[libc][math][c23] Add C23 math functions ilogbf128, logbf128, and llogb(f|l|f128). #82144
Conversation
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
|
@llvm/pr-subscribers-libc Author: None (lntue) ChangesPatch is 49.19 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/82144.diff 41 Files Affected:
diff --git a/libc/config/gpu/entrypoints.txt b/libc/config/gpu/entrypoints.txt
index b333c6be144627..c9010b85ae7902 100644
--- a/libc/config/gpu/entrypoints.txt
+++ b/libc/config/gpu/entrypoints.txt
@@ -197,6 +197,8 @@ set(TARGET_LIBM_ENTRYPOINTS
libc.src.math.ilogbf
libc.src.math.ldexp
libc.src.math.ldexpf
+ libc.src.math.llogb
+ libc.src.math.llogbf
libc.src.math.llrint
libc.src.math.llrintf
libc.src.math.llround
diff --git a/libc/config/linux/aarch64/entrypoints.txt b/libc/config/linux/aarch64/entrypoints.txt
index 6e194682df4bfc..c61f30a46c664b 100644
--- a/libc/config/linux/aarch64/entrypoints.txt
+++ b/libc/config/linux/aarch64/entrypoints.txt
@@ -306,6 +306,9 @@ set(TARGET_LIBM_ENTRYPOINTS
libc.src.math.ilogb
libc.src.math.ilogbf
libc.src.math.ilogbl
+ libc.src.math.llogb
+ libc.src.math.llogbf
+ libc.src.math.llogbl
libc.src.math.llrint
libc.src.math.llrintf
libc.src.math.llrintl
@@ -387,7 +390,10 @@ if(LIBC_COMPILER_HAS_FLOAT128)
libc.src.math.fmaxf128
libc.src.math.fminf128
libc.src.math.frexpf128
+ libc.src.math.ilogbf128
libc.src.math.ldexpf128
+ libc.src.math.llogbf128
+ libc.src.math.logbf128
libc.src.math.roundf128
libc.src.math.sqrtf128
libc.src.math.truncf128
diff --git a/libc/config/linux/arm/entrypoints.txt b/libc/config/linux/arm/entrypoints.txt
index 9bacfab7b0e5ae..a8116bd5e96d8a 100644
--- a/libc/config/linux/arm/entrypoints.txt
+++ b/libc/config/linux/arm/entrypoints.txt
@@ -182,6 +182,9 @@ set(TARGET_LIBM_ENTRYPOINTS
libc.src.math.ldexp
libc.src.math.ldexpf
libc.src.math.ldexpl
+ libc.src.math.llogb
+ libc.src.math.llogbf
+ libc.src.math.llogbl
libc.src.math.llrint
libc.src.math.llrintf
libc.src.math.llrintl
diff --git a/libc/config/linux/riscv/entrypoints.txt b/libc/config/linux/riscv/entrypoints.txt
index 71ff4bcfc35195..7a4a128865e52b 100644
--- a/libc/config/linux/riscv/entrypoints.txt
+++ b/libc/config/linux/riscv/entrypoints.txt
@@ -318,6 +318,9 @@ set(TARGET_LIBM_ENTRYPOINTS
libc.src.math.ldexp
libc.src.math.ldexpf
libc.src.math.ldexpl
+ libc.src.math.llogb
+ libc.src.math.llogbf
+ libc.src.math.llogbl
libc.src.math.llrint
libc.src.math.llrintf
libc.src.math.llrintl
@@ -396,7 +399,10 @@ if(LIBC_COMPILER_HAS_FLOAT128)
libc.src.math.fmaxf128
libc.src.math.fminf128
libc.src.math.frexpf128
+ libc.src.math.ilogbf128
libc.src.math.ldexpf128
+ libc.src.math.llogbf128
+ libc.src.math.logbf128
libc.src.math.roundf128
libc.src.math.sqrtf128
libc.src.math.truncf128
diff --git a/libc/config/linux/x86_64/entrypoints.txt b/libc/config/linux/x86_64/entrypoints.txt
index beb7a5ed448ded..39dd052be4d771 100644
--- a/libc/config/linux/x86_64/entrypoints.txt
+++ b/libc/config/linux/x86_64/entrypoints.txt
@@ -355,6 +355,9 @@ set(TARGET_LIBM_ENTRYPOINTS
libc.src.math.ldexp
libc.src.math.ldexpf
libc.src.math.ldexpl
+ libc.src.math.llogb
+ libc.src.math.llogbf
+ libc.src.math.llogbl
libc.src.math.llrint
libc.src.math.llrintf
libc.src.math.llrintl
@@ -435,7 +438,10 @@ if(LIBC_COMPILER_HAS_FLOAT128)
libc.src.math.fmaxf128
libc.src.math.fminf128
libc.src.math.frexpf128
+ libc.src.math.ilogbf128
libc.src.math.ldexpf128
+ libc.src.math.llogbf128
+ libc.src.math.logbf128
libc.src.math.roundf128
libc.src.math.sqrtf128
libc.src.math.truncf128
diff --git a/libc/docs/math/index.rst b/libc/docs/math/index.rst
index c586fe6664e27f..b16bda44f33cae 100644
--- a/libc/docs/math/index.rst
+++ b/libc/docs/math/index.rst
@@ -185,6 +185,8 @@ Basic Operations
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| ilogbl | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
+| ilogf128 | |check| | |check| | | |check| | | | | | | | | |
++--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| ldexp | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| ldexpf | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
@@ -193,6 +195,14 @@ Basic Operations
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| ldexpf128 | |check| | |check| | | |check| | | | | | | | | |
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
+| llogb | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
++--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
+| llogbf | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
++--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
+| llogbl | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
++--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
+| llogf128 | |check| | |check| | | |check| | | | | | | | | |
++--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| llrint | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| llrintf | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
@@ -211,6 +221,8 @@ Basic Operations
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| logbl | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
+| logf128 | |check| | |check| | | |check| | | | | | | | | |
++--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| lrint | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
+--------------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+
| lrintf | |check| | |check| | |check| | |check| | |check| | | | |check| | |check| | |check| | | |
diff --git a/libc/include/llvm-libc-macros/CMakeLists.txt b/libc/include/llvm-libc-macros/CMakeLists.txt
index 225885d3a9b085..2aff76b9b84df5 100644
--- a/libc/include/llvm-libc-macros/CMakeLists.txt
+++ b/libc/include/llvm-libc-macros/CMakeLists.txt
@@ -83,6 +83,8 @@ add_macro_header(
math_macros
HDR
math-macros.h
+ DEPENDS
+ .limits_macros
)
add_macro_header(
diff --git a/libc/include/llvm-libc-macros/math-macros.h b/libc/include/llvm-libc-macros/math-macros.h
index 136670e665e6b0..54d2c57700fe04 100644
--- a/libc/include/llvm-libc-macros/math-macros.h
+++ b/libc/include/llvm-libc-macros/math-macros.h
@@ -9,6 +9,8 @@
#ifndef __LLVM_LIBC_MACROS_MATH_MACROS_H
#define __LLVM_LIBC_MACROS_MATH_MACROS_H
+#include "limits-macros.h"
+
#define MATH_ERRNO 1
#define MATH_ERREXCEPT 2
@@ -19,6 +21,9 @@
#define FP_ILOGB0 (-__INT_MAX__ - 1)
#define FP_ILOGBNAN __INT_MAX__
+#define FP_LLOGB0 (-__LONG_MAX__ - 1)
+#define FP_LLOGBNAN __LONG_MAX__
+
#define isfinite(x) __builtin_isfinite(x)
#define isinf(x) __builtin_isinf(x)
#define isnan(x) __builtin_isnan(x)
diff --git a/libc/spec/stdc.td b/libc/spec/stdc.td
index 5fdcec7dbbdfcf..316461f5a6d462 100644
--- a/libc/spec/stdc.td
+++ b/libc/spec/stdc.td
@@ -414,6 +414,12 @@ def StdC : StandardSpec<"stdc"> {
FunctionSpec<"ilogb", RetValSpec<IntType>, [ArgSpec<DoubleType>]>,
FunctionSpec<"ilogbf", RetValSpec<IntType>, [ArgSpec<FloatType>]>,
FunctionSpec<"ilogbl", RetValSpec<IntType>, [ArgSpec<LongDoubleType>]>,
+ GuardedFunctionSpec<"ilogbf128", RetValSpec<IntType>, [ArgSpec<Float128Type>], "LIBC_COMPILER_HAS_FLOAT128">,
+
+ FunctionSpec<"llogb", RetValSpec<LongType>, [ArgSpec<DoubleType>]>,
+ FunctionSpec<"llogbf", RetValSpec<LongType>, [ArgSpec<FloatType>]>,
+ FunctionSpec<"llogbl", RetValSpec<LongType>, [ArgSpec<LongDoubleType>]>,
+ GuardedFunctionSpec<"llogbf128", RetValSpec<LongType>, [ArgSpec<Float128Type>], "LIBC_COMPILER_HAS_FLOAT128">,
FunctionSpec<"ldexp", RetValSpec<DoubleType>, [ArgSpec<DoubleType>, ArgSpec<IntType>]>,
FunctionSpec<"ldexpf", RetValSpec<FloatType>, [ArgSpec<FloatType>, ArgSpec<IntType>]>,
@@ -435,6 +441,7 @@ def StdC : StandardSpec<"stdc"> {
FunctionSpec<"logb", RetValSpec<DoubleType>, [ArgSpec<DoubleType>]>,
FunctionSpec<"logbf", RetValSpec<FloatType>, [ArgSpec<FloatType>]>,
FunctionSpec<"logbl", RetValSpec<LongDoubleType>, [ArgSpec<LongDoubleType>]>,
+ GuardedFunctionSpec<"logbf128", RetValSpec<Float128Type>, [ArgSpec<Float128Type>], "LIBC_COMPILER_HAS_FLOAT128">,
FunctionSpec<"modf", RetValSpec<DoubleType>, [ArgSpec<DoubleType>, ArgSpec<DoublePtr>]>,
FunctionSpec<"modff", RetValSpec<FloatType>, [ArgSpec<FloatType>, ArgSpec<FloatPtr>]>,
diff --git a/libc/src/__support/FPUtil/ManipulationFunctions.h b/libc/src/__support/FPUtil/ManipulationFunctions.h
index 9e760a28f42d75..8dd9156ce6b4b3 100644
--- a/libc/src/__support/FPUtil/ManipulationFunctions.h
+++ b/libc/src/__support/FPUtil/ManipulationFunctions.h
@@ -71,54 +71,79 @@ LIBC_INLINE T copysign(T x, T y) {
return xbits.get_val();
}
-template <typename T, cpp::enable_if_t<cpp::is_floating_point_v<T>, int> = 0>
-LIBC_INLINE int ilogb(T x) {
- // TODO: Raise appropriate floating point exceptions and set errno to the
- // an appropriate error value wherever relevant.
- FPBits<T> bits(x);
- if (bits.is_zero()) {
- return FP_ILOGB0;
- } else if (bits.is_nan()) {
- return FP_ILOGBNAN;
- } else if (bits.is_inf()) {
- return INT_MAX;
+template <typename T> struct IntLogbConstants;
+
+template <> struct IntLogbConstants<int> {
+ LIBC_INLINE_VAR static constexpr int FP_LOGB0 = FP_ILOGB0;
+ LIBC_INLINE_VAR static constexpr int FP_LOGBNAN = FP_ILOGBNAN;
+ LIBC_INLINE_VAR static constexpr int T_MAX = INT_MAX;
+ LIBC_INLINE_VAR static constexpr int T_MIN = INT_MIN;
+};
+
+template <> struct IntLogbConstants<long> {
+ LIBC_INLINE_VAR static constexpr long FP_LOGB0 = FP_ILOGB0;
+ LIBC_INLINE_VAR static constexpr long FP_LOGBNAN = FP_ILOGBNAN;
+ LIBC_INLINE_VAR static constexpr long T_MAX = INT_MAX;
+ LIBC_INLINE_VAR static constexpr long T_MIN = INT_MIN;
+};
+
+template <typename T, typename U>
+LIBC_INLINE constexpr cpp::enable_if_t<cpp::is_floating_point_v<U>, T>
+intlogb(U x) {
+ FPBits<U> bits(x);
+ if (LIBC_UNLIKELY(bits.is_zero() || bits.is_inf_or_nan())) {
+ set_errno_if_required(EDOM);
+ raise_except_if_required(FE_INVALID);
+
+ if (bits.is_zero())
+ return IntLogbConstants<T>::FP_LOGB0;
+ if (bits.is_nan())
+ return IntLogbConstants<T>::FP_LOGBNAN;
+ // bits is inf.
+ return IntLogbConstants<T>::T_MAX;
}
- NormalFloat<T> normal(bits);
+ DyadicFloat<FPBits<U>::STORAGE_LEN> normal(bits.get_val());
+ int exponent = normal.get_unbiased_exponent();
// The C standard does not specify the return value when an exponent is
// out of int range. However, XSI conformance required that INT_MAX or
// INT_MIN are returned.
// NOTE: It is highly unlikely that exponent will be out of int range as
// the exponent is only 15 bits wide even for the 128-bit floating point
// format.
- if (normal.exponent > INT_MAX)
- return INT_MAX;
- else if (normal.exponent < INT_MIN)
- return INT_MIN;
- else
- return normal.exponent;
+ if (LIBC_UNLIKELY(exponent > IntLogbConstants<T>::T_MAX ||
+ exponent < IntLogbConstants<T>::T_MIN)) {
+ set_errno_if_required(ERANGE);
+ raise_except_if_required(FE_INVALID);
+ return exponent ? IntLogbConstants<T>::T_MAX : IntLogbConstants<T>::T_MIN;
+ }
+
+ return static_cast<T>(exponent);
}
template <typename T, cpp::enable_if_t<cpp::is_floating_point_v<T>, int> = 0>
-LIBC_INLINE T logb(T x) {
+LIBC_INLINE constexpr T logb(T x) {
FPBits<T> bits(x);
- if (bits.is_zero()) {
- // TODO(Floating point exception): Raise div-by-zero exception.
- // TODO(errno): POSIX requires setting errno to ERANGE.
- return FPBits<T>::inf(Sign::NEG).get_val();
- } else if (bits.is_nan()) {
- return x;
- } else if (bits.is_inf()) {
- // Return positive infinity.
+ if (LIBC_UNLIKELY(bits.is_zero() || bits.is_inf_or_nan())) {
+ if (bits.is_nan())
+ return x;
+
+ raise_except_if_required(FE_DIVBYZERO);
+
+ if (bits.is_zero()) {
+ set_errno_if_required(ERANGE);
+ return FPBits<T>::inf(Sign::NEG).get_val();
+ }
+ // bits is inf.
return FPBits<T>::inf().get_val();
}
- NormalFloat<T> normal(bits);
- return static_cast<T>(normal.exponent);
+ DyadicFloat<FPBits<T>::STORAGE_LEN> normal(bits.get_val());
+ return static_cast<T>(normal.get_unbiased_exponent());
}
template <typename T, cpp::enable_if_t<cpp::is_floating_point_v<T>, int> = 0>
-LIBC_INLINE T ldexp(T x, int exp) {
+LIBC_INLINE constexpr T ldexp(T x, int exp) {
FPBits<T> bits(x);
if (LIBC_UNLIKELY((exp == 0) || bits.is_zero() || bits.is_inf_or_nan()))
return x;
diff --git a/libc/src/__support/FPUtil/dyadic_float.h b/libc/src/__support/FPUtil/dyadic_float.h
index 7797c57b96fd85..14bc73433097b7 100644
--- a/libc/src/__support/FPUtil/dyadic_float.h
+++ b/libc/src/__support/FPUtil/dyadic_float.h
@@ -79,6 +79,11 @@ template <size_t Bits> struct DyadicFloat {
return *this;
}
+ // Assume that it is already normalized. Output the unbiased exponent.
+ LIBC_INLINE constexpr int get_unbiased_exponent() const {
+ return exponent + (Bits - 1);
+ }
+
// Assume that it is already normalized.
// Output is rounded correctly with respect to the current rounding mode.
template <typename T,
diff --git a/libc/src/math/CMakeLists.txt b/libc/src/math/CMakeLists.txt
index 05ce51e8fc6503..1b2fe78182b212 100644
--- a/libc/src/math/CMakeLists.txt
+++ b/libc/src/math/CMakeLists.txt
@@ -145,6 +145,12 @@ add_math_entrypoint_object(hypotf)
add_math_entrypoint_object(ilogb)
add_math_entrypoint_object(ilogbf)
add_math_entrypoint_object(ilogbl)
+add_math_entrypoint_object(ilogbf128)
+
+add_math_entrypoint_object(llogb)
+add_math_entrypoint_object(llogbf)
+add_math_entrypoint_object(llogbl)
+add_math_entrypoint_object(llogbf128)
add_math_entrypoint_object(ldexp)
add_math_entrypoint_object(ldexpf)
@@ -166,6 +172,7 @@ add_math_entrypoint_object(logf)
add_math_entrypoint_object(logb)
add_math_entrypoint_object(logbf)
add_math_entrypoint_object(logbl)
+add_math_entrypoint_object(logbf128)
add_math_entrypoint_object(llrint)
add_math_entrypoint_object(llrintf)
diff --git a/libc/src/math/generic/CMakeLists.txt b/libc/src/math/generic/CMakeLists.txt
index 07885ae6a3e242..f4b7fcc82fb8cc 100644
--- a/libc/src/math/generic/CMakeLists.txt
+++ b/libc/src/math/generic/CMakeLists.txt
@@ -969,10 +969,10 @@ add_entrypoint_object(
ilogb.cpp
HDRS
../ilogb.h
+ COMPILE_OPTIONS
+ -O3
DEPENDS
libc.src.__support.FPUtil.manipulation_functions
- COMPILE_OPTIONS
- -O2
)
add_entrypoint_object(
@@ -981,10 +981,10 @@ add_entrypoint_object(
ilogbf.cpp
HDRS
../ilogbf.h
+ COMPILE_OPTIONS
+ -O3
DEPENDS
libc.src.__support.FPUtil.manipulation_functions
- COMPILE_OPTIONS
- -O2
)
add_entrypoint_object(
@@ -993,10 +993,72 @@ add_entrypoint_object(
ilogbl.cpp
HDRS
../ilogbl.h
+ COMPILE_OPTIONS
+ -O3
DEPENDS
libc.src.__support.FPUtil.manipulation_functions
+)
+
+add_entrypoint_object(
+ ilogbf128
+ SRCS
+ ilogbf128.cpp
+ HDRS
+ ../ilogbf128.h
COMPILE_OPTIONS
- -O2
+ -O3
+ DEPENDS
+ libc.src.__support.macros.properties.float
+ libc.src.__support.FPUtil.manipulation_functions
+)
+
+add_entrypoint_object(
+ llogb
+ SRCS
+ llogb.cpp
+ HDRS
+ ../llogb.h
+ COMPILE_OPTIONS
+ -O3
+ DEPENDS
+ libc.src.__support.FPUtil.manipulation_functions
+)
+
+add_entrypoint_object(
+ llogbf
+ SRCS
+ llogbf.cpp
+ HDRS
+ ../llogbf.h
+ COMPILE_OPTIONS
+ -O3
+ DEPENDS
+ libc.src.__support.FPUtil.manipulation_functions
+)
+
+add_entrypoint_object(
+ llogbl
+ SRCS
+ llogbl.cpp
+ HDRS
+ ../llogbl.h
+ COMPILE_OPTIONS
+ -O3
+ DEPENDS
+ libc.src.__support.FPUtil.manipulation_functions
+)
+
+add_entrypoint_object(
+ llogbf128
+ SRCS
+ llogbf128.cpp
+ HDRS
+ ../llogbf128.h
+ COMPILE_OPTIONS
+ -O3
+ DEPENDS
+ libc.src.__support.macros.properties.float
+ libc.src.__support.FPUtil.manipulation_functions
)
add_entrypoint_object(
@@ -1044,8 +1106,8 @@ add_entrypoint_object(
COMPILE_OPTIONS
-O3
DEPENDS
- libc.src.__support.macros.properties.float
- libc.src.__support.FPUtil.manipulation_functions
+ libc.src.__support.macros.properties.float
+ libc.src.__support.FPUtil.manipulation_functions
)
add_object_library(
@@ -1229,10 +1291,10 @@ add_entrypoint_object(
logb.cpp
HDRS
../logb.h
+ COMPILE_OPTIONS
+ -O3
DEPENDS
libc.src.__support.FPUtil.manipulation_functions
- COMPILE_OPTIONS
- -O2
)
add_entrypoint_object(
@@ -1241,10 +1303,10 @@ add_entrypoint_object(
logbf.cpp
HDRS
../logbf.h
+ COMPILE_OPTIONS
+ -O3
DEPENDS
libc.src.__support.FPUtil.manipulation_functions
- COMPILE_OPTIONS
- -O2
)
add_entrypoint_object(
@@ -1253,10 +1315,22 @@ add_entrypoint_object(
logbl.cpp
HDRS
../logbl.h
+ COMPILE_OPTIONS
+ -O3
DEPENDS
libc.src.__support.FPUtil.manipulation_functions
+)
+
+add_entrypoint_object(
+ logbf128
+ SRCS
+ logbf128.cpp
+ HDRS
+ ../logbf128.h
COMPILE_OPTIONS
- -O2
+ -O3
+ DEPENDS
+ libc.src.__support.FPUtil.manipulation_functions
)
add_entrypoint_object(
diff --git a/libc/src/math/generic/ilogb.cpp b/libc/src/math/generic/ilogb.cpp
index 4e5f7d9642b4aa..7e4f66970c5d3c 100644
--- a/libc/src/math/generic/ilogb.cpp
+++ b/libc/src/math/generic/ilogb.cpp
@@ -12,6 +12,6 @@
namespace LIBC_NAMESPACE {
-LLVM_LIBC_FUNCTION(int, ilogb, (double x)) { return fputil::ilogb(x); }
+LLVM_LIBC_FUNCTION(int, ilogb, (double x)) { return fputil::intlogb<int>(x); }
} // namespace LIBC_NAMESPACE
diff --git a/libc/src/math/generic/ilogbf.cpp b/libc/src/math/generic/ilogbf.cpp
index ca15879bc25fe5..422788cec9e04f 100644
--- a/libc/src/math/generic/ilogbf.cpp
+++ b/libc/src/math/generic/ilogbf.cpp
@@ -12,6 +12,6 @@
namespace LIBC_NAMESPACE {
-LLVM_LIBC_FUNCTION(int, ilogbf, (float x)) { return fputil::ilogb(x); }
+LLVM_LIBC_FUNCTION(int, ilogbf, (float x)) { return fputil::intlogb<int>(x); }
} // namespace LIBC_NAMES...
[truncated]
|
jhuber6
reviewed
Feb 18, 2024
| COMPILE_OPTIONS | ||
| -O2 | ||
| -O3 |
There was a problem hiding this comment.
Unrelated, but what is the logic behind overriding the optimization level?
There was a problem hiding this comment.
just to make them consistent with the rest of the math functions
Comment on lines
+127
to
+130
| if (LIBC_UNLIKELY(bits.is_zero() || bits.is_inf_or_nan())) { | ||
| if (bits.is_nan()) | ||
| return x; |
There was a problem hiding this comment.
Suggested change
| if (LIBC_UNLIKELY(bits.is_zero() || bits.is_inf_or_nan())) { | |
| if (bits.is_nan()) | |
| return x; | |
| if (LIBC_UNLIKELY(bits.is_nan())) | |
| return x; | |
| if (LIBC_UNLIKELY(bits.is_zero() || bits.is_inf())) { |
There was a problem hiding this comment.
Since the test for is_inf_or_nan() almost the same as is_nan(), while the test for is_inf() is totally different, putting the is_nan() branch out intercepts with the separation of the single LIKELY / UNLIKELY pair, the generated codes are 10-15% worse than the current form.
Comment on lines
114
to
118
| if (LIBC_UNLIKELY(exponent > IntLogbConstants<T>::T_MAX || | ||
| exponent < IntLogbConstants<T>::T_MIN)) { | ||
| set_errno_if_required(ERANGE); | ||
| raise_except_if_required(FE_INVALID); | ||
| return exponent ? IntLogbConstants<T>::T_MAX : IntLogbConstants<T>::T_MIN; |
There was a problem hiding this comment.
This seems like a change of logic (which is maybe ok, but is it intentional)?
For example, consider the case where exponent < INT_MIN. Previously, this would return INT_MIN. Now it looks like it will return INT_MAX?
There was a problem hiding this comment.
Fixed to match with the old implementation. This case is mentioned but unspecified in the standard.
nickdesaulniers
approved these changes
Feb 27, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.