linux-kernel - Re: [PATCH v1 0/2] perf: Drop leaked kernel samples

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <9b3259e8-1d07-adb1-a57c-8b276c133aa6@linux.intel.com>
Date:   Sat, 16 Jun 2018 08:50:11 +0800
From:   "Jin, Yao" <yao.jin@...ux.intel.com>
To:     robert@...llahan.org, Kyle Huey <me@...ehuey.com>
Cc:     acme@...nel.org, jolsa@...nel.org,
        "Peter Zijlstra (Intel)" <peterz@...radead.org>,
        Ingo Molnar <mingo@...hat.com>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        open list <Linux-kernel@...r.kernel.org>,
        Vince Weaver <vincent.weaver@...ne.edu>,
        Will Deacon <will.deacon@....com>,
        Stephane Eranian <eranian@...gle.com>,
        Namhyung Kim <namhyung@...nel.org>, ak@...ux.intel.com,
        kan.liang@...el.com, yao.jin@...el.com
Subject: Re: [PATCH v1 0/2] perf: Drop leaked kernel samples

Hi All,

This patch raised many questions, I was prepared. :)

I'd like to try another proposal that it adds a special flag in the 
returned perf_sample_data to indicate the perf binary that this sample 
is a leaked sample.

For example, create a new PERF_SAMPLE_RETURN_LEAKAGE in 
perf_event_sample_format.

In perf_prepare_sample(),

if (event->attr.exclude_kernel && !user_mode(regs))
	data->type |= PERF_SAMPLE_RETURN_LEAKAGE;

Now all the samples are kept and the leaked kernel samples are tagged 
with PERF_SAMPLE_RETURN_LEAKAGE.

In perf binary, it filters out the samples with 
PERF_SAMPLE_RETURN_LEAKAGE. It needs perf binary modification but rr 
doesn't need to be changed.

I don't 0-stuffing some fields because:

1. Keeping the skid info should allow us to look at that if we have 
interesting later.

2. Not sure if 0-stuffing some fields has potential conflicts with some 
applications.

Is this proposal reasonable?

Thanks
Jin Yao

On 6/16/2018 1:34 AM, Robert O'Callahan wrote:
> On Fri, Jun 15, 2018 at 10:16 AM, Kyle Huey <me@...ehuey.com> wrote:
>>
>> If you want a sysctl for your own reasons that's fine. But we don't
>> want a sysctl. We want to work without any further configuration.
> 
> Also toggling a sysctl would require root privileges, but rr does not
> currently need to run as root. Thus rr users would have to either
> permanently change their system configuration (and every extra
> configuration step is a pain), or run rr as root so rr can toggle the
> sysctl itself. Running rr as root is highly undesirable.
> 
> Rob
>