That is right. I should be able to just printk the base memory of my
buffer and use lcrash to look at the buffers. The only hassle being
that my buffers are in binary. So, I would like to do some processing
on then on a crash. But it may also not be a good idea considering
that the machine just crashed. But then, all I would want to do is
consult some in-memory lookup table to convert the binary to ASCII and
just printk the processed string. I could be doing that before lkcd
starts backing up the kernel log buffer.
Like any problem, there are too many possible solutions. Just ought to
find out the least painful one.
On 6/2/05, Bob Montgomery <bob.montgomery@...> wrote:
> On Thu, 2005-06-02 at 09:55 -0700, Hariharan Subramanian wrote:
> > Hi Troy,
> > Thanks for quick response. I have a logging system that doesent use
> > printk. So stuff get logged into ring buffers that I maintain. I want
> > to dump these buffers on a crash just like lkcd currently dumps the
> > kernel log buffer.
> If your logs are allocated as kernel memory, they should already be
> dumped out by lkcd on a crash, and should be in the dump file. Don't
> you just need to find them and print them from lcrash or crash after
> you've recovered the dump?
> Bob Montgomery, HP