For *ToFixed, in debug mode, assert that the value is in range.

include/core/SkFixed.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 SkFixed_DEFINED
 #define SkFixed_DEFINED
 
@@ skipping 11 matching lines @@
 typedef int32_t             SkFixed;
 #define SK_Fixed1           (1 << 16)
 #define SK_FixedHalf        (1 << 15)
 #define SK_FixedMax         (0x7FFFFFFF)
 #define SK_FixedMin         (-SK_FixedMax)
 #define SK_FixedPI          (0x3243F)
 #define SK_FixedSqrt2       (92682)
 #define SK_FixedTanPIOver8  (0x6A0A)
 #define SK_FixedRoot2Over2  (0xB505)
 
-#define SkFixedToFloat(x)   ((x) * 1.52587890625e-5f)
-#define SkFloatToFixed(x)   ((SkFixed)((x) * SK_Fixed1))
+#define SkFixedToFloat(x)         ((x) * 1.52587890625e-5f)

[mtklein, 2016/03/23 02:16:41]

This may no longer need its new indentation.

+
+///////////////////////////////////////////////////////////////////////////////
+// ASM alternatives for our portable versions.
+
+#if defined(SK_CPU_ARM32)
+    /* This guy does not handle NaN or other obscurities, but is faster than
+       than (int)(x*65536).  When built on Android with -Os, needs forcing
+       to inline or we lose the speed benefit.
+    */
+    SK_ALWAYS_INLINE SkFixed SkFloatToFixed_arm(float x)
+    {
+        int32_t y, z;
+        asm("movs    %1, %3, lsl #1         \n"
+            "mov     %2, #0x8E              \n"
+            "sub     %1, %2, %1, lsr #24    \n"
+            "mov     %2, %3, lsl #8         \n"
+            "orr     %2, %2, #0x80000000    \n"
+            "mov     %1, %2, lsr %1         \n"
+            "it cs                          \n"
+            "rsbcs   %1, %1, #0             \n"
+            : "=r"(x), "=&r"(y), "=&r"(z)
+            : "r"(x)
+            : "cc"
+            );
+        return y;
+    }
+    inline SkFixed SkFixedMul_arm(SkFixed x, SkFixed y)
+    {
+        int32_t t;
+        asm("smull  %0, %2, %1, %3          \n"
+            "mov    %0, %0, lsr #16         \n"
+            "orr    %0, %0, %2, lsl #16     \n"
+            : "=r"(x), "=&r"(y), "=r"(t)
+            : "r"(x), "1"(y)
+            :
+            );
+        return x;
+    }
+
+    #define SkFixedMul(x, y)           SkFixedMul_arm(x, y)
+    #define SkFloatToFixed_Unsafe(x)   SkFloatToFixed_arm(x)
+#else
+    inline SkFixed SkFixedMul_longlong(SkFixed a, SkFixed b) {
+        return (SkFixed)((int64_t)a * b >> 16);
+    }
+
+    #define SkFixedMul(x, y)           SkFixedMul_longlong(x, y)
+    #define SkFloatToFixed_Unsafe(x)   ((SkFixed)((x) * SK_Fixed1))
+#endif
+
+///////////////////////////////////////////////////////////////////////////////
+
+static inline SkFixed SkFloatToFixed(float x) {
+    const SkFixed result = SkFloatToFixed_Unsafe(x);
+    SkASSERT(truncf(x * SK_Fixed1) == static_cast<float>(result));
+    return result;
+}
 
 // Pins over/under flows to SK_FixedMax/SK_FixedMin (slower than just a cast).
 static inline SkFixed SkFloatPinToFixed(float x) {
     x *= SK_Fixed1;
     // Casting float to int outside the range of the target type (int32_t) is undefined behavior.
     if (x >= SK_FixedMax) return SK_FixedMax;
     if (x <= SK_FixedMin) return SK_FixedMin;
     const SkFixed result = static_cast<SkFixed>(x);
     SkASSERT(truncf(x) == static_cast<float>(result));
     return result;
 }
 
-#ifdef SK_DEBUG
-    static inline SkFixed SkFloatToFixed_Check(float x) {
-        int64_t n64 = (int64_t)(x * SK_Fixed1);
-        SkFixed n32 = (SkFixed)n64;
-        SkASSERT(n64 == n32);
-        return n32;
-    }
-#else
-    #define SkFloatToFixed_Check(x) SkFloatToFixed(x)
-#endif
+#define SkFixedToDouble(x)         ((x) * 1.52587890625e-5)
+#define SkDoubleToFixed_Unsafe(x)  ((SkFixed)((x) * SK_Fixed1))
 
-#define SkFixedToDouble(x)  ((x) * 1.52587890625e-5)
-#define SkDoubleToFixed(x)  ((SkFixed)((x) * SK_Fixed1))
+static inline SkFixed SkDoubleToFixed(double x) {
+    const SkFixed result = SkDoubleToFixed_Unsafe(x);
+    SkASSERT(trunc(x * SK_Fixed1) == static_cast<double>(result));
+    return result;
+}
 
 // Pins over/under flows to SK_FixedMax/SK_FixedMin (slower than just a cast).
 static inline SkFixed SkDoublePinToFixed(double x) {
     x *= SK_Fixed1;
     // Casting double to int outside the range of the target type (int32_t) is undefined behavior.
     if (x >= SK_FixedMax) return SK_FixedMax;
     if (x <= SK_FixedMin) return SK_FixedMin;
     const SkFixed result = static_cast<SkFixed>(x);
     SkASSERT(trunc(x) == static_cast<double>(result));
     return result;
@@ skipping 30 matching lines @@
 
 // Blink layout tests are baselined to Clang optimizing through undefined behavior in SkDivBits.
 #if defined(SK_SUPPORT_LEGACY_DIVBITS_UB)
     #define SkFixedDiv(numer, denom) SkDivBits(numer, denom, 16)
 #else
     // The divide may exceed 32 bits. Clamp to a signed 32 bit result.
     #define SkFixedDiv(numer, denom) \
         SkToS32(SkTPin<int64_t>((SkLeftShift((int64_t)numer, 16) / denom), SK_MinS32, SK_MaxS32))
 #endif
 
-//////////////////////////////////////////////////////////////////////////////////////////////////////
-// Now look for ASM overrides for our portable versions (should consider putting this in its own file)
-
-inline SkFixed SkFixedMul_longlong(SkFixed a, SkFixed b) {
-    return (SkFixed)((int64_t)a * b >> 16);
-}
-#define SkFixedMul(a,b)     SkFixedMul_longlong(a,b)
-
-
-#if defined(SK_CPU_ARM32)
-    /* This guy does not handle NaN or other obscurities, but is faster than
-       than (int)(x*65536).  When built on Android with -Os, needs forcing
-       to inline or we lose the speed benefit.
-    */
-    SK_ALWAYS_INLINE SkFixed SkFloatToFixed_arm(float x)
-    {
-        int32_t y, z;
-        asm("movs    %1, %3, lsl #1         \n"
-            "mov     %2, #0x8E              \n"
-            "sub     %1, %2, %1, lsr #24    \n"
-            "mov     %2, %3, lsl #8         \n"
-            "orr     %2, %2, #0x80000000    \n"
-            "mov     %1, %2, lsr %1         \n"
-            "it cs                          \n"
-            "rsbcs   %1, %1, #0             \n"
-            : "=r"(x), "=&r"(y), "=&r"(z)
-            : "r"(x)
-            : "cc"
-            );
-        return y;
-    }
-    inline SkFixed SkFixedMul_arm(SkFixed x, SkFixed y)
-    {
-        int32_t t;
-        asm("smull  %0, %2, %1, %3          \n"
-            "mov    %0, %0, lsr #16         \n"
-            "orr    %0, %0, %2, lsl #16     \n"
-            : "=r"(x), "=&r"(y), "=r"(t)
-            : "r"(x), "1"(y)
-            :
-            );
-        return x;
-    }
-    #undef SkFixedMul
-    #define SkFixedMul(x, y)        SkFixedMul_arm(x, y)
-
-    #undef SkFloatToFixed
-    #define SkFloatToFixed(x)  SkFloatToFixed_arm(x)
-#endif
-
 ///////////////////////////////////////////////////////////////////////////////
 
 typedef int64_t SkFixed3232;   // 32.32
 
-#define SkIntToFixed3232(x)       (SkLeftShift((SkFixed3232)(x), 32))
-#define SkFixed3232ToInt(x)       ((int)((x) >> 32))
-#define SkFixedToFixed3232(x)     (SkLeftShift((SkFixed3232)(x), 16))
-#define SkFixed3232ToFixed(x)     ((SkFixed)((x) >> 16))
-#define SkFloatToFixed3232(x)     ((SkFixed3232)((x) * (65536.0f * 65536.0f)))
+#define SK_Fixed3232_1                (static_cast<SkFixed3232>(1) << 32)
+#define SkIntToFixed3232(x)           (SkLeftShift((SkFixed3232)(x), 32))
+#define SkFixed3232ToInt(x)           ((int)((x) >> 32))
+#define SkFixedToFixed3232(x)         (SkLeftShift((SkFixed3232)(x), 16))
+#define SkFixed3232ToFixed(x)         ((SkFixed)((x) >> 16))
+#define SkFloatToFixed3232_Unsafe(x)  (static_cast<SkFixed3232>((x) * SK_Fixed3232_1))
+
+static inline SkFixed3232 SkFloatToFixed3232(float x) {
+    const SkFixed3232 result = SkFloatToFixed3232_Unsafe(x);
+    SkASSERT(truncf(x * SK_Fixed3232_1) == static_cast<float>(result));
+    return result;
+}
 
 #define SkScalarToFixed3232(x)    SkFloatToFixed3232(x)
 
 ///////////////////////////////////////////////////////////////////////////////
 
 // 64bits wide, with a 16bit bias. Useful when accumulating lots of 16.16 so
 // we don't overflow along the way
 typedef int64_t Sk48Dot16;
 
 #define Sk48Dot16FloorToInt(x)    static_cast<int>((x) >> 16)
 
 static inline float Sk48Dot16ToScalar(Sk48Dot16 x) {
     return static_cast<float>(x * 1.5258789e-5);  // x * (1.0f / (1 << 16))
 }
 #define SkFloatTo48Dot16(x)       (static_cast<Sk48Dot16>((x) * (1 << 16)))
 
 #define SkScalarTo48Dot16(x)      SkFloatTo48Dot16(x)
 
 #endif