[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20220221234610.0d23e2e0@plasteblaster>
Date: Mon, 21 Feb 2022 23:46:10 +0100
From: "Dr. Thomas Orgis" <thomas.orgis@...-hamburg.de>
To: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
CC: <linux-kernel@...r.kernel.org>, <stable@...r.kernel.org>,
Balbir Singh <bsingharora@...il.com>,
"Eric W. Biederman" <ebiederm@...ssion.com>,
"Sudip Mukherjee" <sudipm.mukherjee@...il.com>
Subject: Re: [PATCH 5.4 32/80] taskstats: Cleanup the use of task->exit_code
Am Mon, 21 Feb 2022 09:49:12 +0100
schrieb Greg Kroah-Hartman <gregkh@...uxfoundation.org>:
> As best as I can figure the intent is to return task->exit_code after
> a task exits. The field is returned with per task fields, so the
> exit_code of the entire process is not wanted.
I wondered about the use of exit_code, too, when preparing my patch
that introduces ac_tgid and the AGROUP flag to identify the first and
last tasks of a task group/process, see
https://lkml.org/lkml/2022/2/18/887
With the information about the position of this task in the group,
users can take some meaning from the exit code (individual kills?). The
old style ensured that you got one exit code per process.
I addressing ac_exitcode fits together with my patch, while increasing
the version of taskstats helps clients that then can know that
ac_exitcode now has a different meaning. Right now this is a change
under the hood and you can just guess (or have to know from the kernel
version).
Alrighty then,
Thomas
--
Dr. Thomas Orgis
HPC @ Universität Hamburg
Powered by blists - more mailing lists