Re: [PATCH 0/5] arm: set of big endian related fixes for armeb (v7)

[Victor Kamensky <victor dot kamensky at linaro dot org>]

Hi Yao,

Thank you for your help and attention! Please see response inline.

On 21 October 2014 00:35, Yao Qi <yao@codesourcery.com> wrote:
> Victor Kamensky <victor.kamensky@linaro.org> writes:
>
>> Note there are several failures remain that are not addressed by this
>> series. I.e valgrind related test cases all failed, because valgrind does
>> not support armeb target. There are other failures in a bit obscure
>> places like gdb.dwarf2/implptrpiece.exp, gdb.python/py-value-cc.exp,
>> but over all number of passes/failures significantly improved in armeb
>> case.
>
> Out of my curiosity, do you plan to address the rest of fails you
> mentioned above?

Not sure yet. Maybe on the second pass. Currently I am working
on aarch64 BE vs LE gdb testsuite result differences. There are few
issues in BE aarch64 case that affect gdb functionality visible to gdb
user.

I believe that remaining differences in ARM V7 BE test results most
likely are issues in test cases. And I am not sure whether it is more
general BE vs LE issues. I would probably give it a spin on BE ppc
and see how tests in question behave there.

Problems related to valgrind will not be addressed any time soon. It
seems it is huge thing to make valgrind to work in ARM V7 BE case.

Thanks,
Victor

> --
> Yao (éå)