Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
OpenHarmony
Third Party Openssl
提交
61ba602a
T
Third Party Openssl
项目概览
OpenHarmony
/
Third Party Openssl
大约 1 年 前同步成功
通知
9
Star
18
Fork
1
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
T
Third Party Openssl
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
提交
61ba602a
编写于
10月 08, 2013
作者:
A
Andy Polyakov
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
sha/asm/sha*-mb-x86_64.pl: commentary update.
上级
c45a48c1
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
18 addition
and
14 deletion
+18
-14
crypto/sha/asm/sha1-mb-x86_64.pl
crypto/sha/asm/sha1-mb-x86_64.pl
+9
-7
crypto/sha/asm/sha256-mb-x86_64.pl
crypto/sha/asm/sha256-mb-x86_64.pl
+9
-7
未找到文件。
crypto/sha/asm/sha1-mb-x86_64.pl
浏览文件 @
61ba602a
...
...
@@ -12,20 +12,22 @@
# naturally limited to 4 on pre-AVX2 processors and to 8 on
# AVX2-capable processors such as Haswell.
#
# this +aesni(
*) sha1 aesni-sha1 gain
# this +aesni(
i) sha1 aesni-sha1 gain(iv)
# -------------------------------------------------------------------
# Westmere(
**
) 10.4/n +1.28=3.88(n=4) 5.44 6.58 +70%
# Atom(
**
) 18.9/n +3.93=8.66(n=4) 10.0 14.0 +62%
# Westmere(
ii
) 10.4/n +1.28=3.88(n=4) 5.44 6.58 +70%
# Atom(
ii
) 18.9/n +3.93=8.66(n=4) 10.0 14.0 +62%
# Sandy Bridge (8.16 +5.15=13.3)/n 4.99 5.98 +80%
# Ivy Bridge (8.03 +5.14=13.2)/n 4.60 5.54 +68%
# Haswell(
***
) (8.96 +5.00=14.0)/n 3.57 4.55 +160%
# Haswell(
iii
) (8.96 +5.00=14.0)/n 3.57 4.55 +160%
# Bulldozer (9.75 +5.76=15.5)/n 5.95 6.37 +64%
#
# (
*
) multi-block CBC encrypt with 128-bit key;
# (
**
) (HASH+AES)/n does not apply to Westmere for n>3 and Atom,
# (
i
) multi-block CBC encrypt with 128-bit key;
# (
ii
) (HASH+AES)/n does not apply to Westmere for n>3 and Atom,
# because of lower AES-NI instruction throughput;
# (
***
) "this" is for n=8, when we gather twice as much data, result
# (
iii
) "this" is for n=8, when we gather twice as much data, result
# for n=4 is 7.98+4.44=12.4;
# (iv) improvement coefficients in real-life application are somewhat
# lower and range from 30% to 100% (on Haswell);
$flavour
=
shift
;
$output
=
shift
;
...
...
crypto/sha/asm/sha256-mb-x86_64.pl
浏览文件 @
61ba602a
...
...
@@ -12,21 +12,23 @@
# naturally limited to 4 on pre-AVX2 processors and to 8 on
# AVX2-capable processors such as Haswell.
#
# this +aesni(
*) sha256 aesni-sha256 gain
# this +aesni(
i) sha256 aesni-sha256 gain(iv)
# -------------------------------------------------------------------
# Westmere(
**
) 23.3/n +1.28=7.11(n=4) 12.3 +3.75=16.1 +126%
# Atom(
**
) 39.1/n +3.93=13.7(n=4) 20.8 +5.69=26.5 +93%
# Westmere(
ii
) 23.3/n +1.28=7.11(n=4) 12.3 +3.75=16.1 +126%
# Atom(
ii
) 39.1/n +3.93=13.7(n=4) 20.8 +5.69=26.5 +93%
# Sandy Bridge (20.5 +5.15=25.7)/n 11.6 13.0 +103%
# Ivy Bridge (20.4 +5.14=25.5)/n 10.3 11.6 +82%
# Haswell(
***
) (21.0 +5.00=26.0)/n 7.80 8.79 +170%
# Haswell(
iii
) (21.0 +5.00=26.0)/n 7.80 8.79 +170%
# Bulldozer (21.6 +5.76=27.4)/n 13.6 13.7 +100%
#
# (
*
) multi-block CBC encrypt with 128-bit key;
# (
**
) (HASH+AES)/n does not apply to Westmere for n>3 and Atom,
# (
i
) multi-block CBC encrypt with 128-bit key;
# (
ii
) (HASH+AES)/n does not apply to Westmere for n>3 and Atom,
# because of lower AES-NI instruction throughput, nor is there
# AES-NI-SHA256 stitch for these processors;
# (
***
) "this" is for n=8, when we gather twice as much data, result
# (
iii
) "this" is for n=8, when we gather twice as much data, result
# for n=4 is 20.3+4.44=24.7;
# (iv) improvement coefficients in real-life application are somewhat
# lower and range from 75% to 130% (on Haswell);
$flavour
=
shift
;
$output
=
shift
;
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录