[PATCH 1/1] powerpc/crash: Use NMI context for printk after crashing other CPUs

[PATCH 1/1] powerpc/crash: Use NMI context for printk after crashing other CPUs

[Leonardo Bras]

Currently, if printk lock (logbuf_lock) is held by other thread during
crash, there is a chance of deadlocking the crash on next printk, and
blocking a possibly desired kdump.

After sending IPI to all other CPUs, make printk enter in NMI context,
as it will use per-cpu buffers to store the message, and avoid locking
logbuf_lock.

Signed-off-by: Leonardo Bras <leonardo@linux.ibm.com>
---
 arch/powerpc/kexec/crash.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/powerpc/kexec/crash.c b/arch/powerpc/kexec/crash.c
index d488311efab1..9b73e3991bf4 100644
--- a/arch/powerpc/kexec/crash.c
+++ b/arch/powerpc/kexec/crash.c
@@ -115,6 +115,7 @@ static void crash_kexec_prepare_cpus(int cpu)
 
 	crash_send_ipi(crash_ipi_callback);
 	smp_wmb();
+	printk_nmi_enter();
 
 again:
 	/*
-- 
2.25.1

Re: [PATCH 1/1] powerpc/crash: Use NMI context for printk after crashing other CPUs

[Michael Ellerman]

Leonardo Bras <leonardo@linux.ibm.com> writes:
> Currently, if printk lock (logbuf_lock) is held by other thread during
> crash, there is a chance of deadlocking the crash on next printk, and
> blocking a possibly desired kdump.
>
> After sending IPI to all other CPUs, make printk enter in NMI context,
> as it will use per-cpu buffers to store the message, and avoid locking
> logbuf_lock.
>
> Signed-off-by: Leonardo Bras <leonardo@linux.ibm.com>
> ---
>  arch/powerpc/kexec/crash.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/arch/powerpc/kexec/crash.c b/arch/powerpc/kexec/crash.c
> index d488311efab1..9b73e3991bf4 100644
> --- a/arch/powerpc/kexec/crash.c
> +++ b/arch/powerpc/kexec/crash.c
> @@ -115,6 +115,7 @@ static void crash_kexec_prepare_cpus(int cpu)

Added context:

	printk(KERN_EMERG "Sending IPI to other CPUs\n");

	if (crash_wake_offline)
		ncpus = num_present_cpus() - 1;

>  
>  	crash_send_ipi(crash_ipi_callback);
>  	smp_wmb();
> +	printk_nmi_enter();
  
Why did you decide to put it there, rather than at the start of
default_machine_crash_shutdown() like I did?

The printk() above could have already deadlocked if another CPU is stuck
with the logbuf lock held.

cheers

Re: [PATCH 1/1] powerpc/crash: Use NMI context for printk after crashing other CPUs

[Leonardo Bras]

[-- Attachment #1: Type: text/plain, Size: 827 bytes --]

On Wed, 2020-04-08 at 22:13 +1000, Michael Ellerman wrote:
[...]
> Added context:
> 
> 	printk(KERN_EMERG "Sending IPI to other CPUs\n");
> 
> 	if (crash_wake_offline)
> 		ncpus = num_present_cpus() - 1;
> 
> >  
> >  	crash_send_ipi(crash_ipi_callback);
> >  	smp_wmb();
> > +	printk_nmi_enter();
>   
> Why did you decide to put it there, rather than at the start of
> default_machine_crash_shutdown() like I did?
> 
> The printk() above could have already deadlocked if another CPU is stuck
> with the logbuf lock held.

Oh, I thought the CPUs would start crashing after crash_send_ipi(), so
only printk() after that would possibly deadlock.

I was not able to see how the printk() above would deadlock, but I see
no problem adding that at the start of the function.

Best regards,
Leonardo Bras

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]