Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:3 as permitted sender) client-ip=2620:137:e000::3:3;
From:   Michael Ellerman <mpe@ellerman.id.au>
To:     Kunwu Chan <chentao@kylinos.cn>,
        Christophe Leroy <christophe.leroy@csgroup.eu>,
        "npiggin@gmail.com" <npiggin@gmail.com>
Cc:     "kunwu.chan@hotmail.com" <kunwu.chan@hotmail.com>,
        "linuxppc-dev@lists.ozlabs.org" <linuxppc-dev@lists.ozlabs.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] powerpc/mm: Fix null-pointer dereference in
 pgtable_cache_add
In-Reply-To: <9d871364-7baa-4daf-8b0c-3fbfbede6fdb@kylinos.cn>
References: <20231122090026.11728-1-chentao@kylinos.cn>
 <32077b74-7335-4f4d-8858-c53c820150d0@csgroup.eu>
 <9d871364-7baa-4daf-8b0c-3fbfbede6fdb@kylinos.cn>
Date:   Tue, 28 Nov 2023 22:32:21 +1100
Message-ID: <87y1ei2jwa.fsf@mail.lhotse>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Kunwu Chan <chentao@kylinos.cn> writes:
> Hi Christophe,
>
> Thanks for your reply.
> It's my bad. According your reply, i read the code in=20
> sysfs_do_create_link_sd.There is a null pointer check indeed.
>
> My intention was to check null pointer after memory allocation.
> Whether we can add a comment here for someone like me, the null pointer=20
> check is no need here?

I don't mind there being a NULL check for name.

But the code shouldn't silently return if name can't be allocated.
Notice that if we can't create the cache we *panic*. A failure to
allocate name, which causes us to skip the cache creation, needs to also
panic.

cheers

> On 2023/11/24 23:17, Christophe Leroy wrote:
>>=20
>>=20
>> Le 22/11/2023 =C3=A0 10:00, Kunwu Chan a =C3=A9crit=C2=A0:
>>> [Vous ne recevez pas souvent de courriers de chentao@kylinos.cn. D=C3=
=A9couvrez pourquoi ceci est important =C3=A0 https://aka.ms/LearnAboutSend=
erIdentification ]
>>>
>>> kasprintf() returns a pointer to dynamically allocated memory
>>> which can be NULL upon failure. Ensure the allocation was successful
>>> by checking the pointer validity.
>>=20
>> Are you sure this is needed ? Did you check what happens what name is NU=
LL ?
>>=20
>> If I followed stuff correctly, I end up in function
>> sysfs_do_create_link_sd() which already handles the NULL name case which
>> a big hammer warning.
>>=20
>>>
>>> Signed-off-by: Kunwu Chan <chentao@kylinos.cn>
>>> ---
>>>    arch/powerpc/mm/init-common.c | 2 ++
>>>    1 file changed, 2 insertions(+)
>>>
>>> diff --git a/arch/powerpc/mm/init-common.c b/arch/powerpc/mm/init-commo=
n.c
>>> index 119ef491f797..0884fc601c46 100644
>>> --- a/arch/powerpc/mm/init-common.c
>>> +++ b/arch/powerpc/mm/init-common.c
>>> @@ -139,6 +139,8 @@ void pgtable_cache_add(unsigned int shift)
>>>
>>>           align =3D max_t(unsigned long, align, minalign);
>>>           name =3D kasprintf(GFP_KERNEL, "pgtable-2^%d", shift);
>>> +       if (!name)
>>> +               return;
>>>           new =3D kmem_cache_create(name, table_size, align, 0, ctor(sh=
ift));
>>>           if (!new)
>>>                   panic("Could not allocate pgtable cache for order %d"=
, shift);
>>> --
>>> 2.34.1
>>>