend.hpp

@@ -1511,10 +1511,14 @@ class Conv2d : public UnaryBlock {
             b = params["bias"];
         }
         if (direct) {
-            #if defined(SD_USE_CUDA) || defined(SD_USE_SYCL) || defined(SD_USE_METAL) || defined(SD_USE_OPENCL)
-                return ggml_nn_conv_2d(ctx, x, w, b, stride.second, stride.first, padding.second, padding.first, dilation.second, dilation.first);
+            #if defined(SD_USE_CONV2D_DIRECT)
+                #if defined(SD_USE_CUDA) || defined(SD_USE_SYCL) || defined(SD_USE_METAL) || defined(SD_USE_OPENCL)
+                    return ggml_nn_conv_2d(ctx, x, w, b, stride.second, stride.first, padding.second, padding.first, dilation.second, dilation.first);
+                #else
+                    return ggml_nn_conv_2d_direct(ctx, x, w, b, stride.second, stride.first, padding.second, padding.first, dilation.second, dilation.first);
+                #endif
             #else
-                return ggml_nn_conv_2d_direct(ctx, x, w, b, stride.second, stride.first, padding.second, padding.first, dilation.second, dilation.first);
+                return ggml_nn_conv_2d(ctx, x, w, b, stride.second, stride.first, padding.second, padding.first, dilation.second, dilation.first);
             #endif
         } else {
             return ggml_nn_conv_2d(ctx, x, w, b, stride.second, stride.first, padding.second, padding.first, dilation.second, dilation.first);