https://treeherder.mozilla.org/logviewer.html#?job_id=7474990&repo=fx-team 14:40:23 INFO - 5929 INFO TEST-UNEXPECTED-FAIL | devtools/client/inspector/markup/test/browser_markup_load_01.js | The markup view is rendering - Got 0, expected 1
This started failing when I landed bug 1244120 I think. I'll take a look at this one.
Assignee: nobody → pbrosset
Status: NEW → ASSIGNED
Priority: -- → P2
Review commit: https://reviewboard.mozilla.org/r/37143/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/37143/
Attachment #8724750 - Flags: review?(gl)
Just making the test more safe by waiting for the markup-view to be loaded fully fixes it. I have re-triggered many many of the corresponding dt jobs on this try push: https://treeherder.mozilla.org/#/jobs?repo=try&revision=c27ca9cf87da&group_state=expanded
Comment on attachment 8724750 [details] MozReview Request: Bug 1250523 - Wait for the markup-view to be loaded in browser_markup_load_01.js; r=gl https://reviewboard.mozilla.org/r/37143/#review33687
Attachment #8724750 - Flags: review?(gl) → review+
Status: ASSIGNED → RESOLVED
Last Resolved: 3 years ago
status-firefox47: --- → fixed
Resolution: --- → FIXED
Target Milestone: --- → Firefox 47
This is still happening.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Inspector bugs triage. Filter on CLIMBING SHOES. Very frequent failures last week -> P1 Taking this one as Pbro will be on PTO next week.
Assignee: pbrosset → jdescottes
Status: REOPENED → ASSIGNED
Priority: P2 → P1
If anyone happens to care, here are some pointers on where to look at. tl;dr: After reload, the new root node does not have a BODY element as a descendant. HTML node is selected but it's parent node is different from the root node. In the case when this test fails, the following sequence of events occurs after reload in the test: 1) A newRoot mutation gives inspector a new, previously unseen #document node with ID server1.conn44.domnode81 as the root node. 2) The new root handler goes on calling getDefaultNodeForSelection which tries to look for the IMG and BODY nodes as a descendant of the new root node but does not find either 3) It falls back to asking for the document element and receives HTML element that has a parent of server1.conn44.domnode86 which is different from the received new root node. Eventually importNode is called for the HTML element and its parent but since the parent node of the HTML node was server1.conn44.domnode86 and the root node is server1.conn44.domnode81,  never passes and the markupview stays empty forever. I've attached the relevant parts from a log extra with debugging information. The full log can be found from  (dunno how long it stays around)  https://dxr.mozilla.org/mozilla-central/source/devtools/client/inspector/markup/markup.js#851  http://firstname.lastname@example.org/try-linux64-debug/try_ubuntu64_vm-debug_test-mochitest-devtools-chrome-1-bm53-tests1-linux64-build1483.txt.gz
Component: Developer Tools → Developer Tools: Inspector
Adjusting the priority down since this doesn't fail as often as it used to now.
Priority: P1 → P3
Assignee: jdescottes → nobody
Status: ASSIGNED → NEW
You need to log in before you can comment on or make changes to this bug.