[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Lightning] Many tests failed on x86_32 systems
From: |
Paulo César Pereira de Andrade |
Subject: |
Re: [Lightning] Many tests failed on x86_32 systems |
Date: |
Fri, 15 Sep 2017 13:26:13 -0400 |
2017-09-15 12:31 GMT-04:00 Antonio Trande <address@hidden>:
[...]
>>>> It fails with default rpm %optflags due to generating invalid code, when
>>>> not properly updating the pointer where lightning writes instructions to,
>>>> and generating invalid opcodes.
>>> I will disable testing.
>>> On koji, they tend to fail.
>>
>> Please keep the tests during build. I will help with any failure.
>> If it were not for the tests, it would not be noticed the need for
>> -fno-strict-aliasing in CFLAGS.
>>
>
> Okay; it's fine using
>
> for libs:
> make %{?_smp_mflags} V=1 CFLAGS="%{optflags} -fno-strict-aliasing"
>
> for tests:
> make %{?_smp_mflags} V=1 CFLAGS="-g -O2 -fPIC -fno-strict-aliasing"
>
> ?
Yes. It is fine. I will check later if it is a real strict aliasing
issue (most
likely is), but at first I noticed only "jit_bnei" implementation incorrect.
> --
> Antonio Trande
> sagitter AT fedoraproject dot org
> See my vCard.
Thanks,
Paulo