Re: [Bugme-new] [Bug 12562] New: High overhead while switching or synchronizing threads on different cores

[Peter Zijlstra <peterz@infradead.org>]

On Tue, 2009-02-03 at 07:29 +1300, Michael Kerrisk wrote:
> Hi Peter,
> 
> On Tue, Feb 3, 2009 at 1:15 AM, Peter Zijlstra <peterz@infradead.org> wrote:
> > On Mon, 2009-02-02 at 09:55 +0100, Peter Zijlstra wrote:
> >
> >> Ah, that was a good clue, apparently all you need to so it use
> >> random_r() and provide your own state and all should be well.
> >
> > Michael, would it make sense to add the random_r() family to the "SEE
> > ALSO" section of the random() man page?
> >
> > (Admittedly, my random() manpage is ancient: 2008-03-07, so it might be
> > this is already the case, in which case, ignore me :)
> 
> (Up-to-date version of the pages can always be found online at the
> location in the .sig.)

Ah, I'll try to remember that.

> Well, the man page already had this text under notes:
> 
>        This  function  should  not  be  used  in  cases where multiple
>        threads use random() and the behavior should  be  reproducible.
>        Use random_r(3) for that purpose.

Yeah, but I found it eventually, but I generally don't read a full
manpage when I'm looking for related functions, only the SEE ALSO
section.

> But it certainly doesn't hurt to have random_r(3) also listed under
> the SEE ALSO, and I've added it for man-pages-3.18.

Thanks.