linux-kernel - Re: [PATCHv3 0/3] perf tool: Add new event group management

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]

Date:	Mon, 2 Jul 2012 22:04:21 -0300
From:	Arnaldo Carvalho de Melo <acme@...hat.com>
To:	Namhyung Kim <namhyung@...nel.org>
Cc:	Jiri Olsa <jolsa@...hat.com>, a.p.zijlstra@...llo.nl,
	mingo@...e.hu, paulus@...ba.org, cjashfor@...ux.vnet.ibm.com,
	fweisbec@...il.com, linux-kernel@...r.kernel.org,
	tglx@...utronix.de, andi@...stfloor.org, drepper@...il.com,
	eranian@...gle.com
Subject: Re: [PATCHv3 0/3] perf tool: Add new event group management

Em Tue, Jul 03, 2012 at 09:50:19AM +0900, Namhyung Kim escreveu:
> On Mon, 02 Jul 2012 23:20:34 +0900, Namhyung Kim wrote:
> > We could pass such information to perf report explicitly, But I guess it
> > 'd better if perf report did it for me automagically by detecting group
> > relations.
> 
> Oh I think we can do it by parsing the command line (again) at the perf
> report time. In that way, no need to change the file format. I'm gonna
> give it a shot later.

Its there already...

[acme@...icio ~]$ perf record -e cycles,cache-misses sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.016 MB perf.data (~713 samples) ]
[acme@...icio ~]$ perf report --stdio | head -20
# ========
# captured on: Mon Jul  2 22:02:00 2012
# hostname : felicio.ghostprotocols.net
# os release : 3.4.0-rc5+
# perf version : 3.5.rc1.104.gd3076383
# arch : x86_64
# nrcpus online : 4
# nrcpus avail : 4
# cpudesc : Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz
# cpuid : GenuineIntel,6,42,7
# total memory : 8089800 kB
# cmdline : /home/acme/bin/perf record -e cycles,cache-misses sleep 1 
# event : name = cycles, type = 0, config = 0x0, config1 = 0x0, config2
# = 0x0, excl_usr = 0, excl_kern = 0, id = { 10, 11, 12, 13 }
# event : name = cache-misses, type = 0, config = 0x3, config1 = 0x0,
# config2 = 0x0, excl_usr = 0, excl_kern = 0, id = { 14, 15, 16, 17 }
# HEADER_CPU_TOPOLOGY info available, use -I to display
# HEADER_NUMA_TOPOLOGY info available, use -I to display
# ========
#
# Samples: 19  of event 'cycles'
# Event count (approx.): 2442170
[acme@...icio ~]$ 

I.e. for perf report it is ok already, the user can see what was asked
for, exactly as specified on the command line.

What I discussed with Jiri was that it would be good to have a
programatic way of regenerating the perf_evlist instance from the
perf_event_attr and other feature bits.

I.e. the counterpart to perf_evsel being created from the
perf_event_attr in the perf.data file.

- Arnaldo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/