Re: [PATCH v1] mm: soft-offline: check return value in second __get_any_page() call

[Naoya Horiguchi]

On Fri, Jan 08, 2016 at 03:36:26PM -0800, Andrew Morton wrote:
> On Fri, 8 Jan 2016 07:51:59 +0000 Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx> wrote:
> 
> > >   [   52.600579]  [<ffffffff811bd18c>] SyS_madvise+0x6bc/0x6f0
> > >   [   52.600579]  [<ffffffff8104d0ac>] ? fpu__restore_sig+0xcc/0x320
> > >   [   52.600579]  [<ffffffff810a0003>] ? do_sigaction+0x73/0x1b0
> > >   [   52.600579]  [<ffffffff8109ceb2>] ? __set_task_blocked+0x32/0x70
> > >   [   52.600579]  [<ffffffff81652757>] entry_SYSCALL_64_fastpath+0x12/0x6a
> > >   [   52.600579] Code: 8b fc ff ff 5b 5d c3 48 89 df e8 b0 fa ff ff 48 89 df 31 f6 e8 c6 7d ff ff 5b 5d c3 48 c7 c6 08 54 a2 81 48 89 df e8 a4 c5 01 00 <0f> 0b 66 90 66 66 66 66 90 55 48 89 e5 41 55 41 54 53 48 8b 47
> > >   [   52.600579] RIP  [<ffffffff8118998c>] put_page+0x5c/0x60
> > >   [   52.600579]  RSP <ffff88007c213e00>
> > > 
> > > The root cause resides in get_any_page() which retries to get a refcount of
> > > the page to be soft-offlined. This function calls put_hwpoison_page(), expecting
> > > that the target page is putback to LRU list. But it can be also freed to buddy.
> > > So the second check need to care about such case.
> > > 
> > > Fixes: af8fae7c0886 ("mm/memory-failure.c: clean up soft_offline_page()")
> > > Signed-off-by: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
> > > Cc: stable@xxxxxxxxxxxxxxx # v3.9+
> 
> Please don't top-post.  I manually fixed it here.

sorry, I keep this rule in mind.

> > Sorry, I forgot to notice that this specific problem is already fixed in
> > mmotm with patch "mm: hwpoison: adjust for new thp refcounting", but
> > considering backporting to -stable, it's easier to handle this separately.
> > 
> > So Andrew, could you separate out the code of this patch from
> > "mm: hwpoison: adjust for new thp refcounting"?
> 
> I don't understand what you're asking for.  Please be very
> specific and carefully identify patches by filename or Subject:.

OK, so what I really wanted is that (1) applying this patch just before
http://ozlabs.org/~akpm/mmots/broken-out/mm-hwpoison-adjust-for-new-thp-refcounting.patch
and (2) removing the following chunk from the mm-hwpoison-adjust-for-new-thp-refcounting.patch:

@@ -1575,7 +1540,7 @@ static int get_any_page(struct page *pag
 		 * Did it turn free?
 		 */
 		ret = __get_any_page(page, pfn, 0);
-		if (!PageLRU(page)) {
+		if (ret == 1 && !PageLRU(page)) {
 			/* Drop page reference which is from __get_any_page() */
 			put_hwpoison_page(page);
 			pr_info("soft_offline: %#lx: unknown non LRU page type %lx\n",

Thanks,
Naoya Horiguchi