[PATCH V3] sched/rt, powerpc: Prepare for PREEMPT_RT

[PATCH V3] sched/rt, powerpc: Prepare for PREEMPT_RT

[Wang Qing]

PREEMPT_RT is a separate preemption model, CONFIG_PREEMPT will
 be disabled when CONFIG_PREEMPT_RT is enabled,  so we need
to add CONFIG_PREEMPT_RT output to __die().

Signed-off-by: Wang Qing <wangqing@vivo.com>

Changes in v3:
 - Fix typo issue.

Changes in v2:
 - Modify as Christophe suggested.
---
 arch/powerpc/kernel/traps.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c
index 5006dcb..dec7b81
--- a/arch/powerpc/kernel/traps.c
+++ b/arch/powerpc/kernel/traps.c
@@ -262,10 +262,11 @@ static int __die(const char *str, struct pt_regs *regs, long err)
 {
 	printk("Oops: %s, sig: %ld [#%d]\n", str, err, ++die_counter);
 
-	printk("%s PAGE_SIZE=%luK%s%s%s%s%s%s %s\n",
+	printk("%s PAGE_SIZE=%luK%s%s%s%s%s%s%s %s\n",
 	       IS_ENABLED(CONFIG_CPU_LITTLE_ENDIAN) ? "LE" : "BE",
 	       PAGE_SIZE / 1024, get_mmu_str(),
 	       IS_ENABLED(CONFIG_PREEMPT) ? " PREEMPT" : "",
+	       IS_ENABLED(CONFIG_PREEMPT_RT) ? " PREEMPT_RT" : "",
 	       IS_ENABLED(CONFIG_SMP) ? " SMP" : "",
 	       IS_ENABLED(CONFIG_SMP) ? (" NR_CPUS=" __stringify(NR_CPUS)) : "",
 	       debug_pagealloc_enabled() ? " DEBUG_PAGEALLOC" : "",
-- 
2.7.4

Re: [PATCH V3] sched/rt, powerpc: Prepare for PREEMPT_RT

[Christophe Leroy]

Le 10/11/2020 à 09:53, Wang Qing a écrit :
> PREEMPT_RT is a separate preemption model, CONFIG_PREEMPT will
>   be disabled when CONFIG_PREEMPT_RT is enabled,  so we need
> to add CONFIG_PREEMPT_RT output to __die().
> 
> Signed-off-by: Wang Qing <wangqing@vivo.com>

Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>

> 
> Changes in v3:
>   - Fix typo issue.
> 
> Changes in v2:
>   - Modify as Christophe suggested.
> ---
>   arch/powerpc/kernel/traps.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c
> index 5006dcb..dec7b81
> --- a/arch/powerpc/kernel/traps.c
> +++ b/arch/powerpc/kernel/traps.c
> @@ -262,10 +262,11 @@ static int __die(const char *str, struct pt_regs *regs, long err)
>   {
>   	printk("Oops: %s, sig: %ld [#%d]\n", str, err, ++die_counter);
>   
> -	printk("%s PAGE_SIZE=%luK%s%s%s%s%s%s %s\n",
> +	printk("%s PAGE_SIZE=%luK%s%s%s%s%s%s%s %s\n",
>   	       IS_ENABLED(CONFIG_CPU_LITTLE_ENDIAN) ? "LE" : "BE",
>   	       PAGE_SIZE / 1024, get_mmu_str(),
>   	       IS_ENABLED(CONFIG_PREEMPT) ? " PREEMPT" : "",
> +	       IS_ENABLED(CONFIG_PREEMPT_RT) ? " PREEMPT_RT" : "",
>   	       IS_ENABLED(CONFIG_SMP) ? " SMP" : "",
>   	       IS_ENABLED(CONFIG_SMP) ? (" NR_CPUS=" __stringify(NR_CPUS)) : "",
>   	       debug_pagealloc_enabled() ? " DEBUG_PAGEALLOC" : "",
> 

Re: [PATCH V3] sched/rt, powerpc: Prepare for PREEMPT_RT

[Michael Ellerman]

Wang Qing <wangqing@vivo.com> writes:
> PREEMPT_RT is a separate preemption model, CONFIG_PREEMPT will
>  be disabled when CONFIG_PREEMPT_RT is enabled,  so we need
> to add CONFIG_PREEMPT_RT output to __die().
>
> Signed-off-by: Wang Qing <wangqing@vivo.com>

Something fairly similar was posted previously.

That time I said:

  I don't think there's any point adding the "_RT" to the __die() output
  until/if we ever start supporting PREEMPT_RT.

  https://lore.kernel.org/linuxppc-dev/87d0ext4q3.fsf@mpe.ellerman.id.au/
   

And I think I still feel the same way. It's not clear powerpc will ever
support PREEMPT_RT, so this would just be confusing to people. And
potentially someone will then send a patch to remove it as dead code.

cheers



> diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c
> index 5006dcb..dec7b81
> --- a/arch/powerpc/kernel/traps.c
> +++ b/arch/powerpc/kernel/traps.c
> @@ -262,10 +262,11 @@ static int __die(const char *str, struct pt_regs *regs, long err)
>  {
>  	printk("Oops: %s, sig: %ld [#%d]\n", str, err, ++die_counter);
>  
> -	printk("%s PAGE_SIZE=%luK%s%s%s%s%s%s %s\n",
> +	printk("%s PAGE_SIZE=%luK%s%s%s%s%s%s%s %s\n",
>  	       IS_ENABLED(CONFIG_CPU_LITTLE_ENDIAN) ? "LE" : "BE",
>  	       PAGE_SIZE / 1024, get_mmu_str(),
>  	       IS_ENABLED(CONFIG_PREEMPT) ? " PREEMPT" : "",
> +	       IS_ENABLED(CONFIG_PREEMPT_RT) ? " PREEMPT_RT" : "",
>  	       IS_ENABLED(CONFIG_SMP) ? " SMP" : "",
>  	       IS_ENABLED(CONFIG_SMP) ? (" NR_CPUS=" __stringify(NR_CPUS)) : "",
>  	       debug_pagealloc_enabled() ? " DEBUG_PAGEALLOC" : "",
> -- 
> 2.7.4