opencl: refine logic for selecting Adreno kernels

Author: lhez

ggml/src/ggml-opencl/ggml-opencl.cpp

@@ -1970,8 +1970,14 @@ static enum ggml_status ggml_backend_opencl_buffer_init_tensor(ggml_backend_buff
 
 // The optimized gemm and gemv kernels are used for large matrices without batch.
 // tensor is the quantized weights matrix.
-inline bool use_adreno_kernels(const ggml_tensor *tensor) {
-    return tensor->ne[0] >= 512 && tensor->ne[1] >= 512 &&
+inline bool use_adreno_kernels(const ggml_backend_opencl_context *backend_ctx, const ggml_tensor *tensor) {
+    int64_t threshold_ne0 = 512;
+    int64_t threshold_ne1 = 512;
+    if (backend_ctx->adreno_cl_compiler_version.major <= 38) {
+        threshold_ne0 = 128;
+        threshold_ne1 = 128;
+    }
+    return tensor->ne[0] >= threshold_ne0 && tensor->ne[1] >= threshold_ne1 &&
             tensor->ne[2] == 1 && tensor->ne[3] == 1;
 }
 
@@ -2049,7 +2055,7 @@ static void ggml_backend_opencl_buffer_set_tensor(ggml_backend_buffer_t buffer,
         cl_kernel kernel = backend_ctx->kernel_convert_block_q4_0;
 
         // The optimized kernels need weights in natural order, so unshuffle.
-        if (use_adreno_kernels(tensor)) {
+        if (use_adreno_kernels(backend_ctx, tensor)) {
             kernel = backend_ctx->kernel_convert_block_q4_0_noshuffle;
         }
     #else
@@ -2073,7 +2079,7 @@ static void ggml_backend_opencl_buffer_set_tensor(ggml_backend_buffer_t buffer,
     #ifdef GGML_OPENCL_USE_ADRENO_KERNELS
         // Only do transpose for large, non batched matrix
         // TODO: use preallocated images instead of sub-buffer then image
-        if (use_adreno_kernels(tensor)) {
+        if (use_adreno_kernels(backend_ctx, tensor)) {
         // <----------------------------------------------------------------------------------> //
         // start transpose
         // <----------------------------------------------------------------------------------> //
@@ -3512,7 +3518,7 @@ static void ggml_cl_mul_mat(ggml_backend_t backend, const ggml_tensor * src0, co
 #ifdef GGML_OPENCL_USE_ADRENO_KERNELS
     cl_context context = backend_ctx->context;
 
-    if (ne01 && ne1 && use_adreno_kernels(src0)) {
+    if (ne01 && ne1 && use_adreno_kernels(backend_ctx, src0)) {
 
     // init CL objects
     // <--------------------------------------------> //