Subject: Re: [tip:perf/urgent] perf/core: Fix the perf_cpu_time_max_percent
 check
To: Vince Weaver <vincent.weaver@maine.edu>,
        Peter Zijlstra <peterz@infradead.org>
References: <1487829879-56237-1-git-send-email-tanxiaojun@huawei.com>
 <tip-1572e45a924f254d9570093abde46430c3172e3d@git.kernel.org>
 <58B13BFD.2050304@huawei.com>
 <20170225095117.GU6500@twins.programming.kicks-ass.net>
 <alpine.DEB.2.20.1703131432371.12942@macbook-air>
CC: <jolsa@redhat.com>, <alexander.shishkin@linux.intel.com>,
        <eranian@google.com>, <mingo@kernel.org>,
        <linux-kernel@vger.kernel.org>
From: Tan Xiaojun <tanxiaojun@huawei.com>
Message-ID: <58C74CEA.4030402@huawei.com>
Date: Tue, 14 Mar 2017 09:52:42 +0800
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:38.0) Gecko/20100101
 Thunderbird/38.6.0
MIME-Version: 1.0
In-Reply-To: <alpine.DEB.2.20.1703131432371.12942@macbook-air>
Content-Type: text/plain; charset="windows-1252"
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1434
Lines: 44

On 2017/3/14 2:35, Vince Weaver wrote:
> On Sat, 25 Feb 2017, Peter Zijlstra wrote:
> 
>> On Sat, Feb 25, 2017 at 04:10:37PM +0800, Tan Xiaojun wrote:
>>
>>> 2)If it is, where we will fix it more appropriate, perf_fuzzer(not set
>>> 0 or 100) or kernel(limit 1 to 99), or maybe it is the bug of
>>> hardware(too many hardware interruptions)?
>>
>> I think the best would be if the fuzzer would not set 0,100, those are
>> clearly 'unsafe' settings and you pretty much get to keep the pieces.
>>
>> I would like to preserve these settings for people that 'know' what
>> they're doing and are willing to take the risk, but clearly, when you
>> take the guard-rails off, things can come apart.
> 
> sorry for the delay responding, these e-mails ended up in the spam folder
> somehow.
> 
> I could add a new "avoid stupid things as root" flag for the perf_fuzzer.
> 
> Besides this issue, are there other known things to skip?
> 
> Generally running a fuzzer as root can be a bad idea which is why I don't 
> test that use case very often.
> I think there were other issues in the past, like certain ftrace 
> combinations being known to lock the system.
> 
> Vince
> 

It would be better if you could add such a flag to the perf_fuzzer. And I
have not found any other problems yet.

By the way. Use Non-root user to test is OK, and they do not have permission
to configure these parameters.

Thank you for your reply.

Xiaojun.

> .
>