1123921 - (containerless-root) Turn on containerless scrolling for root scroll frames on desktop

Status: RESOLVED FIXED

(Core :: Layout, defect)

Description

[David Anderson [:dvander] - inactive, e-mail if emergency]

Most recent try run: https://treeherder.mozilla.org/#/jobs?repo=try&revision=1122813f4e2e

Comment 1

[David Anderson [:dvander] - inactive, e-mail if emergency]

Attachment: desktop

The sanity test doesn't test anything interesting so I just removed it. The other two pass with APZ, and to change them to use reftest-async-scroll-y we'd need to have APZ on anyway. So I'm just skipping them without APZ.

Comment 2

[Kartikaya Gupta (email:[e-mail])]

Comment on attachment 8552946 [details] [diff] [review]
desktop

Review of attachment 8552946 [details] [diff] [review]:
-----------------------------------------------------------------

::: layout/reftests/bugs/reftest.list
@@ +1592,5 @@
>  == 589615-1b.html 589615-1-ref.html
>  == 589672-1.html 589672-1-ref.html
>  != 589682-1.html 589682-1-notref.html
> +skip-if(!browserIsRemote) skip-if(!asyncPanZoom) == 593243-1.html 593243-1-ref.html # bug 593168
> +skip-if(!browserIsRemote) skip-if(!asyncPanZoom) == 593243-2.html 593243-2-ref.html # bug 593168

If you can get rid of the skip-if(!browserIsRemote) clause that would be nice; I suspect the only reason that's there is to isolate async pan zoom platforms, and now you're doing that explicitly.

::: modules/libpref/init/all.js
@@ +573,4 @@
>  pref("gfx.hidpi.enabled", 2);
>  #endif
>  
> +#if !defined(ANDROID) && !defined(MOZ_B2G)

ANDROID is defined on both fennec and b2g device builds. MOZ_WIDGET_ANDROID and MOZ_WIDGET_GONK would isolate fennec and b2g device builds. MOZ_B2G i think is true on b2g desktop builds as well as b2g device builds. I'm not sure which combination you're actually looking for here - do the B2G desktop builds pass?

Comment 3

[Kartikaya Gupta (email:[e-mail])]

(In reply to David Anderson [:dvander] from comment #1)
> The sanity test doesn't test anything interesting so I just removed it. The
> other two pass with APZ, and to change them to use reftest-async-scroll-y
> we'd need to have APZ on anyway. So I'm just skipping them without APZ.

Also for posterity I'd like to point out that the test-displayport-2.html *does* test something interesting.

Comment 4

[David Anderson [:dvander] - inactive, e-mail if emergency]

remote:   https://hg.mozilla.org/integration/mozilla-inbound/rev/48171dc1a535

Comment 5

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/48171dc1a535

Comment 6

[Kartikaya Gupta (email:[e-mail])]

Closing because it's annoying to leave-open bugs across trains.

Comment 7

[Botond Ballo [:botond] [away until Oct 19]]

Adjusting title to reflect revised scope.