On 04/28/2011 06:17 PM, Joseph S. Myers wrote:
This patch cleans up the libgcj-disabling case statement at toplevel,
removing the default disabling for *-*-* (which allows other empty
cases to be removed) and disabling based on architecture, or most
cases of architecture-OS combination where the OS previously had
libgcj built for other architectures.
I left the logic restricting libgcj on Darwin to certain targets, but
removed that relating to FreeBSD and NetBSD targets (given the removal
of support for a.out NetBSD, and that there was never such support for
a.out FreeBSD, I doubt the reliability of the lists of targets that
were present). target-libffi disabling for arm-*-elf* | arm*-*-eabi*
was removed (given that there*is* libffi support for those targets),
and target-zlib disabling for hppa*64*-*-linux* was changed to disable
Java.
This is the last cleanup I plan for this libgcj-disabling logic. This
patch is*not* conservatively safe; the idea is that if it turns out
that there are problems building some Java library on a
libffi-supporting target where the library was not previously built,
appropriate disabling should be added *with a comment explaining the
problem and probably pointing to a relevant PR*.
OK to commit?
Ok.
Paolo
