syscall statx not implemented on powerpc

[Michael Ellerman]

Chris Packham <Chris.Packham at alliedtelesis.co.nz> writes:

> On 15/03/17 00:57, Michael Ellerman wrote:
>> Chris Packham <Chris.Packham at alliedtelesis.co.nz> writes:
>>> On 13/03/17 21:52, Chandan Rajendra wrote:
>>>> On Monday, March 13, 2017 03:33:07 AM Chris Packham wrote:
>>>>> I've just attempted to build a powerpc kernel from 4.11-rc2 using a
>>>>> custom defconfig (available on request) and I'm hitting the following
>>>>> error in the early stages of compilation.
>>>>>
>>>>> <stdin>:1325:2: error: #warning syscall statx not implemented [-Werror=cpp]
>>>>>
>>>>> Same thing seems to happen with mpc85xx_basic_defconfig.
>>>>>
>>>>> I don't actually need this syscall so I'd be happy to turn something off
>>>>> to get things building. I did a quick search and couldn't see anything
>>>>> on linuxppc-dev but google keeps correcting "statx" to "stats" so I
>>>>> could have missed it.
>>>>
>>>> The upstream commit
>>>> (https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=a528d35e8bfcc521d7cb70aaf03e1bd296c8493f)
>>>> that introduces the statx syscall provides a test program. I will wire-up the
>>>> syscall on ppc64, run that test program and post the patch if the test program
>>>> works well.
>>>
>>> Thanks, I'd be happy to test a patch here.
>>>
>>> In the meantime I worked around the build issue by adding __INGORE_statx
>>> to checksyscalls.sh.
>>
>> Is it actually breaking the build? It's meant to be a warning, and
>> that's all I get.
>
> It breaks my builds for ppc and arm. We do have KCFLAGS=-Werror in our 
> environment but KCPPFLAGS= is not set so I'm not sure why it's stopping 
> compilation.

OK, I guess that must be it.

We turn on -Werror for all of arch/powerpc by default (unless
CONFIG_PPC_DISABLE_WERROR=y), but the syscall checking script is not
under arch/powerpc so shouldn't be affected by that.

cheers