Re: patch for boost macro

On 08/04/15 15:17, Peter Simons wrote:
> Hi Jaap,
> 
>  > First, also when 'x86_64' is detected as host architecture/cpu then
>  > i386 should be used as fallback architecture.
> 
> why do you think that should be so? Personally, I'd be rather unhappy if
> the macro would pick up a set of libraries that don't match the platform
> I'm building for?

I'm not sure if you are referring to the difference between host/target
architecture. (Indeed, I suppose host should be replaced by target
architecture. I'm simply not used to thinking about cross-compiling.)

Alternatively, regarding the issue of i386 vs. x86_64: I'm running
Debian i386 userland with an amd64 kernel on x86_64 hardware. The
autoconf macro detected my host architecture as x86_64 and thus could
not find my native i386 boost libraries.

My patch is meant to fix precisely this issue by searching for i386 as a
fallback when x86_64 is detected, but maybe I'm doing it the wrong way
and there is a method to detect the userland architecture?

Best,
Jaap