1676618 - JS Debugger frequently fails to attach to mochitests

Status: RESOLVED FIXED

(DevTools :: Debugger, defect, P3)

Description

[Mark Banner (:standard8)]

Attachment: Test log from a failed run.

What were you doing?

1. Add a debugger; statement just before this line on current master: https://searchfox.org/mozilla-central/source/toolkit/components/search/tests/xpcshell/head_search.js#34
2. Run one of the search tests

./mach mochitest browser/components/search/test/browser/browser_searchbar_focus_timing.js --jsdebugger

What happened?

The javascript debugger often fails to connect and start debugging.

This morning, it happened around 5-10 times in a row, then suddenly kicked in. I then retried it after a couple of minutes, and it was successful for a few runs, and then started failing again.

I think I have also seen this on xpcshell-tests a week or so ago.

What should have happened?

The js debugger should have connected properly.

Anything else we should know?

Please see attached for a log. In particular, in the failing cases, I'm seeing these error messages appearing just before the test starts:

 0:06.45 GECKO(15721) > console.error: (new Error("Resource \"server1.conn0.contentProcess15730/source15\" already exists, cannot insert {\"id\":\"server1.conn0.contentProcess15730/source15\",\"actor\":\"server1.conn0.contentProcess15730/source15\",\"thread\":\"server1.conn0.contentProcess15730/thread3\",\"source\":\"sourceURL-resource://gre/modules/ComponentUtils.jsm\",\"isBlackBoxed\":false,\"sourceMapBaseURL\":\"resource://gre/modules/ComponentUtils.jsm\",\"sourceMapURL\":null,\"url\":\"resource://gre/modules/ComponentUtils.jsm\",\"breakpointPositions\":{},\"breakableLines\":null}", "resource://devtools/client/debugger/src/utils/resource/core.js", 42))
 0:06.46 GECKO(15721) JavaScript warning: resource://devtools/shared/builtin-modules.js, line 196: debuggee 'resource://devtools/shared/base-loader.js:289' would run
 0:06.56 GECKO(15721) > JavaScript error: resource://devtools/client/debugger/src/utils/resource/core.js, line 42: Error: Resource "server1.conn0.contentProcess15732/source6" already exists, cannot insert {"id":"server1.conn0.contentProcess15732/source6","actor":"server1.conn0.contentProcess15732/source6","thread":"server1.conn0.contentProcess15732/thread3","source":"sourceURL-resource://gre/modules/ComponentUtils.jsm","isBlackBoxed":false,"sourceMapBaseURL":"resource://gre/modules/ComponentUtils.jsm","sourceMapURL":null,"url":"resource://gre/modules/ComponentUtils.jsm","breakpointPositions":{},"breakableLines":null}
 0:07.34 GECKO(15721) > JavaScript error: resource://devtools/client/debugger/src/utils/resource/core.js, line 42: Error: Resource "server1.conn0.contentProcess15730/source158" already exists, cannot insert {"id":"server1.conn0.contentProcess15730/source158","actor":"server1.conn0.contentProcess15730/source158","thread":"server1.conn0.contentProcess15730/thread3","source":"sourceURL-resource:///actors/ContentSearchChild.jsm","isBlackBoxed":false,"sourceMapBaseURL":"resource:///actors/ContentSearchChild.jsm","sourceMapURL":null,"url":"resource:///actors/ContentSearchChild.jsm","breakpointPositions":{},"breakableLines":null}

Comment 1

[Mark Banner (:standard8)]

I'm not convinced about those errors - I've just had them on a successful run as well...

> JavaScript error: resource://devtools/client/debugger/src/utils/resource/core.js, line 42: Error: Resource "server1.conn0.contentProcess20266/source31" already exists, cannot insert {"id":"server1.conn0.contentProcess20266/source31","actor":"server1.conn0.contentProcess20266/source31","thread":"server1.conn0.contentProcess20266/thread3","source":"sourceURL-resource://devtools/server/connectors/js-window-actor/DevToolsFrameChild.jsm","isBlackBoxed":false,"sourceMapBaseURL":"resource://devtools/server/connectors/js-window-actor/DevToolsFrameChild.jsm","sourceMapURL":null,"url":"resource://devtools/server/connectors/js-window-actor/DevToolsFrameChild.jsm","breakpointPositions":{},"breakableLines":null}
 0:06.37 GECKO(20259) > JavaScript error: resource://devtools/client/debugger/src/utils/resource/core.js, line 42: Error: Resource "server1.conn0.contentProcess20266/source57" already exists, cannot insert {"id":"server1.conn0.contentProcess20266/source57","actor":"server1.conn0.contentProcess20266/source57","thread":"server1.conn0.contentProcess20266/thread3","source":"sourceURL-resource://devtools/client/jsonview/Sniffer.jsm","isBlackBoxed":false,"sourceMapBaseURL":"resource://devtools/client/jsonview/Sniffer.jsm","sourceMapURL":null,"url":"resource://devtools/client/jsonview/Sniffer.jsm","breakpointPositions":{},"breakableLines":null}
 0:06.41 GECKO(20259) JavaScript warning: resource://devtools/shared/builtin-modules.js, line 196: debuggee 'resource://devtools/shared/base-loader.js:289' would run
 0:06.44 GECKO(20259) > JavaScript error: resource://devtools/client/debugger/src/utils/resource/core.js, line 42: Error: Resource "server1.conn0.contentProcess20266/source121" already exists, cannot insert {"id":"server1.conn0.contentProcess20266/source121","actor":"server1.conn0.contentProcess20266/source121","thread":"server1.conn0.contentProcess20266/thread3","source":"sourceURL-resource://devtools/server/actors/webconsole.js","isBlackBoxed":false,"sourceMapBaseURL":"resource://devtools/server/actors/webconsole.js","sourceMapURL":null,"url":"resource://devtools/server/actors/webconsole.js","breakpointPositions":{},"breakableLines":null}
 0:06.67 GECKO(20259) > JavaScript error: resource://devtools/client/debugger/src/utils/resource/core.js, line 42: Error: Resource "server1.conn0.contentProcess20266/source131" already exists, cannot insert {"id":"server1.conn0.contentProcess20266/source131","actor":"server1.conn0.contentProcess20266/source131","thread":"server1.conn0.contentProcess20266/thread3","source":"sourceURL-resource://gre/modules/PrivateBrowsingUtils.jsm","isBlackBoxed":false,"sourceMapBaseURL":"resource://gre/modules/PrivateBrowsingUtils.jsm","sourceMapURL":null,"url":"resource://gre/modules/PrivateBrowsingUtils.jsm","breakpointPositions":{},"breakableLines":null}
 0:07.70 GECKO(20259) > JavaScript error: resource://devtools/client/debugger/src/utils/resource/core.js, line 42: Error: Resource "server1.conn0.contentProcess20266/source160" already exists, cannot insert {"id":"server1.conn0.contentProcess20266/source160","actor":"server1.conn0.contentProcess20266/source160","thread":"server1.conn0.contentProcess20266/thread3","source":"sourceURL-resource:///actors/ContentSearchChild.jsm","isBlackBoxed":false,"sourceMapBaseURL":"resource:///actors/ContentSearchChild.jsm","sourceMapURL":null,"url":"resource:///actors/ContentSearchChild.jsm","breakpointPositions":{},"breakableLines":null}

Comment 2

[Jan Honza Odvarko [:Honza] (always need-info? me)]

Thanks for the report!

I can reproduce the bug using STRs from comment #0

It works for me for one of the DevTools tests
STR

1. Add debugger; keyword at this line https://searchfox.org/mozilla-central/rev/02cb78667e87ccc42fea5edc6f3f2dd2edd6ecd5/devtools/client/netmonitor/test/browser_net_api-calls.js#23
2. Run mach test devtools/client/netmonitor/test/browser_net_api-calls.js --jsdebugger
3. The debugger opens and hits the breakpoint.

Honza

Comment 3

[Julian Descottes [:jdescottes]]

I am consistently getting this this morning when adding a debugger; statement before https://searchfox.org/mozilla-central/source/devtools/client/framework/test/browser_toolbox_error_count.js#29 and running

./mach test devtools/client/framework/test/browser_toolbox_error_count.js --jsdebugger

Comment 4

[Julian Descottes [:jdescottes]]

Looks like getting rid of the checks at https://searchfox.org/mozilla-central/rev/c54c71ecbd9e64cafc0df3b596e206ac4072cb91/devtools/client/debugger/src/utils/resource/core.js#49-60 makes the debugger statements work again. Question is, are those checks necessary (probably?) and if yes why do we get in there too many times.

Comment 5

[Julian Descottes [:jdescottes]]

Attachment: Bug 1676618 - [devtools] Do not create legacy listeners twice for same target

Comment 6

[Julian Descottes [:jdescottes]]

Attachment: Bug 1676618 - [devtools] Only return emitted targets from TargetList::getAllTargets

Comment 7

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/0435260ce644
[devtools] Do not create legacy listeners twice for same target r=ochameau,nchevobbe

Comment 8

[Natalia Csoregi [:nataliaCs]]

https://hg.mozilla.org/mozilla-central/rev/0435260ce644

Comment 9

[Julian Descottes [:jdescottes]]

Note that this is broken again, I'm going to look at this in Bug 1687645.