try simplest code: inline whenever vaguely sensible

include/core/SkUtils.h

 /*
  * Copyright 2006 The Android Open Source Project
  *
  * Use of this source code is governed by a BSD-style license that can be
  * found in the LICENSE file.
  */
 
 #ifndef SkUtils_DEFINED
 #define SkUtils_DEFINED
 
@@ skipping 13 matching lines @@
 //     code was not helpful; it's got to be here outside.
 //   - NEON code generation for ARMv8 with GCC 4.9 is terrible,
 //     making the NEON code ~8x slower that just a serial loop.
 
 /** Similar to memset(), but it assigns a 16bit value into the buffer.
     @param buffer   The memory to have value copied into it
     @param value    The 16bit value to be copied into buffer
     @param count    The number of times value should be copied into the buffer.
 */
 static inline void sk_memset16(uint16_t buffer[], uint16_t value, int count) {
-#if defined(SK_CPU_ARM64)
-    while (count --> 0) { *buffer++ = value; } return;
-#elif defined(SK_CPU_ARM32)
-    if (count <= 10) { while (count --> 0) { *buffer++ = value; } return; }
+#if defined(SK_ARM_HAS_NEON) || SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE2
+    while (count --> 0) { *buffer++ = value; }
+#else
+    SkOpts::memset16(buffer, value, count);
 #endif
-    SkOpts::memset16(buffer, value, count);
 }
 
 /** Similar to memset(), but it assigns a 32bit value into the buffer.
     @param buffer   The memory to have value copied into it
     @param value    The 32bit value to be copied into buffer
     @param count    The number of times value should be copied into the buffer.
 */
 static inline void sk_memset32(uint32_t buffer[], uint32_t value, int count) {
-#if defined(SK_CPU_ARM64)
-    while (count --> 0) { *buffer++ = value; } return;
-#elif defined(SK_CPU_ARM32)
-    if (count <= 10) { while (count --> 0) { *buffer++ = value; } return; }
+#if defined(SK_ARM_HAS_NEON) || SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE2
+    while (count --> 0) { *buffer++ = value; }
+#else
+    SkOpts::memset32(buffer, value, count);
 #endif
-    SkOpts::memset32(buffer, value, count);
 }
 
 
 ///////////////////////////////////////////////////////////////////////////////
 
 #define kMaxBytesInUTF8Sequence     4
 
 #ifdef SK_DEBUG
     int SkUTF8_LeadByteToCount(unsigned c);
 #else
@@ skipping 60 matching lines @@
     }
     ~SkAutoTrace() {
         SkDebugf("--- trace: %s Leave\n", fLabel);
     }
 private:
     const char* fLabel;
 };
 #define SkAutoTrace(...) SK_REQUIRE_LOCAL_VAR(SkAutoTrace)
 
 #endif