Re: [Qemu-devel] [PATCH for-2.0 V3] tests/acpi-test: do not run iasl on big endian machines

[Alexey Kardashevskiy]

On 03/24/2014 10:02 PM, Andreas Färber wrote:
> Am 23.03.2014 10:49, schrieb Michael S. Tsirkin:
>> On Fri, Mar 21, 2014 at 12:16:53AM +0100, Paolo Bonzini wrote:
>>> Il 20/03/2014 23:33, Marcel Apfelbaum ha scritto:
>>>> I've seen something like that somewhere, but I didn't quite like it.
>>>> I was looking for something more elegant as I was *almost* sure
>>>> this kind of solution will not pass the reviews :)
>>>>
>>>> But maybe I'll try this, let's see what happens,
>>>
>>> If all you're looking for is bigendian (disabling iasl disassembly
>>> on bigendian makes sense), your patch v2 is fine.
>>>
>>> Assembling ASL on bigendian is supported by at least Fedora and
>>> Debian (and hence Ubuntu).
>>>
>>> Paolo
>>
>>
>> At this point I'm confused.
>> If iasl compiler is broken, we should detect and fix that.
>> It might be ok to just detect endian-ness as a quick work-around.
>> BTW configure already has code to detect endian-ness:
>> if test "$bigendian" = "yes" ; then
>>   echo "HOST_WORDS_BIGENDIAN=y" >> $config_host_mak
>> fi
> 
> Careful, this is about the endianness of the built target binary, which
> may be different from the endianness of the build system. However I
> would hope the tests will not be executed for cross-builds.
> 
> Alexey, you're using cross-builds for ppc, right? Have you ever tested
> running make check there?


"make check" on what? ppc64-softmmu target compiled on x86_64? That fails
because of errors like this:

tests/check-qjson: tests/check-qjson: cannot execute binary file
make: *** [check-tests/check-qstring] Error 1



-- 
Alexey