Re: Toplevel cleanup: disable Java when libffi not supported

[Andrew Haley <aph at redhat dot com>]

On 04/27/2011 05:33 PM, Joseph S. Myers wrote:
> This patch, relative to a tree with
> <http://gcc.gnu.org/ml/gcc-patches/2011-04/msg02123.html> applied,
> continues the cleanup of toplevel cases relating to disabling Java or
> Java libraries by arranging for Java to be disabled (via
> unsupported_languages) on targets not supporting libffi, and removing
> many cases in the Java-disabling case statement that are no longer
> needed, either because they relate to targets not supporting libffi or
> because they relate to targets no longer supported in GCC so it
> doesn't matter what the Java configuration is for them.
> 
> The ideal would be for information about libffi-supported targets to
> come from a configure fragment in the libffi/ directory, as per item 4
> in <http://gcc.gnu.org/ml/gcc/2011-03/msg00486.html>.  Given that at
> present there is some divergence between the libffi repository and the
> GCC copy of libffi, however, now may not be the best time to add such
> a fragment in GCC; instead, I put a minimal case statement (covering
> only targets supported by both libffi and GCC) at toplevel.
> 
> Where Java-disabling-related cases did more than just disabling Java
> or ${libgcj} (for example, explicitly disabling target-boehm-gc or
> target-libffi) I left them in for now even if not strictly needed with
> the separate libffi case statement.
> 
> This patch is intended to be safe, in that the only way it should
> change behavior for targets supported by GCC is to disable Java in
> some cases where the libraries would not build because of lack of
> libffi support (whether or not the libraries and front end were
> previously enabled in such cases).
> 
> OK to commit?

That looks like a good improvement.  Fine by me.

Andrew.