Re: [PATCH] timer: Fix jiffies wrap behavior of round_jiffies*()

[Andrew Morton <akpm@linux-foundation.org>]

On Tue, 21 May 2013 20:43:50 +0200 Bart Van Assche <bart.vanassche@gmail.com> wrote:

> Make sure that the round_jiffies*() functions return a time that is
> in the future when the jiffies counter is about to wrap.

Actually "when the jiffies counter has recently wrapped".

I assume this was found by inspection?

> @@ -149,9 +149,7 @@ static unsigned long round_jiffies_common(unsigned long j, int cpu,
>  	/* now that we have rounded, subtract the extra skew again */
>  	j -= cpu * 3;
> -	if (j <= jiffies) /* rounding ate our timeout entirely; */

This used to be a very common bug - we've weeded out many instances but
I'm sure more still remain.  We could perhaps have a checkpatch rule
which looks for comparisons against jiffes (and any other
time-measuring variables we can detect) and tells people to use
time_after() and friends, or time_is_*_jiffies().

> -		return original;
> -	return j;
> +	return time_is_after_jiffies(j) ? j : original;
>  }

Your email client mangles patches, btw.