[libc][math] Refactor expf16 implementation to header-only in src/__support/math folder. #147428
+203
−125
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -12,5 +12,6 @@ | |
| #include "libc_common.h" | ||
|
|
||
| #include "math/expf.h" | ||
| #include "math/expf16.h" | ||
|
|
||
| #endif // LLVM_LIBC_SHARED_MATH_H | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,23 @@ | ||
| //===-- Shared expf16 function ----------------------------------*- 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 | ||
| // | ||
| //===----------------------------------------------------------------------===// | ||
|
|
||
| #ifndef LLVM_LIBC_SHARED_MATH_EXPF16_H | ||
| #define LLVM_LIBC_SHARED_MATH_EXPF16_H | ||
|
|
||
| #include "shared/libc_common.h" | ||
| #include "src/__support/math/expf16.h" | ||
|
|
||
| namespace LIBC_NAMESPACE_DECL { | ||
| namespace shared { | ||
|
|
||
| using math::expf16; | ||
|
|
||
| } // namespace shared | ||
| } // namespace LIBC_NAMESPACE_DECL | ||
|
|
||
| #endif // LLVM_LIBC_SHARED_MATH_EXPF16_H | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,134 @@ | ||
| //===-- Implementation header for expf16 ------------------------*- 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 | ||
| // | ||
| //===----------------------------------------------------------------------===// | ||
|
|
||
| #ifndef LLVM_LIBC_SRC___SUPPORT_MATH_EXPF16_H | ||
| #define LLVM_LIBC_SRC___SUPPORT_MATH_EXPF16_H | ||
|
|
||
| #include "hdr/errno_macros.h" | ||
| #include "hdr/fenv_macros.h" | ||
| #include "src/__support/FPUtil/FEnvImpl.h" | ||
| #include "src/__support/FPUtil/FPBits.h" | ||
| #include "src/__support/FPUtil/PolyEval.h" | ||
| #include "src/__support/FPUtil/cast.h" | ||
| #include "src/__support/FPUtil/except_value_utils.h" | ||
| #include "src/__support/FPUtil/rounding_mode.h" | ||
| #include "src/__support/common.h" | ||
| #include "src/__support/macros/config.h" | ||
| #include "src/__support/macros/optimization.h" | ||
| #include "src/math/generic/expxf16.h" | ||
|
|
||
| namespace LIBC_NAMESPACE_DECL { | ||
|
|
||
| namespace math { | ||
|
|
||
| static constexpr float16 expf16(float16 x) { | ||
bassiounix marked this conversation as resolved.
Show resolved
Hide resolved
|
||
| #ifndef LIBC_MATH_HAS_SKIP_ACCURATE_PASS | ||
| constexpr fputil::ExceptValues<float16, 2> EXPF16_EXCEPTS_LO = {{ | ||
| // (input, RZ output, RU offset, RD offset, RN offset) | ||
| // x = 0x1.de4p-8, expf16(x) = 0x1.01cp+0 (RZ) | ||
| {0x1f79U, 0x3c07U, 1U, 0U, 0U}, | ||
| // x = 0x1.73cp-6, expf16(x) = 0x1.05cp+0 (RZ) | ||
| {0x25cfU, 0x3c17U, 1U, 0U, 0U}, | ||
| }}; | ||
|
|
||
| constexpr fputil::ExceptValues<float16, 3> EXPF16_EXCEPTS_HI = {{ | ||
| // (input, RZ output, RU offset, RD offset, RN offset) | ||
| // x = 0x1.c34p+0, expf16(x) = 0x1.74cp+2 (RZ) | ||
| {0x3f0dU, 0x45d3U, 1U, 0U, 1U}, | ||
| // x = -0x1.488p-5, expf16(x) = 0x1.ebcp-1 (RZ) | ||
| {0xa922U, 0x3bafU, 1U, 0U, 0U}, | ||
| // x = -0x1.55p-5, expf16(x) = 0x1.ebp-1 (RZ) | ||
| {0xa954U, 0x3bacU, 1U, 0U, 0U}, | ||
| }}; | ||
| #endif // !LIBC_MATH_HAS_SKIP_ACCURATE_PASS | ||
|
|
||
| using FPBits = fputil::FPBits<float16>; | ||
| FPBits x_bits(x); | ||
|
|
||
| uint16_t x_u = x_bits.uintval(); | ||
| uint16_t x_abs = x_u & 0x7fffU; | ||
|
|
||
| // When 0 < |x| <= 2^(-5), or |x| >= 12, or x is NaN. | ||
| if (LIBC_UNLIKELY(x_abs <= 0x2800U || x_abs >= 0x4a00U)) { | ||
| // exp(NaN) = NaN | ||
| if (x_bits.is_nan()) { | ||
| if (x_bits.is_signaling_nan()) { | ||
| fputil::raise_except_if_required(FE_INVALID); | ||
| return FPBits::quiet_nan().get_val(); | ||
| } | ||
|
|
||
| return x; | ||
| } | ||
|
|
||
| // When x >= 12. | ||
| if (x_bits.is_pos() && x_abs >= 0x4a00U) { | ||
| // exp(+inf) = +inf | ||
| if (x_bits.is_inf()) | ||
| return FPBits::inf().get_val(); | ||
|
|
||
| switch (fputil::quick_get_round()) { | ||
| case FE_TONEAREST: | ||
| case FE_UPWARD: | ||
| fputil::set_errno_if_required(ERANGE); | ||
| fputil::raise_except_if_required(FE_OVERFLOW); | ||
| return FPBits::inf().get_val(); | ||
| default: | ||
| return FPBits::max_normal().get_val(); | ||
| } | ||
| } | ||
|
|
||
| // When x <= -18. | ||
| if (x_u >= 0xcc80U) { | ||
| // exp(-inf) = +0 | ||
| if (x_bits.is_inf()) | ||
| return FPBits::zero().get_val(); | ||
|
|
||
| fputil::set_errno_if_required(ERANGE); | ||
| fputil::raise_except_if_required(FE_UNDERFLOW | FE_INEXACT); | ||
|
|
||
| switch (fputil::quick_get_round()) { | ||
| case FE_UPWARD: | ||
| return FPBits::min_subnormal().get_val(); | ||
| default: | ||
| return FPBits::zero().get_val(); | ||
| } | ||
| } | ||
|
|
||
| // When 0 < |x| <= 2^(-5). | ||
| if (x_abs <= 0x2800U && !x_bits.is_zero()) { | ||
| #ifndef LIBC_MATH_HAS_SKIP_ACCURATE_PASS | ||
| if (auto r = EXPF16_EXCEPTS_LO.lookup(x_u); LIBC_UNLIKELY(r.has_value())) | ||
| return r.value(); | ||
| #endif // !LIBC_MATH_HAS_SKIP_ACCURATE_PASS | ||
|
|
||
| float xf = x; | ||
| // Degree-3 minimax polynomial generated by Sollya with the following | ||
| // commands: | ||
| // > display = hexadecimal; | ||
| // > P = fpminimax(expm1(x)/x, 2, [|SG...|], [-2^-5, 2^-5]); | ||
| // > 1 + x * P; | ||
| return fputil::cast<float16>( | ||
| fputil::polyeval(xf, 0x1p+0f, 0x1p+0f, 0x1.0004p-1f, 0x1.555778p-3f)); | ||
| } | ||
| } | ||
|
|
||
| #ifndef LIBC_MATH_HAS_SKIP_ACCURATE_PASS | ||
| if (auto r = EXPF16_EXCEPTS_HI.lookup(x_u); LIBC_UNLIKELY(r.has_value())) | ||
| return r.value(); | ||
| #endif // !LIBC_MATH_HAS_SKIP_ACCURATE_PASS | ||
|
|
||
| // exp(x) = exp(hi + mid) * exp(lo) | ||
| auto [exp_hi_mid, exp_lo] = exp_range_reduction(x); | ||
| return fputil::cast<float16>(exp_hi_mid * exp_lo); | ||
| } | ||
|
|
||
| } // namespace math | ||
|
|
||
| } // namespace LIBC_NAMESPACE_DECL | ||
|
|
||
| #endif // LLVM_LIBC_SRC___SUPPORT_MATH_EXPF16_H | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This needs to be guarded with
LIBC_TYPES_HAS_FLOAT16