[distcc] Re: issue fix proposal for distcc and intel C++ 7.1

[Martin Pool]

On  1 Jul 2003, Markus Werle <numerical.simulation at web.de> wrote:
> Hi!
> 
> After struggling a lot with the C programming language
> (yes, I am a pure C++ programmer and I think by now that
> these  two languages have nothing in common), I finally found a
> solution that makes intel C++, distcc _and_ autotools happy
> at once.
> 
> I introduced a new environment variable
> DISTCC_ADDITIONAL_REMOTE_COMPILER_FLAGS
> which I set to "-xc++"

[cc'd to list]

Alternatively, distcc could just *always* generate the appropriate -x
option depending on the input filename.   -x is supported since gcc
2.95 at least.

Doing so would remove the need for the server to set the extension of
the temporary filename.  That would be quite nice, because in the long
term I would like to move gcc-specific heuristics out of the server
into only the client.

Doing so would make distcc explicitly depend on gcc and similar
compilers, but in truth it is close to doing so already, and it would
be possible to add modes for other compilers.

-- 
Martin