Re: [Xen-devel] [xen-unstable-smoke test] 114005: regressions - trouble: blocked/fail

[Bhupinder Thakur <bhupinder.thakur@xxxxxxxxxx>]

Hi,

On 4 October 2017 at 17:56, Julien Grall <julien.grall@xxxxxxx> wrote:
>
>
> On 04/10/17 12:47, osstest service owner wrote:
>> flight 114005 xen-unstable-smoke real [real]
>> http://logs.test-lab.xenproject.org/osstest/logs/114005/
>>
>> Regressions :-(
>>
>> Tests which did not succeed and are blocking,
>> including tests which could not be run:
>>   build-amd64                   6 xen-build                fail REGR. vs. 
>> 113972
>
> I think this was the error discussed on IRC:
>
> /home/osstest/build.114005.build-amd64/xen/stubdom/cross-root-x86_64/x86_64-xen-elf/include/string.h:24:8:
>  error: conflicting types for 'memcpy'
>  _PTR   _EXFUN(memcpy,(_PTR, const _PTR, size_t));
>         ^
> In file included from 
> /home/osstest/build.114005.build-amd64/xen/extras/mini-os-remote/include/console.h:44:0,
>                  from main.c:11:
> /home/osstest/build.114005.build-amd64/xen/stubdom/include/xen/io/console.h:44:485:
>  note: previous implicit declaration of 'memcpy' was here
>  DEFINE_XEN_FLEX_RING(xencons);
>
> I am not sure how to fix that one.
>
It seems that grub/main.c does not include string.h before including
console.h which could be causing this implicit declaration. Ideally,
ring.h should include string.h explicitly as it depends on string.h.

>
>>   build-armhf                   6 xen-build                fail REGR. vs. 
>> 113972
>
> daemon/io.c: In function 'console_evtchn_unmask':
> daemon/io.c:1050:18: error: cast from pointer to integer of different size 
> [-Werror=pointer-to-int-cast]
>   long long now = (long long)data;
>                   ^
> daemon/io.c: In function 'handle_io':
> daemon/io.c:1346:53: error: cast to pointer from integer of different size 
> [-Werror=int-to-pointer-cast]
>     console_iter_void_arg2(d, console_evtchn_unmask, (void *)now);
>
>                                                      ^
>
> It seems that 32-bit build has been built test it. I will send a patch for it.
>
Instead of using a generic void *, I can use the long long type
itself. I think there was a review comment regarding that.

Regards,
Bhupinder

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel