On 7/28/2010 6:25 PM, Paolo Bonzini wrote:
On 07/22/2010 08:36 PM, Charles Wilson wrote:
@@ -5777,11 +5812,13 @@ func_mode_link ()
for deplib in $dependency_libs; do
newdependency_libs="$deplib $newdependency_libs"
if $opt_preserve_dup_deps ; then
+ func_resolve_sysroot "$deplib"
+ deplib=$func_resolve_sysroot_result
case "$tmp_libs " in
*" $deplib "*) func_append specialdeplibs " $deplib" ;;
esac
+ func_append tmp_libs " $deplib"
fi
- func_append tmp_libs " $deplib"
done
if test "$link_all_deplibs" != no; then
Any reason why you moved the func_append within the if?
No, that's a mistake. The rebase was kinda difficult, because since last
July all the foo="$foo $other" stuff got replaced with func_append. So,
the before-and-after diffs were a little hard to interpret; I tried to
be pretty careful, but I'm not surprised that a mistake got thru.
Do you think this might be the reason for the new:041 failure?