This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: 3 libstdc++ tests fail at random
- From: Jakub Jelinek <jakub at redhat dot com>
- To: Paolo Carlini <paolo dot carlini at oracle dot com>
- Cc: "H.J. Lu" <hjl dot tools at gmail dot com>, GCC Development <gcc at gcc dot gnu dot org>, libstdc++ <libstdc++ at gcc dot gnu dot org>, Jonathan Wakely <jwakely dot gcc at gmail dot com>
- Date: Sat, 7 Dec 2013 18:09:05 +0100
- Subject: Re: 3 libstdc++ tests fail at random
- Authentication-results: sourceware.org; auth=none
- References: <CAMe9rOoAVxu49za0BXeFXjy8sg2ifhE1gnnujFv5sG1ZD0nccQ at mail dot gmail dot com> <52A35020 dot 5090608 at oracle dot com>
- Reply-to: Jakub Jelinek <jakub at redhat dot com>
On Sat, Dec 07, 2013 at 05:43:12PM +0100, Paolo Carlini wrote:
> On 12/07/2013 04:48 PM, H.J. Lu wrote:
> >I have been seeing 3 libstdc++ tests:
> >
> >FAIL: 17_intro/headers/c++200x/stdc+
> >+.cc (test for excess errors)
> >FAIL: 17_intro/headers/c++200x/stdc++_multiple_inclusion.cc (test for
> >excess errors)
These two failed for me just once so far, in i686-linux build,
the errors were the same for both, perhaps related to PR58627?
Just a random guess, but until that PR is fixed, if the compiler is using
ggc_freed memory, really anything can happen.
tree check: expected tree_list, have error_mark in dump_template_bindings, at cp/error.c:274
_M_head(const _Head_base& __b) noexcept { return __b._M_head_impl; }
^
0x8bcc083 tree_check_failed(tree_node const*, char const*, int, char const*, ...)
../../gcc/tree.c:9190
0x831d349 tree_check
../../gcc/tree.h:2702
0x831d349 dump_template_bindings
../../gcc/cp/error.c:274
0x831d349 dump_substitution
../../gcc/cp/error.c:1386
0x831e3fd dump_function_decl
../../gcc/cp/error.c:1515
0x8320c19 decl_to_string
../../gcc/cp/error.c:2829
0x8320c19 cp_printer
../../gcc/cp/error.c:3430
0x92c7c89 pp_format(pretty_printer*, text_info*)
../../gcc/pretty-print.c:616
0x92c840d pp_format_verbatim(pretty_printer*, text_info*)
../../gcc/pretty-print.c:675
0x92c8471 pp_verbatim(pretty_printer*, char const*, ...)
../../gcc/pretty-print.c:871
0x8311530 print_instantiation_full_context
../../gcc/cp/error.c:3199
0x8311530 maybe_print_instantiation_context
../../gcc/cp/error.c:3343
0x831f4eb cp_diagnostic_starter
../../gcc/cp/error.c:3027
0x92c5c0a diagnostic_report_diagnostic(diagnostic_context*, diagnostic_info*)
../../gcc/diagnostic.c:791
0x92c63c9 internal_error(char const*, ...)
../../gcc/diagnostic.c:1129
0x8bcc083 tree_check_failed(tree_node const*, char const*, int, char const*, ...)
../../gcc/tree.c:9190
0x840928a tree_check
../../gcc/tree.h:2702
0x840928a decl_is_template_id
../../gcc/cp/mangle.c:291
0x840dba6 write_nested_name
../../gcc/cp/mangle.c:949
0x840e320 write_name
../../gcc/cp/mangle.c:866
> >FAIL: 30_threads/async/async.cc execution test
async.exe: /usr/src/gcc/libstdc++-v3/testsuite/30_threads/async/async.cc:66: void test02(): Assertion `status == std::future_status::ready' failed.
This one fails pretty frequently for me, 6 times out of last 12 i686-linux
bootstraps/regtests, starting with ~ Nov 21th bootstrap (but before that
I did one only on Nov 15th, so there is a big window). Neither of the above
ones were seen in x86_64 regtests (but, I'm not testing multilib, instead
doing two parallel full bootstraps and primary multilib regtests
(x86_64-linux and i686-linux)).
Jakub