Re: [PATCH printk v3 4/6] printk: remove NMI tracking

From: "Russell King (Oracle)" <linux@armlinux.org.uk>
To: Petr Mladek <pmladek@suse.com>
Cc: John Ogness <john.ogness@linutronix.de>,
	Sergey Senozhatsky <senozhatsky@chromium.org>,
	Steven Rostedt <rostedt@goodmis.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	linux-kernel@vger.kernel.org,
	Michael Ellerman <mpe@ellerman.id.au>,
	Benjamin Herrenschmidt <benh@kernel.crashing.org>,
	Paul Mackerras <paulus@samba.org>, Ingo Molnar <mingo@redhat.com>,
	Marc Zyngier <maz@kernel.org>,
	Valentin Schneider <valentin.schneider@arm.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Pekka Enberg <penberg@kernel.org>,
	Mike Rapoport <rppt@kernel.org>,
	"Wolfram Sang (Renesas)" <wsa+renesas@sang-engineering.com>,
	Anshuman Khandual <anshuman.khandual@arm.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Frederic Weisbecker <frederic@kernel.org>,
	Kees Cook <keescook@chromium.org>,
	Masahiro Yamada <masahiroy@kernel.org>,
	Nathan Chancellor <nathan@kernel.org>,
	Sami Tolvanen <samitolvanen@google.com>,
	Alexei Starovoitov <ast@kernel.org>,
	Nick Terrell <terrelln@fb.com>,
	Chris Wilson <chris@chris-wilson.co.uk>,
	Vlastimil Babka <vbabka@suse.cz>,
	linux-arm-kernel@lists.infradead.org,
	linuxppc-dev@lists.ozlabs.org
Subject: Re: [PATCH printk v3 4/6] printk: remove NMI tracking
Date: Fri, 25 Jun 2021 14:34:25 +0100	[thread overview]
Message-ID: <20210625133425.GX22278@shell.armlinux.org.uk> (raw)
In-Reply-To: <YNXNx68CXrI9QpHD@alley>

On Fri, Jun 25, 2021 at 02:36:23PM +0200, Petr Mladek wrote:
> On Thu 2021-06-24 13:17:46, John Ogness wrote:
> > All NMI contexts are handled the same as the safe context: store the
> > message and defer printing. There is no need to have special NMI
> > context tracking for this. Using in_nmi() is enough.
> > 
> > Signed-off-by: John Ogness <john.ogness@linutronix.de>
> > Reviewed-by: Petr Mladek <pmladek@suse.com>
> > ---
> >  arch/arm/kernel/smp.c       |  2 --
> >  arch/powerpc/kexec/crash.c  |  3 ---
> >  include/linux/hardirq.h     |  2 --
> >  include/linux/printk.h      | 12 ------------
> >  init/Kconfig                |  5 -----
> >  kernel/printk/internal.h    |  6 ------
> >  kernel/printk/printk_safe.c | 37 +------------------------------------
> >  kernel/trace/trace.c        |  2 --
> >  8 files changed, 1 insertion(+), 68 deletions(-)
> > 
> > diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
> > index 74679240a9d8..0dd2d733ad62 100644
> > --- a/arch/arm/kernel/smp.c
> > +++ b/arch/arm/kernel/smp.c
> > @@ -668,9 +668,7 @@ static void do_handle_IPI(int ipinr)
> >  		break;
> >  
> >  	case IPI_CPU_BACKTRACE:
> > -		printk_nmi_enter();
> >  		nmi_cpu_backtrace(get_irq_regs());
> > -		printk_nmi_exit();
> 
> It looks to me that in_nmi() returns false here. As a result,
> nmi_cpu_backtrace() might newly call consoles immediately.
> 
> If I recall correctly, arm does not have a proper NMI.
> And this is just some special case of a "normal" IRQ.
> 
> And indeed, nmi_enter() is called only from handle_fiq_as_nmi()
> and it is just a boiler plate.
> 
> If I am right, we should replace printk_nmi_enter() with
> printk_safe_enter_irqsave(flags) or so.
> 
> Even better solution might be to call this within
> nmi_enter()/nmi_exit(). But I am not sure if this is what
> the arm people want.

As I seem to recall, the guy in ARM Ltd who was working on this seemed
to drift away and it never got finished - however, I've always carried
platform specific hacks in my tree to make this work from FIQ on the
platforms I cared about:

http://git.armlinux.org.uk/cgit/linux-arm.git/commit/?h=fiq

Not suitable for mainline like that. I'm not aware of anyone working on
it now.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!