Sk4f: floor() via int32_t roundtrip.

src/opts/SkNx_neon.h

Index: src/opts/SkNx_neon.h
diff --git a/src/opts/SkNx_neon.h b/src/opts/SkNx_neon.h
index cdc4615849556ff185afd59f15c82e8d5a5320b5..641e9d296b54014e124ab306a24df328bc1544f9 100644
--- a/src/opts/SkNx_neon.h
+++ b/src/opts/SkNx_neon.h
@@ -11,14 +11,13 @@
 #define SKNX_IS_FAST
 
 // ARMv8 has vrndmq_f32 to floor 4 floats.  Here we emulate it:
-//   - round by adding (1<<23) with our sign, then subtracting it;
-//   - if that rounded value is bigger than our input, subtract 1.
+//   - roundtrip through integers via truncation
+//   - subtract 1 if that's too big (possible for negative values).
+// This restricts the domain of our inputs to a maximum somehwere around 2^31.  Seems plenty big.
 static inline float32x4_t armv7_vrndmq_f32(float32x4_t v) {
-    auto sign = vandq_u32((uint32x4_t)v, vdupq_n_u32(1<<31));
-    auto bias = (float32x4_t)(vorrq_u32((uint32x4_t)vdupq_n_f32(1<<23), sign));
-    auto rounded = vsubq_f32(vaddq_f32(v, bias), bias);
-    auto too_big = vcgtq_f32(rounded, v);
-    return vsubq_f32(rounded, (float32x4_t)vandq_u32(too_big, (uint32x4_t)vdupq_n_f32(1)));
+    auto roundtrip = vcvtq_f32_s32(vcvtq_s32_f32(v));
+    auto too_big = vcgtq_f32(roundtrip, v);
+    return vsubq_f32(roundtrip, (float32x4_t)vandq_u32(too_big, (uint32x4_t)vdupq_n_f32(1)));
 }
 
 // Well, this is absurd.  The shifts require compile-time constant arguments.