Received-SPF: pass (google.com: domain of linux-ext4-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <000000000000d3a33205add2f7b2@google.com> <20200828100755.GG7072@quack2.suse.cz>
 <20200831100340.GA26519@quack2.suse.cz> <CAHk-=wivRS_1uy326sLqKuwerbL0APyKYKwa+vWVGsQg8sxhLw@mail.gmail.com>
 <alpine.LSU.2.11.2011231928140.4305@eggly.anvils>
In-Reply-To: <alpine.LSU.2.11.2011231928140.4305@eggly.anvils>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Mon, 23 Nov 2020 20:53:20 -0800
Message-ID: <CAHk-=whYO5v09E8oHoYQDn7qqV0hBu713AjF+zxJ9DCr1+WOtQ@mail.gmail.com>
Subject: Re: kernel BUG at fs/ext4/inode.c:LINE!
To:     Hugh Dickins <hughd@google.com>
Cc:     Jan Kara <jack@suse.cz>,
        syzbot <syzbot+3622cea378100f45d59f@syzkaller.appspotmail.com>,
        Andreas Dilger <adilger.kernel@dilger.ca>,
        Ext4 Developers List <linux-ext4@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        syzkaller-bugs <syzkaller-bugs@googlegroups.com>,
        "Theodore Ts'o" <tytso@mit.edu>, Linux-MM <linux-mm@kvack.org>,
        Oleg Nesterov <oleg@redhat.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        "Kirill A. Shutemov" <kirill@shutemov.name>,
        Nicholas Piggin <npiggin@gmail.com>,
        Alex Shi <alex.shi@linux.alibaba.com>, Qian Cai <cai@lca.pw>,
        Christoph Hellwig <hch@infradead.org>,
        "Darrick J. Wong" <darrick.wong@oracle.com>,
        Matthew Wilcox <willy@infradead.org>,
        William Kucharski <william.kucharski@oracle.com>,
        Jens Axboe <axboe@kernel.dk>,
        linux-fsdevel <linux-fsdevel@vger.kernel.org>,
        linux-xfs <linux-xfs@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Nov 23, 2020 at 8:07 PM Hugh Dickins <hughd@google.com> wrote:
>
> Then on crashing a second time, realized there's a stronger reason against
> that approach.  If my testing just occasionally crashes on that check,
> when the page is reused for part of a compound page, wouldn't it be much
> more common for the page to get reused as an order-0 page before reaching
> wake_up_page()?  And on rare occasions, might that reused page already be
> marked PageWriteback by its new user, and already be waited upon?  What
> would that look like?
>
> It would look like BUG_ON(PageWriteback) after wait_on_page_writeback()
> in write_cache_pages() (though I have never seen that crash myself).

So looking more at the patch, I started looking at this part:

> +       writeback = TestClearPageWriteback(page);
> +       /* No need for smp_mb__after_atomic() after TestClear */
> +       waiters = PageWaiters(page);
> +       if (waiters) {
> +               /*
> +                * Writeback doesn't hold a page reference on its own, relying
> +                * on truncation to wait for the clearing of PG_writeback.
> +                * We could safely wake_up_page_bit(page, PG_writeback) here,
> +                * while holding i_pages lock: but that would be a poor choice
> +                * if the page is on a long hash chain; so instead choose to
> +                * get_page+put_page - though atomics will add some overhead.
> +                */
> +               get_page(page);
> +       }

and thinking more about this, my first reaction was "but that has the
same race, just a smaller window".

And then reading the comment more, I realize you relied on the i_pages
lock, and that this odd ordering was to avoid the possible latency.

But what about the non-mapping case? I'm not sure how that happens,
but this does seem very fragile.

I'm wondering why you didn't want to just do the get_page()
unconditionally and early. Is avoiding the refcount really such a big
optimization?

            Linus