Re: [PATCH] dump_stack: convert generic dump_stack into a weak symbol

[Steven Rostedt <rostedt@goodmis.org>]

On Fri, 9 Mar 2018 10:50:55 +0100
Petr Mladek <pmladek@suse.com> wrote:

> On Wed 2018-03-07 20:48:56, Sergey Senozhatsky wrote:
> > On (03/07/18 11:57), Arnd Bergmann wrote:  
> > > Since lib/dump_stack.c is mandatory then, I would suggest making it
> > > obj-y and moving it out of lib/ into kernel/printk/.  
> > 
> > Totally agree on obj-y. And tend to agree on moving lib/dump_stack to
> > kernel/printk. lib/dump_stack depends on CONFIG_PRINTK and is partially
> > in printk.h. So it _mostly_ seems like the right place after all.  
> 
> I agree that we should define dump_stack.o as obj-y. Also it makes
> sense to move it to kernel/printk/. As Sergey said: it depends on
> CONFIG_PRINTK and is in printk.h.
> 
> Steven, do you agree or would you suggest another location?
> It was you who suggested to move the several functions from printk.c
> to dump_stack.c. IMHO, the motivation was to fix responsibilities
> (maintainer-ship) of this stuff. This will not really happen if
> we move the file to printk directory.
>

I still would like to keep it in lib. If blackfin is going to be nuked
in 4.17, and nds32 is the only one that is causing a issue. I would
highly recommend fixing nds32 and have Stephen stop building blackfin.

-- Steve