linux-kernel - Re: [RESEND PATCH 1/1] perf arm-spe: report all SPE records as "all" events

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] [day] [month] [year] [list]

Date:   Thu, 25 Nov 2021 20:49:05 +0800
From:   Leo Yan <leo.yan@...aro.org>
To:     German Gomez <german.gomez@....com>
Cc:     linux-kernel@...r.kernel.org, linux-perf-users@...r.kernel.org,
        acme@...nel.org, James Clark <james.clark@....com>,
        Mark Rutland <mark.rutland@....com>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Jiri Olsa <jolsa@...hat.com>,
        Namhyung Kim <namhyung@...nel.org>,
        John Garry <john.garry@...wei.com>,
        Will Deacon <will@...nel.org>,
        Mathieu Poirier <mathieu.poirier@...aro.org>,
        linux-arm-kernel@...ts.infradead.org
Subject: Re: [RESEND PATCH 1/1] perf arm-spe: report all SPE records as "all"
 events

On Thu, Nov 25, 2021 at 11:54:16AM +0000, German Gomez wrote:
> Hi Leo,
> 
> On 25/11/2021 07:53, Leo Yan wrote:
> > On Wed, Nov 17, 2021 at 02:28:32PM +0000, German Gomez wrote:
> >> [...]
> > After checked the event types, I think "other" samples would include
> > below raw event types:
> >
> >   EV_EXCEPTION_GEN
> >   EV_RETIRED
> >   EV_NOT_TAKEN
> >   EV_ALIGNMENT
> >   EV_PARTIAL_PREDICATE
> >   EV_EMPTY_PREDICATE
> >
> > I am just wander if we can use sample.transaction to store these event
> > types, otherwise, we cannot distinguish the event type for the samples.
> 
> I'm not familiar with the meaning of transactions in this context, but
> I agree giving visibility of these remaining events is a good idea. I'm
> just unsure where to place them from the available --itrace options.

Please take a look in my another reply, I think it's good to consider
to use event type + sample fields (flags/data_src/transactions) for
better expressing samples.

> Regarding the "all" events, we thought having quick access to a  global
> histogram of all the spe events is useful, and from all the --itrace
> options it seemed to fit best under --itrace=o.

Okay, I agree this is a distinct requirement and we can use this way
to generate for all SPE records.  But the question is if a user reviews
"all" events, how user can easily understand the samples if only bases
on fields sample::addr, e.g. we will have no idea if a sample is a
memory operation or a branch operation, and I think the samples for
memory operations will always set sample.addr as 0.

Thanks,
Leo