Re: [PATCH 7/8] irqdomain: trivial pr_fmt conversion.

[Paul Mundt <lethal@linux-sh.org>]

On Fri, May 18, 2012 at 11:34:55PM -0700, Joe Perches wrote:
> On Sat, 2012-05-19 at 15:11 +0900, Paul Mundt wrote:
> > Convert to pr_fmt before things start to get out of hand and some
> > janitors start getting overly excited.
> 
> ;)  OK then, here's some trivia:
> 
> > diff --git a/kernel/irq/irqdomain.c b/kernel/irq/irqdomain.c
> []
> > @@ -1,3 +1,5 @@
> > +#define pr_fmt(fmt)  "irq: " fmt
> 
> "irqdomain: " (KBUILD_MODNAME) may be more sensible
> 
Perhaps. The point of the patch was to consolidate rather than change
behaviour. With a consolidation in place it's possible to change it to
whatever people are happy with.

> []
> > @@ -410,23 +412,23 @@ unsigned int irq_create_mapping(struct irq_domain *domain,
> >  	unsigned int hint;
> >  	int virq;
> >  
> > -	pr_debug("irq: irq_create_mapping(0x%p, 0x%lx)\n", domain, hwirq);
> > +	pr_debug("irq_create_mapping(0x%p, 0x%lx)\n", domain, hwirq);
> 
> "%s: ...", __func__, etc...
> 
Unrelated to pr_fmt conversion. In any event, these debug calls are
unlikely to hang around for very long.

> >  	if (domain = NULL) {
> > -		printk(KERN_WARNING "irq_create_mapping called for"
> > -		       " NULL domain, hwirq=%lx\n", hwirq);
> > +		pr_warning("irq_create_mapping called for"
> > +			   " NULL domain, hwirq=%lx\n", hwirq);
> 
> 		pr_warn("%s: called with NULL domain, hwirq=%lx\n", hwirq);
> 
No. pr_warning() matches the log level, pr_warn() does not. A quick grep
also shoes more pr_warning() users than pr_warn(), so it sounds like you
have your work cut out for you. Good luck.