Re: arch/arm64/kernel/acpi.c:99:30: sparse: sparse: incorrect type in return expression (different address spaces)

From: Tom Lendacky <thomas.lendacky@amd.com>
To: Andy Shevchenko <andy.shevchenko@gmail.com>,
	kernel test robot <lkp@intel.com>,
	"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
Cc: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>,
	kbuild-all@lists.01.org,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: arch/arm64/kernel/acpi.c:99:30: sparse: sparse: incorrect type in return expression (different address spaces)
Date: Wed, 8 Jul 2020 08:14:25 -0500	[thread overview]
Message-ID: <5b3d1bb9-445c-729f-abe3-dae5fd5af3c2@amd.com> (raw)
In-Reply-To: <CAHp75VeE=BrT5haBt8EL+ARGKah6D48XznTPf8RGseyXi=jRcw@mail.gmail.com>

On 6/22/20 4:55 PM, Andy Shevchenko wrote:
> On Tue, Jun 23, 2020 at 12:42 AM Andy Shevchenko
> <andy.shevchenko@gmail.com> wrote:
>>
>> On Mon, Jun 22, 2020 at 11:53 PM kernel test robot <lkp@intel.com> wrote:
>>>
>>> tree:   https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.kernel.org%2Fpub%2Fscm%2Flinux%2Fkernel%2Fgit%2Ftorvalds%2Flinux.git&amp;data=02%7C01%7Cthomas.lendacky%40amd.com%7C22287276fb644db559c908d816f70d4f%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637284597621062483&amp;sdata=Kc%2Fk%2B39BMLFJW6MwcqHxMUkygKU7DJdTITq25UdCGr0%3D&amp;reserved=0 master
>>> head:   dd0d718152e4c65b173070d48ea9dfc06894c3e5
>>> commit: 670d0a4b10704667765f7d18f7592993d02783aa sparse: use identifiers to define address spaces
>>> date:   4 days ago
>>> config: arm64-randconfig-s031-20200622 (attached as .config)
>>> compiler: aarch64-linux-gcc (GCC) 9.3.0
>>> reproduce:
>>>         # apt-get install sparse
>>>         # sparse version: v0.6.2-rc1-18-g27caae40-dirty
>>>         git checkout 670d0a4b10704667765f7d18f7592993d02783aa
>>>         # save the attached .config to linux build tree
>>>         make W=1 C=1 ARCH=arm64 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
>>>
>>> If you fix the issue, kindly add following tag as appropriate
>>> Reported-by: kernel test robot <lkp@intel.com>
>>>
>>>
>>> sparse warnings: (new ones prefixed by >>)
>>>
>>>>> arch/arm64/kernel/acpi.c:99:30: sparse: sparse: incorrect type in return expression (different address spaces) @@     expected void [noderef] __iomem * @@     got void * @@
>>>>> arch/arm64/kernel/acpi.c:99:30: sparse:     expected void [noderef] __iomem *
>>>    arch/arm64/kernel/acpi.c:99:30: sparse:     got void *
>>>>> arch/arm64/kernel/acpi.c:107:24: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected void *addr @@     got void [noderef] __iomem *map @@
>>>    arch/arm64/kernel/acpi.c:107:24: sparse:     expected void *addr
>>>>> arch/arm64/kernel/acpi.c:107:24: sparse:     got void [noderef] __iomem *map
>>
>> ...
>>
>>> 6c9a58e84e59a2 Andy Shevchenko 2017-07-18   94  void __init __iomem *__acpi_map_table(unsigned long phys, unsigned long size)
>>> 6c9a58e84e59a2 Andy Shevchenko 2017-07-18  102  void __init __acpi_unmap_table(void __iomem *map, unsigned long size)
>>
>> Seems like a new sparse fixed some sparse issues and reveals that the
>> above was a mistake and should be reverted.
> 
> Either this patch
> f7750a79568788473c5e8092ee58a52248f34329 x86, mpparse, x86/acpi,
> x86/PCI, x86/dmi, SFI: Use memremap() for RAM mappings
> shouldn't change to memremap(), or

This patch is required for proper functioning of SME/SEV support and
should not be reverted.

Thanks,
Tom

> 
> void __iomem __ref *acpi_os_map_iomem(acpi_physical_address phys,
> acpi_size size);
> void __ref acpi_os_unmap_iomem(void __iomem *virt, acpi_size size);
> 
> Should be cleaned so they either not working with __iomem, or in
> special cases force address space for above mentioned functions.
> 