OK.
Jason
Hi,
today I wanted to add a testcase and noticed again that in quite a few
cases we still use a second error call for the "previous
declaration"-type message. Thus the below, consisting mostly of
testsuite adjustments, the code changes themselves should be largely
obvious.
Tested x86_64-lin
OK.
Jason
Hi,
while working on the last patch I noticed a few more cases of warning_at
and permerror pairs where it makes sense to use inform.
Tested x86_64-linux.
Thanks,
Paolo.
///
/cp
2013-05-15 Paolo Carlini
* name-lookup.c (pushdecl_maybe_friend_1): Replace pai
