Re: pipe/page fault oddness.

[Linus Torvalds]

On Tue, Sep 30, 2014 at 9:05 AM, Dave Jones <davej@xxxxxxxxxx> wrote:
>
> I left it spinning overnight in case someone wanted me to probe it
> further, so I haven't tried reproducing it yet.  It took ~12 hours
> yesterday before it got in that state.  I'll restart it, and tell it
> to only use pipe fd's, which might speed things up a little.

Actually, if you haven't restarted it yet, do a few more "Sysrq-T"'s
to see if the stack below the page fault ever changes. Is it *always*
that "second access by fault_in_pages_writeable()" or migth there be
some looping going on in copy_page_to_iter() after all? (Quite
frankly, I don't see how such looping could happen with a good
compiler, and 4.8.3 should be good, but just in case).

It's an unlikely scenario, so if you already rebooted, not a big deal.

> model name      : Intel(R) Core(TM) i5-4670T CPU @ 2.30GHz

Ok, that pretty much throws the "maybe AMD did something different"
theory out of the water. Something subtler. Still worth trying the
revert.

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