[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944 --- Comment #3 from Georg-Johann Lay gjl at gcc dot gnu.org 2011-05-11 09:25:37 UTC --- Created attachment 24222 -- http://gcc.gnu.org/bugzilla/attachment.cgi?id=24222 pr48944.diff I get the following output from configure *** This configuration is not supported in the following subdirectories: target-libmudflap target-libgomp target-libssp target-libiberty target-libobjc (Any other directories should still work fine.) And the respective line in the toplevel Makefile reads: TARGET_CONFIGDIRS = libgcc libquadmath boehm-gc libffi zlib With the attached patch applied the result is as expected and stuff looks like indicated by Joseph.
[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944
--- Comment #4 from joseph at codesourcery dot com joseph at codesourcery dot
com 2011-05-11 10:23:53 UTC ---
On Wed, 11 May 2011, gjl at gcc dot gnu.org wrote:
I get the following output from configure
*** This configuration is not supported in the following subdirectories:
target-libmudflap target-libgomp target-libssp target-libiberty
target-libobjc
(Any other directories should still work fine.)
And the respective line in the toplevel Makefile reads:
TARGET_CONFIGDIRS = libgcc libquadmath boehm-gc libffi zlib
I think you'll need to debug why that happens. The toplevel configure
script has logic that looks at what runtime libraries are required by what
languages, and disables those that are only required by disabled languages
- and that logic works for me, and I haven't changed it at all. It should
see from java/config-lang.in and go/config-lang.in that libffi is only
needed by those languages, see that neither is being built, and disable
libffi independently of the target.
With the attached patch applied the result is as expected and stuff looks like
indicated by Joseph.
I don't think this is appropriate. A basic principle of my changes was
that disabling ${libgcj} is generally not the right thing to do,
especially based on the target architecture (rather than OS), and that any
new disabling should have a comment explaining *why* the feature is
disabled, and pointing to a relevant PR. Furthermore, for new disablings
it's better to set unsupported_languages instead of disabling the
libraries, unless you have some reason why it's useful to build the Java
compiler for this target without the libraries.
In this case, unsupported_languages is already set to include Java because
of the lack of libffi support, so no further change should be needed; as I
said, you need to debug why it doesn't work. It would not be wrong to add
noconfigdirs=$noconfigdirs target-libffi
in the same case that disables Java for targets not supporting libffi (see
my previous comments that you should be able to override
unsupported_languages with explicit --enable-languages configure options)
but there's still the question of why you're getting boehm-gc and zlib
built for your target.
[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944 --- Comment #5 from Georg-Johann Lay gjl at gcc dot gnu.org 2011-05-11 12:23:00 UTC --- I followed the recommendations in Optimize disk usage of http://gcc.gnu.org/wiki/SvnSetup and applied it to the following directories: libada gnattools gcc/ada gcc/testsuite/ada libstdc++-v3 libjava libgfortran libgo gcc/java gcc/go So the new configure setup don't support mapping non-needed subdirs to emptydir as the previous configure. After adding libffi boehm-gc zlib the build passes without errors, but the host zlib then is not build (configure adds --with-system-zlib to HOST_CONFIGARGS). Is there a list of compatible directory sets that can be used according to Optimize disk usage?
[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944 --- Comment #6 from joseph at codesourcery dot com joseph at codesourcery dot com 2011-05-11 13:08:05 UTC --- On Wed, 11 May 2011, gjl at gcc dot gnu.org wrote: I followed the recommendations in Optimize disk usage of http://gcc.gnu.org/wiki/SvnSetup and applied it to the following directories: libada gnattools gcc/ada gcc/testsuite/ada libstdc++-v3 libjava libgfortran libgo gcc/java gcc/go So the new configure setup don't support mapping non-needed subdirs to emptydir as the previous configure. Indeed, if you're removing front end directories you'll need to deal with removing library directories as well. Is there a list of compatible directory sets that can be used according to Optimize disk usage? Not that I know of. I think that disk usage optimization is pretty obsolete by now (and we similarly ought to eliminate the separate release/snapshot tarballs for particular languages, just having the main monolithic one) - and there are other ways removing front end directories could break things.
[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944 Georg-Johann Lay gjl at gcc dot gnu.org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||INVALID --- Comment #7 from Georg-Johann Lay gjl at gcc dot gnu.org 2011-05-11 17:40:15 UTC --- Closed as resolved+invalid because PR is an artifact of omitting some directories.
[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944 Georg-Johann Lay gjl at gcc dot gnu.org changed: What|Removed |Added Target||avr Priority|P3 |P2 Host||i586-suse-linux Build||i586-suse-linux Severity|normal |blocker
[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944 Richard Guenther rguenth at gcc dot gnu.org changed: What|Removed |Added CC||jsm28 at gcc dot gnu.org --- Comment #1 from Richard Guenther rguenth at gcc dot gnu.org 2011-05-10 10:30:05 UTC --- Sounds like some of the toplevel configure re-org may have caused this? C and C++ should not result in libffi being built (I think).
[Bug libffi/48944] build error: libffi has not been ported to avr-unknown-none.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48944 --- Comment #2 from joseph at codesourcery dot com joseph at codesourcery dot com 2011-05-10 15:10:57 UTC --- When I configure for this target I get: *** This configuration is not supported in the following subdirectories: target-libmudflap target-libgomp target-libssp target-libiberty target-libstdc++-v3 gnattools target-libada target-libgfortran target-libgo target-libffi target-zlib target-libjava target-libobjc target-boehm-gc (Any other directories should still work fine.) and then the generated Makefile contains: TARGET_CONFIGDIRS = libgcc libquadmath Do you get some different list of unsupported directories (not including libffi) and of directories to build (including libffi), or do you get the same lists that I do but still find it tries to build libffi later?
