[RFC] Do we care about binary compatibility of code produced by cross-compilers?

[Paolo Carlini]

Paolo Bonzini wrote:
>> ... actually, if all the checks presently doing link-tests consistently
>> get a parameter allowing to disable the test completely and also to set
>> its value, I'm thinking with proper documentation nobody could be
>> *really* unhappy with the change... Do you agree?
>>     
> The cache file is a way to set the value of a test.
>   
Sorry, but I'm not sure to understand how your sentence relates to my 
question. Can you expand a bit about that? To be more clear on my side, 
besides the first two or three sentences of my previous message, I left 
the issue about running GCC_CHECK_TLS alone. In the sentences above I 
was just describing the effect of moving checks possibly doing 
link-tests to the set of tests run unconditionally, thus even when 
non-native,
*and* at the same time making sure that all those tests can be run with 
three possible arguments (the usual ones): yes, no, auto.

Is that a kind of setup which we would find better than the current one, 
with tests possibly doing link-tests run only when native and 
hard-coding for crosses? Any objections / willingness to help to the 
idea of changing GCC_CHECK_UNWIND_GETIPINFO and AM_ICONV (-> GCC_ICONV) 
consistently with above?

Paolo.