Received-SPF: pass (google.com: domain of linux-kernel+bounces-60795-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
From: David Laight <David.Laight@ACULAB.COM>
To: 'Guenter Roeck' <linux@roeck-us.net>, Helge Deller <deller@gmx.de>
CC: "James E . J . Bottomley" <James.Bottomley@HansenPartnership.com>,
	"linux-parisc@vger.kernel.org" <linux-parisc@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>, "Charlie
 Jenkins" <charlie@rivosinc.com>, Palmer Dabbelt <palmer@rivosinc.com>
Subject: RE: [PATCH] parisc: Fix csum_ipv6_magic on 32-bit systems
Thread-Topic: [PATCH] parisc: Fix csum_ipv6_magic on 32-bit systems
Thread-Index: AQHaXFWRp9eVLmaVBEGREYhx6VwfXbEFKwHA
Date: Sun, 11 Feb 2024 13:57:01 +0000
Message-ID: <dae9f9c428a242368d47ba29561e9455@AcuMS.aculab.com>
References: <20240210191556.3761064-1-linux@roeck-us.net>
In-Reply-To: <20240210191556.3761064-1-linux@roeck-us.net>
Accept-Language: en-GB, en-US
Precedence: bulk
MIME-Version: 1.0
Content-Language: en-US
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

From: Guenter Roeck
> Sent: 10 February 2024 19:16
>=20
> Calculating the IPv6 checksum on 32-bit systems missed overflows when
> adding the proto+len fields into the checksum. This results in the
> following unit test failure.
>=20
>     # test_csum_ipv6_magic: ASSERTION FAILED at lib/checksum_kunit.c:506
>     Expected ( u64)csum_result =3D=3D ( u64)expected, but
>         ( u64)csum_result =3D=3D 46722 (0xb682)
>         ( u64)expected =3D=3D 46721 (0xb681)
>     not ok 5 test_csum_ipv6_magic
>=20
> This is probably rarely seen in the real world because proto+len are
> usually small values which will rarely result in overflows when calculati=
ng
> the checksum. However, the unit test code uses large values for the lengt=
h
> field, causing the test to fail.

Isn't length limited by the protocol encoding?
So this is really a bug in the unit tests for using a length that
it too large for the function?

=09David

>=20
> Fix the problem by adding the missing carry into the final checksum.
>=20
> Cc: Charlie Jenkins <charlie@rivosinc.com>
> Cc: Palmer Dabbelt <palmer@rivosinc.com>
> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> ---
>  arch/parisc/include/asm/checksum.h | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>=20
> diff --git a/arch/parisc/include/asm/checksum.h b/arch/parisc/include/asm=
/checksum.h
> index f705e5dd1074..e619e67440db 100644
> --- a/arch/parisc/include/asm/checksum.h
> +++ b/arch/parisc/include/asm/checksum.h
> @@ -163,7 +163,8 @@ static __inline__ __sum16 csum_ipv6_magic(const struc=
t in6_addr *saddr,
>  "=09ldw,ma=09=094(%2), %7\n"=09/* 4th daddr */
>  "=09addc=09=09%6, %0, %0\n"
>  "=09addc=09=09%7, %0, %0\n"
> -"=09addc=09=09%3, %0, %0\n"=09/* fold in proto+len, catch carry */
> +"=09addc=09=09%3, %0, %0\n"=09/* fold in proto+len */
> +"=09addc=09=090, %0, %0\n"=09/* add carry */
>=20
>  #endif
>  =09: "=3Dr" (sum), "=3Dr" (saddr), "=3Dr" (daddr), "=3Dr" (len),
> --
> 2.39.2
>=20

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1=
PT, UK
Registration No: 1397386 (Wales)