Re: perf, kprobes: fuzzer generates huge number of WARNings

[Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>]

On 2015/07/07 13:00, Vince Weaver wrote:
> On Tue, 7 Jul 2015, Masami Hiramatsu wrote:
> 
>> On 2015/07/07 6:27, Vince Weaver wrote:
>>> Hello
>>>
>>> I've been working on trying to get the perf_fuzzer to start fuzzing the 
>>> PERF_EVENT_IOC_SET_BPF so I've added some really hackish kprobe support.
>>>
>>> However before I can get to the BPF testing the kprobe code generates a 
>>> constant stream of WARNINGS which make the machine more or less useless 
>>> until I stop it.  I've included a small selection here.
>>>
>>> Is this expected?
>>
>> Did you get same message without BPF hack? And also, could you tell us
>> your kernel version and configuration?
> 
> It's a Hawell machine running 4.2-rc1.  I can post the .config if it's of 
> interest.

Yes, I'm interested in.

> Well the BPF hack is in the fuzzer, not the kernel.  And it's not really a 
> hack, it just turned out to be a huge pain to figure out how to 
> manually create a valid BPF program in conjunction with a valid kprobe 
> event.
> 
> I did have to sprinkle printks in the kprobe and bpf code to find out 
> where various EINVAL returns were coming from, so potentially this is just 
> a problem of printks happening where they shouldn't.  I'll remove those 
> changes and try to reproduce this tomorrow.

OK, and also, if you have a chance, please run the ftracetest as below.

$ cd tools/testing/selftest/ftrace/
$ sudo ./ftracetest

This will do a series of basic tests on ftrace and report it.

> This is possibly a long standing issue, until now I never ran the fuzzer 
> as root so these particular code paths weren't tested.

Thanks,

-- 
Masami HIRAMATSU
Linux Technology Research Center, System Productivity Research Dept.
Center for Technology Innovation - Systems Engineering
Hitachi, Ltd., Research & Development Group
E-mail: masami.hiramatsu.pt@hitachi.com