Message-ID: <20200429142106.GG28637@dhcp22.suse.cz>
Date:   Wed, 29 Apr 2020 16:21:06 +0200
From:   Michal Hocko <mhocko@...nel.org>
To:     Tetsuo Handa <penguin-kernel@...ove.sakura.ne.jp>
Cc:     Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
        Petr Mladek <pmladek@...e.com>,
        Steven Rostedt <rostedt@...dmis.org>,
        linux-kernel@...r.kernel.org, Dmitry Safonov <dima@...sta.com>,
        Yafang Shao <laoar.shao@...il.com>
Subject: Re: [PATCH] printk: Add loglevel for "do not print to consoles".

On Wed 29-04-20 01:23:15, Tetsuo Handa wrote:
> On 2020/04/29 0:45, Michal Hocko wrote:
> > On Tue 28-04-20 22:11:19, Tetsuo Handa wrote:
> >> Existing KERN_$LEVEL allows a user to determine whether he/she wants that message
> >> to be printed on consoles (even if it spams his/her operation doing on consoles), and
> >> at the same time constrains that user whether that message is saved to log files.
> >> KERN_NO_CONSOLES allows a user to control whether he/she wants that message to be
> >> saved to log files (without spamming his/her operation doing on consoles).
> > 
> > I understand that. But how do I know whether the user considers the
> > particular information important enough to be dumped on the console.
> > This sounds like a policy in the kernel to me.
> 
> I'm still unable to understand your question.

I am trying to say that KERN_NO_CONSOLES resembles more a policy than a
priority. Because I as a developer have no idea whether the message is
good enough for console or not.

> >                                                I simply cannot forsee
> > any console configuration to tell whether my information is going to
> > swamp the console to no use or not.
> 
> Neither can I.
> 
> >                                     Compare that to KERN_$LEVEL instead.
> > I know that an information is of low/high importance. It is the user
> > policy to decide and base some filtering on top of that priority.
> 
> Whether to use KERN_NO_CONSOLES is not per-importance basis but per-content basis.
> 
> Since both pr_info("[%7d] %5d %5d %8lu %8lu %8ld %8lu         %5hd %s\n", ...) from dump_tasks() and
> pr_info("oom-kill:constraint=%s,nodemask=%*pbl", ...) from dump_oom_summary() use KERN_INFO importance,
> existing KERN_$LEVEL-based approach cannot handle these messages differently. Since changing the former to
> e.g. KERN_DEBUG will cause userspace to discard the messages, we effectively can't change KERN_$LEVEL.

I believe we are free to change kernel log levels as we find a fit. I
was not aware that KERN_DEBUG messages are automatically filtered out.
Even if this is the case then this doesn't really disallow admins to
allow KERN_DEBUG into log files. Dump of the oom eligible tasks is
arguably a debugging output anyway. So I disagree with your statement.

> If the kernel allows the former to use KERN_NO_CONSOLES in addition to KERN_INFO, the administrator can
> select from two choices: printing "both the former and the latter" or "only the latter" to consoles.

I am not really familiar with all the possibilities admins have when
setting filtering for different consoles but KERN_NO_CONSOLES sounds
rather alien to the existing priority based approach. You can fine tune
priorities and that is all right because they should be reflecting
importance. But global no-consoles doesn't really fit in here because
each console might require a different policy but the marking is
unconditional and largely unaware of existing consoles.
-- 
Michal Hocko
SUSE Labs

Powered by blists - more mailing lists