Build regressions/improvements in v5.17-rc1

Randy Dunlap rdunlap at infradead.org
Tue Jan 25 01:29:17 UTC 2022



On 1/24/22 17:23, Felix Kuehling wrote:
> 
> Am 2022-01-24 um 14:11 schrieb Randy Dunlap:
>> On 1/24/22 10:55, Geert Uytterhoeven wrote:
>>> Hi Alex,
>>>
>>> On Mon, Jan 24, 2022 at 7:52 PM Alex Deucher <alexdeucher at gmail.com> wrote:
>>>> On Mon, Jan 24, 2022 at 5:25 AM Geert Uytterhoeven <geert at linux-m68k.org> wrote:
>>>>> On Sun, 23 Jan 2022, Geert Uytterhoeven wrote:
>>>>>>   + /kisskb/src/drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c: error: control reaches end of non-void function [-Werror=return-type]:  => 1560:1
>>>> I don't really see what's going on here:
>>>>
>>>> #ifdef CONFIG_X86_64
>>>> return cpu_data(first_cpu_of_numa_node).apicid;
>>>> #else
>>>> return first_cpu_of_numa_node;
>>>> #endif
>>> Ah, the actual failure causing this was not included:
>>>
>>> In file included from /kisskb/src/arch/x86/um/asm/processor.h:41:0,
>>>                   from /kisskb/src/include/linux/mutex.h:19,
>>>                   from /kisskb/src/include/linux/kernfs.h:11,
>>>                   from /kisskb/src/include/linux/sysfs.h:16,
>>>                   from /kisskb/src/include/linux/kobject.h:20,
>>>                   from /kisskb/src/include/linux/pci.h:35,
>>>                   from
>>> /kisskb/src/drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:25:
>>> /kisskb/src/drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c: In
>>> function 'kfd_cpumask_to_apic_id':
>>> /kisskb/src/arch/um/include/asm/processor-generic.h:103:18: error:
>>> called object is not a function or function pointer
>>>   #define cpu_data (&boot_cpu_data)
>>>                    ^
>>> /kisskb/src/drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1556:9:
>>> note: in expansion of macro 'cpu_data'
>>>    return cpu_data(first_cpu_of_numa_node).apicid;
>>>           ^
>>> /kisskb/src/drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1560:1:
>>> error: control reaches end of non-void function [-Werror=return-type]
>>>   }
>>>   ^
>> ah yes, UML.
>> I have a bunch of UML fixes that I have been hesitant to post.
>>
>> This is one of them.
>> What do people think about this?
> 
> Does it make sense to configure a UML kernel with a real device driver in the first place? Or should we just prevent enabling amdgpu for UML with a Kconfig dependency?
> 

Hi,

Your option IMO. I have seen both opinions given.
I also meant to reply that someone could just add
	depends on !UML
for this device, like you are suggesting.

I'm fine with it either way.

thanks.

> 
>>
>> thanks.
>>
>> ---
>> From: Randy Dunlap <rdunlap at infradead.org>
>>
>>
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1556:9: note: in expansion of macro ‘cpu_data’
>>    return cpu_data(first_cpu_of_numa_node).apicid;
>>           ^~~~~~~~
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1560:1: error: control reaches end of non-void function [-Werror=return-type]
>>
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_crat.c: In function ‘kfd_fill_iolink_info_for_cpu’:
>> ../arch/um/include/asm/processor-generic.h:103:19: error: called object is not a function or function pointer
>>   #define cpu_data (&boot_cpu_data)
>>                    ~^~~~~~~~~~~~~~~
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_crat.c:1688:27: note: in expansion of macro ‘cpu_data’
>>    struct cpuinfo_x86 *c = &cpu_data(0);
>>                             ^~~~~~~~
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_crat.c:1691:7: error: dereferencing pointer to incomplete type ‘struct cpuinfo_x86’
>>    if (c->x86_vendor == X86_VENDOR_AMD)
>>         ^~
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_crat.c:1691:23: error: ‘X86_VENDOR_AMD’ undeclared (first use in this function); did you mean ‘X86_VENDOR_ANY’?
>>    if (c->x86_vendor == X86_VENDOR_AMD)
>>                         ^~~~~~~~~~~~~~
>>                         X86_VENDOR_ANY
>>
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_crat.c: In function ‘kfd_create_vcrat_image_cpu’:
>> ../drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_crat.c:1742:11: warning: unused variable ‘entries’ [-Wunused-variable]
>>    uint32_t entries = 0;
>>
>> Signed-off-by: Randy Dunlap <rdunlap at infradead.org>
>> ---
>>   drivers/gpu/drm/amd/amdkfd/kfd_crat.c     |    6 +++---
>>   drivers/gpu/drm/amd/amdkfd/kfd_topology.c |    2 +-
>>   2 files changed, 4 insertions(+), 4 deletions(-)
>>
>> --- linux-next-20220107.orig/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
>> +++ linux-next-20220107/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
>> @@ -1552,7 +1552,7 @@ static int kfd_cpumask_to_apic_id(const
>>       first_cpu_of_numa_node = cpumask_first(cpumask);
>>       if (first_cpu_of_numa_node >= nr_cpu_ids)
>>           return -1;
>> -#ifdef CONFIG_X86_64
>> +#if defined(CONFIG_X86_64) && !defined(CONFIG_UML)
>>       return cpu_data(first_cpu_of_numa_node).apicid;
>>   #else
>>       return first_cpu_of_numa_node;
>> --- linux-next-20220107.orig/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
>> +++ linux-next-20220107/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
>> @@ -1679,7 +1679,7 @@ static int kfd_fill_mem_info_for_cpu(int
>>       return 0;
>>   }
>>   -#ifdef CONFIG_X86_64
>> +#if defined(CONFIG_X86_64) && !defined(CONFIG_UML)
>>   static int kfd_fill_iolink_info_for_cpu(int numa_node_id, int *avail_size,
>>                   uint32_t *num_entries,
>>                   struct crat_subtype_iolink *sub_type_hdr)
>> @@ -1738,7 +1738,7 @@ static int kfd_create_vcrat_image_cpu(vo
>>       struct crat_subtype_generic *sub_type_hdr;
>>       int avail_size = *size;
>>       int numa_node_id;
>> -#ifdef CONFIG_X86_64
>> +#if defined(CONFIG_X86_64) && !defined(CONFIG_UML)
>>       uint32_t entries = 0;
>>   #endif
>>       int ret = 0;
>> @@ -1803,7 +1803,7 @@ static int kfd_create_vcrat_image_cpu(vo
>>               sub_type_hdr->length);
>>             /* Fill in Subtype: IO Link */
>> -#ifdef CONFIG_X86_64
>> +#if defined(CONFIG_X86_64) && !defined(CONFIG_UML)
>>           ret = kfd_fill_iolink_info_for_cpu(numa_node_id, &avail_size,
>>                   &entries,
>>                   (struct crat_subtype_iolink *)sub_type_hdr);
>>
>>

-- 
~Randy


More information about the amd-gfx mailing list