abandon SK_IGNORE_TREAT_AS_SPRITE as a flag, and use SK_IGNORE_TRANS_CLAMP_FIX
for the fixes in SkDraw.cpp and SkBitmapProcState.cpp

Also, change the call in SkDraw.cpp to pass 0 for subpixel bits, to match the
old logic.... we should have a runtime fix for this in the future, so a caller
*can* get subpixel-filtered images w/ just translate.



git-svn-id: http://skia.googlecode.com/svn/trunk@7065 2bbb7eff-a529-9590-31e7-b0007b416f81
diff --git a/src/core/SkDraw.cpp b/src/core/SkDraw.cpp
index 0700353..07ffe01 100644
--- a/src/core/SkDraw.cpp
+++ b/src/core/SkDraw.cpp
@@ -1114,7 +1114,7 @@
     go ahead and treat it as if it were, so that subsequent code can go fast.
  */
 static bool just_translate(const SkMatrix& matrix, const SkBitmap& bitmap) {
-#ifdef SK_IGNORE_TREAT_AS_SPRITE
+#ifdef SK_IGNORE_TRANS_CLAMP_FIX
     SkMatrix::TypeMask mask = matrix.getType();
 
     if (mask & (SkMatrix::kAffine_Mask | SkMatrix::kPerspective_Mask)) {
@@ -1132,7 +1132,9 @@
     // if we got here, we're either kTranslate_Mask or identity
     return true;
 #else
-    return SkTreatAsSpriteFilter(matrix, bitmap.width(), bitmap.height());
+    unsigned bits = 0;  // TODO: find a way to allow the caller to tell us to
+                        // respect filtering.
+    return SkTreatAsSprite(matrix, bitmap.width(), bitmap.height(), bits);
 #endif
 }