Wed Nov 6 08:27:00 GMT 2002
>>>>> "Ranjit" == Ranjit Mathew <firstname.lastname@example.org> writes:
>> Suppose you have a unified build tree where newlib is built as part of
>> the tree. In this situation, newlib is a target library, a sibling of
>> libgcj. In this case, at the time that the libgcj configure script
>> runs, newlib hasn't yet been built, and so link tests will fail.
Ranjit> From looking at the top-level GCC Makefile, it seems that
Ranjit> newlib will be built before libjava - though it's not a direct
Ranjit> dependency, isn't it assured that newlib would be built before
The problem is that all the target configure scripts are run before
any target libraries are built. So when libjava is configured, newlib
has been configured but not built.
Ranjit> Post GCC 3.2, this conditional now reads "with_newlib" instead
Ranjit> of "with_cross_host" - this even more incorrect. IMHO, we should
Ranjit> stick with the conditional for "with_cross_host".
Great! Write up a clean patch to use this variable instead, and I'll
put it in.
Ranjit> 1. The GCC toplevel configure says --with-cross-host is deprecated
Ranjit> and target libraries should use some other mechanism, but doesn't
Ranjit> suggest any alternative. What's wrong with using this flag?
Ranjit> Does this sound OK? Did I miss out anything? (apart from
Ranjit> the details that is)
I think this summary is ok. It seems like the only real change we
need is to special-case newlib, and have other crosses ("host-x-host"
crosses) do compile- and link- tests.
More information about the Java