Retool tests #88

viega · 2024-07-18T03:14:44Z

Retool the test system to fork by default, so that if a test crashes the runner will still complete.

the runner will still complete.

ee7

Tests seemed to be crashing locally. Checking now with 258c1e9.

ee7

Ah, it looks like setting CON4M_ROOT is once again required, to avoid a crash. Am I doing something dumb? I'd removed it from my shell config after e0a0219:

Fix the issue w/ CON4M_ROOT being necessary to run tests.

Should we remove CON4M_ROOT from the workflow, such that requiring it would produce a CI failure?

libcon4m/.github/workflows/test.yml

Line 14 in e0a0219

CON4M_ROOT: .

With 258c1e9 and a clean working directory:

$ rm -rf build_dev
$ ./dev profile dev
$ CC=clang ./dev compile
$ cd build_dev
$ printenv CON4M_ROOT
$ CON4M_ROOT=../ ./c4test
[...]
Passed 60 out of 60 run tests.
 Con4m testing PASSED.
$ ./c4test
zsh: segmentation fault (core dumped)  ./c4test
$ ./c4test /path/to/libcon4m/tests/basic01.c4m
zsh: segmentation fault (core dumped)  ./c4test /path/to/libcon4m/tests/basic01.c4m
$ valgrind --track-origins=yes --leak-check=no -s -- ./c4test /path/to/libcon4m/tests/basic01.c4m
[...]
Running test 1: /path/to/libcon4m/tests/basic01.c4m
Processing module /path/to/libcon4m/tests/basic01.c4m
info: Done processing: /path/to/libcon4m/tests/basic01.c4m
****STARTING PROGRAM EXECUTION*****
****PROGRAM EXECUTION FINISHED*****
[...]
/path/to/libcon4m/tests/basic01.c4m exited with return code: 0.
Finished test 1: (/path/to/libcon4m/tests/basic01.c4m). PASSED.
Passed 1 out of 1 run tests.
 Con4m testing PASSED.
[...]
ERROR SUMMARY: 15 errors from 9 contexts (suppressed: 0 from 0)

1 errors in context 1 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x138D15: _c4m_str_find (../src/adts/string.c:1146)
   by 0x1395FF: c4m_str_xsplit (../src/adts/string.c:1277)
   by 0x193ACF: c4m_resolve_path (../src/util/path.c:168)
   by 0x11B226: c4m_con4m_root (../src/core/init.c:134)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
   by 0x4F08DC3: call_init (libc-start.c:145)
   by 0x4F08DC3: __libc_start_main@@GLIBC_2.34 (libc-start.c:347)
   by 0x117724: (below main) (in /path/to/libcon4m/build_dev/c4test)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


1 errors in context 2 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x138D15: _c4m_str_find (../src/adts/string.c:1146)
   by 0x1395FF: c4m_str_xsplit (../src/adts/string.c:1277)
   by 0x193ACF: c4m_resolve_path (../src/util/path.c:168)
   by 0x11B19A: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
   by 0x4F08DC3: call_init (libc-start.c:145)
   by 0x4F08DC3: __libc_start_main@@GLIBC_2.34 (libc-start.c:347)
   by 0x117724: (below main) (in /path/to/libcon4m/build_dev/c4test)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


1 errors in context 3 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x48A4334: UnknownInlinedFun (utf8proc.c:135)
   by 0x48A4334: utf8proc_iterate (utf8proc.c:125)
   by 0x13568B: c4m_internal_utf8_set_codepoint_count (../src/adts/string.c:17)
   by 0x13A3A4: utf8_init (../src/adts/string.c:662)
   by 0x133F1F: _c4m_new (../src/core/object.c:476)
   by 0x1940ED: c4m_app_path (../src/util/path.c:464)
   by 0x11B192: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
   by 0x4F08DC3: call_init (libc-start.c:145)
   by 0x4F08DC3: __libc_start_main@@GLIBC_2.34 (libc-start.c:347)
   by 0x117724: (below main) (in /path/to/libcon4m/build_dev/c4test)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


2 errors in context 4 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x48A44DD: utf8proc_encode_char (utf8proc.c:180)
   by 0x1365ED: c4m_to_utf8 (../src/adts/string.c:569)
   by 0x1936EA: internal_normalize_and_join (../src/util/path.c:64)
   by 0x193AD7: c4m_resolve_path (../src/util/path.c:167)
   by 0x11B19A: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
   by 0x4F08DC3: call_init (libc-start.c:145)
   by 0x4F08DC3: __libc_start_main@@GLIBC_2.34 (libc-start.c:347)
   by 0x117724: (below main) (in /path/to/libcon4m/build_dev/c4test)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


2 errors in context 5 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x48A4334: UnknownInlinedFun (utf8proc.c:135)
   by 0x48A4334: utf8proc_iterate (utf8proc.c:125)
   by 0x135ECD: c4m_to_utf32 (../src/adts/string.c:605)
   by 0x13946B: c4m_str_xsplit (../src/adts/string.c:1257)
   by 0x193ACF: c4m_resolve_path (../src/util/path.c:168)
   by 0x11B19A: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
   by 0x4F08DC3: call_init (libc-start.c:145)
   by 0x4F08DC3: __libc_start_main@@GLIBC_2.34 (libc-start.c:347)
   by 0x117724: (below main) (in /path/to/libcon4m/build_dev/c4test)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


2 errors in context 6 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x48A44DD: utf8proc_encode_char (utf8proc.c:180)
   by 0x1365ED: c4m_to_utf8 (../src/adts/string.c:569)
   by 0x1910AC: assemble_formatted_result (../src/util/format.c:524)
   by 0x1907BD: c4m_str_vformat (../src/util/format.c:575)
   by 0x1911AA: c4m_base_format (../src/util/format.c:591)
   by 0x1913DF: _c4m_cstr_format (../src/util/format.c:616)
   by 0x193889: internal_normalize_and_join (../src/util/path.c:109)
   by 0x193AD7: c4m_resolve_path (../src/util/path.c:167)
   by 0x11B19A: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


2 errors in context 7 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x190FC2: assemble_formatted_result (../src/util/format.c:506)
   by 0x1907BD: c4m_str_vformat (../src/util/format.c:575)
   by 0x1911AA: c4m_base_format (../src/util/format.c:591)
   by 0x1913DF: _c4m_cstr_format (../src/util/format.c:616)
   by 0x193889: internal_normalize_and_join (../src/util/path.c:109)
   by 0x193AD7: c4m_resolve_path (../src/util/path.c:167)
   by 0x11B19A: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
   by 0x4F08DC3: call_init (libc-start.c:145)
   by 0x4F08DC3: __libc_start_main@@GLIBC_2.34 (libc-start.c:347)
   by 0x117724: (below main) (in /path/to/libcon4m/build_dev/c4test)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


2 errors in context 8 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x48A4334: UnknownInlinedFun (utf8proc.c:135)
   by 0x48A4334: utf8proc_iterate (utf8proc.c:125)
   by 0x135ECD: c4m_to_utf32 (../src/adts/string.c:605)
   by 0x190F72: assemble_formatted_result (../src/util/format.c:500)
   by 0x1907BD: c4m_str_vformat (../src/util/format.c:575)
   by 0x1911AA: c4m_base_format (../src/util/format.c:591)
   by 0x1913DF: _c4m_cstr_format (../src/util/format.c:616)
   by 0x193889: internal_normalize_and_join (../src/util/path.c:109)
   by 0x193AD7: c4m_resolve_path (../src/util/path.c:167)
   by 0x11B19A: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)


2 errors in context 9 of 9:
Conditional jump or move depends on uninitialised value(s)
   at 0x484B9B8: strlen (vg_replace_strmem.c:505)
   by 0x13A2B5: utf8_init (../src/adts/string.c:644)
   by 0x133F1F: _c4m_new (../src/core/object.c:476)
   by 0x1940ED: c4m_app_path (../src/util/path.c:464)
   by 0x11B192: c4m_con4m_root (../src/core/init.c:130)
   by 0x11B405: c4m_system_module_path (../src/core/init.c:155)
   by 0x11BDF5: c4m_init_path (../src/core/init.c:180)
   by 0x11BB08: c4m_init (../src/core/init.c:295)
   by 0x4F08DC3: call_init (libc-start.c:145)
   by 0x4F08DC3: __libc_start_main@@GLIBC_2.34 (libc-start.c:347)
   by 0x117724: (below main) (in /path/to/libcon4m/build_dev/c4test)
 Uninitialised value was created by a stack allocation
   at 0x194024: c4m_app_path (../src/util/path.c:458)

ERROR SUMMARY: 15 errors from 9 contexts (suppressed: 0 from 0)

viega · 2024-07-18T13:26:20Z

I don't get that crash; I'll look at it in a bit.

viega · 2024-07-18T13:30:07Z

Okay, I see what the issue is; it was there before, and makes sense that it'd only trigger w/ ASAN / UBSAN. One sec.

ee7 · 2024-07-18T13:33:31Z

It seemed to be the other way around: the crash didn't happen when I enabled sanitizers. The steps in #88 (review) were without sanitizers.

Will check the new commit in a sec.

viega · 2024-07-18T13:37:23Z

Either way would make sense, and I'm 100% sure it was the problem.

ee7

Thanks. Confirmed that 3fd9dec fixed what I was seeing.

Nit: Using a relative path like ./c4test ../tests/basic01.c4m seems to hang - is that supposed to be supported?

src/compiler/scope.c

…nt was to be able to debug when a test does crash...

Retool the test system to fork by default, so that if a test crashes,

afda051

the runner will still complete.

viega force-pushed the jtv/dev branch from c4471ef to afda051 Compare July 18, 2024 03:21

viega requested a review from orangematt July 18, 2024 03:23

ee7 self-requested a review July 18, 2024 12:04

Make the dev output sane

258c1e9

ee7 reviewed Jul 18, 2024

View reviewed changes

Forgot to add the explicit null for one of my 3 readlink calls :(

3fd9dec

ee7 approved these changes Jul 18, 2024

View reviewed changes

src/compiler/scope.c Outdated Show resolved Hide resolved

Add forkless tests by default to the debug build, since the whole poi…

65e5c15

…nt was to be able to debug when a test does crash...

viega force-pushed the jtv/dev branch from c49a880 to 65e5c15 Compare July 18, 2024 13:42

viega merged commit ae7a185 into main Jul 18, 2024
3 checks passed

viega deleted the jtv/dev branch July 18, 2024 13:44

viega restored the jtv/dev branch July 18, 2024 23:10

viega deleted the jtv/dev branch July 21, 2024 22:37

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Retool tests #88

Retool tests #88

viega commented Jul 18, 2024

ee7 left a comment

ee7 left a comment •

edited

Loading

viega commented Jul 18, 2024

viega commented Jul 18, 2024

ee7 commented Jul 18, 2024

viega commented Jul 18, 2024

ee7 left a comment •

edited

Loading

Retool tests #88

Retool tests #88

Conversation

viega commented Jul 18, 2024

ee7 left a comment

Choose a reason for hiding this comment

ee7 left a comment • edited Loading

Choose a reason for hiding this comment

viega commented Jul 18, 2024

viega commented Jul 18, 2024

ee7 commented Jul 18, 2024

viega commented Jul 18, 2024

ee7 left a comment • edited Loading

Choose a reason for hiding this comment

ee7 left a comment •

edited

Loading

ee7 left a comment •

edited

Loading