Re: [PATCH] x86, tsc: downgrade message for fast calibration failure

[Paul Bolle <pebolle@tiscali.nl>]

On Wed, 2012-09-26 at 12:09 +0200, Paul Bolle wrote:
> When fast calibration of the TSC fails an error is printed:
>     tsc: Fast TSC calibration failed
> 
> This message was printed at default level (ie, as a warning) before
> commit c767a54ba0657e52e6edaa97cbe0b0a8bf1c1655 ("x86/debug: Add
> KERN_<LEVEL> to bare printks, convert printks to pr_<level>"). But it is
> more appropriate to print it at informational level. If fast calibration
> fails that only means an alternative calibration strategy will be tried,
> and if that strategy also fails only a warning is printed. Besides,
> there's nothing one can really do after noticing a fast calibration
> failure.
> 
> Signed-off-by: Paul Bolle <pebolle@tiscali.nl>

What's the status of this patch? Did anyone (besides Joe) have a look at
it? (I'm fine with the wordsmithing Joe suggested, by the way.)

> 0) This is a rather verbose commit explanation. For trivialities like
> these I try to write a very short explanation. But here the verbosity is
> meant to obfuscate my complete ignorance of the (calibration of the)
> TSC.
> 
> 1) By the way, it's debatable whether this message should be printed at
> informational or at notice level. But the difference between KERN_NOTICE
> and KERN_INFO is too subtle for me. I even find the difference between
> KERN_ALERT and KERN_CRIT, and between KERN_ERR and KERN_WARNING a bit
> subtle. Are there guidelines for choosing between those levels?
> 
>  arch/x86/kernel/tsc.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
> index cfa5d4f..4f258ae 100644
> --- a/arch/x86/kernel/tsc.c
> +++ b/arch/x86/kernel/tsc.c
> @@ -374,7 +374,7 @@ static unsigned long quick_pit_calibrate(void)
>  			goto success;
>  		}
>  	}
> -	pr_err("Fast TSC calibration failed\n");
> +	pr_info("Fast TSC calibration failed\n");
>  	return 0;
>  
>  success:


Paul Bolle