TEST-UNEXPECTED-FAIL | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | an unexpected uncaught JS exception reported through window.onerror - TypeError: domRules is null at resource:///modules/devtools/CssLogic.jsm:632 https://tbpl.mozilla.org/?tree=Fx-Team&rev=c77b869c2025 https://tbpl.mozilla.org/php/getParsedLog.php?id=17481911&tree=Fx-Team TEST-INFO | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | START Remove an element with the delete key TEST-PASS | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Node 18 should exist. TEST-PASS | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Node 18 should not exist. TEST-PASS | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Should be able to undo. TEST-PASS | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Node 18 should exist. TEST-PASS | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Should be able to redo. TEST-UNEXPECTED-FAIL | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | an unexpected uncaught JS exception reported through window.onerror - TypeError: domRules is null at resource:///modules/devtools/CssLogic.jsm:632 Stack trace: JS frame :: chrome://mochikit/content/tests/SimpleTest/SimpleTest.js :: simpletestOnerror :: line 1066 native frame :: <unknown filename> :: <TOP_LEVEL> :: line 0 TEST-INFO | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Console message: [JavaScript Warning: "XUL box for span element contained an inline #text child, forcing all its children to be wrapped in a block." {file: "chrome://browser/content/devtools/layoutview/view.xhtml" line: 0}] TEST-PASS | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Node 18 should not exist. TEST-INFO | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | END Remove an element with the delete key TEST-INFO | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | Console message: [JavaScript Error: "TypeError: domRules is null" {file: "resource:///modules/devtools/CssLogic.jsm" line: 632}] INFO TEST-END | chrome://mochitests/content/browser/browser/devtools/markupview/test/browser_inspector_markup_edit.js | finished in 1638ms

(TBPL & OrangeFactor just require the intermittent-failure keyword in order for bug suggestions to work; [orange] is depreciated - dev.platform post coming shortly)

Mike, do you see anything obvious here? The failing code is related to your recent changes in browser_inspector_markup_edit.js: https://github.com/joewalker/devtools-window/commit/354ad6fc925f6897cfd677a8f826c7226d43461d https://github.com/joewalker/devtools-window/commit/2a87fcf1d52aac8cb189c68ac4802e7e66f92458

I think this is line 369 of https://github.com/joewalker/devtools-window/commit/354ad6fc925f6897cfd677a8f826c7226d43461d#L0L371 I had to add executeSoon here because the undo/redo manager takes time before undo will work properly. Obviously, some kind of event would work better.

(In reply to Michael Ratcliffe [:miker] [:mratcliffe] from comment #19) > I think this is line 369 of > https://github.com/joewalker/devtools-window/commit/ > 354ad6fc925f6897cfd677a8f826c7226d43461d#L0L371 > > I had to add executeSoon here because the undo/redo manager takes time > before undo will work properly. Obviously, some kind of event would work > better. What is asynchronous here? Why do we need executeSoon in the first place?

Making a change, undoing and redoing in rapid succession can currently cause this problem (the undo/redo manager is not finished updating itself). If the undo/redo manager fired an event to signify that it is ready we could overcome this. What is asynchronous here? Why do we need executeSoon in the first place? I have no idea which parts of the undo/redo manager are async. All I know is that the process I describe throws that error if you don't use executeSoon.

I think I have a fix for that (part of bug 814889).

Unfortunately this is caused by the mutation observer. We call Highlighter.jsm::updateInfobar() using Selection.jsm:87 (this.emit("attribute-changed")) Because the MutationObserver gathers *batches* of events and then periodically calls the _onMutations listener (only then emitting "attribute-changed") we have the async issue. We can wait for markupmutation before calling executeSoon and it should work.

No new reports for 3 weeks? Anyhow, waiting for "markupmutation" before setTimeout can't hurt. This is what we always did before calling undo and redo ... we never had issues at that point but only used setTimeout after redo.

(In reply to Michael Ratcliffe [:miker] [:mratcliffe] from comment #42) > Created attachment 703236 [details] [diff] [review] > Wait for "markupmutation" event before calling setTimeout > > No new reports for 3 weeks? I made a lot of changes in this test. > Anyhow, waiting for "markupmutation" before setTimeout can't hurt. This is > what we always did before calling undo and redo ... we never had issues at > that point but only used setTimeout after redo. Are you sure all the tests cause a mutation? Why do we need to be in a second event loop to call test.after?

I probably fixed this bug while fixing bug 827990 or bug 814889. Let's re-open if we run into new oranges.