Re: native_save_fl taking good amount of time in profile

[Liu Zhiyou <liuzhiyou.cs@gmail.com>]

于 2012年05月26日 07:10, Xin Tong 写道:
> i do not know its callers from the profile. one thing i am suspicious
> of is that my kernel debug-info is wrong. by the way, what does the
> function do ?
>
> Thanks
>
> Xin
>
>
> On Fri, May 25, 2012 at 6:27 PM, Andi Kleen<andi@firstfloor.org>  wrote:
>> Xin Tong<xerox.time.tech@gmail.com>  writes:
>>
>>> I have collected a profile on the whole system. But the only thing
>>> that is really running is a mysql database server. I have noticed that
>>> there is a function in kernel that took 5% of the overall running
>>> time. what does native_save_fl do ? does it make sense that
>>> native_save_fl takes the most amount of time among all kernel
>>> functions ?
>> What are the callers (-g)? It does not make sense normally.
>> -Andi
Xin Tong,

Use perf record -g, then it will give you a call-graph.
$ man perf-record
-g, --call-graph
Do call-graph (stack chain/backtrace) recording.
$ man pref-report
-g [type,min[,limit],order], --call-graph
Display call chains using type, min percent threshold, optional print 
limit and order. type can be either:

· flat: single column, linear exposure of call chains.

· graph: use a graph tree, displaying absolute overhead rates.

· fractal: like graph, but displays relative rates. Each branch of the 
tree is considered as a new profiled object.

order can be either:
- callee: callee based call graph.
- caller: inverted caller based call graph.

Default: fractal,0.5,callee.

Regards,
LIU Zhiyou