Re: [PATCH v3 3/3] microblaze: generate uapi header and system call table files

[Michal Simek]

On 08. 11. 18 14:44, Arnd Bergmann wrote:
> On Mon, Nov 5, 2018 at 11:26 AM Firoz Khan <firoz.khan@xxxxxxxxxx> wrote:
>> On Mon, 5 Nov 2018 at 14:46, Michal Simek <monstr@xxxxxxxxx> wrote:
>>> On 26. 10. 18 7:09, Firoz Khan wrote:
> 
>>>> -     .long sys_rseq
>>>> +#include <asm/syscall_table.h>
>>>> +#undef __SYSCALL
>>>>
>>>
>>> I still don't think this is correct.
>>>
>>> I have done these steps
>>>
>>> make mrproper
>>> make ARCH=microblaze headers_install INSTALL_HDR_PATH=/tmp/ && move
>>> Applied 1/3
>>> make ARCH=microblaze headers_install INSTALL_HDR_PATH=/tmp/ && move
>>> Applied 2/3
>>> make ARCH=microblaze headers_install INSTALL_HDR_PATH=/tmp/ && move
>>> Applied 3/3
>>> make ARCH=microblaze headers_install INSTALL_HDR_PATH=/tmp/ && move
>>>
>>> And compare all these versions. Till 2/3 all is fine but when 3/3 is
>>> applied
>>>
>>> include/asm/unistd.h contains
>>> #include <asm/unistd_32.h>
>>>
>>> Then include/asm/unistd_32.h contains
>>> #include <asm-generic/unistd_32.h>
>>> which is different syscalls table.
>>>
>>> I didn't play with it for a long time but removing generic unistd_32.h
>>> is fixing it. Please take a look.
>>
>> Sure, I'll check and update you ASAP.
>>
>>>
>>> diff --git a/arch/microblaze/include/uapi/asm/Kbuild
>>> b/arch/microblaze/include/uapi/asm/Kbuild
>>> index cda864f4b62e..2c6a6bffea32 100644
>>> --- a/arch/microblaze/include/uapi/asm/Kbuild
>>> +++ b/arch/microblaze/include/uapi/asm/Kbuild
>>> @@ -28,4 +28,3 @@ generic-y += termbits.h
>>>  generic-y += termios.h
>>>  generic-y += types.h
>>>  generic-y += ucontext.h
>>> -generic-y += unistd_32.h
>>
>> AFAIK, If I apply the above fix; the generated file - unistd_32.h will
>> get removed.
>> No longer present in arch/microblaze/include/generated/uapi/asm.
> 
> It certainly must be 'generated-y', not 'generic-y', since you are
> generating an architecture specific file, not taking the one
> from asm-generic. I think this was correct in an earlier version.

yep. It was correct in v1. It has changed in v2.

I have retested and it looks good with that.

Thanks,
Michal


-- 
Michal Simek, Ing. (M.Eng), OpenPGP -> KeyID: FE3D1F91
w: www.monstr.eu p: +42-0-721842854
Maintainer of Linux kernel - Xilinx Microblaze
Maintainer of Linux kernel - Xilinx Zynq ARM and ZynqMP ARM64 SoCs
U-Boot custodian - Xilinx Microblaze/Zynq/ZynqMP SoCs

Attachment: signature.asc
Description: OpenPGP digital signature