Fix NEON sqrt for 32-bit, add prsqrt.
With !406, we accidentally broke arm 32-bit NEON builds, since
`vsqrt_f32` is only available for 64-bit.
Here we add back the `rsqrt` implementation for 32-bit, relying
on a `prsqrt` implementation with better handling of edge cases.
Note that several of the 32-bit NEON packet tests are currently
failing - either due to denormal handling (NEON versions flush
to zero, but scalar paths don't) or due to accuracy (e.g. sin/cos).
diff --git a/Eigen/src/Core/GenericPacketMath.h b/Eigen/src/Core/GenericPacketMath.h
index f8c7826..bc0fe39 100644
--- a/Eigen/src/Core/GenericPacketMath.h
+++ b/Eigen/src/Core/GenericPacketMath.h
@@ -684,7 +684,7 @@
/** \internal \returns the square-root of \a a (coeff-wise) */
template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
-Packet psqrt(const Packet& a) { EIGEN_USING_STD(sqrt); return sqrt(a); }
+Packet psqrt(const Packet& a) { return numext::sqrt(a); }
/** \internal \returns the reciprocal square-root of \a a (coeff-wise) */
template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
diff --git a/Eigen/src/Core/arch/NEON/PacketMath.h b/Eigen/src/Core/arch/NEON/PacketMath.h
index 9715bf4..f77a18a 100644
--- a/Eigen/src/Core/arch/NEON/PacketMath.h
+++ b/Eigen/src/Core/arch/NEON/PacketMath.h
@@ -202,6 +202,7 @@
HasLog = 1,
HasExp = 1,
HasSqrt = 1,
+ HasRsqrt = 1,
HasTanh = EIGEN_FAST_MATH,
HasErf = EIGEN_FAST_MATH,
HasBessel = 0, // Issues with accuracy.
@@ -3329,8 +3330,42 @@
return res;
}
+template<> EIGEN_STRONG_INLINE Packet4f prsqrt(const Packet4f& a) {
+ // Compute approximate reciprocal sqrt.
+ Packet4f x = vrsqrteq_f32(a);
+ // Do Newton iterations for 1/sqrt(x).
+ x = vmulq_f32(vrsqrtsq_f32(vmulq_f32(a, x), x), x);
+ x = vmulq_f32(vrsqrtsq_f32(vmulq_f32(a, x), x), x);
+ const Packet4f infinity = pset1<Packet4f>(NumTraits<float>::infinity());
+ return pselect(pcmp_eq(a, pzero(a)), infinity, x);
+}
+
+template<> EIGEN_STRONG_INLINE Packet2f prsqrt(const Packet2f& a) {
+ // Compute approximate reciprocal sqrt.
+ Packet2f x = vrsqrte_f32(a);
+ // Do Newton iterations for 1/sqrt(x).
+ x = vmul_f32(vrsqrts_f32(vmul_f32(a, x), x), x);
+ x = vmul_f32(vrsqrts_f32(vmul_f32(a, x), x), x);
+ const Packet2f infinity = pset1<Packet2f>(NumTraits<float>::infinity());
+ return pselect(pcmp_eq(a, pzero(a)), infinity, x);
+}
+
+// Unfortunately vsqrt_f32 is only available for A64.
+#if EIGEN_ARCH_ARM64
template<> EIGEN_STRONG_INLINE Packet4f psqrt(const Packet4f& _x){return vsqrtq_f32(_x);}
template<> EIGEN_STRONG_INLINE Packet2f psqrt(const Packet2f& _x){return vsqrt_f32(_x); }
+#else
+template<> EIGEN_STRONG_INLINE Packet4f psqrt(const Packet4f& a) {
+ const Packet4f infinity = pset1<Packet4f>(NumTraits<float>::infinity());
+ const Packet4f is_zero_or_inf = por(pcmp_eq(a, pzero(a)), pcmp_eq(a, infinity));
+ return pselect(is_zero_or_inf, a, pmul(a, prsqrt(a)));
+}
+template<> EIGEN_STRONG_INLINE Packet2f psqrt(const Packet2f& a) {
+ const Packet2f infinity = pset1<Packet2f>(NumTraits<float>::infinity());
+ const Packet2f is_zero_or_inf = por(pcmp_eq(a, pzero(a)), pcmp_eq(a, infinity));
+ return pselect(is_zero_or_inf, a, pmul(a, prsqrt(a)));
+}
+#endif
//---------- bfloat16 ----------
// TODO: Add support for native armv8.6-a bfloat16_t
@@ -3722,6 +3757,7 @@
HasLog = 1,
HasExp = 1,
HasSqrt = 1,
+ HasRsqrt = 1,
HasTanh = 0,
HasErf = 0
};
@@ -3933,6 +3969,17 @@
template<> EIGEN_STRONG_INLINE Packet2d pset1frombits<Packet2d>(uint64_t from)
{ return vreinterpretq_f64_u64(vdupq_n_u64(from)); }
+template<> EIGEN_STRONG_INLINE Packet2d prsqrt(const Packet2d& a) {
+ // Compute approximate reciprocal sqrt.
+ Packet2d x = vrsqrteq_f64(a);
+ // Do Newton iterations for 1/sqrt(x).
+ x = vmulq_f64(vrsqrtsq_f64(vmulq_f64(a, x), x), x);
+ x = vmulq_f64(vrsqrtsq_f64(vmulq_f64(a, x), x), x);
+ x = vmulq_f64(vrsqrtsq_f64(vmulq_f64(a, x), x), x);
+ const Packet2d infinity = pset1<Packet2d>(NumTraits<double>::infinity());
+ return pselect(pcmp_eq(a, pzero(a)), infinity, x);
+}
+
template<> EIGEN_STRONG_INLINE Packet2d psqrt(const Packet2d& _x){ return vsqrtq_f64(_x); }
#endif // EIGEN_ARCH_ARM64
@@ -3978,6 +4025,7 @@
HasLog = 0,
HasExp = 0,
HasSqrt = 1,
+ HasRsqrt = 1,
HasErf = EIGEN_FAST_MATH,
HasBessel = 0, // Issues with accuracy.
HasNdtri = 0,
diff --git a/test/packetmath.cpp b/test/packetmath.cpp
index ae7168f..ceafb90 100644
--- a/test/packetmath.cpp
+++ b/test/packetmath.cpp
@@ -504,6 +504,7 @@
data1[i] = numext::abs(internal::random<Scalar>());
}
CHECK_CWISE1_IF(PacketTraits::HasSqrt, numext::sqrt, internal::psqrt);
+ CHECK_CWISE1_IF(PacketTraits::HasRsqrt, numext::rsqrt, internal::prsqrt);
}
// Notice that this definition works for complex types as well.
@@ -532,7 +533,7 @@
CHECK_CWISE1_IF(PacketTraits::HasLog, std::log, internal::plog);
CHECK_CWISE1_IF(PacketTraits::HasLog, log2, internal::plog2);
- CHECK_CWISE1_IF(PacketTraits::HasRsqrt, 1 / std::sqrt, internal::prsqrt);
+ CHECK_CWISE1_IF(PacketTraits::HasRsqrt, numext::rsqrt, internal::prsqrt);
for (int i = 0; i < size; ++i) {
data1[i] = Scalar(internal::random<double>(-1, 1) * std::pow(10., internal::random<double>(-3, 3)));
