Re: Memory hotplug softlock issue

[Baoquan He]

Hi,

On 11/20/18 at 02:38pm, Vlastimil Babka wrote:
> On 11/20/18 6:44 AM, Hugh Dickins wrote:
> > [PATCH] mm: put_and_wait_on_page_locked() while page is migrated
> > 
> > We have all assumed that it is essential to hold a page reference while
> > waiting on a page lock: partly to guarantee that there is still a struct
> > page when MEMORY_HOTREMOVE is configured, but also to protect against
> > reuse of the struct page going to someone who then holds the page locked
> > indefinitely, when the waiter can reasonably expect timely unlocking.
> > 
> > But in fact, so long as wait_on_page_bit_common() does the put_page(),
> > and is careful not to rely on struct page contents thereafter, there is
> > no need to hold a reference to the page while waiting on it.  That does
> 
> So there's still a moment where refcount is elevated, but hopefully
> short enough, right? Let's see if it survives Baoquan's stress testing.

Yes, I applied Hugh's patch 8 hours ago, then our QE Ping operated on
that machine, after many times of hot removing/adding, the endless
looping during mirgrating is not seen any more. The test result for
Hugh's patch is positive. I even suggested Ping increasing the memory
pressure to "stress -m 250", it still succeeded to offline and remove.

So I think this patch works to solve the issue. Thanks a lot for your
help, all of you. 

High, will you post a formal patch in a separate thread?

Meanwhile we found sometime onlining page may not add back all memory
blocks on one memory board, then hot removing/adding them will cause
kernel panic. I will investigate further and collect information, see if
it's a kernel issue or udev issue.

Thanks
Baoquan

> 
> > mean that this case cannot go back through the loop: but that's fine for
> > the page migration case, and even if used more widely, is limited by the
> > "Stop walking if it's locked" optimization in wake_page_function().
> > 
> > Add interface put_and_wait_on_page_locked() to do this, using negative
> > value of the lock arg to wait_on_page_bit_common() to implement it.
> > No interruptible or killable variant needed yet, but they might follow:
> > I have a vague notion that reporting -EINTR should take precedence over
> > return from wait_on_page_bit_common() without knowing the page state,
> > so arrange it accordingly - but that may be nothing but pedantic.
> > 
> > shrink_page_list()'s __ClearPageLocked(): that was a surprise! this
> > survived a lot of testing before that showed up.  It does raise the
> > question: should is_page_cache_freeable() and __remove_mapping() now
> > treat a PG_waiters page as if an extra reference were held?  Perhaps,
> > but I don't think it matters much, since shrink_page_list() already
> > had to win its trylock_page(), so waiters are not very common there: I
> > noticed no difference when trying the bigger change, and it's surely not
> > needed while put_and_wait_on_page_locked() is only for page migration.
> > 
> > Signed-off-by: Hugh Dickins <hughd@xxxxxxxxxx>
> > ---
> 
> ...
> 
> > @@ -1100,6 +1111,17 @@ static inline int wait_on_page_bit_common(wait_queue_head_t *q,
> >  			ret = -EINTR;
> >  			break;
> >  		}
> > +
> > +		if (lock < 0) {
> > +			/*
> > +			 * We can no longer safely access page->flags:
> 
> Hmm...
> 
> > +			 * even if CONFIG_MEMORY_HOTREMOVE is not enabled,
> > +			 * there is a risk of waiting forever on a page reused
> > +			 * for something that keeps it locked indefinitely.
> > +			 * But best check for -EINTR above before breaking.
> > +			 */
> > +			break;
> > +		}
> >  	}
> >  
> >  	finish_wait(q, wait);
> 
> ... the code continues by:
> 
>         if (thrashing) {
>                 if (!PageSwapBacked(page))
> 
> So maybe we should not set 'thrashing' true when lock < 0?
> 
> Thanks!
> Vlastimil