Re: System crash with perf_fuzzer (kernel: 5.0.0-rc3)

[Ravi Bangoria <ravi.bangoria@linux.ibm.com>]

Hi Jiri,

On 2/1/19 1:13 PM, Jiri Olsa wrote:
> On Thu, Jan 31, 2019 at 09:27:11AM +0100, Jiri Olsa wrote:
>> On Wed, Jan 30, 2019 at 07:36:48PM +0100, Jiri Olsa wrote:
>>
>> SNIP
>>
>>> diff --git a/kernel/events/core.c b/kernel/events/core.c
>>> index 280a72b3a553..22ec63a0782e 100644
>>> --- a/kernel/events/core.c
>>> +++ b/kernel/events/core.c
>>> @@ -4969,6 +4969,26 @@ static void __perf_event_period(struct perf_event *event,
>>>  	}
>>>  }
>>>  
>>> +static int check_period(struct perf_event *event, u64 value)
>>> +{
>>> +	u64 sample_period_attr = event->attr.sample_period;
>>> +	u64 sample_period_hw   = event->hw.sample_period;
>>> +	int ret;
>>> +
>>> +	if (event->attr.freq) {
>>> +		event->attr.sample_freq = value;
>>> +	} else {
>>> +		event->attr.sample_period = value;
>>> +		event->hw.sample_period = value;
>>> +	}
>>
>> hm, I think we need to check the period without changing the event,
>> because we don't disable pmu, so it might get picked up by bts code
>>
>> will check
> 
> with attached patch I did not trigger the fuzzer crash
> for over a day now, could you guys try?

I ran fuzzer for couple of hours but I didn't see any crash with
your previous patch.

I'll try this newer one as well.

Thanks.