[sprdlinux] 答复: some info about likely/unlikely
- From: Chunyan Zhang (张春艳) <Chunyan.Zhang@xxxxxxxxxxxxxx>
- To: Orson Zhai <orsonzhai@xxxxxxxxx>, Aijun Sun (孙爱军) <Aijun.Sun@xxxxxxxxxxxxxx>
- Date: Mon, 22 Dec 2014 02:52:00 +0000
Who can copy/paste this webpage for us, please?
http://bitsup.blogspot.sg/2008/04/linux-kernel-likely-not-measured.html
Best regards
Chunyan
________________________________________
发件人: Orson Zhai <orsonzhai@xxxxxxxxx>
发送时间: 2014年12月20日 1:20
收件人: Chunyan Zhang (张春艳)
抄送: sprdlinux@xxxxxxxxxxxxx
主题: some info about likely/unlikely
前几天你提到Greg K.H. 不推荐使用likely/unlikely，我看到这篇文章佐证了他
的观点
http://bitsup.blogspot.sg/2008/04/linux-kernel-likely-not-measured.html
（需翻墙）
作者失望的发现，likely/unlikely没有带来任何实质的好处。通过这2个宏达到性
能优化，完全是程序员yy而已。
在Linux Newbie的FAQ结论里也提到，
> You should use it only in cases when the likeliest branch is very
very very likely, or when the unlikeliest branch is very very very
unlikely.
作者连用了3个very来表达其强烈的感情....
http://kernelnewbies.org/FAQ/LikelyUnlikely
P.S.
Grep K.H.除了是社区的2号人物外，也是LDD3的作者之一。
Other related posts:
