Bug 1367568 part 2: Rewrite nsChangeHint_ReflowHintsForBSizeChange in terms of nsChangeHint_AllReflowHints. r=bz

This patch doesn't affect behavior at all.  It's simply adjusting the way we
express a bitmask, without modifying the actual value of that bitmask.  The
intent of this change is to make this bitmask's definition as similar as
possible to the definition for its ISize (inline-axis) analog, so that it's
easy to compare the two.

MozReview-Commit-ID: FUJnBQm8hmJ

--HG--
extra : rebase_source : 400972b9af447721183769b603eedb9a9fd4c939

layout/base/nsChangeHint.h

@@ -414,13 +414,11 @@ static_assert(!(nsChangeHint_Hints_AlwaysHandledForDescendants &
 // NOTE: You might think that BSize changes could exclude
 // nsChangeHint_ClearAncestorIntrinsics (which is inline-axis specific), but we
 // do need to send it, to clear cached results from CSS Flex measuring reflows.
-//
-// XXXdholbert The next patch rewrites this hint to be similar to the ISize one.
 #define nsChangeHint_ReflowHintsForBSizeChange            \
-  nsChangeHint(nsChangeHint_NeedReflow |                  \
-               nsChangeHint_UpdateComputedBSize |         \
-               nsChangeHint_ReflowChangesSizeOrPosition | \
-               nsChangeHint_ClearAncestorIntrinsics)
+  nsChangeHint((nsChangeHint_AllReflowHints |             \
+                nsChangeHint_UpdateComputedBSize) &       \
+               ~(nsChangeHint_ClearDescendantIntrinsics | \
+                 nsChangeHint_NeedDirtyReflow))
 
 #define NS_STYLE_HINT_REFLOW \
   nsChangeHint(NS_STYLE_HINT_VISUAL | nsChangeHint_AllReflowHints)