Unit tests running on Windows sometimes claim RenderPadding violates the intrinsic protocol contract #19696
Labels
a: tests
"flutter test", flutter_test, or one of our tests
c: flake
Tests that sometimes, but not always, incorrectly pass
customer: mulligan (g3)
framework
flutter/packages/flutter repository. See also f: labels.
platform-windows
Building on or for Windows specifically
Comments
Hixie
added
a: tests
|
I just saw this with RenderPadding again, though a different test... |
This was referenced Aug 4, 2018
|
Those two were also RenderPadding... |
|
Not sure if this is the same issue - just seen a similar failure on my PR on a Windows build, but it's a different test ( |
|
Interesting, that one is RenderPadding too... |
|
I have a CL in g3 that is reliably producing this exception. I was just changing the route transitions to use The |
|
@alanrussian Are you able to reduce that to a single main.dart you can upload here? Being able to reproduce it has been the issue here... |
|
I haven't tried but that sounds unlikely. If you need that to go forward, you can remove customer blocker and I guess I can suppress the test I have that's failing for the time being... |
|
It would really help to get a standalone main.dart. I'll drop the customer blocker label from this until you have time to put it together for us, and you can suppress the test at the moment --- feel free to tag again if you disagree. Thanks! |
|
Just witnessed this with https://gist.github.com/tvolkert/9cb808b5a2ef8f9248eef048887cb8d2 |
|
The issue cropped up again today after an engine roll, #24345. Failure was on Cirrus/Windows: @liyuqian pointed out that all Cirrus tests have to be passed before the merging of an auto-roll, so having the same test fail afterwards seems odd. Re-running the tests until they pass :-). |
|
The test runnings will continue until the build succeeds! Here's another instance, again on Cirrus/Windows: https://cirrus-ci.com/task/4917027888168960 |
|
Also happened to me on Cirrus Windows https://cirrus-ci.com/task/5637587874086912 |
|
I saw this again: https://cirrus-ci.com/task/4667445715402752 |
gspencergoog
changed the title
|
I've seen a bunch of these lately, and they all happen only on Windows, usually with Some Cirrus logs from the same PR where it failed twice in a row and then succeeded: Another recent example: |
gspencergoog
added a commit
that referenced
this issue
May 16, 2019
Date picker tests have been flaky on Windows for a long time, and repeatedly require restarting of bots. In the interest of productivity, disabling them on Windows. Related to #19696
Hixie
added a commit
to Hixie/flutter
that referenced
this issue
Jun 13, 2019
kiku-jw
pushed a commit
to kiku-jw/flutter
that referenced
this issue
Jun 14, 2019
Date picker tests have been flaky on Windows for a long time, and repeatedly require restarting of bots. In the interest of productivity, disabling them on Windows. Related to flutter#19696
|
https://gist.github.com/aam/cbf08d577d5226305072eaa2077083b1 captures output of |
|
Tracing the code further i bet the bug is in the TextPainter or dart:ui's Paragraph logic. We should probably add some asserts down in TextPainter that verifies that the intrinsic values we get from there make sense. |
dart-bot
pushed a commit
to dart-lang/sdk
that referenced
this issue
Jul 11, 2019
…en if it can't add method extractor. The fact that function resolution returned a target led to incorrect range merging for polymorphic instance call sites. Fixes flutter/flutter#19696 Fixes #37348 Change-Id: If6c53dd322f9788df241baf7071b98906336d2e5 Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/108860 Commit-Queue: Alexander Aprelev <aam@google.com> Reviewed-by: Ryan Macnak <rmacnak@google.com>
|
Reported to be fixed by dart-lang/sdk@7139a2a, which will roll into Flutter soon! |
|
cc @aam this is fixed right? |
Right. |
|
This thread has been automatically locked since there has not been any recent activity after it was closed. If you are still experiencing a similar issue, please open a new bug, including the output of |
flutter-triage-bot
bot
added
the
c: flake
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Seen on Appveyor for a PR that had only documentation changes:
The text was updated successfully, but these errors were encountered: