logging_FeedbackReport mostly green, I see one failure on cyan: 
https://stainless.corp.google.com/search?test=%5Elogging%5C_&exclude_non_release=true&exclude_cts=true&col=build&row=test&view=matrix&first_date=2018-06-20&last_date=2018-06-26

Devices reported in these feedback records are 
- eve(2) for M68 and 69, 
- kevin for M69 
- n..i - for M69

Test results for these are green - https://screenshot.googleplex.com/o3N66Fr6RCA



For one of the board(kevin) feedback there is a section with some errors

board-specific=/bin/sh: 1: /usr/share/userfeedback/scripts/get_board_specific_info: not found
buddyinfo=Node 0, zone      DMA   3434   1647   4949   6123   2859   1217    398    131     48      9      1
cbi_info=<multiline>
---------- START ----------
Cannot stat /run/lock.
Trying fallback directory: /tmp
Cannot find I2C adapter
Unable to establish host communication
Couldn't find EC
/usr/share/userfeedback/scripts/cbi_info: ERROR: Failed to read type 0
Cannot stat /run/lock.
Trying fallback directory: /tmp
Cannot find I2C adapter
Unable to establish host communication
Couldn't find EC
/usr/share/userfeedback/scripts/cbi_info: ERROR: Failed to read type 1
Cannot stat /run/lock.
Trying fallback directory: /tmp
Cannot find I2C adapter
Unable to establish host communication
Couldn't find EC
/usr/share/userfeedback/scripts/cbi_info: ERROR: Failed to read type 2
[0]
[1]
[2]
---------- END ----------

I'm having this issue on my Pixelbook (and am the submitter of one of the reports above).

It does work in Guest mode, so we're thinking this *may* be an issue with an extension. Not sure why this is coming up with M68 Beta, as it hasn't been an issue before.

Just kidding, now I can't get it to work, even in guest mode.

Feedback report with logs (sent from Settings menu instead):
https://listnr.corp.google.com/report/85525064051

I flashed the same version you have 68.0.3440.40 / 10718.34.0 on my pixelbook, and I can't repro the issue. Alt+Shift+i works as expected.

Questions to help me diagnose:
- Did you map one of your keys to anything else from the keyboard settings?
- Are you using a non-US keyboard layout (like US-EXT for example)?

For the crash in report (2): It has the following logs:

2018-06-22T12:07:55.358790-07:00 INFO session_manager[1468]: [INFO:session_manager_service.cc(201)] SessionManagerService exiting
2018-06-22T12:07:55.364936-07:00 INFO crash_reporter[7271]: Processing selinux violation: dropped sample
2018-06-22T12:07:55.366007-07:00 ERR crash_reporter[7274]: Failed to get PID of ARC container
2018-06-22T12:07:55.366179-07:00 WARNING crash_reporter[7274]: [user] Received crash notification for supplied_init[2809] sig 6, user 655360 (handling)
2018-06-22T12:07:55.367186-07:00 ERR crash_reporter[7273]: Failed to get PID of ARC container
2018-06-22T12:07:55.367354-07:00 WARNING crash_reporter[7273]: [user] Received crash notification for supplied_init[2808] sig 6, user 655360 (handling)
2018-06-22T12:07:55.368468-07:00 INFO crash_reporter[7274]: Using supplied UID 655360 for crashed process [2809] due to error reading status file
2018-06-22T12:07:55.369353-07:00 INFO crash_reporter[7273]: Using supplied UID 655360 for crashed process [2808] due to error reading status file
2018-06-22T12:07:55.370696-07:00 ERR crash_reporter[7273]: Path /proc/2808 does not exist
2018-06-22T12:07:55.371234-07:00 INFO crash_reporter[7273]: Skipped converting core file to minidump due to unusable proc files
2018-06-22T12:07:55.371253-07:00 INFO crash_reporter[7273]: Leaving core file at /var/spool/crash/supplied_init.20180622.120755.2808.core due to conversion error
2018-06-22T12:07:55.371271-07:00 INFO crash_reporter[7273]: Writing conversion problems as separate crash report.
2018-06-22T12:07:55.373825-07:00 ERR crash_reporter[7274]: Path /proc/2809 does not exist
2018-06-22T12:07:55.374254-07:00 INFO crash_reporter[7274]: Skipped converting core file to minidump due to unusable proc files
2018-06-22T12:07:55.374277-07:00 INFO crash_reporter[7274]: Leaving core file at /var/spool/crash/supplied_init.20180622.120755.2809.core due to conversion error
2018-06-22T12:07:55.374294-07:00 INFO crash_reporter[7274]: Writing conversion problems as separate crash report.
2018-06-22T12:08:20.039985-07:00 WARNING crash_reporter[10258]: [user] Received crash notification for <email: 1>ne[7935] sig 11, user 656373 (ignoring - PID filtered out)
2018-06-22T12:08:20.040846-07:00 WARNING crash_reporter[10258]: [ARC] Received crash notification for <email: 1>ne[7935] sig 11, user 656373 (handling)
2018-06-22T12:08:20.044248-07:00 INFO crash_reporter[10258]: State of crashed process [7935]: D (disk sleep)
2018-06-22T12:08:20.102663-07:00 INFO crash_reporter[10279]: libminijail[10279]: mount /dev/log -> /dev/log type ''
2018-06-22T12:08:20.127854-07:00 INFO crash_reporter[10279]: Processing selinux violation: dropped sample
2018-06-22T12:08:20.166628-07:00 INFO crash_reporter[10258]: Stored minidump to /home/user/a805f4caaf37a49e09428985c1cacf8c0c8a21f7/crash/android_hardware_drm_1_1_service_widevine.20180622.120820.7935.dmp


+elijah, do you have any idea what "android_hardware_drm_1_1_service_widevine" or "supplied_init" are? This report is coming from a "kevin-arcnext-release/R69-10800.0.0" device, which might be related to issue b/110716828.

Crash in report (3) has the following logs:

2018-06-22T11:30:32.352540+02:00 WARNING crash_reporter[15911]: Received crash notification for chrome[3803] user 1000 (called directly)
2018-06-22T11:30:32.397784+02:00 INFO session_manager[1682]: [INFO:child_exit_dispatcher.cc(59)] Handling 15911 exit.
2018-06-22T11:30:32.397796+02:00 ERR session_manager[1682]: [ERROR:child_exit_dispatcher.cc(67)]   Exited with signal 9
2018-06-22T11:30:32.399652+02:00 INFO kernel: [ 2645.662734] traps: chrome[3803] trap invalid opcode ip:5961513c2446 sp:7fff2778d660 error:0 in chrome[59614a680000+94d0000]
2018-06-22T11:30:32.404586+02:00 INFO crash_reporter[15932]: libminijail[15932]: mount /dev/log -> /dev/log type ''
2018-06-22T11:30:32.423364+02:00 WARNING crash_reporter[15932]: [user] Received crash notification for chrome[3803] sig 4, user 1000 (ignoring call by kernel - chrome crash; waiting for chrome to call us directly)
2018-06-22T11:30:32.423531+02:00 WARNING crash_reporter[15932]: [ARC] Received crash notification for chrome[3803] sig 4, user 1000 (ignoring - crash origin is not ARC)


Unfortunately, I can't see links to crash reports from these feedback reports. +sling@ +yyushkina@ Did the ability to link crash reports (happened within an hour from the time the report is filed) from the feedback report regress?

Devlin, for reports (1) and (2), I'm afraid that there might be a regression in `EventRouter::DispatchEventWithLazyListener()` https://cs.chromium.org/chromium/src/extensions/browser/api/feedback_private/feedback_private_api.cc?q=FeedbackPrivateAPI::RequestFeedbackForFlow&sq=package:chromium&g=0&l=167

The EventRouter code hasn't changed significantly in months, so I wouldn't guess it's a regression there.

https://chromium.googlesource.com/chromium/src/+log/d0e917c2b35bf8bb2aab02c6e4735024562cbd35/extensions/browser/event_router.cc

Unfortunately, I don't really know how to read these logs - is there anything in particular I should be lookin gat?

Correction in #12: I meant to say for reports (1) and (4).

Devlin, I looked at the logs and couldn't find anything that stood out, but the fact that report (4) says it always fails the first time the (Alt+Shift+i) shortcut is invoked might mean that `DispatchEventWithLazyListener()` it's failing on the first attempt. This brought back memories of issue 754329. 

May be worth an email to sling@ - I doubt he'll check this. I am not aware of any changes that'd cause this regression but I'm also not in the loop.

Entirely possible.  It'd be worth checking prefs-internals to see what the saved preferences look like for it.  Do we have reliable repro steps?

I can't repro this myself.

trumbull@ and igo@ could you please share the contents of your "chrome://prefs-internals"?

Here's my chrome://prefs-internals contents (available to Googlers): https://drive.google.com/file/d/1ww2jsS4h4tknDliehC8RyAhK8nDL1o9z/view?usp=sharing

No clear repro steps.

Since updating to M68 Beta, I've been able to get Alt-shift-i working twice:
(1) one time after nothing particular
(2) once after updating, restarting, and using guest mode

When I experience this issue, restarting doesn't seem to do anything.

Thanks, trumbull@

Devlin, trumbull's prefs-internal doesn't have an entry at all for the feedback extension ID "gfdkimpbcpahaombhbimeihdjnejgicl" nor any listener for the "feedbackPrivate.onFeedbackRequested" event.

@20: Looking at the file, it actually does - it's just that there's so many policy-blocked extensions that google drive doesn't load the whole thing so Ctrl+F doesn't work. :)

It looks like the extension is installed (has an entry) and enabled (state: 1), and is correctly registered for events ("events": [ "feedbackPrivate.onFeedbackRequested", "runtime.onMessageExternal" ]).

(Not pasting the whole preference here for privacy reasons.)

So it *looks* like the extension entry is right - which leaves us with possibilities of bugs in the feedback private code (to listen for and dispatch the event) or in the extension event dispatching code itself.

One hacky way we might be able to guess:
chrome::ShowFeedbackPage() (which calls into the extension API to dispatch the event) emits a histogram when it's called [1].  So if you go to chrome://histograms, you should see entries there.  A basic test might be go to chrome://histograms and see what the entries for Feedback.RequestSource are.  Then, hit the keyboard chord multiple times, and if the histogram counter for that source increases without the window appearing, then it sounds like it's probably a bug in the event dispatching code.  If the counter doesn't increase, it seems like it's a bug in the feedback code.  It's pretty rough, but it *may* work?

(Though obviously, if we can get a consistent repro, it would be much better :))

[1] https://chromium.googlesource.com/chromium/src/+/ed47f31beff24c6610486dbc958948cee7a68fd3/chrome/browser/feedback/show_feedback_page.cc#39

Thanks, Devlin! Sorry I should have scrolled the page till the end to force load it entirely.

trumbull@ Could you please follow Devlin's instructions in #21?

Searched the chrome://histograms page for "Feedback" and only found "Accessibility.CrosSpokenFeedback".

Tapped alt-shift-i a bunch of times and then searched again. No change.

trumbull@ Please make sure you click the "Refresh" button at the top of the "chrome://histograms" page after every time you invoke the "Alt+Shift+i" shortcut.

Thanks for the clarification. I tried again, invoking alt-shift-i both before and after hitting refresh a few times and no change. 

Still don't see any entries for feedback (other than "Accessibility.CrosSpokenFeedback)

Thanks for the data, trumbell@!

Judging from #23 and #25, sounds like maybe the feedback code is never getting to the point where the event is dispatched?  Unfortunately, that means that I'm likely not going to be much help, but hopefully it's a starting point.

trumbull@ Did you remap any of your modifier keys in the keyboard settings: "chrome://settings/keyboard-overlay" differently?

No. 

Launcher, Ctrl, Alt, Escape, and Backspace are all mapped the same way.

I have "Treat top-row keys as function keys" as OFF and "Enable auto-repeat" as ON.

Sometimes I turn on Accessibility settings for testing, but it's off now.

Also, I just tried and now Alt-shift-i is working for me (on 68.0.3440.59).

Glad to know it's working now. I'm not sure what was the issue since I was unable to repro.

Closing since it's no longer an issue. Please reopen if it comes up again.