squiid/llvm
1
0
Fork 0
Code Issues Pull requests Activity

[libc] Use nearest_integer instructions to improve expm1f performance.

Browse source 
Use nearest_integer instructions to improve expf performance.

Performance tests with CORE-MATH's perf tool:

Before the patch:
```
$ ./perf.sh expm1f
LIBC-location: /home/lnt/experiment/llvm/llvm-project/build/projects/libc/lib/libllvmlibc.a
GNU libc version: 2.31
GNU libc release: stable
CORE-MATH reciprocal throughput   : 10.096
System LIBC reciprocal throughput : 44.036
LIBC reciprocal throughput        : 11.575

$ ./perf.sh expm1f --latency
LIBC-location: /home/lnt/experiment/llvm/llvm-project/build/projects/libc/lib/libllvmlibc.a
GNU libc version: 2.31
GNU libc release: stable
CORE-MATH latency   : 42.239
System LIBC latency : 122.815
LIBC latency        : 50.122
```
After the patch:
```
$ ./perf.sh expm1f
LIBC-location: /home/lnt/experiment/llvm/llvm-project/build/projects/libc/lib/libllvmlibc.a
GNU libc version: 2.31
GNU libc release: stable
CORE-MATH reciprocal throughput   : 10.046
System LIBC reciprocal throughput : 43.899
LIBC reciprocal throughput        : 9.179

$ ./perf.sh expm1f --latency
LIBC-location: /home/lnt/experiment/llvm/llvm-project/build/projects/libc/lib/libllvmlibc.a
GNU libc version: 2.31
GNU libc release: stable
CORE-MATH latency   : 42.078
System LIBC latency : 120.488
LIBC latency        : 41.528
```

Reviewed By: zimmermann6

Differential Revision: https://reviews.llvm.org/D130502
This commit is contained in:
Tue Ly 2022-07-25 13:44:46 -04:00
parent 91ee672062
commit 628fbbef81
3 changed files with 7 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
libc/docs/math.rst
View file

@ -203,7 +203,7 @@ Performance
+--------------+-----------+-------------------+-----------+-------------------+-------------------------------------+------------+-------------------------+--------------+---------------+
| exp2f | 25 | 8 | 81 | 37 | :math:`[-10, 10]` | Ryzen 1700 | Ubuntu 20.04 LTS x86_64 | Clang 12.0.0 | FMA |
+--------------+-----------+-------------------+-----------+-------------------+-------------------------------------+------------+-------------------------+--------------+---------------+
| expm1f | 14 | 53 | 59 | 146 | :math:`[-10, 10]` | Ryzen 1700 | Ubuntu 20.04 LTS x86_64 | Clang 12.0.0 | FMA |
| expm1f | 9 | 44 | 42 | 121 | :math:`[-10, 10]` | Ryzen 1700 | Ubuntu 20.04 LTS x86_64 | Clang 12.0.0 | FMA |
+--------------+-----------+-------------------+-----------+-------------------+-------------------------------------+------------+-------------------------+--------------+---------------+
| fmodf | 73 | 263 | - | - | [MIN_NORMAL, MAX_NORMAL] | i5 mobile | Ubuntu 20.04 LTS x86_64 | Clang 12.0.0 | |
| +-----------+-------------------+-----------+-------------------+-------------------------------------+------------+-------------------------+--------------+---------------+

1
libc/src/math/generic/CMakeLists.txt
View file

@ -515,6 +515,7 @@ add_entrypoint_object(
.common_constants
libc.src.__support.FPUtil.fputil
libc.src.__support.FPUtil.multiply_add
libc.src.__support.FPUtil.nearest_integer
libc.src.__support.FPUtil.polyeval
libc.include.math
COMPILE_OPTIONS

8
libc/src/math/generic/expm1f.cpp
View file

@ -13,6 +13,8 @@
#include "src/__support/FPUtil/FMA.h"
#include "src/__support/FPUtil/FPBits.h"
#include "src/__support/FPUtil/PolyEval.h"
#include "src/__support/FPUtil/multiply_add.h"
#include "src/__support/FPUtil/nearest_integer.h"
#include "src/__support/common.h"
#include <errno.h>
@ -133,10 +135,10 @@ LLVM_LIBC_FUNCTION(float, expm1f, (float x)) {
// generated by Sollya.
// x_hi = hi + mid.
int x_hi = static_cast<int>(x * 0x1.0p7f + (xbits.get_sign() ? -0.5f : 0.5f));
float kf = fputil::nearest_integer(x * 0x1.0p7f);
int x_hi = static_cast<int>(kf);
// Subtract (hi + mid) from x to get lo.
x -= static_cast<float>(x_hi) * 0x1.0p-7f;
double xd = static_cast<double>(x);
double xd = static_cast<double>(fputil::multiply_add(kf, -0x1.0p-7f, x));
x_hi += 104 << 7;
// hi = x_hi >> 7
double exp_hi = EXP_M1[x_hi >> 7];