RE: printk: Add process name information to printk() output.

[김창기]

> On Fri, Sep 04, 2020 at 06:31:13PM +0900,   â   wrote:
> > > On Fri, Sep 04, 2020 at 05:24:38PM +0900, Changki Kim wrote:
> > > > Printk() meesages are the most basic and useful debug method.
> > > > However, additional information needs in multi-processor.
> > > > If we add messages with processor id and process name, we can find
> > > > a problem only with messages when the problem occurs with H/W IP or
> CPU.
> > > > This is very useful in narrowing down the scope of the problems.
> > > >
> > > > Therefore, instead of trying to help buffering, let's try to help
> > > > reconstructing messages by saving caller information as of calling
> > > > log_store() and adding it as "[$processor_id: $process_name:
> > > $thread_id]"
> > > > upon printing to consoles.
> > > >
> > > > Some examples for console output:
> > > >
> > > >   [    0.059580] [0:      swapper/0:    1] CPU: All CPU(s) started at
> > EL1i
> > > >   [    2.153157] [5:           init:    1] migov: complete to probe
> > migov
> > > >   [    7.441317] [4:           init:    1] NET: Registered protocol
> > family 39
> > > >   [   22.793389] [5:  kworker/u16:1:  184] SMP: stopping secondary
> CPUs
> > > >
> > > > Signed-off-by: Changki Kim <changki.kim@xxxxxxxxxxx>
> > > > ---
> > > >  kernel/printk/printk.c            | 20 ++++++++++++++++++--
> > > >  kernel/printk/printk_ringbuffer.h |  7 +++++++
> > > >  lib/Kconfig.debug                 | 16 ++++++++++++++++
> > > >  3 files changed, 41 insertions(+), 2 deletions(-)
> > >
> > > Isn't the CPU number already present if you want it today?  Why add
> > > it again?
> > >
> > > thanks,
> > >
> > > greg k-h
> >
> > CPU number isn't present now. If I enabled PRINTK_CALLER, then it is
> > present cpu id or thread id.
> 
> Great, that should give you what you need here, right?
> 
> > The config isn't always present cpu id.
> 
> What do you mean?

When PRINTK_CALLER is enabled, cpu id or thread id are present. not always cpu id.
> 
> > We need cpu id each each message.
> 
> Why each message?  What does that provide you?

In mobile project, Sometimes we can meet the problem in the physically cpu.
It is important to know cpu id for H/W Issue.
> 
> And if PRINTK_CALLER does not work properly, please fix it.  I don't see
> how it would not work any different from your patch, you shouldn't
> duplicate things.

PRINTK_CALLER worked well.
> 
> thanks,
> 
> greg k-h