mthom
commented
3 years ago I get the same crash.
Closed pmoura closed 1 year ago
mthom
commented
3 years ago I get the same crash.
pmoura
commented
3 years ago The crash with the searching example is similar:
thread 'main' panicked at 'internal error: entered unreachable code', src/machine/compile.rs:728:17
stack backtrace:
0: 0x103bb34fe - <std::sys_common::backtrace::_print::DisplayBacktrace as core::fmt::Display>::fmt::h04abbaabf148650a
1: 0x103bef83e - core::fmt::write::h14dac7cadec1cc70
2: 0x103bb2e1a - std::io::Write::write_fmt::hfaf2e10dfdcc61d8
3: 0x103bd0369 - std::panicking::default_hook::{{closure}}::h350fee0bf60f2674
4: 0x103bcfedb - std::panicking::default_hook::h0b4e3bc46e6dcb8d
5: 0x103bd08fa - std::panicking::rust_panic_with_hook::h8cdc0a575f4a5a7b
6: 0x103bb3c99 - std::panicking::begin_panic_handler::{{closure}}::h7a7b30fd1c313876
7: 0x103bb3648 - std::sys_common::backtrace::__rust_end_short_backtrace::h2e099be83c81509d
8: 0x103bd0473 - _rust_begin_unwind
9: 0x103c0458f - core::panicking::panic_fmt::h4f63e4f6b62b650b
10: 0x103c044e7 - core::panicking::panic::hb3c06f51e2e58141
11: 0x10394528b - scryer_prolog::machine::compile::thread_choice_instr_at_to::h5ace4fd13a0b3d92
12: 0x10394796a - scryer_prolog::machine::compile::append_compiled_clause::h3b427530b04d913a
13: 0x10382106d - scryer_prolog::machine::compile::<impl scryer_prolog::machine::load_state::LoadState>::incremental_compile_clause::hbb760f797a9d1334
14: 0x10387d208 - scryer_prolog::machine::loader::<impl scryer_prolog::machine::Machine>::compile_assert::h4c72d8054cc4a1e3
15: 0x103885bcc - scryer_prolog::machine::Machine::handle_toplevel_command::hcbc4d2457d98f1d8
16: 0x10387f03f - scryer_prolog::machine::Machine::run_module_predicate::h665a482ee83b04f2
17: 0x10387fdc0 - scryer_prolog::machine::Machine::run_top_level::h98ad3e5a02b065d9
18: 0x103755f5e - scryer_prolog::main::hd9ecefa6a4ac075a
19: 0x10375a3ca - std::sys_common::backtrace::__rust_begin_short_backtrace::h1cd4f2d4668f330b
20: 0x10375a3ec - std::rt::lang_start::{{closure}}::hd97e1ceb299400ab
21: 0x103bd50a0 - std::rt::lang_start_internal::hf4b96bfc8c02c8b0
22: 0x1037560c9 - _main{lgtunit(loader)} appears to work after the latest commit but {searching(loader)} just hangs also seems to load correctly now.
pmoura
commented
3 years ago Same results here: both lgtunit and searching load fine. But trying to run the tests for the searching example (using the goal {searching(tester)}) crashes. Same with tests for a smaller example. E.g.
?- {ack(tester)}.
failing code:
1984080 | dynamic_else 48338, inf, 142
1984081 | execute $lgt_ctg_super_call_nv/64039/3, 0
1984082 | retry_me_else 80
1984083 | get_structure $lgt_ctg_super_call_/3, X1
1984084 | unify_variable X3
1984085 | unify_variable X4
1984086 | unify_variable X5
1984087 | get_structure $lgt_ctg_super_call_nv/3, X2
1984088 | unify_value X3
1984089 | unify_value X4
1984090 | unify_value X5
1984091 | proceed
1984092 | dynamic_else 48339, inf, 0
1984093 | switch_on_term 1, IndexingCodePtr::DynamicExternal(17), IndexingCodePtr::Internal(1), IndexingCodePtr::Fail, IndexingCodePtr::Failswitch_on_constant 2dynamic(2)dynamic(142)
1984094 | dynamic_internal_else 48339, 48343, fail(0)
1984095 | get_constant user, A1
1984096 | get_value X2, A4
1984097 | get_constant true, A5
1984098 | execute true/5/0, 0
1984099 | retry_me_else 63
1984100 | get_structure $lgt_db_lookup_cache_/5, X1
1984101 | unify_constant user
1984102 | unify_variable X3
1984103 | unify_void 1
1984104 | unify_value X3
1984105 | unify_constant true
1984106 | get_constant true, A2
1984107 | proceed
1984108 | dynamic_else 48340, inf, 0
1984109 | rev_jmp_by 16
1984110 | dynamic_internal_else 48340, inf, 10
1984111 | get_constant tests, A1
1984112 | get_structure passed_/1, X2
1984113 | unify_variable X6
1984114 | get_constant lgtunit, A3
1984115 | get_structure $tests#0.passed_#1/2, X4
1984116 | unify_value X6
1984117 | unify_void 1
1984118 | get_constant true, A5
1984119 | execute true/5/0, 0
1984120 | rev_jmp_by 26
1984121 | rev_jmp_by 0
1984122 | try_me_else 0
1984123 | rev_jmp_by 1953401
1984124 | retry_me_else 14
1984125 | get_structure $lgt_db_lookup_cache_/5, X1
1984126 | unify_constant tests
1984127 | unify_variable X3
1984128 | unify_constant lgtunit
1984129 | unify_variable X4
1984130 | unify_constant true
1984131 | get_constant true, A2
1984132 | get_structure passed_/1, X3
1984133 | unify_variable X5
1984134 | get_structure $tests#0.passed_#1/2, X4
1984135 | unify_value X5
1984136 | unify_void 1
1984137 | proceed
1984138 | rev_jmp_by 152
1984139 | rev_jmp_by 0
1984140 | dynamic_else 48341, inf, 0
1984141 | switch_on_term 1, IndexingCodePtr::DynamicExternal(1), IndexingCodePtr::Internal(1), IndexingCodePtr::Fail, IndexingCodePtr::Failswitch_on_constant 3
1984142 | dynamic_internal_else 48341, inf, 0
1984143 | get_constant 0, A1
1984144 | execute true/5/0, 0
1984145 | try_me_else 0
1984146 | rev_jmp_by 1953424
1984147 | try_me_else 6
1984148 | get_structure $tests#0.passed_#1/2, X1
1984149 | unify_constant 0
1984150 | unify_void 1
1984151 | get_constant true, A2
1984152 | proceed
1984153 | rev_jmp_by 29
1984154 | rev_jmp_by 0
1984155 | dynamic_internal_else 48342, inf, fail(0)
1984156 | get_structure failed_/1, X1
1984157 | unify_variable X4
1984158 | get_structure $tests#0.failed_#1/2, X3
1984159 | unify_value X4
thread 'main' panicked at 'internal error: entered unreachable code', src/machine/compile.rs:742:17
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
✘-101 ~ With that latest fix, I'm able to run the ack example tests both manually and automated:
?- {ack(tester)}.
%
% tests started at 0-00-00, 00:00:00
%
% running tests from object tests
% file: /Users/pmoura/logtalk/examples/ack/tests.lgt
%
% ack_1: success (in 0.0 seconds)
% ack_2: success (in 0.0 seconds)
% ack_3: success (in 0.0 seconds)
%
% 3 tests: 0 skipped, 3 passed, 0 failed
% completed tests from object tests
%
%
% clause coverage ratio and covered clauses per entity predicate
%
% ack: ack/3 - 3/3 - (all)
% ack: 3 out of 3 clauses covered, 100.0% coverage
%
% 1 entity declared as covered containing 3 clauses
% 1 out of 1 entity covered, 100.0% entity coverage
% 3 out of 3 clauses covered, 100.0% clause coverage
%
% tests ended at 0-00-00, 00:00:00
%
true
; ...
?-
$ cd logtalk/examples/ack/
$ logtalk_tester -p scryer
% Batch testing started @ 2021-05-05 20:47:28
% Logtalk version: 3.46.0-stable
% Scryer Prolog version: 0.8.128
%
% logtalk/examples/ack
% 3 tests: 0 skipped, 3 passed, 0 failed
% completed tests from object tests
% clause coverage 100.0%
%
% 1 test sets: 1 completed, 0 skipped, 0 broken, 0 timeouts, 0 crashes
% 3 tests: 0 skipped, 3 passed, 0 failed
%
% Batch testing ended @ 2021-05-05 20:49:04This is promising 🙂 But current Logtalk support for Scryer Prolog is a bit of a hack due to missing standard (ISO or de facto) predicates but specially a much more complete operating-system/file-system interface. I will open later a separate issue for the later.
mthom
commented
3 years ago I noticed some other issues with Logtalk on Scryer that are better addressed after I finish the new heap representation and add GC, since many of the currently problematic parts are being revamped anyway.
pmoura
commented
3 years ago Running all tests in the Logtalk distribution, I get:
$ cd logtalk
$ logtalk_tester -p scryer -g "set_logtalk_flag(clean,off)" -w -t 360
...
% 509 test sets: 374 completed, 93 skipped, 26 broken, 9 timeouts, 7 crashes
% 7919 tests: 2496 skipped, 4558 passed, 865 failedMost of the test failures are caused by a small number of common issues. I will try to isolate and report them.
pmoura
commented
3 years ago One of the crashes is with the random library tests:
$ scryerlgt
...
?- {random(tester)}.
%
% tests started at 0-00-00, 00:00:00
%
% running tests from object tests(random)
% file: /Users/pmoura/logtalk/library/random/tests.lgt
%
% random_random_1_01: success (in 0.0 seconds)
% random_between_3_01: success (in 0.0 seconds)
% random_between_3_02: success (in 0.0 seconds)
% random_member_2_01: success (in 0.0 seconds)
% random_select_3_01: success (in 0.0 seconds)
! random_enumerate_2_01: failure (in 0.0 seconds)
! quick check test failure (at test 1 after 0 shrinks with starting seed seed(3172,9814,20125)): random_enumerate_2([5,9,2,3,1,8,4,0,7,6,9,8,4,1,2,3,5,0,7,6,4,1,7,2,3,8,0,6,9,5,4,2,3,9,6,5,8,1,0,7,6,1,8,2,7,4,9,5,0,3,1,7,4,0,9,6,2,5,8,3,6,5,9,8,7,2,3,0,1,4,5,0,4,1,6,7,3,9,8,2,4,6,0,2,8,1,5,3,7,9,4,6,0,9,1,5,8,7,3,2,8,9,3,6,0,4,5,2,7,1,3,7,6,0,4,9,5,2,1,8,5,2,8,0,1,6,7,4,3,9,4,1,2,3,5,6,8,0,7,9,5,3,2,0,6,9,1,8,7,4,0,1,8,3,5,4,6,7,9,2,2,0,5,7,1,8,6,4,3,9,4,1,6,3,0,8,7,2,5,9,6,7,4,5,3,9,2,8,1,0,8,3,9,0,1,2,7,4,6,5,9,6,1,5,8,7,0,2,3,4,5,3,1,2,4,9,6,8,7,0,3,5,9,4,1,0,6,2,8,7,2,9,5,1,0,4,8,7,6,3,0,4,7,3,8,1,2,9,6,5,9,8,5,2,3,6,1,7,4,0,2,0,8,4,3,9,1,6,7,5,2,0,3,6,4,9,5,8,1,7,5,6,9,7,2,8,3,1,4,0,1,6,9,8,3,7,0,4,5,2,6,9,5,8,2,7,1,3,0,4,8,0,9,6,3,5,7,1,2,4,2,8,7,1,3,0,5,4,9,6,8,1,0,6,7,9,3,2,5,4,5,3,1,2,9,0,4,7,8,6,1,4,6,5,3,0,9,2,7,8,1,3,7,4,8,0,5,2,6,9,4,0,2,8,7,6,1,5,3,9,2,4,5,3,6,8,9,7,1,0,7,9,8,5,3,1,4,0,2,6,5,9,2,0,3,6,8,1,7,4,3,7,5,2,9,4,1,0,8,6,2,6,4,0,1,3,7,9,8,5,3,4,6,8,0,7,5,2,1,9,3,1,8,7,0,9,2,5,6,4,9,8,2,1,6,0,4,3,7,5,8,5,6,9,2,0,3,7,1,4,4,8,3,6,2,0,9,5,7,1,0,1,9,3,7,8,5,6,4,2,0,6,1,8,7,5,9,2,4,3,9,2,7,8,0,1,6,4,3,5,4,3,2,5,8,1,9,0,7,6,4,5,9,2,7,3,6,8,0,1,4,2,9,3,5,0,6,8,1,7,2,7,0,6,1,5,8,9,4,3,4,6,8,9,1,3,7,5,0,2,5,3,6,0,2,4,9,8,1,7])
! in file /Users/pmoura/logtalk/library/random/tests.lgt at or above line -1
% random_permutation_2_01: success (in 0.0 seconds)
% random_sequence_4_01: success (in 0.0 seconds)
% random_sequence_4_02: success (in 0.0 seconds)
% random_sequence_4_03: success (in 0.0 seconds)
% random_sequence_4_04: success (in 0.0 seconds)
% random_set_4_01: success (in 0.0 seconds)
% random_set_4_02: success (in 0.0 seconds)
% random_set_4_03: success (in 0.0 seconds)
% random_set_4_04: success (in 0.0 seconds)
% random_set_4_05: success (in 0.0 seconds)
% random_set_4_06: success (in 0.0 seconds)
% random_get_seed_1_01: success (in 0.0 seconds)
% random_set_seed_1_01: success (in 0.0 seconds)
thread 'main' panicked at 'called `Result::unwrap()` on an `Err` value: other (not out-of-memory)', /Users/pmoura/.cargo/registry/src/github.com-1ecc6299db9ec823/slice-deque-0.3.0/src/lib.rs:571:38
stack backtrace:
0: 0x1083854fe - <std::sys_common::backtrace::_print::DisplayBacktrace as core::fmt::Display>::fmt::h04abbaabf148650a
1: 0x1083c183e - core::fmt::write::h14dac7cadec1cc70
2: 0x108384e1a - std::io::Write::write_fmt::hfaf2e10dfdcc61d8
3: 0x1083a2369 - std::panicking::default_hook::{{closure}}::h350fee0bf60f2674
4: 0x1083a1edb - std::panicking::default_hook::h0b4e3bc46e6dcb8d
5: 0x1083a28fa - std::panicking::rust_panic_with_hook::h8cdc0a575f4a5a7b
6: 0x108385cc5 - std::panicking::begin_panic_handler::{{closure}}::h7a7b30fd1c313876
7: 0x108385648 - std::sys_common::backtrace::__rust_end_short_backtrace::h2e099be83c81509d
8: 0x1083a2473 - _rust_begin_unwind
9: 0x1083d658f - core::panicking::panic_fmt::h4f63e4f6b62b650b
10: 0x1083d60f5 - core::result::unwrap_failed::hc2a56d017349147e
11: 0x107fea192 - <slice_deque::SliceDeque<T> as slice_deque::SpecExtend<T,I>>::spec_extend::hddbbbd4bff81b2af
12: 0x108117b6a - scryer_prolog::machine::compile::prepend_compiled_clause::h4d54ac4a4038ba43
13: 0x107ff3758 - scryer_prolog::machine::compile::<impl scryer_prolog::machine::load_state::LoadState>::incremental_compile_clause::hbb760f797a9d1334
14: 0x1080dd038 - scryer_prolog::machine::compile::<impl scryer_prolog::machine::loader::Loader<TS>>::compile_clause_clauses::h5ad13622eaec9ac1
15: 0x10804f466 - scryer_prolog::machine::loader::<impl scryer_prolog::machine::Machine>::compile_assert::h4c72d8054cc4a1e3
16: 0x108054ddf - scryer_prolog::machine::Machine::handle_toplevel_command::hcbc4d2457d98f1d8
17: 0x10805104f - scryer_prolog::machine::Machine::run_module_predicate::h665a482ee83b04f2
18: 0x108051dd0 - scryer_prolog::machine::Machine::run_top_level::h98ad3e5a02b065d9
19: 0x107f27f5e - scryer_prolog::main::hd9ecefa6a4ac075a
20: 0x107f2c3ca - std::sys_common::backtrace::__rust_begin_short_backtrace::h1cd4f2d4668f330b
21: 0x107f2c3ec - std::rt::lang_start::{{closure}}::hd97e1ceb299400ab
22: 0x1083a70a0 - std::rt::lang_start_internal::hf4b96bfc8c02c8b0
23: 0x107f280c9 - _main
✘-101
14:22 $```After fixing a bug in the hack for missing functionality in Scryer Prolog, re-running the tests shows:
% 510 test sets: 376 completed, 93 skipped, 25 broken, 9 timeouts, 7 crashes
% 7922 tests: 2496 skipped, 4872 passed, 554 failed123 of the 554 failures are due to test goals succeeding non-deterministically when determinism is expected. The actual number of tests is also lower (~7100). There's a weird bug in Scryer where backtracking results in the same answer being repeated multiple times, which in turn results in tests being reported multiple times. For example:
?- {ack(loader)}.
% [ /Users/pmoura/logtalk/examples/ack/ack.lgt loaded ]
% [ /Users/pmoura/logtalk/examples/ack/loader.lgt loaded ]
% (0 warnings)
true
; ...
?- ack::ack(2, 4, V).
V = 11
; V = 11
; V = 11
; V = 11
; V = 11
; V = 11
; V = 11
; V = 11
; V = 11.
?- With 56bbcb1cb3b0d18b03177ade9e0e2d77f38ba7a3, we get running all tests in the Logtalk distribution:
$ logtalk_tester -p scryer -g "set_logtalk_flag(clean,off)" -w -t 360
...
% 510 test sets: 380 completed, 93 skipped, 23 broken, 5 timeouts, 9 crashes
% 5504 tests: 129 skipped, 5083 passed, 292 failedOf those 292 test failures, 210 are for standards compliance tests. Experience tells that most failures are due to a relatively small number of bugs.
mthom
commented
3 years ago Still a ways to go, but that's great progress!
The current Logtalk git version includes preliminary support for Scryer Prolog, including a
scryerlgtshell script to start Logtalk. The script requires the current Scryer Prolog git version. For now, the main goal would be run Logtalk's standards compliance tests. But loading thelgtunittool or a large example (e.g.searching) always ends with a crash (at least on macOS). To reproduce these crashes, after installing the current Logtalk git version, try:Or:
Not sure how much valuer is in the backtrace but I get: