[libc][NFC] Add NOLINT annotations at call sites to immintrin functions.
These annotations are intended to be temporary while we understand why clang-tidy is not able to treat them as builtin exceptions.
This commit is contained in:
parent
2ff795a462
commit
e5d104b280
|
@ -26,11 +26,11 @@ __attribute__((target(
|
|||
"fma"))) static inline cpp::EnableIfType<cpp::IsSame<T, float>::Value, T>
|
||||
fma(T x, T y, T z) {
|
||||
float result;
|
||||
__m128 xmm = _mm_load_ss(&x);
|
||||
__m128 ymm = _mm_load_ss(&y);
|
||||
__m128 zmm = _mm_load_ss(&z);
|
||||
__m128 r = _mm_fmadd_ss(xmm, ymm, zmm);
|
||||
_mm_store_ss(&result, r);
|
||||
__m128 xmm = _mm_load_ss(&x); // NOLINT
|
||||
__m128 ymm = _mm_load_ss(&y); // NOLINT
|
||||
__m128 zmm = _mm_load_ss(&z); // NOLINT
|
||||
__m128 r = _mm_fmadd_ss(xmm, ymm, zmm); // NOLINT
|
||||
_mm_store_ss(&result, r); // NOLINT
|
||||
return result;
|
||||
}
|
||||
|
||||
|
@ -39,11 +39,11 @@ __attribute__((target(
|
|||
"fma"))) static inline cpp::EnableIfType<cpp::IsSame<T, double>::Value, T>
|
||||
fma(T x, T y, T z) {
|
||||
double result;
|
||||
__m128d xmm = _mm_load_sd(&x);
|
||||
__m128d ymm = _mm_load_sd(&y);
|
||||
__m128d zmm = _mm_load_sd(&z);
|
||||
__m128d r = _mm_fmadd_sd(xmm, ymm, zmm);
|
||||
_mm_store_sd(&result, r);
|
||||
__m128d xmm = _mm_load_sd(&x); // NOLINT
|
||||
__m128d ymm = _mm_load_sd(&y); // NOLINT
|
||||
__m128d zmm = _mm_load_sd(&z); // NOLINT
|
||||
__m128d r = _mm_fmadd_sd(xmm, ymm, zmm); // NOLINT
|
||||
_mm_store_sd(&result, r); // NOLINT
|
||||
return result;
|
||||
}
|
||||
|
||||
|
|
|
@ -25,11 +25,11 @@ namespace fputil {
|
|||
template <>
|
||||
__attribute__((target("fma"))) inline float
|
||||
polyeval(float x, float a0, float a1, float a2, float a3) {
|
||||
__m128 xmm = _mm_set1_ps(x);
|
||||
__m128 a13 = _mm_set_ps(0.0f, x, a3, a1);
|
||||
__m128 a02 = _mm_set_ps(0.0f, 0.0f, a2, a0);
|
||||
__m128 xmm = _mm_set1_ps(x); // NOLINT
|
||||
__m128 a13 = _mm_set_ps(0.0f, x, a3, a1); // NOLINT
|
||||
__m128 a02 = _mm_set_ps(0.0f, 0.0f, a2, a0); // NOLINT
|
||||
// r = (0, x^2, a3*x + a2, a1*x + a0)
|
||||
__m128 r = _mm_fmadd_ps(a13, xmm, a02);
|
||||
__m128 r = _mm_fmadd_ps(a13, xmm, a02); // NOLINT
|
||||
// result = (a3*x + a2) * x^2 + (a1*x + a0)
|
||||
return fma(r[2], r[1], r[0]);
|
||||
}
|
||||
|
@ -37,11 +37,11 @@ polyeval(float x, float a0, float a1, float a2, float a3) {
|
|||
template <>
|
||||
__attribute__((target("fma"))) inline double
|
||||
polyeval(double x, double a0, double a1, double a2, double a3) {
|
||||
__m256d xmm = _mm256_set1_pd(x);
|
||||
__m256d a13 = _mm256_set_pd(0.0, x, a3, a1);
|
||||
__m256d a02 = _mm256_set_pd(0.0, 0.0, a2, a0);
|
||||
__m256d xmm = _mm256_set1_pd(x); // NOLINT
|
||||
__m256d a13 = _mm256_set_pd(0.0, x, a3, a1); // NOLINT
|
||||
__m256d a02 = _mm256_set_pd(0.0, 0.0, a2, a0); // NOLINT
|
||||
// r = (0, x^2, a3*x + a2, a1*x + a0)
|
||||
__m256d r = _mm256_fmadd_pd(a13, xmm, a02);
|
||||
__m256d r = _mm256_fmadd_pd(a13, xmm, a02); // NOLINT
|
||||
// result = (a3*x + a2) * x^2 + (a1*x + a0)
|
||||
return fma(r[2], r[1], r[0]);
|
||||
}
|
||||
|
@ -52,14 +52,14 @@ polyeval(double x, double a0, double a1, double a2, double a3) {
|
|||
template <>
|
||||
__attribute__((target("fma"))) inline float
|
||||
polyeval(float x, float a0, float a1, float a2, float a3, float a4, float a5) {
|
||||
__m128 xmm = _mm_set1_ps(x);
|
||||
__m128 a25 = _mm_set_ps(0.0f, x, a5, a2);
|
||||
__m128 a14 = _mm_set_ps(0.0f, 0.0f, a4, a1);
|
||||
__m128 a03 = _mm_set_ps(0.0f, 0.0f, a3, a0);
|
||||
__m128 xmm = _mm_set1_ps(x); // NOLINT
|
||||
__m128 a25 = _mm_set_ps(0.0f, x, a5, a2); // NOLINT
|
||||
__m128 a14 = _mm_set_ps(0.0f, 0.0f, a4, a1); // NOLINT
|
||||
__m128 a03 = _mm_set_ps(0.0f, 0.0f, a3, a0); // NOLINT
|
||||
// r1 = (0, x^2, a5*x + a4, a2*x + a1)
|
||||
__m128 r1 = _mm_fmadd_ps(a25, xmm, a14);
|
||||
__m128 r1 = _mm_fmadd_ps(a25, xmm, a14); // NOLINT
|
||||
// r2 = (0, x^3, (a5*x + a4)*x + a3, (a2*x + a1)*x + a0
|
||||
__m128 r2 = _mm_fmadd_ps(r1, xmm, a03);
|
||||
__m128 r2 = _mm_fmadd_ps(r1, xmm, a03); // NOLINT
|
||||
// result = ((a5*x + a4)*x + a3) * x^3 + ((a2*x + a1)*x + a0)
|
||||
return fma(r2[2], r2[1], r2[0]);
|
||||
}
|
||||
|
@ -68,14 +68,14 @@ template <>
|
|||
__attribute__((target("fma"))) inline double
|
||||
polyeval(double x, double a0, double a1, double a2, double a3, double a4,
|
||||
double a5) {
|
||||
__m256d xmm = _mm256_set1_pd(x);
|
||||
__m256d a25 = _mm256_set_pd(0.0, x, a5, a2);
|
||||
__m256d a14 = _mm256_set_pd(0.0, 0.0, a4, a1);
|
||||
__m256d a03 = _mm256_set_pd(0.0, 0.0, a3, a0);
|
||||
__m256d xmm = _mm256_set1_pd(x); // NOLINT
|
||||
__m256d a25 = _mm256_set_pd(0.0, x, a5, a2); // NOLINT
|
||||
__m256d a14 = _mm256_set_pd(0.0, 0.0, a4, a1); // NOLINT
|
||||
__m256d a03 = _mm256_set_pd(0.0, 0.0, a3, a0); // NOLINT
|
||||
// r1 = (0, x^2, a5*x + a4, a2*x + a1)
|
||||
__m256d r1 = _mm256_fmadd_pd(a25, xmm, a14);
|
||||
__m256d r1 = _mm256_fmadd_pd(a25, xmm, a14); // NOLINT
|
||||
// r2 = (0, x^3, (a5*x + a4)*x + a3, (a2*x + a1)*x + a0
|
||||
__m256d r2 = _mm256_fmadd_pd(r1, xmm, a03);
|
||||
__m256d r2 = _mm256_fmadd_pd(r1, xmm, a03); // NOLINT
|
||||
// result = ((a5*x + a4)*x + a3) * x^3 + ((a2*x + a1)*x + a0)
|
||||
return fma(r2[2], r2[1], r2[0]);
|
||||
}
|
||||
|
|
Loading…
Reference in New Issue