1852724 - Set titlebar state early to avoid further mShell/mContainer window reparent

Status: RESOLVED FIXED

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

Description

[Martin Stránský [:stransky] (ni? me)]

Set titlebar state early to avoid further mShell/mContainer window reparent.

Comment 1

[Martin Stránský [:stransky] (ni? me)]

Attachment: Bug 1852724 [Linux] Call nsWindow::SetDrawsInTitlebar() early to avoid mShell/Container widget reparent r?emilio

Comment 2

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/a28fc8d4ff32
[Linux] Call nsWindow::SetDrawsInTitlebar() early to avoid mShell/Container widget reparent r=emilio

Comment 3

[Sandor Molnar[:smolnar]]

https://hg.mozilla.org/mozilla-central/rev/a28fc8d4ff32

Comment 4

[Geoff Lankow (:darktrojan)]

This change gives me windows without a title bar in Thunderbird. Not the main window where we draw our own title bar, but things like the Developer Toolbox and Compose windows.

Comment 5

[Andrej (:aglavic)]

(In reply to Pulsebot from comment #2)

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/a28fc8d4ff32
[Linux] Call nsWindow::SetDrawsInTitlebar() early to avoid mShell/Container
widget reparent r=emilio

== Change summary for alert #39560 (as of Fri, 15 Sep 2023 18:44:43 GMT) ==

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
59%	tp5o XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,861.48 -> 1,085,148.60
59%	tp5o_webext XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,860.65 -> 1,085,000.20
59%	tp5o_webext XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,270.80 -> 1,084,932.50
3%	tresize	linux1804-64-shippable-qr	e10s fission stylo webrender	31.96 -> 31.06

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=39560

Comment 6

[Martin Stránský [:stransky] (ni? me)]

(In reply to Andrej Glavic (:aglavic) from comment #5)

(In reply to Pulsebot from comment #2)

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/a28fc8d4ff32
[Linux] Call nsWindow::SetDrawsInTitlebar() early to avoid mShell/Container
widget reparent r=emilio

== Change summary for alert #39560 (as of Fri, 15 Sep 2023 18:44:43 GMT) ==

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
59%	tp5o XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,861.48 -> 1,085,148.60
59%	tp5o_webext XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,860.65 -> 1,085,000.20
59%	tp5o_webext XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,270.80 -> 1,084,932.50
3%	tresize	linux1804-64-shippable-qr	e10s fission stylo webrender	31.96 -> 31.06

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=39560

Do I understand correctly that it comes with significant performance gain? Can you give me more info what the test actually check? I may consider to use some narrowed optimized variant of the patch if it's so significant gain here.

Thanks.

Comment 7

[Andrej (:aglavic)]

(In reply to Martin Stránský [:stransky] (ni? me) from comment #6)

(In reply to Andrej Glavic (:aglavic) from comment #5)

(In reply to Pulsebot from comment #2)

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/a28fc8d4ff32
[Linux] Call nsWindow::SetDrawsInTitlebar() early to avoid mShell/Container
widget reparent r=emilio

== Change summary for alert #39560 (as of Fri, 15 Sep 2023 18:44:43 GMT) ==

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
59%	tp5o XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,861.48 -> 1,085,148.60
59%	tp5o_webext XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,860.65 -> 1,085,000.20
59%	tp5o_webext XRes	linux1804-64-shippable-qr	e10s fission stylo webrender	2,654,270.80 -> 1,084,932.50
3%	tresize	linux1804-64-shippable-qr	e10s fission stylo webrender	31.96 -> 31.06

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=39560

Do I understand correctly that it comes with significant performance gain? Can you give me more info what the test actually check? I may consider to use some narrowed optimized variant of the patch if it's so significant gain here.

Thanks.

Hi :stransky

This ticket caused 1852918 which because of not displaying the title bar, artificially improved the metrics
See my comment in 1853972 which was initially identified as a regression but just brought back the metric to it's expected value
Sorry for the confusion