Em 16-10-2015 10:20, Pierre Labastie escreveu:
> On 16/10/2015 12:41, Fernando de Oliveira wrote:
>> Em 16-10-2015 02:37, Pierre Labastie escreveu:
>>> Hi,
>>>
>>> Just want to share something everybody may be knowing already...
>>>
>>> In the past years, we have seen more and more kernel messages cluttering
>>> the screen (when mounting an ext2 system, when setting up network, etc).
>>> It culminates with the LVM tests, where it is almost impossible to work
>>> on another console during the tests (since the kernel messages go the
>>> the console you are working on, not the console where you started the
>>> test).
>>>
>>> The command "dmesg -D" can be used to avoid those annoyances. Normally,
>>> critical messages should still show up through the sysklogd system.
>>>
>>> Pierre
>>>
>> It is good discussing this. First time I wrote to help a client, IIRC,
>> DJ warned me, because I instructed a user to change a bootscript. :-)
>>
>> Found in man 2 syslog.
>>
>> Kernel constant Level value Meaning
>> KERN_EMERG 0 System is unusable
>> KERN_ALERT 1 Action must be taken immediately
>> KERN_CRIT 2 Critical conditions
>> KERN_ERR 3 Error conditions
>> KERN_WARNING 4 Warning conditions
>> KERN_NOTICE 5 Normal but significant condition
>> KERN_INFO 6 Informational
>> KERN_DEBUG 7 Debug-level messages
>>
>> I use /etc/sysconfig/rc.site:
>>
>> $ grep LOGLEVEL /etc/sysconfig/rc.site
>> #LOGLEVEL=5
>> LOGLEVEL=3
>>
>>
>> Perhaps it would do with level 4.
> Thanks Fernando,
>
> You gave me pointers to where to look for. Actually, it seems to
> me that there is a small issue in rc.site: it has:
>
> "#LOGLEVEL=5"
>
> and I would imagine that it means the default log level is 5. But the
> rc script has:
>
> dmesg -n "${LOGLEVEL:-7}"
>
> which means the default level is 7. I'd suggest put "#LOGLEVEL=7" in
> rc.site. There is a thread in 2010 (most relevant message is
> http://lists.linuxfromscratch.org/pipermail/lfs-dev/2010-February/063518.html)
>
> where Bruce explains why the log level should be 7.
>
> Now, for BLFS, we always assume that the user uses the defaults,
> so that log level is 7 (all messages printed except DEBUG). In this
> case, running the LVM test just renders any virtual console unusable
> (I have not tried on an X display). What do you think if change the test
> command to:
> <command>dmesg -D;make -k check;dmesg -E</command>
>
> Pierre
>
Both are OK, for me.
But I had seen in man dmesg:
{{{
-n, --console-level level
Set the level at which printing of messages is done to the con‐
sole. The level is a level number or abbreviation of the level
name. For all supported levels see the --help output.
For example, -n 1 or -n alert prevents all messages, except
emergency (panic) messages, from appearing on the console. All
levels of messages are still written to /proc/kmsg, so sys‐
logd(8) can still be used to control exactly where kernel mes‐
sages appear. When the -n option is used, dmesg will not print
or clear the kernel ring buffer.
}}}
Therefore, it is more useful to having default in the rc script and
rc.site as 4, explaining somwhere (if it isn't already done, that all
messages are still logged in /dev/kmsg, and can be read with, e.g.:
dd if=/dev/kmsg iflag=nonblock
I believe they are also in /var/log/sys.log and /var/log/kern.log.
But while this is not discussed, for the status quo, at least the LVM
tests could be modified as you suggested.
Almost OT: I am impressed how your tests results improved. What have you
done?
--
[]s,
Fernando, aka Sísifo
--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
