Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From:   Orlando Chamberlain <orlandoch.dev@gmail.com>
Mime-Version: 1.0 (1.0)
Subject: Re: drivers/platform/x86/apple-gmux.c:276:15: sparse: sparse: cast to restricted __be32
Date:   Tue, 16 May 2023 21:31:16 +1000
Message-Id: <D62B0A4F-6A2E-495D-B3AE-4CEAEE172E39@gmail.com>
References: <56f2f65b-150d-3cf6-95f9-2dd8232d2717@redhat.com>
Cc:     kernel test robot <lkp@intel.com>, oe-kbuild-all@lists.linux.dev,
        linux-kernel@vger.kernel.org
In-Reply-To: <56f2f65b-150d-3cf6-95f9-2dd8232d2717@redhat.com>
To:     Hans de Goede <hdegoede@redhat.com>
Precedence: bulk

Hi,

> On 16 May 2023, at 8:27 pm, Hans de Goede <hdegoede@redhat.com> wrote:
>=20
> =EF=BB=BFHi,
>=20
>> On 5/16/23 12:16, kernel test robot wrote:
>> tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.gi=
t master
>> head:   f1fcbaa18b28dec10281551dfe6ed3a3ed80e3d6
>> commit: 0c18184de990e63f708b090bcb9fc6c0fbc427cd platform/x86: apple-gmux=
: support MMIO gmux on T2 Macs
>> date:   9 weeks ago
>> config: i386-randconfig-s001-20230515
>> compiler: gcc-11 (Debian 11.3.0-12) 11.3.0
>> reproduce:
>>        # apt-get install sparse
>>        # sparse version: v0.6.4-39-gce1a6720-dirty
>>        # https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.g=
it/commit/?id=3D0c18184de990e63f708b090bcb9fc6c0fbc427cd
>>        git remote add linus https://git.kernel.org/pub/scm/linux/kernel/g=
it/torvalds/linux.git
>>        git fetch --no-tags linus master
>>        git checkout 0c18184de990e63f708b090bcb9fc6c0fbc427cd
>>        # save the config file
>>        mkdir build_dir && cp config build_dir/.config
>>        make W=3D1 C=3D1 CF=3D'-fdiagnostic-prefix -D__CHECK_ENDIAN__' O=3D=
build_dir ARCH=3Di386 olddefconfig
>>        make W=3D1 C=3D1 CF=3D'-fdiagnostic-prefix -D__CHECK_ENDIAN__' O=3D=
build_dir ARCH=3Di386 SHELL=3D/bin/bash drivers/platform/x86/
>>=20
>> If you fix the issue, kindly add following tag where applicable
>> | Reported-by: kernel test robot <lkp@intel.com>
>> | Link: https://lore.kernel.org/oe-kbuild-all/202305161712.5l3f4iI4-lkp@i=
ntel.com/
>=20
> <snip>
>=20
>> vim +276 drivers/platform/x86/apple-gmux.c
>>=20
>>   265   =20
>>   266    static u32 gmux_mmio_read32(struct apple_gmux_data *gmux_data, i=
nt port)
>>   267    {
>>   268        u32 val;
>>   269   =20
>>   270        mutex_lock(&gmux_data->index_lock);
>>   271        gmux_mmio_wait(gmux_data);
>>   272        iowrite8((port & 0xff), gmux_data->iomem_base + GMUX_MMIO_PO=
RT_SELECT);
>>   273        iowrite8(GMUX_MMIO_READ | sizeof(val),
>>   274            gmux_data->iomem_base + GMUX_MMIO_COMMAND_SEND);
>>   275        gmux_mmio_wait(gmux_data);
>>> 276        val =3D be32_to_cpu(ioread32(gmux_data->iomem_base));
>=20
> Ok, so sparse does not like this line.
>=20
>>   277        mutex_unlock(&gmux_data->index_lock);
>>   278   =20
>>   279        return val;
>>   280    }
>>   281   =20
>>   282    static void gmux_mmio_write32(struct apple_gmux_data *gmux_data,=
 int port,
>>   283                       u32 val)
>>   284    {
>>   285        mutex_lock(&gmux_data->index_lock);
>>> 286        iowrite32(cpu_to_be32(val), gmux_data->iomem_base);
>=20
> Nor this line. But this is what we want (convert to/from be32 to CPU
> when reading/writing).
>=20
> There is iowrite32be() but that always unconditionally swabs
> the byte order independent of the CPU byte-order.
>=20
> Now this is an x86 driver so always swapping is fine, still
> I wonder if there is a better option here then using
> iowrite32be() and ioread32be() ?

Thanks for finding those functions. I can't think of anything better unless t=
here's a simple way to change sparse's rules so that it will be happy with c=
ode like this.

Regards,
Orlando

>=20
> Regards,
>=20
> Hans
>=20
>=20
>=20
>=20
>>   287        iowrite8(port & 0xff, gmux_data->iomem_base + GMUX_MMIO_PORT=
_SELECT);
>>   288        iowrite8(GMUX_MMIO_WRITE | sizeof(val),
>>   289            gmux_data->iomem_base + GMUX_MMIO_COMMAND_SEND);
>>   290        gmux_mmio_wait(gmux_data);
>>   291        mutex_unlock(&gmux_data->index_lock);
>>   292    }
>>   293   =20
>>=20
>=20