#225 | 2879048-225-log-cacheGet-extend.patch | 9.24 KB | Anonymous (not verified) |
|
#224 | 2879048-224-config-cid-field.storage.entity_test_mulrev_changed.field_test_text.patch | 1.4 KB | Anonymous (not verified) |
|
#224 | 2879048-224-config-cid-field.storage.entity_test_mul_changed.field_test_text.patch | 1.4 KB | Anonymous (not verified) |
|
#224 | 2879048-224-config-cid-field.storage.entity_test_mul_langcode_key.field_test_text.patch | 1.4 KB | Anonymous (not verified) |
|
#224 | 2879048-224-config-cid-field.storage.entity_test_mulrev.field_test_text.patch | 1.39 KB | Anonymous (not verified) |
|
#224 | 2879048-224-config-cid-field.storage.entity_test_mul.field_test_text.patch | 1.39 KB | Anonymous (not verified) |
|
#223 | 2879048-222-config-cids-p3.patch | 1.53 KB | Anonymous (not verified) |
|
#223 | 2879048-222-config-cids-p2.patch | 1.45 KB | Anonymous (not verified) |
|
#223 | 2879048-222-config-cids-p1.patch | 1.44 KB | Anonymous (not verified) |
|
#222 | 2879048-222-config-cids-non-field.patch | 1.57 KB | Anonymous (not verified) |
|
#222 | 2879048-222-config-cids-field.patch | 1.75 KB | Anonymous (not verified) |
|
#221 | 2879048-220-config-cids-non-field.patch | 1.59 KB | Anonymous (not verified) |
|
#221 | 2879048-220-config-cids-field.patch | 1.77 KB | Anonymous (not verified) |
|
#220 | 2879048-220-config-cids-2.patch | 1.99 KB | Anonymous (not verified) |
|
#219 | 2879048-219-cids-by-equal-default_config_hash.patch | 1.63 KB | Anonymous (not verified) |
|
#218 | 2879048-217-bins-special-deleteAll-config.patch | 1.27 KB | Anonymous (not verified) |
|
#218 | 2879048-217-bins-special-deleteAll-bootstrap.patch | 1.27 KB | Anonymous (not verified) |
|
#218 | 2879048-217-bins-special-deleteAll-discovery.patch | 1.27 KB | Anonymous (not verified) |
|
#217 | 2879048-217-bins-special-deleteAll.patch | 1.37 KB | Anonymous (not verified) |
|
#216 | 2879048-216-bins-special-delete.patch | 1.47 KB | Anonymous (not verified) |
|
#215 | 2879048-215-log-bins.patch | 2.41 KB | Anonymous (not verified) |
|
#214 | 2879048-214-log-bins.patch | 2.42 KB | Anonymous (not verified) |
|
#212 | 2879048-212-bins-deleteAll.patch | 1.3 KB | Anonymous (not verified) |
|
#211 | 2879048-211-cacheGet-false.patch | 1.62 KB | Anonymous (not verified) |
|
#210 | 2879048-210-log-getMultiple-with-spec-cid.patch | 3.89 KB | Anonymous (not verified) |
|
#209 | 2879048-209-test-only-gc-collect-binsearch-2.patch | 1.19 KB | tacituseu |
|
#209 | 2879048-209-test-only-gc-collect-binsearch-1.patch | 1.29 KB | tacituseu |
|
#208 | 2879048-208-log-EntityFieldManager-getFieldMap.patch | 18.39 KB | Anonymous (not verified) |
|
#207 | 2879048-206-log-FieldConfig-save.patch | 15.35 KB | Anonymous (not verified) |
|
#205 | 2879048-205-log-FieldConfig-create.patch | 8.03 KB | Anonymous (not verified) |
|
#203 | 2879048-203-ChainedFastBackend.patch | 2.03 KB | Anonymous (not verified) |
|
#202 | 2879048-202-test-only-without-old-one.patch | 2.16 KB | tacituseu |
|
#201 | 2879048-201-log-setUp-EntityResourceTestBase.patch | 4 KB | Anonymous (not verified) |
|
#200 | 2879048-200-log-setUp.patch | 4.62 KB | Anonymous (not verified) |
|
#199 | 2879048-199-log-resetAll-and-rebuildAll.patch | 3.3 KB | Anonymous (not verified) |
|
#198 | 2879048-198-log-drupal_flush_all_caches.patch | 6.63 KB | Anonymous (not verified) |
|
#196 | 2879048-196-log-calls.patch | 2.06 KB | Anonymous (not verified) |
|
#194 | 2879048-194-no-static.patch | 1.43 KB | tacituseu |
|
#194 | 2879048-194-clone.patch | 1.67 KB | tacituseu |
|
#190 | 2879048-176-test-only-well.patch | 2.71 KB | tacituseu |
|
#189 | 2879048-176-test-only-sorted_graph.patch | 2.75 KB | tacituseu |
|
#189 | 2879048-176-test-only-graph_object.patch | 2.75 KB | tacituseu |
|
#189 | 2879048-176-test-only-graph.patch | 2.75 KB | tacituseu |
|
#188 | 2879048-188-tour.patch | 1.35 KB | Anonymous (not verified) |
|
#188 | 2879048-188-statistics.patch | 1.36 KB | Anonymous (not verified) |
|
#188 | 2879048-188-book.patch | 1.35 KB | Anonymous (not verified) |
|
#188 | 2879048-188-action.patch | 1.35 KB | Anonymous (not verified) |
|
#183 | 2879048-182-test-only-with-random-modules.patch | 1.38 KB | Anonymous (not verified) |
|
#183 | 2879048-182-test-only-without-c74c39e.patch | 34.87 KB | Anonymous (not verified) |
|
#181 | 2879048-181-test-only-with-add-random-modules.patch | 3.46 KB | Anonymous (not verified) |
|
#181 | 2879048-181-test-only-without-entity_test.data_types.schema.yml.patch | 3.26 KB | Anonymous (not verified) |
|
#180 | 2879048-180-test-only-without-2c2e85c.patch | 5.89 KB | tacituseu |
|
#180 | 2879048-180-test-only-without-113c6dc.patch | 5.56 KB | tacituseu |
|
#178 | test-only.patch | 759 bytes | Anonymous (not verified) |
|
#176 | fix.patch | 2.82 KB | Anonymous (not verified) |
|
#176 | test-only.patch | 759 bytes | Anonymous (not verified) |
|
#173 | x100-EntityTestJsonInternalPropertyNormalizerTest-and-129.patch | 2.69 KB | Anonymous (not verified) |
|
#173 | x100-EntityTestJsonInternalPropertyNormalizerTest.patch | 628 bytes | Anonymous (not verified) |
|
#171 | x200-EntityTestHalJsonInternalPropertyNormalizerTest-and-129.patch | 2.7 KB | Anonymous (not verified) |
|
#171 | x200-EntityTestHalJsonInternalPropertyNormalizerTest.patch | 632 bytes | Anonymous (not verified) |
|
#170 | x20-EntityTestHalJsonInternalPropertyNormalizerTest.patch | 631 bytes | Anonymous (not verified) |
|
#167 | 2879048-167-graph-stress.patch | 34 KB | tacituseu |
|
#165 | 2879048-165-graph-stress.patch | 33.92 KB | tacituseu |
|
#163 | 2879048-163-graph-stress.patch | 33.75 KB | tacituseu |
|
#161 | 2879048-161-graph-stress.patch | 33.6 KB | tacituseu |
|
#159 | 2879048-159-graph-stress.patch | 33.59 KB | tacituseu |
|
#157 | 2879048-157-graph-stress.patch | 33.59 KB | tacituseu |
|
#155 | restore-d7c496150e.patch | 147.6 KB | Anonymous (not verified) |
|
#154 | revert-d7c496150e.patch | 108.88 KB | Anonymous (not verified) |
|
#153 | revert_dc5ca923d0_r.patch | 3.78 MB | Anonymous (not verified) |
|
#152 | revert_fdef8555ac_r.patch | 3.82 MB | Anonymous (not verified) |
|
#151 | revert_10a1e15fa3_r.patch | 3.82 MB | Anonymous (not verified) |
|
#150 | revert_d7c496150e_r.patch | 3.67 MB | Anonymous (not verified) |
|
#149 | revert_82f1a1ca6c_r.patch | 4.04 MB | Anonymous (not verified) |
|
#148 | revert_a47d8c49b2_r.patch | 3.55 MB | Anonymous (not verified) |
|
#147 | 2879048-147-revert-yml-modules-and-themes.patch | 7.93 KB | Anonymous (not verified) |
|
#146 | 2879048-130-minus-two-themes.patch | 2.05 KB | Anonymous (not verified) |
|
#145 | revert_7204a00_r.patch | 4.46 MB | Anonymous (not verified) |
|
#143 | revert_50a541b_r.patch | 5.65 MB | Anonymous (not verified) |
|
#142 | revert_50a541b.patch | 5.65 MB | Anonymous (not verified) |
|
#141 | 2879048-141.patch | 1.49 KB | Anonymous (not verified) |
|
#139 | 2879048-98-No-config-user-namespace__ConfigImportUITest.patch | 1.04 KB | Anonymous (not verified) |
|
#139 | 2879048-98-No-config-user-namespace_without_many_themes.patch | 64.64 KB | Anonymous (not verified) |
|
#137 | 2879048-98-No-config-user-namespace-save-graph-sorted-1-add-mod.patch | 1.85 KB | tacituseu |
|
#137 | 2879048-98-No-config-user-namespace-save-graph-1-add-mod.patch | 1.67 KB | tacituseu |
|
#135 | 2879048-98-No-config-user-namespace-save-graph-sorted-1.patch | 1.31 KB | tacituseu |
|
#135 | 2879048-98-No-config-user-namespace-save-graph-1.patch | 1.14 KB | tacituseu |
|
#134 | 2879048-98-No-config-user-namespace-save-graph-sorted.patch | 1.31 KB | tacituseu |
|
#133 | 2879048-98-No-config-user-namespace-save-graph.patch | 1.14 KB | tacituseu |
|
#130 | 2879048-98-No-config-user-namespace.patch | 1.03 KB | tacituseu |
|
#129 | 2879048-98-No-config-user-namespace-keep-graph.patch | 3.68 KB | tacituseu |
|
#126 | x20__2849674-49_rest.patch | 13.42 KB | Anonymous (not verified) |
|
#125 | x20__2849674-49_system.patch | 13.43 KB | Anonymous (not verified) |
|
#124 | 2879048-122-action.patch | 1.6 KB | Anonymous (not verified) |
|
#124 | 2879048-122-filter.patch | 1.6 KB | Anonymous (not verified) |
|
#124 | 2879048-122-system.patch | 1.6 KB | Anonymous (not verified) |
|
#122 | 2879048-122-more-dependencies.patch | 1.64 KB | Anonymous (not verified) |
|
#118 | 2879048-118-action_default_configuration.patch | 1.54 KB | Anonymous (not verified) |
|
#114 | 2879048-98-No-config-user-namespace-NodeCreationTest.patch | 1.05 KB | Anonymous (not verified) |
|
#114 | 2879048-98-No-action__order.patch | 1.61 KB | Anonymous (not verified) |
|
#114 | 2879048-98-No-install__order.patch | 2.69 KB | Anonymous (not verified) |
|
#114 | 2879048-98-action-schema__order.patch | 3.34 KB | Anonymous (not verified) |
|
#112 | interdiff-order.txt | 518 bytes | Anonymous (not verified) |
#112 | x200__2849674-49_change_order.patch | 13.42 KB | Anonymous (not verified) |
|
#112 | x200__2849674-49.patch | 13.42 KB | Anonymous (not verified) |
|
#108 | 2879048-108-system_intead_user.patch | 1.02 KB | Anonymous (not verified) |
|
#108 | 2879048-108-filter_intead_view.patch | 1.02 KB | Anonymous (not verified) |
|
#108 | 2879048-108-other_company.patch | 1.45 KB | Anonymous (not verified) |
|
#108 | 2879048-108-change_order.patch | 1.41 KB | Anonymous (not verified) |
|
#108 | 2879048-108-submodules.patch | 1.45 KB | Anonymous (not verified) |
|
#107 | 2879048-107-user_name-and-req.patch | 1.04 KB | Lendude |
|
#107 | 2879048-107-user_name.patch | 1.04 KB | Lendude |
|
#105 | 2879048-105-only-user.patch | 1.01 KB | Anonymous (not verified) |
|
#102 | 2879048-98-No-config-views-namespace.patch | 1.01 KB | Lendude |
|
#102 | 2879048-98-No-config-user-namespace.patch | 1.03 KB | Lendude |
|
#98 | 2879048-98-No-action.patch | 1.61 KB | Lendude |
|
#98 | 2879048-98-No-install.patch | 2.69 KB | Lendude |
|
#98 | 2879048-98-action-schema.patch | 3.34 KB | Lendude |
|
#97 | 2879048-97-fpt.patch | 1.93 KB | tacituseu |
|
#93 | 2879048-93-test-batch-only-config-and-action-bare-gc-disabled.patch | 3.83 KB | tacituseu |
|
#90 | 2879048-90-2.patch | 3.27 KB | Anonymous (not verified) |
|
#90 | 2879048-90-1.patch | 3.38 KB | Anonymous (not verified) |
|
#88 | 2879048-88-test-batch-only-config-and-action-bare.patch | 3.31 KB | tacituseu |
|
#86 | 2879048-86-test-batch-only-config-and-action-access-as-object+not-typed-as-content-execute.patch | 4.72 KB | tacituseu |
|
#81 | 2879048-81-gc.patch | 5.87 KB | Anonymous (not verified) |
|
#80 | 2879048-80-only-FileFieldWidgetTest.patch | 498 bytes | Anonymous (not verified) |
|
#78 | 2879048-78-test-batch-only-config-and-action-comma.patch | 4.64 KB | tacituseu |
|
#76 | 2879048-76-test-batch-only-config-and-action-concurrency-1.patch | 4.98 KB | tacituseu |
|
#74 | 2879048-74-test-batch-only-config-and-action-access-as-object.patch | 4.69 KB | tacituseu |
|
#72 | interdiff.txt | 509 bytes | Anonymous (not verified) |
#72 | patch-67_and_rebuild_themes.patch | 5.13 KB | Anonymous (not verified) |
|
#67 | 2879048-67-test-batch-only-config-and-plugins.patch | 4.64 KB | tacituseu |
|
#64 | 2879048-64-test-kernel.patch | 2.51 KB | tacituseu |
|
#64 | 2879048-64-test-batch.patch | 6.13 KB | tacituseu |
|
#62 | 2879048-62-test.patch | 8.16 KB | tacituseu |
|
#60 | 2879048-60-test.patch | 8.15 KB | tacituseu |
|
#60 | 2879048-60-patch.patch | 5.75 KB | tacituseu |
|
#57 | 2879048-57.patch | 13.42 KB | Anonymous (not verified) |
|
#56 | x100_2809471-13.patch | 12.3 KB | Anonymous (not verified) |
|
#54 | page_2809471.patch | 11.97 KB | Anonymous (not verified) |
|
#52 | 2809471-13-2.patch | 12.16 KB | Anonymous (not verified) |
|
#52 | 2809471-13-1.patch | 12.28 KB | Anonymous (not verified) |
|
#50 | 2879048-50.patch | 43.93 KB | Anonymous (not verified) |
|
#48 | 2879048-48.patch | 44.04 KB | Anonymous (not verified) |
|
#46 | 2879048-46.patch | 43.67 KB | Anonymous (not verified) |
|
#43 | 2879048-43.patch | 43.03 KB | Anonymous (not verified) |
|
#42 | 2879048-42-full.patch | 42.71 KB | Anonymous (not verified) |
|
#42 | 2879048-42.patch | 43.23 KB | Anonymous (not verified) |
|
#40 | 2879048-40.patch | 43.09 KB | Anonymous (not verified) |
|
#38 | ignore_patch_testing-2879048-38.patch | 27.18 KB | Anonymous (not verified) |
|
#36 | ignore_patch_testing-2879048-36.patch | 26.7 KB | Anonymous (not verified) |
|
#34 | 2879048-33.patch | 35.03 KB | Anonymous (not verified) |
|
#32 | phpunit.patch | 233.32 KB | Anonymous (not verified) |
|
#31 | rebuild.patch | 7.26 KB | Anonymous (not verified) |
|
#29 | 50x_sleep.patch | 7.73 KB | Anonymous (not verified) |
|
#29 | 50x_cache_list.patch | 7.8 KB | Anonymous (not verified) |
|
#29 | 50x_list_null.patch | 7.74 KB | Anonymous (not verified) |
|
#28 | 50x_refresh.patch | 7.74 KB | Anonymous (not verified) |
|
#28 | 50x_reset.patch | 7.74 KB | Anonymous (not verified) |
|
#28 | 50x_rebuild.patch | 7.39 KB | Anonymous (not verified) |
|
#26 | x100_theme_exception.patch | 3.45 KB | Anonymous (not verified) |
|
#25 | x100_theme_alone.patch | 3.41 KB | Anonymous (not verified) |
|
#24 | x100_assert_alone.patch | 1.05 KB | Anonymous (not verified) |
|
#22 | x100_assert.patch | 7.28 KB | Anonymous (not verified) |
|
#21 | assert_default_fix.patch | 6.85 KB | Anonymous (not verified) |
|
#19 | assert_default.patch | 6.82 KB | Anonymous (not verified) |
|
#19 | assert_return.patch | 6.83 KB | Anonymous (not verified) |
|
#17 | assert.patch | 6.79 KB | Anonymous (not verified) |
- 8.4.x:
- PHP 5.6 & MySQL 5.5 22,059 pass, 2 fail
- PHP 7.0.x-dev & MySQL 5.5 22,058 pass, 6 fail
- PHP 7.1.x-dev & MySQL 5.5 22,060 pass
- PHP 7.1 & MySQL 5.5 22,058 pass, 6 fail
- PHP 7 & MySQL 5.5 22,058 pass, 6 fail
- PHP 7 & PostgreSQL 9.1 22,056 pass
- PHP 7 & SQLite 3.14 22,054 pass, 3 fail
|
#15 | use_class_v3.patch | 848.11 KB | Anonymous (not verified) |
|
#13 | use_class_3.patch | 846.52 KB | Anonymous (not verified) |
|
#12 | use_class_v2.patch | 785.4 KB | Anonymous (not verified) |
|
#9 | use_class.patch | 785.75 KB | Anonymous (not verified) |
|
#8 | menu_link_content_does_not_properly_validate_description-2877559-29.patch | 1.74 KB | Dinesh18 |
|
#7 | menu_link_content_does_not_properly_validate_description-2877559-29.patch | 1.74 KB | Dinesh18 |
|
#4 | x10000_runs_2857843.patch | 5.34 KB | Anonymous (not verified) |
|
#4 | 500x10_langs_2857843.patch | 5.34 KB | Anonymous (not verified) |
|
#2 | x1-2857843.patch | 5.06 KB | Anonymous (not verified) |
|
Comments
Comment #1
Anonymous (not verified) CreditAttribution: Anonymous commentedvaplas created an issue. See original summary.
Comment #2
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #4
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #7
Dinesh18 CreditAttribution: Dinesh18 as a volunteer commentedIgnore patch test
Comment #8
Dinesh18 CreditAttribution: Dinesh18 as a volunteer commentedComment #9
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #10
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #12
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #13
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #15
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #17
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #19
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #21
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #22
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #24
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #25
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #26
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #28
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #29
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #31
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #32
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #34
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #36
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #38
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #40
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #42
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #43
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #46
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #48
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #50
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #52
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #54
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #56
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #57
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #60
tacituseu CreditAttribution: tacituseu commentedChecking which part exactly as it doesn't seem to have overlap at all, and with less iterations not to give CI a hard time.
Comment #62
tacituseu CreditAttribution: tacituseu commentedComment #64
tacituseu CreditAttribution: tacituseu commentedMore splitting.
Comment #66
LendudeHa, did not expect that at all!
Comment #67
tacituseu CreditAttribution: tacituseu commentedMe neither, let's take it one step further.
Comment #69
tacituseu CreditAttribution: tacituseu commentedo_O
Comment #70
tacituseu CreditAttribution: tacituseu commentedComment #72
Anonymous (not verified) CreditAttribution: Anonymous commentedAn interesting observation: in the beginning there are failures, and then passing:
I already saw this effect when I was working on another incomprehensible fall (#50, related
with #2808063: LibraryDiscoveryParser::buildByExtension() doesn't validate that extensions exist):
But #48 was a slight deviation, so this is not a permanent behavior:
I wonder how close these cases are?
Comment #74
tacituseu CreditAttribution: tacituseu commentedcould it be ?
Drupal\Core\Action\ActionInterface
:@vaplas: interesting observation indeed.
Comment #76
tacituseu CreditAttribution: tacituseu commentedLet's force concurrency to 1 then and see if only the first one fails.
Comment #78
tacituseu CreditAttribution: tacituseu commentedWouldn't it be fun ;)
Comment #80
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #81
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #83
Anonymous (not verified) CreditAttribution: Anonymous commentedWell, #60 fail with 5.6 and pass with 7.1. There may be some specific behavior, for example:
Or some vendor, like
phpspec
, in which there are many checks likeversion_compare(PHP_VERSION, '7...)
?Comment #84
Anonymous (not verified) CreditAttribution: Anonymous commentedAll versions php 7.x passed tests of #78. Annnd php 5.5 too!) Random luck? Not. Rerun #57 with 100 runs, and pass too. Hence problem only with php 5.6.
Comment #85
tacituseu CreditAttribution: tacituseu commented@vaplas: yes, tried them on #67 too, and all it does is registering an Action plugin, out of ideas at the moment.
Comment #86
tacituseu CreditAttribution: tacituseu commentedComment #88
tacituseu CreditAttribution: tacituseu commentedComment #90
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #93
tacituseu CreditAttribution: tacituseu commentedComment #94
Anonymous (not verified) CreditAttribution: Anonymous commentedMAN!!!
Comment #95
tacituseu CreditAttribution: tacituseu commentedSprinkling core with
gc_disable()
every time 5.6 isn't happy can't be a solution though.Comment #96
Anonymous (not verified) CreditAttribution: Anonymous commentedLet it be so) But it's a good jab in the fight against #2842393: Discover why gc_disable() improves update.php stability. Now the problem is easier to reproduce.
Comment #97
tacituseu CreditAttribution: tacituseu commentedLet's check the old one then.
Comment #98
LendudeWhat's so damn special about this action that it causes all this ????
Wild stabs in the dark....
Comment #102
LendudeOk so, it's the adding a module that kills it....sheeeesh
lets try some empty modules and see if that still kills it....
Comment #103
Anonymous (not verified) CreditAttribution: Anonymous commentedLendude!!! It is epic!
Comment #105
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #106
LendudeYou've got te be kidding me.....
I gotta go now, so @vaplas go crazy with this :)
Comment #107
LendudeOne more shot...
Comment #108
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #109
Lendude@vaplas, @catch just did a ton of commits, so try the original failing file too, the problem might have moved to a different test
Comment #112
Anonymous (not verified) CreditAttribution: Anonymous commentedIt seems the order matters for stable works of Mr. Bot:
Let's check the #2849674-49: Complex job in ViewExecutable::unserialize() causes data corruption 200 runs one then.
Comment #114
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #118
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #119
tacituseu CreditAttribution: tacituseu commentedLooks like CI ran out of free space again on some of those patches.
Comment #120
Anonymous (not verified) CreditAttribution: Anonymous commentedBecause I abused the trust of CI and ran it 200 times in #112. But I saw the tests fall in both versions. Apparently in the schema.yml also something hidden, because changed order in #114 doesn't help.
Comment #121
tacituseu CreditAttribution: tacituseu commentedLooking over latest patches, could it be that the dependency resolver is the one responsible for creating those 'bad' objects https://bugs.php.net/bug.php?id=72286 mentions ?
Comment #122
Anonymous (not verified) CreditAttribution: Anonymous commented#121: Maybe even more so that @xjm emphasized this bug. But it says:
But our tests works with 5.5. Of course, maybe this problem actual only special 5.x versions. In any case, it's too smart for me :)
Comment #123
tacituseu CreditAttribution: tacituseu commented@vaplas: 5.5 is patched/supported by Ubuntu, so might have some fixes that are not in 5.6, but that is only if installed from their repository, not sure what is the state of things now, since AFAIR to support generating stack dumps/backtraces on those coredumps the CI team started custom compiling php from sources.
Comment #124
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #125
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #126
Anonymous (not verified) CreditAttribution: Anonymous commentedDoes anyone understand why adding dependency helps to get rid of random falls? Me not. What if add any independent module?
Comment #127
tacituseu CreditAttribution: tacituseu commentedThe fails are in circular reference garbage collector, my guess would be it either:
1. changes order of objects and the algorithm has easier time collecting them
2, increases reference count so the algorithm doesn't even try to collect it
Also follow up on php versions used for testbots (drupalci_environments): they are based on tarballs directly from https://secure.php.net, except for those suffixed with 'x-dev' which are from https://github.com/php/php-src/, so no Ubuntu/Debian specific patches in them.
Comment #128
Anonymous (not verified) CreditAttribution: Anonymous commented@tacituseu, thank you! Sounds logical for me!
Comment #129
tacituseu CreditAttribution: tacituseu commentedComment #130
tacituseu CreditAttribution: tacituseu commented:D re-uploading control patch from #102
Comment #131
Anonymous (not verified) CreditAttribution: Anonymous commentedWOW! This is awesome!
Comment #132
tacituseu CreditAttribution: tacituseu commented@vaplas: Thanks for the question, it kept bugging me so took a stab at #127.2 and it worked ;).
Comment #133
tacituseu CreditAttribution: tacituseu commentedComment #134
tacituseu CreditAttribution: tacituseu commentedComment #135
tacituseu CreditAttribution: tacituseu commentedComment #136
Anonymous (not verified) CreditAttribution: Anonymous commentedLooks like patches haven't
user_batch_action_test.info.yml
part.Comment #137
tacituseu CreditAttribution: tacituseu commentedThis time with user_batch_action_test included.
Comment #138
tacituseu CreditAttribution: tacituseu commentedYeah forgot to git add ;), but will be good for baseline.
Comment #139
Anonymous (not verified) CreditAttribution: Anonymous commentedAll I could understand is that the drop may occur with graph of themes (although the graph of a successful test is the same as the last graph of a failed test).
I like the idea that the themes are the culprit, because I have a tooth on them due #2849674: Complex job in ViewExecutable::unserialize() causes data corruption :) But maybe I'm wrong. Try our test, leaving only the main themes.
Also check ConfigImportUITest.
Comment #140
tacituseu CreditAttribution: tacituseu commentedThere are 152 calls to
ModuleHandler::buildModuleDependencies()
per test run, it is used for both modules and themes, the very first one is fromThemeHandler::rebuildThemeData()
, and it is in a different order than all the rest so there might be something to this tooth (also observation from #72 that the first ones tend to fail), but the resulting graphs are the same.I don't expect them to differ, just trying to create a test case for
Graph
.You can split the log with:
$ csplit -n 3 -f mod-dep-calls- graph.log '/^20170807/' "{*}"
Comment #141
Anonymous (not verified) CreditAttribution: Anonymous commentedSooner or later the FileFieldWidgetTest will be converted to phpunit. Therefore, it makes sense to prepare in advance. After that, I will create a new issue with move the main points from here.
Comment #142
Anonymous (not verified) CreditAttribution: Anonymous commentedIt seems that we lost this control test :(, because #130 also not work now. Checking the reason: is the changes in the set of modules/themes or the change CI settings? (revert to 2017-08-07, commit hash 50a541b.... + #130 patch).
5Mb changes by 2 month, crazy!
Comment #143
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #145
Anonymous (not verified) CreditAttribution: Anonymous commentedYep, due to list of modules/themes was changed. Before a binary search, one heuristic attempt.
Comment #146
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #147
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #148
Anonymous (not verified) CreditAttribution: Anonymous commentedHm. Revert *.yml of modules and themes not helped :( Okay, I'll try to find a topic after which problem gone. Most likely this will not help, because after it many changes can affect on the graph. But anyway.
Comment #149
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #150
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #151
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #152
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #153
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #154
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #155
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #156
Anonymous (not verified) CreditAttribution: Anonymous commentedWell, #153 red, #150 green. So, after #2569805: For Drupal migration, identify the source module (commit d7c496150e....) the problem again became elusive 😥
I tried to roll back the patch from #2569805:
But it did not help. This confirms, that the problem with php 5.6 has not disappeared, but simply ceased to appear on this test :(
Comment #157
tacituseu CreditAttribution: tacituseu commentedLets try to make some use of those old logs then. Also re-queued #130 with 8.4.x and 8.3.x.
Comment #159
tacituseu CreditAttribution: tacituseu commentedComment #161
tacituseu CreditAttribution: tacituseu commentedYikes... this one, hopefully, without c&p/s&r errors.
Comment #163
tacituseu CreditAttribution: tacituseu commentedComment #165
tacituseu CreditAttribution: tacituseu commentedComment #167
tacituseu CreditAttribution: tacituseu commentedEdit: No luck, previous patches were just figuring out how to get bootstrap.inc included, will try moving into
Drupal\Tests\Core\Extension\ModuleHandlerTest
next.Comment #168
Anonymous (not verified) CreditAttribution: Anonymous commentedIssue with info about this works #2919863: Discover why gc_disable() during non-interactive install improves tests stability
Changed IS, because ton of work has been done, and it will be more correct to make this issue more specialized.
Comment #169
tacituseu CreditAttribution: tacituseu commentedLooks like it moved back to PHP 5.5
https://www.drupal.org/pift-ci-job/803921
https://www.drupal.org/pift-ci-job/804596
https://www.drupal.org/pift-ci-job/804598
HEAD failing inside
EntityTestHalJsonInternalPropertyNormalizerTest
andEntityTestJsonInternalPropertyNormalizerTest
, after #2871591: Allow ComplexData in TypedData to specify computed properties that should be exposed in normalization and other contexts went in (see #2871591-167: Allow ComplexData in TypedData to specify computed properties that should be exposed in normalization and other contexts).Comment #170
Anonymous (not verified) CreditAttribution: Anonymous commentedHah, very interesting! Let's check it!
Comment #171
Anonymous (not verified) CreditAttribution: Anonymous commentedYep, this is definitely a random fail! Let's increase the number of tests, and check it with #129.
Comment #172
Anonymous (not verified) CreditAttribution: Anonymous commentedChaos without #129 (https://dispatcher.drupalci.org/job/drupal_patches/36707/console)
Green works with #129 (https://dispatcher.drupalci.org/job/drupal_patches/36700/console)
@tacituseu, 👏👏👏
Comment #173
Anonymous (not verified) CreditAttribution: Anonymous commentedAnd check Postgresql.
Comment #174
Wim LeersPer #169, #2871591: Allow ComplexData in TypedData to specify computed properties that should be exposed in normalization and other contexts risks being reverted unless we can figure out a work-around for the random fail on 5.5 (which btw is completely unsupported by the PHP team…)
#170 confirmed that this is a random fail. #173 is now attempting to check whether the patch of #129 fixes the random fail or not.
I'll be actively following this issue, ping me if you get stuck on something.
Comment #175
tacituseu CreditAttribution: tacituseu commentedTest bots are having free space issues now (
'Requeued after CI error.'
).Comment #176
Anonymous (not verified) CreditAttribution: Anonymous commented#174: Woot!
#175: not problem, just reduce the number of tests and combine
EntityTestJsonInternalPropertyNormalizerTest
+EntityTestHalJsonInternalPropertyNormalizerTest
If I understand correctly, then adding a couple of random module dependencies will fix the error, but first check that we haven't 'Requeued after CI error.' now.
Comment #178
Anonymous (not verified) CreditAttribution: Anonymous commentedOkay, 20 is too much for Postresql, what about 10?
Comment #179
tacituseu CreditAttribution: tacituseu commented@vaplas:
I think it needs manual cleanup when it gets out of space, but maybe it is enough for the instance to spin down after 2 hours.
Looking over the patch it doesn't seem like it touched anything that would affect
ModuleHandler::buildModuleDependencies()
, since it first failed on daily I checked commit history for that day:113c6dc: #2840392: Enable BigPipe by default in the Standard install profile
2c2e85c: #2695109: Cache bins are not deleted when the module that declares them is uninstalled
3cadcd5: #2903161: Fix incorrect FieldFormatter id for "weight" field in base field definition in display options
1e85083: #2910682: Mark serializer encoders @internal
Comment #180
tacituseu CreditAttribution: tacituseu commentedAttached patch from #178 with specific reverts.
Comment #181
Anonymous (not verified) CreditAttribution: Anonymous commented#2871591: Allow ComplexData in TypedData to specify computed properties that should be exposed in normalization and other contexts contains entity_test.data_types.schema.yml, it can affect?
Comment #183
Anonymous (not verified) CreditAttribution: Anonymous commentedopps, bad patches.
Comment #185
Anonymous (not verified) CreditAttribution: Anonymous commentedLuck with random modules slightly smoothed my shame with the revert.
Comment #186
tacituseu CreditAttribution: tacituseu commentedGreat find :D, and it leads to
ModuleHandler::buildModuleDependencies()
too:1.
FunctionalTestSetupTrait::installModulesFromClassProperty()
:2.
ModuleInstaller::install(array $module_list, $enable_dependencies = TRUE)
:3.
system.module::system_rebuild_module_data()
:Edit: fixed call stack.
Comment #187
Anonymous (not verified) CreditAttribution: Anonymous commentedExcellent! I believe that we should apply #129 for php < 7. At least like workaround, because these random fails can stall the others issues.
Comment #188
Anonymous (not verified) CreditAttribution: Anonymous commentedJust for demonstration, check fix with one random module.
Comment #189
tacituseu CreditAttribution: tacituseu commentedSplitting the patch from #176 to get more clues.
Comment #190
tacituseu CreditAttribution: tacituseu commentedComment #191
Wim LeersWhy are all #189 patches green, but is #190 green?
Comment #192
Anonymous (not verified) CreditAttribution: Anonymous commentedPer #127:
In #189 collected more reference via
In #190 nothing is collected.
Comment #193
tacituseu CreditAttribution: tacituseu commentedThere's also a third option that keeping the references is padding the buffer with safe objects, makes the buffer go over GC_ROOT_BUFFER_MAX_ENTRIES, and the real trouble-maker doesn't get recorded and hence never attempted to be collected. Each of those arrays has hundreds of entries (300-400 AFAIR).
But #188 shows that just messing with this array/dependencies makes it green, so I'd bet on the source of the problem to be near that.
Comment #194
tacituseu CreditAttribution: tacituseu commentedRandom stabs.
Comment #196
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #197
Wim LeersI have to say, it’s super impressive to see the two of you expertly analyzing this! 😮👏👏👏
Comment #198
Anonymous (not verified) CreditAttribution: Anonymous commented@Wim Leers, your post is very inspiring, thank you so much!
The main brain in this analysis is @tacituseu, of course. I posting patches just to keep the company ;)
#196: it's interesting that with fails the
buildModuleDependencies()
works well as well as with passes:Relevant logs are equivalent, so the error occurs after the all
buildModuleDependencies()
calls.Comment #199
Anonymous (not verified) CreditAttribution: Anonymous commentedLooks like @alexpott made a patch with a fix #2921862-18: Segfault on PHP5.5 and PostgreSQL. But it will be nice to catch a specific line. Therefore the continuation.
drupal_flush_all_caches()
works also well. Go next.Comment #200
Anonymous (not verified) CreditAttribution: Anonymous commentedHm.. or i'm or I miss something, or
installDrupal()
works like a charm too. Going to setUp().Comment #201
Anonymous (not verified) CreditAttribution: Anonymous commentedI'm betting that this way will lead to some query to the database. Which collapses because of the bug in PHP. But what exactly?
Comment #202
tacituseu CreditAttribution: tacituseu commentedRe #198: The problem with GC is it will almost never fail at the place the problem was introduced, only later when buffer is filled up and it automatically triggers
gc_collect_cycles()
. Also thought about trying to dissect it by placinggc_collect_cycles()
at various places in code path until it starts to fail in exactly the same place each time.This patch just checks if things move without old workaround.
Comment #203
Anonymous (not verified) CreditAttribution: Anonymous commentedMany thanks for the clarification. Most likely you are right, as always. I just want to finish the dissection for get the final fiasco. And one more suspect.
Comment #204
tacituseu CreditAttribution: tacituseu commented#202: Looks like removing the previous patch from
DrupalKernel::initializeContainer()
makes it go away too ;D.Comment #205
Anonymous (not verified) CreditAttribution: Anonymous commented#204: It is nice!)
But I will continue to divide (I hope that very little remains, and I have already added trash after #202 in #203).
Comment #206
tacituseu CreditAttribution: tacituseu commentedBunch of commits just went in, also test bots are having 'No space left on device' issues again.
Comment #207
Anonymous (not verified) CreditAttribution: Anonymous commented@tacituseu, thanks for the warning, it saved me from panicking!
Comment #208
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #209
tacituseu CreditAttribution: tacituseu commentedComment #210
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #211
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #212
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #213
Anonymous (not verified) CreditAttribution: Anonymous commentedSorry for the bunch of patches without comments. Here are the current research.
#208 detects fail after get cid
'entity_base_field_definitions:entity_test_cache:en'
from cache, when new FieldConfig entity callsave()
.#210 checks, that fail not occurs due to request to value from database. Because
getMultiple()
works good.Maybe problem due to use incorrect value from cache later?
#211 tries to create a new entity without using the cache. At first it seemed to me that it helps, but the retest said otherwise.
Maybe creating FieldConfig entity only first signal of problem with outdate cache?
#212 tries to clear all cache bins at the beginning of the setUp(). And looks like it helps.
So, php 5.5/5.6 bug somehow violated the cache (in the database or references), and it becomes fatal?
Comment #214
Anonymous (not verified) CreditAttribution: Anonymous commentedI know what you're thinking:
I do not know, let's see!)
Comment #215
Anonymous (not verified) CreditAttribution: Anonymous commentedopps, 10 times it more than need. I will cancel these test by manual.
Comment #216
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #217
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #218
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #219
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #220
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #221
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #222
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #223
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #224
Anonymous (not verified) CreditAttribution: Anonymous commentedComment #225
Anonymous (not verified) CreditAttribution: Anonymous commentedIt can only mean one thing, I missed something between the #208 and #210. I have a guess, but first another test of this area.
Comment #226
Anonymous (not verified) CreditAttribution: Anonymous commentedOkay. I can only explain this by bug in gc_collect_cycles. While I was catching a bug without creating new variables (just a pair of global variables with primitive values), I walked confidently in one direction. Example, all fails in #208 occur after cacheGet with cid
'entity_base_field_definitions:entity_test_cache:en'
(I added a few more runs now, and this behavior has remained)But in #225 I added a lot more variables to save the intermediate results, and the fail went to a completely different place. See also #202 explanation from @tacituseu.
Why #224 passed? These cids are not created/used in any way on the scanned area (created and save FieldConfig instance). So, most likely it's just another random action, that prevents the appearance of a bug.
Once again I apologize for all this noise. Now I'm on my own skin checked the insidiousness of gc :(
Comment #227
Anonymous (not verified) CreditAttribution: Anonymous commented#2921862: Segfault on PHP5.5 and PostgreSQL was fixed! And most likely we will use this approach further, if necessary. Do you mind if I close the issue? Because it already contains a tone of patches, interfering with comfortable work.
Comment #228
tacituseu CreditAttribution: tacituseu commented+1 for closing for now, event though it's bound to show up again as long as we're supporting 5.x.
What did we learn, Palmer? ;),
Comment #229
Anonymous (not verified) CreditAttribution: Anonymous commentedxD