Hi Richard,
> As discussed in the PR, let's take the opportunity while bumping the soname
> to add symbol versioning.
great idea: I'd already suggested this back in 2010 when doing the bulk
of the Solaris symbol versioning work
http://gcc.gnu.org/ml/gcc/2010-02/msg00339.html
http://sourceware.org/ml/libffi-discuss/2010/msg00045.html
but there never was a conclusion on the questions I'd raised.
> Versioning is complicated by the fact that there are several pieces of API
> that are "optional" based on the target. If these optional pieces are
> later implemented by targets that currently do not, we can't have these
> extra symbols suddenly appear in the base version -- that voids the promise
> that symbol versioning makes.
>
> So the symbols are split among 4 different versions. Each version's set of
> symbols is either entirely present[*] or entirely absent. Programs using
> the library will only depend on the subset of versions that they use.
>
> Tested on {x86_64,ppc64,aarch64}-linux. Compile tested on mips64el-linux,
> which doesn't implement go closures or complex types.
Unfortunately, the patch broke Solaris bootstrap:
make[2]: Entering directory
'/var/gcc/regression/trunk/12-gcc/build/i386-pc-solaris2.12/libffi'
Makefile:1906: *** missing separator (did you mean TAB instead of 8 spaces?).
Stop.
Fixed as obvious like this, will commit shortly:
2016-03-03 Rainer Orth <r...@cebitec.uni-bielefeld.de>
* Makefile.am (libffi.map-sun): Tabify:
* Makefile.in: Regenerate.
# HG changeset patch
# Parent 3195cd69f93aa48e3c342ea80e6f2660e30f33da
Tabify libffi/Makefile.am
diff --git a/libffi/Makefile.am b/libffi/Makefile.am
--- a/libffi/Makefile.am
+++ b/libffi/Makefile.am
@@ -214,11 +214,11 @@ libffi_version_script = -Wl,-M,libffi.ma
libffi_version_dep = libffi.map-sun
libffi.map-sun : libffi.map $(top_srcdir)/../contrib/make_sunver.pl \
$(libffi_la_OBJECTS) $(libffi_la_LIBADD)
- perl $(top_srcdir)/../contrib/make_sunver.pl libffi.map \
- $(libffi_la_OBJECTS:%.lo=.libs/%.o) \
- `echo $(libffi_la_LIBADD) | \
- sed 's,\([^/ ]*\)\.l\([ao]\),.libs/\1.\2,g'` \
- > $@ || (rm -f $@ ; exit 1)
+ perl $(top_srcdir)/../contrib/make_sunver.pl libffi.map \
+ $(libffi_la_OBJECTS:%.lo=.libs/%.o) \
+ `echo $(libffi_la_LIBADD) | \
+ sed 's,\([^/ ]*\)\.l\([ao]\),.libs/\1.\2,g'` \
+ > $@ || (rm -f $@ ; exit 1)
endif
else
libffi_version_script =
Unfortunately, even with this fixed, all Solaris/x86 tests now fail to
link:
FAIL: libffi.call/closure_fn0.c -W -Wall -Wno-psabi -O0 (test for excess errors)
Excess errors:
Undefined first referenced
symbol in file
ffi_closure_alloc /var/tmp//ccIpq3qc.o
ffi_type_float /var/tmp//ccIpq3qc.o
ffi_type_uint64 /var/tmp//ccIpq3qc.o
ffi_type_sint32 /var/tmp//ccIpq3qc.o
ffi_type_sint16 /var/tmp//ccIpq3qc.o
ffi_type_double /var/tmp//ccIpq3qc.o
ffi_prep_cif /var/tmp//ccIpq3qc.o
ld: fatal: symbol referencing errors
> For gcc7, we really should pull out these m4 macros to new config/ files.
> I didn't really want to touch anything except libffi for now.
Great idea: I'd meant to do this for a long time, but never got around
to it.
Rainer
--
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University
