X-Git-Url: https://git.sesse.net/?p=movit;a=blobdiff_plain;f=fp16.h;h=535eccf36b04c181f08cc432b3fca5785ca54da4;hp=9f7eff926fca594d54198b14b3ec30d834124508;hb=d7def3f9e42e72566d7aedd52ffdeb5851da8314;hpb=82464681eca54bf11be3ca53eed20841ff6eb204

diff --git a/fp16.h b/fp16.h
index 9f7eff9..535eccf 100644
--- a/fp16.h
+++ b/fp16.h
@@ -1,26 +1,140 @@
 #ifndef _MOVIT_FP16_H
 #define _MOVIT_FP16_H 1
 
+#ifdef __F16C__
+#include <immintrin.h>
+#endif
+
 // Code for converting to and from fp16 (from fp64), without any particular
 // machine support, with proper IEEE round-to-even behavior (and correct
 // handling of NaNs and infinities). This is needed because some OpenGL
 // drivers don't properly round off when asked to convert data themselves.
 //
-// These routines are not particularly fast.
+// These routines are originally written by Fabian Giesen, and released by
+// him into the public domain;
+// see https://fgiesen.wordpress.com/2012/03/28/half-to-float-done-quic/.
+// They are quite fast, and can be vectorized if need be; of course, using
+// the f16c instructions (see below) will be faster still.
 
 namespace movit {
 
-typedef unsigned int fp32_int_t;
-typedef unsigned short fp16_int_t;
+// structs instead of ints, so that they are not implicitly convertible.
+struct fp32_int_t {
+	unsigned int val;
+};
+struct fp16_int_t {
+	unsigned short val;
+};
+
+#ifdef __F16C__
+
+// Use the f16c instructions from Haswell if available (and we know that they
+// are at compile time).
+static inline float fp16_to_fp32(fp16_int_t x)
+{
+	return _cvtsh_ss(x.val);
+}
+
+static inline fp16_int_t fp32_to_fp16(float x)
+{
+	fp16_int_t ret;
+	ret.val = _cvtss_sh(x, 0);
+	return ret;
+}
+
+#else
+
+union fp32 {
+	float f;
+	unsigned int u;
+};
+
+static inline float fp16_to_fp32(fp16_int_t h)
+{
+	fp32 magic; magic.u = 113 << 23;
+	unsigned int shifted_exp = 0x7c00 << 13;  // exponent mask after shift
+	fp32 o;
+
+	// mantissa+exponent
+	unsigned int shifted = (h.val & 0x7fff) << 13;
+	unsigned int exponent = shifted & shifted_exp;
+
+	// exponent cases
+	o.u = shifted;
+	if (exponent == 0) {  // Zero / Denormal
+		o.u += magic.u;
+		o.f -= magic.f;
+	} else if (exponent == shifted_exp) {  // Inf/NaN
+		o.u += (255 - 31) << 23;
+	} else {
+		o.u += (127 - 15) << 23;
+	}
+
+	o.u |= (h.val & 0x8000) << 16;  // copy sign bit
+	return o.f;
+}
+
+static inline fp16_int_t fp32_to_fp16(float x)
+{
+	fp32 f; f.f = x;
+	unsigned int f32infty = 255 << 23;
+	unsigned int f16max = (127 + 16) << 23;
+	fp32 denorm_magic; denorm_magic.u = ((127 - 15) + (23 - 10) + 1) << 23;
+	unsigned int sign_mask = 0x80000000u;
+	fp16_int_t o = { 0 };
+
+	unsigned int sign = f.u & sign_mask;
+	f.u ^= sign;
+
+	// NOTE all the integer compares in this function can be safely
+	// compiled into signed compares since all operands are below
+	// 0x80000000. Important if you want fast straight SSE2 code
+	// (since there's no unsigned PCMPGTD).
+
+	if (f.u >= f16max) {  // result is Inf or NaN (all exponent bits set)
+		o.val = (f.u > f32infty) ? 0x7e00 : 0x7c00; // NaN->qNaN and Inf->Inf
+	} else {  // (De)normalized number or zero
+		if (f.u < (113 << 23)) {  // resulting FP16 is subnormal or zero
+			// use a magic value to align our 10 mantissa bits at the bottom of
+			// the float. as long as FP addition is round-to-nearest-even this
+			// just works.
+			f.f += denorm_magic.f;
+
+			// and one integer subtract of the bias later, we have our final float!
+			o.val = f.u - denorm_magic.u;
+		} else {
+			unsigned int mant_odd = (f.u >> 13) & 1; // resulting mantissa is odd
+
+			// update exponent, rounding bias part 1
+			f.u += (unsigned(15 - 127) << 23) + 0xfff;
+			// rounding bias part 2
+			f.u += mant_odd;
+			// take the bits!
+			o.val = f.u >> 13;
+		}
+	}
+
+	o.val |= sign >> 16;
+	return o;
+}
+
+#endif
+
+// Overloads for use in templates.
+static inline float to_fp32(double x) { return x; }
+static inline float to_fp32(float x) { return x; }
+static inline float to_fp32(fp16_int_t x) { return fp16_to_fp32(x); }
+
+template<class T> inline T from_fp32(float x);
+template<> inline double from_fp32<double>(float x) { return x; }
+template<> inline float from_fp32<float>(float x) { return x; }
+template<> inline fp16_int_t from_fp32<fp16_int_t>(float x) { return fp32_to_fp16(x); }
 
-double fp16_to_fp64(fp16_int_t x);
-fp16_int_t fp64_to_fp16(double x);
+template<class From, class To>
+inline To convert_float(From x) { return from_fp32<To>(to_fp32(x)); }
 
-// These are not very useful by themselves, but are implemented using the same
-// code as the fp16 ones (just with different constants), so they are useful
-// for verifying against the FPU in unit tests.
-double fp32_to_fp64(fp32_int_t x);
-fp32_int_t fp64_to_fp32(double x);
+template<class Same>
+inline Same convert_float(Same x) { return x; }
 
 }  // namespace movit