linux-kernel - Re: [PATCH] FS: timerfd: Fix unexpected return value of timerfd

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] [day] [month] [year] [list]

Message-ID: <alpine.DEB.2.21.1908191752580.2147@nanos.tec.linutronix.de>
Date:   Mon, 19 Aug 2019 17:59:00 +0200 (CEST)
From:   Thomas Gleixner <tglx@...utronix.de>
To:     Arul Jeniston <arul.jeniston@...il.com>
cc:     viro@...iv.linux.org.uk, linux-fsdevel@...r.kernel.org,
        linux-kernel@...r.kernel.org, arul_mc@...l.com
Subject: Re: [PATCH] FS: timerfd: Fix unexpected return value of timerfd_read
 function.

On Mon, 19 Aug 2019, Arul Jeniston wrote:

> hi Tglx,
> > But for the above scenario:
> >
> > ktime_get()
> >         do {
> >                 seq = read_seqcount_begin(&tk_core.seq);
> >                 base = tk->tkr_mono.base;
> >                 nsecs = timekeeping_get_ns(&tk->tkr_mono);
> >
> >         } while (read_seqcount_retry(&tk_core.seq, seq));
> >
> > So if the interrupt which updates the timekeeper hits in the middle of
> > timekeeping_get_ns() then the result is discarded because the sequence
> > count changed and read_seqcount_retry() returns true. So it takes another
> > round which will be perfectly aligned with the updated time keeper.
> >
> 
> Do you mean to say the timekeeper  updates always happen from ktime_get?
> My point was, when one thread is in ktime_get other thread/isr updates
> timekeeper from different flow.

Timekeeper updates happen of course NOT from ktime_get(), but ktime_get()
is protected against concurrent updates via the seqcount. Simplified
without all the required barriers etc.

ktime_get()

    do {
       seq = tk->seq;
       if (seq & 1)
          continue;
       base = tk->base;
       nsec = get_nsec();
    while (seq != tk->seq);

update()

   tk->seq++;
   update_data();
   tk-<seq++;

It does not matter whether the update is an interrupt on the same CPU which
hits ktime_get() or whether it happens concurrent on a different CPU.

ktime_get() can never use inconsistent tk data for calculating the time.

Thanks,

	tglx