DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 46E795D687
Date: Sat, 1 Jul 2017 11:47:03 -0300
From: Arnaldo Carvalho de Melo <acme@redhat.com>
To: Arun Kalyanasundaram <arunkaly@google.com>
Cc: tzanussi@gmail.com, linux-kernel@vger.kernel.org,
        David Carrillo-Cisneros <davidcc@google.com>, acme@kernel.org
Subject: Re: perf script: Question: Python trace processing script contains
 the tid of the process in the common_pid attribute
Message-ID: <20170701144703.GA5504@redhat.com>
References: <CAGjB_Bp0haKX+YPC5117yK440mTK1ZW3WBubt616hbKaQNxd6w@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAGjB_Bp0haKX+YPC5117yK440mTK1ZW3WBubt616hbKaQNxd6w@mail.gmail.com>
User-Agent: Mutt/1.5.20 (2009-12-10)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1137
Lines: 28

Em Fri, Jun 30, 2017 at 03:40:57PM -0700, Arun Kalyanasundaram escreveu:
> The handlers in the python script generated from "perf script" have an
> attribute: common_pid. This attribute contains the tid of the process
> instead of its pid. I would like to know if this is the expected behavior.
> There are no other attributes in the Python handler that provide the pid
> and knowing the process id is useful to be able to group all samples that
> belong to the same process that generated different threads.

Humm, you have:

def process_event(param_dict):
        event_attr = param_dict["attr"]
        sample     = param_dict["sample"]
        raw_buf    = param_dict["raw_buf"]
        comm       = param_dict["comm"]
        name       = param_dict["ev_name"]

And then, on sample you have (from a recent python script for processing
Intel PT samples):

def print_common_start(comm, sample, name):
       ts = sample["time"]
       cpu = sample["cpu"]
       pid = sample["pid"]
       tid = sample["tid"]
       print "%16s %5u/%-5u [%03u] %9u.%09u %7s:" % (comm, pid, tid, cpu, ts / 1000000000, ts %1000000000, name),

- Arnaldo