New issue
Advanced search Search tips

Issue 891584 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Oct 4
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 1
Type: Bug



Sign in to add a comment

Auto-hidden Disconnect Window is reshown w/o local input

Project Member Reported by joedow@chromium.org, Oct 3

Issue description

A recent update to the Me2Me disconnect window for Windows was to allow it to auto-hide after several seconds of inactivity on the local machine.  After shipping the change we started receiving reports of the window being reshown when there is no local input.

I suspect this could be due to a HID device sending spurious input events.  More investigation will be needed to know for sure.
 
Project Member

Comment 1 by bugdroid1@chromium.org, Oct 3

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/df242b026028cda8d2c3564f0e5129059d7bb7bd

commit df242b026028cda8d2c3564f0e5129059d7bb7bd
Author: Joe Downing <joedow@chromium.org>
Date: Wed Oct 03 04:17:36 2018

Ignore mouse messages if the position or button states don't change

We've received reports that the disconnect window is appearing at
odd times after the auto-hide timeout has taken effect.  I suspect
there could be hardware which is emitting spurious HID events which
is triggering the reshow logic.  My assumption is that this behavior
is't egregious, otherwise the user would notice the mouse cursor
moving or errant keypresses.  Thus I want to prevent triggering the
reshow logic if we receive a RAWINPUT event where the cursor has not
moved and no button states have changed.

Bug:  891584 

Change-Id: Id9efd72f24f8caf0e67b1a717fee2bbafaab55af
Reviewed-on: https://chromium-review.googlesource.com/c/1258316
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Cr-Commit-Position: refs/heads/master@{#596121}
[modify] https://crrev.com/df242b026028cda8d2c3564f0e5129059d7bb7bd/remoting/host/input_monitor/local_mouse_input_monitor_win.cc

Project Member

Comment 2 by bugdroid1@chromium.org, Oct 3

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/edc50bc7d6901596d8579db67470edd17e827e9c

commit edc50bc7d6901596d8579db67470edd17e827e9c
Author: Joe Downing <joedow@chromium.org>
Date: Wed Oct 03 19:02:53 2018

Do not reshow the disconnect window for small mouse movements

This change adds a fudge factor of ~1px to the logic which determines
whether to reshow the disconnect window after it has been hidden.
I suspect the reports of the window being reshown are due to small
accumulations for the cursor position or potentially just raw input
events where the cursor does not change.  We want to avoid reshowing
the dialog in cases such as this because it is quite annoying for the
remote user.

Bug:  891584 

Ignore mouse messages if the position or button states don't change

We've received reports that the disconnect window is appearing at
odd times after the auto-hide timeout has taken effect.  I suspect
there could be hardware which is emitting spurious HID events which
is triggering the reshow logic.  My assumption is that this behavior
is't egregious, otherwise the user would notice the mouse cursor
moving or errant keypresses.  Thus I want to prevent triggering the
reshow logic if we receive a RAWINPUT event where the cursor has not
moved and no button states have changed.

Change-Id: I310f2cb4a5f30b0b7e14bf7221d250030492add8
Reviewed-on: https://chromium-review.googlesource.com/c/1258979
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Cr-Commit-Position: refs/heads/master@{#596308}
[modify] https://crrev.com/edc50bc7d6901596d8579db67470edd17e827e9c/remoting/host/disconnect_window_win.cc

Labels: Merge-Request-70
I'd like to request a merge for these two CLs into M70 if we aren't past the point where merges are allowed.

This change only affects the Chrome Remote Desktop host.  It addresses a problem we found during our 10% push.  Merging now would allow us to release this a couple of weeks earlier than waiting for M71.
Project Member

Comment 4 by sheriffbot@chromium.org, Oct 3

Labels: -Merge-Request-70 Merge-Review-70 Hotlist-Merge-Review
This bug requires manual review: We are only 12 days from stable.
Please contact the milestone owner if you have questions.
Owners: benmason@(Android), kariahda@(iOS), geohsu@(ChromeOS), abdulsyed@(Desktop)

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
Is this independent of the chrome release cycle?
Our team chooses release candidate builds from the release branch (to ensure the lower level components (base/net/etc) we rely on are stable.

We aren't tied to the specific Dev/Beta/Stable channel release process for Windows/Mac/Linux.  ChromeOS follows this process but the change I would like to merge only affects our Windows host.
Labels: -Merge-Review-70 Merge-Approved-70
Thanks!
Project Member

Comment 9 by bugdroid1@chromium.org, Oct 4

Labels: -merge-approved-70 merge-merged-3538
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/4b399f91a3ac7c75daf69674466e17853a2b19e8

commit 4b399f91a3ac7c75daf69674466e17853a2b19e8
Author: Joe Downing <joedow@chromium.org>
Date: Thu Oct 04 18:33:24 2018

Ignore mouse messages if the position or button states don't change

We've received reports that the disconnect window is appearing at
odd times after the auto-hide timeout has taken effect.  I suspect
there could be hardware which is emitting spurious HID events which
is triggering the reshow logic.  My assumption is that this behavior
is't egregious, otherwise the user would notice the mouse cursor
moving or errant keypresses.  Thus I want to prevent triggering the
reshow logic if we receive a RAWINPUT event where the cursor has not
moved and no button states have changed.

Bug:  891584 

Change-Id: Id9efd72f24f8caf0e67b1a717fee2bbafaab55af
Reviewed-on: https://chromium-review.googlesource.com/c/1258316
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#596121}(cherry picked from commit df242b026028cda8d2c3564f0e5129059d7bb7bd)
Reviewed-on: https://chromium-review.googlesource.com/c/1262419
Reviewed-by: Joe Downing <joedow@chromium.org>
Cr-Commit-Position: refs/branch-heads/3538@{#857}
Cr-Branched-From: 79f7c91a2b2a2932cd447fa6f865cb6662fa8fa6-refs/heads/master@{#587811}
[modify] https://crrev.com/4b399f91a3ac7c75daf69674466e17853a2b19e8/remoting/host/input_monitor/local_mouse_input_monitor_win.cc

Project Member

Comment 10 by bugdroid1@chromium.org, Oct 4

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/3c49fc56ac9a9d5e2db1d615b6cac8c27382539e

commit 3c49fc56ac9a9d5e2db1d615b6cac8c27382539e
Author: Joe Downing <joedow@chromium.org>
Date: Thu Oct 04 18:34:15 2018

Do not reshow the disconnect window for small mouse movements

This change adds a fudge factor of ~1px to the logic which determines
whether to reshow the disconnect window after it has been hidden.
I suspect the reports of the window being reshown are due to small
accumulations for the cursor position or potentially just raw input
events where the cursor does not change.  We want to avoid reshowing
the dialog in cases such as this because it is quite annoying for the
remote user.

Bug:  891584 

Ignore mouse messages if the position or button states don't change

We've received reports that the disconnect window is appearing at
odd times after the auto-hide timeout has taken effect.  I suspect
there could be hardware which is emitting spurious HID events which
is triggering the reshow logic.  My assumption is that this behavior
is't egregious, otherwise the user would notice the mouse cursor
moving or errant keypresses.  Thus I want to prevent triggering the
reshow logic if we receive a RAWINPUT event where the cursor has not
moved and no button states have changed.

Change-Id: I310f2cb4a5f30b0b7e14bf7221d250030492add8
Reviewed-on: https://chromium-review.googlesource.com/c/1258979
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#596308}(cherry picked from commit edc50bc7d6901596d8579db67470edd17e827e9c)
Reviewed-on: https://chromium-review.googlesource.com/c/1262420
Reviewed-by: Joe Downing <joedow@chromium.org>
Cr-Commit-Position: refs/branch-heads/3538@{#858}
Cr-Branched-From: 79f7c91a2b2a2932cd447fa6f865cb6662fa8fa6-refs/heads/master@{#587811}
[modify] https://crrev.com/3c49fc56ac9a9d5e2db1d615b6cac8c27382539e/remoting/host/disconnect_window_win.cc

Labels: Merge-Merged-70-3538
The following revision refers to this bug: 
https://chromium.googlesource.com/chromium/src.git/+/3c49fc56ac9a9d5e2db1d615b6cac8c27382539e

Commit: 3c49fc56ac9a9d5e2db1d615b6cac8c27382539e
Author: joedow@chromium.org
Commiter: joedow@chromium.org
Date: 2018-10-04 18:34:15 +0000 UTC

Do not reshow the disconnect window for small mouse movements

This change adds a fudge factor of ~1px to the logic which determines
whether to reshow the disconnect window after it has been hidden.
I suspect the reports of the window being reshown are due to small
accumulations for the cursor position or potentially just raw input
events where the cursor does not change.  We want to avoid reshowing
the dialog in cases such as this because it is quite annoying for the
remote user.

Bug:  891584 

Ignore mouse messages if the position or button states don't change

We've received reports that the disconnect window is appearing at
odd times after the auto-hide timeout has taken effect.  I suspect
there could be hardware which is emitting spurious HID events which
is triggering the reshow logic.  My assumption is that this behavior
is't egregious, otherwise the user would notice the mouse cursor
moving or errant keypresses.  Thus I want to prevent triggering the
reshow logic if we receive a RAWINPUT event where the cursor has not
moved and no button states have changed.

Change-Id: I310f2cb4a5f30b0b7e14bf7221d250030492add8
Reviewed-on: https://chromium-review.googlesource.com/c/1258979
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#596308}(cherry picked from commit edc50bc7d6901596d8579db67470edd17e827e9c)
Reviewed-on: https://chromium-review.googlesource.com/c/1262420
Reviewed-by: Joe Downing <joedow@chromium.org>
Cr-Commit-Position: refs/branch-heads/3538@{#858}
Cr-Branched-From: 79f7c91a2b2a2932cd447fa6f865cb6662fa8fa6-refs/heads/master@{#587811}
The following revision refers to this bug: 
https://chromium.googlesource.com/chromium/src.git/+/4b399f91a3ac7c75daf69674466e17853a2b19e8

Commit: 4b399f91a3ac7c75daf69674466e17853a2b19e8
Author: joedow@chromium.org
Commiter: joedow@chromium.org
Date: 2018-10-04 18:33:24 +0000 UTC

Ignore mouse messages if the position or button states don't change

We've received reports that the disconnect window is appearing at
odd times after the auto-hide timeout has taken effect.  I suspect
there could be hardware which is emitting spurious HID events which
is triggering the reshow logic.  My assumption is that this behavior
is't egregious, otherwise the user would notice the mouse cursor
moving or errant keypresses.  Thus I want to prevent triggering the
reshow logic if we receive a RAWINPUT event where the cursor has not
moved and no button states have changed.

Bug:  891584 

Change-Id: Id9efd72f24f8caf0e67b1a717fee2bbafaab55af
Reviewed-on: https://chromium-review.googlesource.com/c/1258316
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#596121}(cherry picked from commit df242b026028cda8d2c3564f0e5129059d7bb7bd)
Reviewed-on: https://chromium-review.googlesource.com/c/1262419
Reviewed-by: Joe Downing <joedow@chromium.org>
Cr-Commit-Position: refs/branch-heads/3538@{#857}
Cr-Branched-From: 79f7c91a2b2a2932cd447fa6f865cb6662fa8fa6-refs/heads/master@{#587811}
Status: Fixed (was: Assigned)

Sign in to add a comment