derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
90 lines
3.0 KiB
C++
90 lines
3.0 KiB
C++
//===----------------------------------------------------------------------===//
|
|
//
|
|
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
|
|
// See https://llvm.org/LICENSE.txt for license information.
|
|
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#if __CLC_FPSIZE == 32
|
|
|
|
_CLC_DEF _CLC_OVERLOAD _CLC_CONST __CLC_GENTYPE
|
|
__clc_tanpi_reduced_eval(__CLC_GENTYPE x, __CLC_INTN is_odd) {
|
|
__CLC_GENTYPE s = x * x;
|
|
|
|
__CLC_GENTYPE t = __clc_mad(s, __clc_mad(s, __clc_mad(s, __clc_mad(s,
|
|
__clc_mad(s, __clc_mad(s,
|
|
0x1.7d2bd4p+16f, 0x1.a4d306p+12f), 0x1.435004p+11f), 0x1.4b6926p+9f),
|
|
0x1.451e22p+7f), 0x1.467a9cp+5f), 0x1.4abb6ap+3f);
|
|
|
|
t = x * s * t;
|
|
t = __clc_mad(x, 0x1.921fb6p+1f, t);
|
|
|
|
__CLC_GENTYPE tr = __CLC_FP_LIT(-1.0) / t;
|
|
|
|
return is_odd ? tr : t;
|
|
}
|
|
|
|
#elif __CLC_FPSIZE == 64
|
|
|
|
_CLC_DEF _CLC_OVERLOAD _CLC_CONST __CLC_DOUBLEN
|
|
__clc_tanpi_reduced_eval(__CLC_DOUBLEN x, __CLC_INTN is_odd) {
|
|
__CLC_GENTYPE s = x * x;
|
|
__CLC_GENTYPE t = __clc_mad(s, __clc_mad(s, __clc_mad(s, __clc_mad(s,
|
|
__clc_mad(s, __clc_mad(s, __clc_mad(s, __clc_mad(s,
|
|
__clc_mad(s, __clc_mad(s, __clc_mad(s, __clc_mad(s,
|
|
__clc_mad(s,
|
|
0x1.3fad0a71ea6d1p+32, -0x1.11a76ac97377bp+30), 0x1.ba2bcaca6da1bp+27), -0x1.79e8e2d7aaf57p+22),
|
|
0x1.c1c1102e46eccp+21), 0x1.31291bbcb5588p+19), 0x1.486b2d6bb3db2p+17), 0x1.45be1b46ff156p+15),
|
|
0x1.45f61b419c746p+13), 0x1.45f311045a4ffp+11), 0x1.45f4739a998c7p+9), 0x1.45fff9b243050p+7),
|
|
0x1.466bc6775cf74p+5), 0x1.4abbce625be8bp+3);
|
|
t = x * s * t;
|
|
t = __clc_mad(x, 0x1.921fb54442d18p+1, t);
|
|
|
|
__CLC_GENTYPE tr = __CLC_FP_LIT(-1.0) / t;
|
|
|
|
return __CLC_CONVERT_LONGN(is_odd) ? tr : t;
|
|
}
|
|
|
|
#elif __CLC_FPSIZE == 16
|
|
|
|
_CLC_DEF _CLC_OVERLOAD _CLC_CONST __CLC_HALFN
|
|
__clc_tanpi_reduced_eval(__CLC_HALFN x, __CLC_INTN is_odd) {
|
|
__CLC_HALFN s = x * x;
|
|
|
|
__CLC_HALFN t =
|
|
__clc_mad(s, __clc_mad(s, 0x1.3d8p+8h, 0x1.fe4p+4h), 0x1.508p+3h);
|
|
|
|
t = x * s * t;
|
|
t = __clc_mad(x, 0x1.92p+1h, t);
|
|
|
|
__CLC_HALFN tr = __CLC_FP_LIT(-1.0) / t;
|
|
|
|
return __CLC_CONVERT_SHORTN(is_odd) ? tr : t;
|
|
}
|
|
|
|
#endif
|
|
|
|
_CLC_DEF _CLC_OVERLOAD _CLC_CONST __CLC_GENTYPE __clc_tanpi(__CLC_GENTYPE x) {
|
|
x = __clc_select(x, __CLC_GENTYPE_NAN,
|
|
__CLC_CONVERT_S_GENTYPE(__clc_isinf(x)));
|
|
|
|
__CLC_GENTYPE absx = __clc_fabs(x);
|
|
|
|
__CLC_GENTYPE reduced;
|
|
__CLC_INTN n = __clc_piArgReductionS(&reduced, absx);
|
|
|
|
__CLC_GENTYPE t = __clc_tanpi_reduced_eval(reduced, (n & 1) != 0);
|
|
|
|
__CLC_S_GENTYPE n_1_or_2 = __CLC_CONVERT_S_GENTYPE((n == 1) || (n == 2));
|
|
|
|
__CLC_S_GENTYPE flip = (n_1_or_2 && (reduced == __CLC_FP_LIT(0.0)))
|
|
? __CLC_GENTYPE_S_SIGNBIT
|
|
: (__CLC_S_GENTYPE)0;
|
|
|
|
__CLC_S_GENTYPE result = (__CLC_AS_S_GENTYPE(t) ^ flip) ^
|
|
(__CLC_AS_S_GENTYPE(x) & __CLC_GENTYPE_S_SIGNBIT);
|
|
|
|
return __CLC_AS_GENTYPE(result);
|
|
}
|