2003258 - [Wayland] Submenu position intermittently slightly off

Status: RESOLVED FIXED

(Core :: Widget: Gtk, defect, P3)

Description

[Manuel Bucher [:manuel]]

Attachment: menu-position-off.mp4

I'm unable to do a good mozregression, but I believe it broke somewhere in a9b1de7c0585b7f4b77990b2e6f51f796ed3e4e4...3befa2ffc228fd4d071c8b696872a760954392fc. ./mach mozregression --good 2025-10-30 --bad 2025-11-01

With my last regression I ended up with the regression window: 2c449ac57a9062ead7243b27bb5dceb94b804c5c...d51bf7d6db72d1f039cd874eee72bfdd352f59af

Due to the intermittent nature of the bug, I ended on three different unrelated bugs. So I'm not 100% confident that I've set the correct regression bug. However, it does look at least somewhat related. Sorry if I tagged the wrong regression bug.

I can reproduce on a current build, so it's not fixed by the other regression bug (bug 1998949)

Comment 1

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1992198

:stransky, since you are the author of the regressor, bug 1992198, could you take a look?

For more information, please visit BugBot documentation.

Comment 2

[Martin Stránský [:stransky] (ni? me) (PTO till Aug 03)]

Checked but unable to reproduce. Which system is that? Please attach your about:support.
Thanks.

Comment 3

[Martin Stránský [:stransky] (ni? me) (PTO till Aug 03)]

Can you possibly run it on terminal with MOZ_LOG="WidgetPopup:5" env variable and attach the log here if you can reproduce it?
Thanks.

Comment 4

[Manuel Bucher [:manuel]]

Attachment: log.txt-main.34665.moz_log

Comment 5

[Manuel Bucher [:manuel]]

Attachment: about:support.json

Comment 6

[Manuel Bucher [:manuel]]

I enabled logging from about:logging for producing this log, because its more convenient in combination with ./mach mozregression for a clean profile. Let me know if you need the log from application start. I'd then record again.

Comment 7

[Martin Stránský [:stransky] (ni? me) (PTO till Aug 03)]

I see parent position change according to the log but the screencast shows no parent popup change but the child one. Also according to the log the popup is moved left, not right - I wonder how's that possible.

Comment 8

[Emilio Cobos Álvarez [:emilio]]

Attachment: This reproduces very easily if you use the context menu near the bottom of the screen.

Comment 9

[Steven DeTar [:sdetar]]

Could we get a severity set for this bug
(from REO Fx146)

Comment 10

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1992198

Comment 11

[Emilio Cobos Álvarez [:emilio]]

So I can reproduce this only if I have a 1.5 dpi, does that match your set-up?

Comment 12

[Manuel Bucher [:manuel]]

Yes, my monitors are set to 150% scaling.

Comment 13

[Emilio Cobos Álvarez [:emilio]]

Attachment: Bug 2003258 - Stop rounding popup coordinates to the GTK ceiled scale. r=stransky!

Now that we use unscaled coordinates internally, there's no guarantee
that the coordinates that we set in device pixels end up rounding up to
the multiple of the scale, which can cause menus to drift with scalings
that are >= 1.5.

Comment 14

[Emilio Cobos Álvarez [:emilio]]

Attachment: Bug 2003258 - Stop rounding popup coordinates to the GTK ceiled scale. r=stransky!

Now that we use unscaled coordinates internally, there's no guarantee
that the coordinates that we set in device pixels end up rounding up to
the multiple of the scale, which can cause menus to drift with scalings
that are >= 1.5.

Comment 15

[Martin Stránský [:stransky] (ni? me) (PTO till Aug 03)]

Let's see if we have any regression on move-to-rect scenario and popup reposition:
https://searchfox.org/firefox-main/rev/929fb12c1318659f24965f2908dec6e5fd56d9d9/widget/gtk/nsWindow.cpp#2088
maybe we may update it to scale*2 or so?

Comment 16

[Pulsebot]

Pushed by [e-mail]:
https://github.com/mozilla-firefox/firefox/commit/ab68768044ea
https://hg.mozilla.org/integration/autoland/rev/fba0f8de63ba
Stop rounding popup coordinates to the GTK ceiled scale. r=stransky

Comment 17

[Serban Stanca [:SerbanS]]

https://hg.mozilla.org/mozilla-central/rev/fba0f8de63ba

Comment 18

[BugBot [:suhaib / :marco/ :calixte]]

The patch landed in nightly and beta is affected.
:emilio, is this bug important enough to require an uplift?

• If yes, please nominate the patch for beta approval.
  • See https://wiki.mozilla.org/Release_Management/Requesting_an_Uplift for documentation on how to request an uplift.
• If no, please set status-firefox147 to wontfix.

For more information, please visit BugBot documentation.

Comment 19

[Emilio Cobos Álvarez [:emilio]]

These changes are never risk-free, so let's go with wontfix for now since there are other changes in flight in the area.