Jason Merrill <ja...@redhat.com> wrote:
On 11/3/19 6:42 AM, Kamlesh Kumar wrote:
ChangeLog Entries:
-----------
2019-11-02 Kamlesh Kumar <kamleshbha...@gmail.com>
* cp-demangle.c (d_expr_primary): Handle
nullptr demangling.
* testsuite/demangle-expected: Added test.
This test is failing in at least some places (x86_64-darwin, x86_64-linux)
x86_64-apple-darwin16-gcc -DHAVE_CONFIG_H -g -O2 -I..
-I/src-local/gcc-master/libiberty/testsuite/../../include -o test-demangle
\
/src-local/gcc-master/libiberty/testsuite/test-demangle.c ../libiberty.a
./test-demangle < /src-local/gcc-master/libiberty/testsuite/demangle-expected
FAIL at line 1452, options :
in: _Z3fooILPv0EEvPN9enable_ifIXeqT_LDnEEvE4typeE
out: void foo<(void*)0>(enable_if<((void*)0)==(decltype(nullptr)),
void>::type*)
exp: void foo<(void*)0>(enable_if<((void*)0)==((decltype(nullptr))),
void>::type*)
./test-demangle: 335 tests, 1 failures
Is the extra set of parentheses significant in your expected output?
if so, then the demangling is not working …
… if not, we should amend the test since it causes early failure of
libiberty tests
(I guess this goes unnoticed, because they are not integrated into the
headline
GCC regression tests).
thanks
Iain
