Open
Description
This is tracking issue to refactor all higher math functions in single and double precision currently implemented in LLVM libc to be header-only.
Status:
- acos
- acos
- acosf16
- acosf
- acosh
- acoshf16
- acoshf
- acospif16
- asin
- asin
- asinf16
- asinf
- asinh
- asinhf16
- asinhf
-
asinpi - atan
- atan
- atanf16
- atanf
- atan2
- atan2
- atan2f128
- atan2f
- atan2f_float
- atan2l
-
atan2pi - atanh
- atanhf16
- atanhf
-
atanpi - cbrt
- cbrt
- cbrtf
-
compoundn - cos
- cos
- cosf16
- cosf
- cosh
- coshf16
- coshf
- cospi
- cospif16
- cospif
- dsqrt
- dsqrtf128
- dsqrtl
- erff
-
erfc - exp
- expf [libc][math] Refactor expf implementation to header-only in src/__support/math folder. #143790
- expf16 [libc][math] Refactor expf16 implementation to header-only in src/__support/math folder. #147428
- frexpf128 [libc][math] Refactor frexpf128 implementation to header-only in src/… #147822
- frexpf16
- frexpf
- ldexpf128
- ldexpf16
- ldexpf
- exp10
- exp10
- exp10f16
- exp10f
- exp10m1
- exp10m1f16
- exp10m1f
- exp2
- exp2f16
- exp2f
- exp2
- exp2m1
- exp2m1f16
- exp2m1f
- expm1
- expm1f16
- expm1f
- expm1
- fma
- dfmaf128
- dfmal
- f16fmaf128
- f16fmaf
- f16fma
- f16fmal
- ffmaf128
- ffma
- ffmal
- fmaf16
- fmaf
- fma
- f16sqrt
- f16sqrtf128
- f16sqrtf
- f16sqrt
- f16sqrtl
- fsqrt
- fsqrtf128
- fsqrt
- fsqrtl
- hypot
- hypotf16
- hypotf
- hypot
- lgamma
- lgamma
- lgamma_r
- log
- ilogbf128
- ilogbf16
- ilogbf
- ilogb
- ilogbl
- llogbf128
- llogbf16
- llogbf
- llogb
- llogbl
- logbf128
- logbf16
- logbf
- logb
- logbl
- logf16
- logf
- log
- log10
- log10f16
- log10f
- log10
-
log10p1 - log1p
- log1p
- log1pf
- log2
- log2f16
- log2f
- log2
-
log2p1 -
logp1 - pow
- pow
- powf
- powi*
- powi
- powif
-
pown -
powr -
rootn -
rsqrt - sin
- sin
- sinf16
- sinf
- sincos
- sincosf
- sincos
- sinh
- sinhf16
- sinhf
- sinh
- sinpi
- sinpi
- sinpif
- sqrt
- sqrtf128
- sqrtf16
- sqrtf
- sqrt
- sqrtl
- tan
- tanf16
- tanf
- tan
- tanh
- tanhf16
- tanhf
- tanh
- tanpif16
- tgamma
- tgamma
- tgammaf