Re: [PATCH 2/2] printk: always report lost messages on serial console

[Sergey Senozhatsky]

On (01/03/17 15:55), Petr Mladek wrote:
[..]
> This causes the opposite problem. We might print a message that was supposed
> to be suppressed.

so what? yes, we print a message that otherwise would have been suppressed.
not a big deal. at all. we are under high printk load and the best thing
we can do is to report "we are losing the messages" straight ahead. the
next 'visible' message may be seconds/minutes/forever away. think of a
printk() flood of messages with suppressed loglevel coming from CPUA-CPUX,
big enough to drain all 'visible' loglevel messages from CPUZ. we are
back to problem "a".

thus I want a simple bool flag and a simple rule: we see something - we say it.


[..]
> The best solution seems to be to print the warning with
> the next visible message.

not sure.

	-ss