Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [RFC PATCH v2 15/16] mm/hwpoison-inject: Rip off duplicated
 checks
To:     Oscar Salvador <osalvador@suse.de>, n-horiguchi@ah.jp.nec.com
Cc:     mhocko@kernel.org, mike.kravetz@oracle.com, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org
References: <20191017142123.24245-1-osalvador@suse.de>
 <20191017142123.24245-16-osalvador@suse.de>
From:   David Hildenbrand <david@redhat.com>
Organization: Red Hat GmbH
Message-ID: <1aa1f09a-1210-d0bc-86ac-9674828bff49@redhat.com>
Date:   Mon, 21 Oct 2019 11:40:39 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.1.1
MIME-Version: 1.0
In-Reply-To: <20191017142123.24245-16-osalvador@suse.de>
Content-Language: en-US
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 17.10.19 16:21, Oscar Salvador wrote:
> memory_failure() already performs the same checks, so leave it
> to the main routine.
>=20
> Signed-off-by: Oscar Salvador <osalvador@suse.de>
> ---
>   mm/hwpoison-inject.c | 33 +++------------------------------
>   1 file changed, 3 insertions(+), 30 deletions(-)
>=20
> diff --git a/mm/hwpoison-inject.c b/mm/hwpoison-inject.c
> index 0c8cdb80fd7d..fdcca3df4283 100644
> --- a/mm/hwpoison-inject.c
> +++ b/mm/hwpoison-inject.c
> @@ -14,49 +14,22 @@ static struct dentry *hwpoison_dir;
>   static int hwpoison_inject(void *data, u64 val)
>   {
>   =09unsigned long pfn =3D val;
> -=09struct page *p;
> -=09struct page *hpage;
> -=09int err;
>  =20
>   =09if (!capable(CAP_SYS_ADMIN))
>   =09=09return -EPERM;
>  =20
> -=09if (!pfn_valid(pfn))
> -=09=09return -ENXIO;
> -
> -=09p =3D pfn_to_page(pfn);
> -=09hpage =3D compound_head(p);
> -
> -=09if (!hwpoison_filter_enable)
> -=09=09goto inject;
> -
> -=09shake_page(hpage, 0);
> -=09/*
> -=09 * This implies unable to support non-LRU pages.
> -=09 */
> -=09if (!PageLRU(hpage) && !PageHuge(p))
> -=09=09return 0;
> -
> -=09/*
> -=09 * do a racy check to make sure PG_hwpoison will only be set for
> -=09 * the targeted owner (or on a free page).
> -=09 * memory_failure() will redo the check reliably inside page lock.
> -=09 */
> -=09err =3D hwpoison_filter(hpage);
> -=09if (err)
> -=09=09return 0;
> -
> -inject:
>   =09pr_info("Injecting memory failure at pfn %#lx\n", pfn);
>   =09return memory_failure(pfn, 0);
>   }
>  =20

I explored somewhere already why this code was added:


commit 31d3d3484f9bd263925ecaa341500ac2df3a5d9b
Author: Wu Fengguang <fengguang.wu@intel.com>
Date:   Wed Dec 16 12:19:59 2009 +0100

    HWPOISON: limit hwpoison injector to known page types
   =20
    __memory_failure()'s workflow is
   =20
            set PG_hwpoison
            //...
            unset PG_hwpoison if didn't pass hwpoison filter
   =20
    That could kill unrelated process if it happens to page fault on the
    page with the (temporary) PG_hwpoison. The race should be big enough to
    appear in stress tests.
   =20
    Fix it by grabbing the page and checking filter at inject time.  This
    also avoids the very noisy "Injecting memory failure..." messages.
   =20
    - we don't touch madvise() based injection, because the filters are
      generally not necessary for it.
    - if we want to apply the filters to h/w aided injection, we'd better t=
o
      rearrange the logic in __memory_failure() instead of this patch.
   =20
    AK: fix documentation, use drain all, cleanups


You should justify why it is okay to do rip that code out now.
It's not just duplicate checks.

Was the documented race fixed?
Will we fix the race within memory_failure() later?
Don't we care?

Also, you should add that this fixes the access of uninitialized memmaps
now and makes the interface work correctly with devmem.

--=20

Thanks,

David / dhildenb