linux-kernel - Re: [PATCH 1/1] cputime: Make the reported utime+stime correspond to the actual runtime.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 7 Jul 2015 14:10:56 +0200
From:	Fredrik Markström <fredrik.markstrom@...il.com>
To:	Peter Zijlstra <peterz@...radead.org>
Cc:	Frederic Weisbecker <fweisbec@...il.com>, mingo@...hat.com,
	linux-kernel@...r.kernel.org, Rik van Riel <riel@...hat.com>,
	Jason Low <jason.low2@...com>
Subject: Re: [PATCH 1/1] cputime: Make the reported utime+stime correspond to
 the actual runtime.

Just to let you know, I've tested your last patch and it solves all my
original problems (which is should since the code is functionally
equivalent).

/Fredrik

On Tue, Jul 7, 2015 at 10:09 AM, Peter Zijlstra <peterz@...radead.org> wrote:
> On Tue, Jul 07, 2015 at 09:59:54AM +0200, Peter Zijlstra wrote:
>> > > + /*
>> > > +  * Make sure stime doesn't go backwards; this preserves monotonicity
>> > > +  * for utime because rtime is monotonic.
>> > > +  *
>> > > +  *  utime_i+1 = rtime_i+1 - stime_i
>> > > +  *            = rtime_i+1 - (rtime_i - stime_i)
>> > > +  *            = (rtime_i+1 - rtime_i) + stime_i
>> > > +  *            >= stime_i
>> > > +  */
>
> Argh, just noticed I messed that up, it should read:
>
> +       /*
> +        * Make sure stime doesn't go backwards; this preserves monotonicity
> +        * for utime because rtime is monotonic.
> +        *
> +        *  utime_i+1 = rtime_i+1 - stime_i
> +        *            = rtime_i+1 - (rtime_i - utime_i)
> +        *            = (rtime_i+1 - rtime_i) + utime_i
> +        *            >= utime_i
> +        */
>
> I got some [us] confusion. Typing is hard.
>
> So we compute: utime = rtime - stime, which we'll denote as:
>
>         utime_i+1 = rtime_i+1 - stime_i
>
> since: stime_i + utime_i = rtime_i, we can write: stime_i = rtime_i -
> utime_i and substitute in the above:
>
>                  = rtime_i+1 - (rtime_i - utime_i)
>
> Rearrange terms:
>
>                 = (rtime_i+1 - rtime_i) + utime_i
>
> And since we have: rtime_i+1 >= rtime_i, which we can write like:
> rtime_i+1 - rtime_i >= 0, substitution gives:
>
>                 >= utime_i
>
> And we've proven that the new utime must be equal or greater than the
> previous utime, because rtime is monotonic.



-- 
/Fredrik
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/