Re: [PATCH 1/2] panic: ratelimit panic messages

[Don Zickus <dzickus@redhat.com>]

On Wed, Jan 05, 2011 at 02:21:08PM +0100, Ingo Molnar wrote:
> 
> * Don Zickus <dzickus@redhat.com> wrote:
> 
> > Sometimes when things go bad, so much spew is coming on the console it is hard
> > to figure out what happened.  This patch allows you to ratelimit the panic
> > messages with the intent that the first panic message will provide the info
> > we need to figure out what happened.
> > 
> > Adds new kernel param 'panic_ratelimit=on/<integer in seconds>'
> > 
> > Signed-off-by: Don Zickus <dzickus@redhat.com>
> > ---
> >  Documentation/kernel-parameters.txt |    6 ++++++
> >  kernel/panic.c                      |   30 ++++++++++++++++++++++++++++++
> >  2 files changed, 36 insertions(+), 0 deletions(-)
> 
> Well, but it's not just panics that can occur repeatedly (in fact panics tend to 
> trigger once), but recursive/repeated oopses.
> 
> So if we try to do some sort of 'show first crash, rate-limit afterwards' thing (i'm 
> still not 100% sure about the wiseness of it) then we should cover all common 
> occurances of system panics/crashes.
> 
> btw., while printk-delay exists and can serve a similar purpose, it's pretty 
> cumbersome for panic ratelimit: it slows down bootups extremely, and it also slows 
> down regular, harmless printks.
> 
> So having some sort of rate-limit for emitting crash screens may make sense, while 
> also preserving the oops counter or at least some other notification that we 'lost' 
> messages.

Ok.  I guess I am feeling a little slow today.  So would moving this to
the kmsg_dump() function be the right place then?

Cheers,
Don

> 
> 	Ingo