Re: [PATCH v2 0/4] lib: list_sort: Various minor improvements

[Rasmus Villemoes <linux@rasmusvillemoes.dk>]

Andrew Morton <akpm@linux-foundation.org> writes:

> On Tue, 24 Jun 2014 12:06:27 +0200 Rasmus Villemoes <linux@rasmusvillemoes.dk> wrote:
>
>> Reading the source of lib/list_sort.c, I came up with a few possible
>> improvements. I think 4/4 may be a bit controversial, but 1/4, 2/4 and
>> 3/4 should be straightforward.
>> 
>
> All looks OK to me.

Thanks.

> We may as well do the pr_foo() conversion as well.  As often happens,
> the results are quite pleasing.
>
> --- a/lib/list_sort.c~lib-list_sortc-convert-to-pr_foo
> +++ a/lib/list_sort.c
> @@ -1,3 +1,6 @@
> +
> +#define pr_fmt(fmt) "list_sort_test: " fmt
> +
>  #include <linux/kernel.h>
>  #include <linux/module.h>
>  #include <linux/list_sort.h>
> @@ -125,9 +128,8 @@ void list_sort(void *priv, struct list_h
>  		}
>  		if (lev > max_lev) {
>  			if (unlikely(lev >= ARRAY_SIZE(part)-1)) {
> -				printk_once(KERN_DEBUG "list passed to"
> -					" list_sort() too long for"
> -					" efficiency\n");
> +				pr_debug_once("list passed to list_sort() too "
> +					      "long for efficiency\n");

Minor comment: Won't this end up saying "list_sort_test: list passed to
...", despite the list coming from a 'real' user? Maybe change the first
#define to '"list_sort: " fmt', the above message to "passed list too
long for efficiency", and redefine pr_fmt right after #ifdef
CONFIG_TEST_LIST_SORT.

Rasmus