Re: [PATCH v1] printk: Declare log_wait as external variable

[Petr Mladek]

On Wed 2020-02-12 10:31:33, Sergey Senozhatsky wrote:
> On (20/02/11 13:43), Petr Mladek wrote:
> >
> > Even better solution might be to move fs/proc/kmsg.c to
> > kernel/printk/proc_kmsg.c and declare printk_log_wait only
> > in kernel/printk/internal.h. I think that this is what
> > Sergey suggested.
> 
> Yes, right.
> 
> > Another great thing would be to extract devkmsg stuff from
> > kernel/printk/printk.c and put it into kernel/printk/dev_kmsg.c.
> 
> Yeah, can do, I would still prefer proc_kmsg to "move in".
> Either both can live in printk.c (won't make it much worse),
> or in kernel/printk/dev_kmsg.c and kernel/printk/proc_kmsg.c

I would prefer to split it:

    + printk.c is already too big and would deserve splitting.

    + The two different kmgs interfaces are confusing on its
      own. IMHO, it will be even more confusing when they
      live in one huge source file.


> I can take a look at dev_ksmg.c/proc_kmsg.c option, unless
> someone else wants to spend their time on this.

It would be lovely from my POV. I am only concerned about
the lockless printk() stuff. I would prefer to avoid creating
too many conflicts in the same merge window. Well, I am
not sure how many conflicts there would be. Adding John
into CC.

Best Regards,
Petr