Closed
Bug 1087038
Opened 10 years ago
Closed 10 years ago
Gaia is closed, perma failures after last m-c -> b-i merge
Categories
(Remote Protocol :: Marionette, defect)
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: kgrandon, Assigned: timdream)
References
Details
(Keywords: regression)
Attachments
(1 file, 2 obsolete files)
2.44 KB,
patch
|
Details | Diff | Splinter Review |
The last m-c -> b-i merge seems to have broken us as usual. We really need to get our Gij suite unhidden. https://treeherder.mozilla.org/ui/#/jobs?repo=b2g-inbound&revision=15099ba111e8 Failures: http://ftp.mozilla.org/pub/mozilla.org/b2g/tinderbox-builds/b2g-inbound-linux64_gecko/1413919356/b2g-inbound_ubuntu64_vm-b2gdt_test-gaia-js-integration-1-bm117-tests1-linux64-build10.txt.gz http://ftp.mozilla.org/pub/mozilla.org/b2g/tinderbox-builds/b2g-inbound-linux64_gecko/1413919356/b2g-inbound_ubuntu64_vm-b2gdt_test-gaia-js-integration-3-bm53-tests1-linux64-build1.txt.gz http://ftp.mozilla.org/pub/mozilla.org/b2g/tinderbox-builds/b2g-inbound-linux64_gecko/1413919356/b2g-inbound_ubuntu64_vm-b2gdt_test-gaia-js-integration-4-bm68-tests1-linux64-build2.txt.gz
Reporter | ||
Comment 1•10 years ago
|
||
We need bug 960072 so this doesn't happen to us in the future.
See Also: → 960072
Reporter | ||
Comment 3•10 years ago
|
||
It looks like bug 1084412 changes findElement which we use under the hood in waitForElement: https://github.com/mozilla-b2g/marionette-helper/blob/b205ef7041ee7148e6aa2200667b052911178f49/index.js#L135
Assignee | ||
Comment 4•10 years ago
|
||
I can reproduce this error locally after updating B2G Desktop: https://treeherder.mozilla.org/ui/logviewer.html#?job_id=676527&repo=b2g-inbound There seems to be a legit regression on the marionette server side -- I can't bypass the error in anyway in waitForElement on the client side, and when the error throws I can clearly see the element is on screen. I am building B2G Desktop to see if I can debug Gecko (instead of the client).
Assignee | ||
Comment 5•10 years ago
|
||
I can confirm that backing out bug 1084412 can make the test passes locally, however I am now suspect that patch is legit and it simply correct the false positive of our tests somehow. Going to insert more throw into Gecko and see what's wrong.
Assignee | ||
Comment 6•10 years ago
|
||
The cause of the failure on comment 4 (and comment 4 along) is indeed due to logic error in bug 1084412. The check in https://hg.mozilla.org/mozilla-central/rev/56e2d9aff637#l2.13 will fail for HTMLFormElement because it comes with a length property === 0. I have a patch ready.
Assignee | ||
Updated•10 years ago
|
Assignee: nobody → timdream
Status: NEW → ASSIGNED
Assignee | ||
Comment 7•10 years ago
|
||
Attachment #8509249 -
Flags: review?(dburns)
Attachment #8509249 -
Flags: feedback?(ahalberstadt)
Assignee | ||
Updated•10 years ago
|
Keywords: regression
Assignee | ||
Comment 8•10 years ago
|
||
https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=1efd6b1b78d5 This is definitely not the complete set involving this patch, but I don't know what to run since bug 1084412 didn't come with a try run...
Assignee | ||
Updated•10 years ago
|
Component: Gaia → Marionette
Product: Firefox OS → Testing
Updated•10 years ago
|
Attachment #8509249 -
Flags: review?(dburns) → review+
Assignee | ||
Comment 9•10 years ago
|
||
Attachment #8509249 -
Attachment is obsolete: true
Attachment #8509249 -
Flags: feedback?(ahalberstadt)
Assignee | ||
Updated•10 years ago
|
Keywords: checkin-needed
Comment 10•10 years ago
|
||
https://hg.mozilla.org/integration/mozilla-inbound/rev/04a18f29f774
Keywords: checkin-needed
Assignee | ||
Comment 11•10 years ago
|
||
(In reply to Carsten Book [:Tomcat] from comment #10) > https://hg.mozilla.org/integration/mozilla-inbound/rev/04a18f29f774 This needs to be backed out ... there is a |throw| for debug that I didn't remove :-/
Flags: needinfo?(cbook)
Keywords: checkin-needed
Assignee | ||
Comment 12•10 years ago
|
||
Attachment #8509375 -
Attachment is obsolete: true
Comment 13•10 years ago
|
||
(In reply to Tim Guan-tin Chien [:timdream] (MoCo-TPE) (please ni?) from comment #11) > (In reply to Carsten Book [:Tomcat] from comment #10) > > https://hg.mozilla.org/integration/mozilla-inbound/rev/04a18f29f774 > > This needs to be backed out ... there is a |throw| for debug that I didn't > remove :-/ this got backouted in remote: https://hg.mozilla.org/integration/mozilla-inbound/rev/4243e118f58d
Flags: needinfo?(cbook)
Comment 14•10 years ago
|
||
(In reply to Tim Guan-tin Chien [:timdream] (MoCo-TPE) (please ni?) from comment #12) > Created attachment 8509427 [details] [diff] [review] > Real patch for commit landed remote: https://hg.mozilla.org/integration/mozilla-inbound/rev/2b1c2663f92a
Assignee | ||
Updated•10 years ago
|
Keywords: checkin-needed
Assignee | ||
Comment 15•10 years ago
|
||
https://treeherder.mozilla.org/ui/#/jobs?repo=mozilla-inbound&revision=2b1c2663f92a
Comment 16•10 years ago
|
||
This is a try that Tim triggered some time ago: https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=0eb1ea3c9007
Comment 17•10 years ago
|
||
Here is another try: https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=8a0b20c634a1 Hopefully we'll get the test results this time.
Reporter | ||
Comment 18•10 years ago
|
||
(In reply to Tim Guan-tin Chien [:timdream] (MoCo-TPE) (please ni?) from comment #15) > https://treeherder.mozilla.org/ui/#/jobs?repo=mozilla- > inbound&revision=2b1c2663f92a If this is the real landing then we are still in trouble here. Looks like we still have quite a few Gij failures.
Reporter | ||
Comment 19•10 years ago
|
||
It seems we are *still* getting failures here. Since the tree is closed, I'm going to recommend that we backout this patch and the patches in bug 1084412. I hate that we have to do this, but FirefoxOS has many deadlines, and currently hundreds of people are blocked from landing patches. Tim/Andrew/David - Is this approach ok with you? If we are ok backing out, we can definitely find gaia people to help update our tests/libraries where necessary.
Flags: needinfo?(timdream)
Flags: needinfo?(dburns)
Flags: needinfo?(ahalberstadt)
Comment 21•10 years ago
|
||
I backed it out here: https://hg.mozilla.org/mozilla-central/rev/305d24174ace I'd ask a sheriff for help getting this to propagate to b-i
Flags: needinfo?(timdream)
Flags: needinfo?(dburns)
Reporter | ||
Comment 22•10 years ago
|
||
(In reply to Andrew Halberstadt [:ahal] from comment #21) > I backed it out here: > https://hg.mozilla.org/mozilla-central/rev/305d24174ace > > I'd ask a sheriff for help getting this to propagate to b-i Thanks for the understanding Andrew. Apologies for this - will make getting our tree un-hidden a top priority. Do we need to backout the patch in this bug as well? https://hg.mozilla.org/integration/mozilla-inbound/rev/2b1c2663f92a
Comment 23•10 years ago
|
||
Yes please Kevin. Maybe it's been backed out with the other back out though (because it was a fix for the previous patch) ?
Comment 24•10 years ago
|
||
Nope, ahal's backout was directly to m-c where comment 15 hadn't been merged to yet. I've backed it out from inbound now and will merge everything around when things are green enough for me to do so. https://hg.mozilla.org/integration/mozilla-inbound/rev/0f76f410b03b
Reporter | ||
Comment 25•10 years ago
|
||
The next issue to look at is another remaining Gij failure: https://tbpl.mozilla.org/php/getParsedLog.php?id=50810640&tree=Mozilla-Central I think the changelog that introduces this is the very next merge: https://tbpl.mozilla.org/?showall=1&rev=367d8d88c2cb
Reporter | ||
Comment 26•10 years ago
|
||
It's hard to say due to the other failures, but I suspect that the fullscreen issue was introduced in this changeset: https://tbpl.mozilla.org/?tree=Mozilla-Inbound&showall=1&jobname=b2g_ubuntu64_vm%20mozilla-inbound%20opt%20test%20gaia-js-integration&rev=2cc8826abef2 Most likely culprit: bug 1082652
Blocks: 1082652
Flags: needinfo?(etienne)
Assignee | ||
Comment 27•10 years ago
|
||
https://treeherder.mozilla.org/ui/#/jobs?repo=mozilla-inbound&revision=6066a2a0766f There are still failures after both bug 1084412 and this bug reverted from m-i. I have just hit retrigger on Gij-3 here ...
Assignee | ||
Comment 28•10 years ago
|
||
We reaches green after backing out bug 1082652. Let's call this fixed.
Status: ASSIGNED → RESOLVED
Closed: 10 years ago
Flags: needinfo?(etienne)
Resolution: --- → FIXED
Assignee | ||
Comment 29•10 years ago
|
||
(In reply to Tim Guan-tin Chien [:timdream] (MoCo-TPE) (please ni?) from comment #28) > We reaches green after backing out bug 1082652. Let's call this fixed. https://tbpl.mozilla.org/?tree=B2g-Inbound&jobname=b2g_ubuntu64_vm%20b2g-inbound%20opt%20test%20gaia-js-integration&rev=fc8e45f09f78
Updated•1 year ago
|
Product: Testing → Remote Protocol
You need to log in
before you can comment on or make changes to this bug.
Description
•