linux-kernel - Re: [sched next] overflowed cpu time for kernel threads in /proc/PID/stat

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]

Message-ID: <20130830230402.GA14760@somewhere>
Date:	Sat, 31 Aug 2013 01:04:04 +0200
From:	Frederic Weisbecker <fweisbec@...il.com>
To:	Sergey Senozhatsky <sergey.senozhatsky@...il.com>
Cc:	Stanislaw Gruszka <sgruszka@...hat.com>,
	Ingo Molnar <mingo@...hat.com>,
	Peter Zijlstra <peterz@...radead.org>,
	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>,
	Borislav Petkov <bp@...en8.de>, linux-kernel@...r.kernel.org
Subject: Re: [sched next] overflowed cpu time for kernel threads in
 /proc/PID/stat

On Wed, Aug 21, 2013 at 06:39:57PM +0300, Sergey Senozhatsky wrote:
> On (08/20/13 17:42), Frederic Weisbecker wrote:
> > On Tue, Aug 20, 2013 at 06:35:50PM +0300, Sergey Senozhatsky wrote:
> > > On (08/20/13 17:15), Frederic Weisbecker wrote:
> > > > On Tue, Aug 20, 2013 at 02:14:26PM +0300, Sergey Senozhatsky wrote:
> > > > Hello,
> > > > 
> > > > Please upload your config.
> > > > 
> > > > I'm adding Stanislaw in Cc in case it relates to cputime scaling.
> > > >
> > > 
> > > please find attached.
> > > 
> > > 	-ss
> > 
> > > #
> > > # CPU/Task time and stats accounting
> > > #
> > > CONFIG_VIRT_CPU_ACCOUNTING=y
> > > CONFIG_VIRT_CPU_ACCOUNTING_GEN=y
> > 
> > So it happens with full dynticks cputime accounting.
> > How reproducable is this? Does it always happen?
> > Does it happen with CONFIG_TICK_CPU_ACCOUNTING as well? (you'll
> > need to deactivate full dynticks.)
> 
> in cputime_adjust() `stime' is greater than `rtime', so `utime = rtime - stime'
> sets wrong value.

But stime should always be below rtime due to the calculation done by scale_stime()
which roughly sums up to:

      stime = (stime / stime + utime) * rtime

So this shouldn't happen.

I couldn't manage to reproduce it though. Can you still trigger it with latest -next?

Thanks.
--
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/