946999 - [Homescreen] With APZ turned on the homescreen is not rendered until you touch the screen in certain conditions

Status: RESOLVED FIXED

(Core :: Panning and Zooming, defect)

Description

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

With APZ turned on, when I use the device with the lockscreen disabled (via Settings -> Lock -> Lockscreen) and when the device starts the homescreen does not repaint until a finger is touching the screen.

I think I'm seeing a similar issue with the call screen once it has been minimized with the home button and then when it is reopened by either clicking on it or hitting the home button.

Comment 1

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Milan, can we fix this one before turning the pref on? It looks very bad on the device, mostly concerned about the call screen.

Comment 2

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Seems like we are missing a content repaint request at http://mxr.mozilla.org/mozilla-central/source/gfx/layers/ipc/AsyncPanZoomController.cpp#1450

Comment 3

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

First issue is that we don't check the viewport height at: http://mxr.mozilla.org/mozilla-central/source/gfx/layers/ipc/AsyncPanZoomController.cpp#1455

Second issue seems to be that we are over zealous to ignore paint request in http://mxr.mozilla.org/mozilla-central/source/gfx/layers/ipc/AsyncPanZoomController.cpp#1246

Digging into it to understand why.

Comment 4

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Attachment: bug948396.patch

Let's be less zealous and more open to viewport changes.

Comment 5

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

Wrong patch ^

Comment 6

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Sigh...

Comment 7

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Attachment: bug946999.patch

Let's see if this one is correct!

Comment 8

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

Comment on attachment 8345328 [details] [diff] [review]
bug946999.patch

Review of attachment 8345328 [details] [diff] [review]:
-----------------------------------------------------------------

r=me with the comments addressed

::: gfx/layers/ipc/AsyncPanZoomController.cpp
@@ +1273,5 @@
>        fabsf(mLastPaintRequestMetrics.mScrollOffset.y -
>              mFrameMetrics.mScrollOffset.y) < EPSILON &&
> +      mFrameMetrics.mZoom == mLastPaintRequestMetrics.mZoom &&
> +      mFrameMetrics.mViewport.width == mLastPaintRequestMetrics.mViewport.width &&
> +      mFrameMetrics.mViewport.height == mLastPaintRequestMetrics.mViewport.height) {

mViewport has floats, so you should do the fabsf(..) < EPSILON comparison

@@ +1474,1 @@
>        needContentRepaint = true;

Add braces here since you're touching the code anyway

Comment 9

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

https://tbpl.mozilla.org/?tree=Try&rev=1bacd4bd77cc

Comment 10

[Carsten Book [:Tomcat]]

Hey Milan, did you really wanted to mark this bug blocking/depend on 942746 - that was a update for shockwave :)

Comment 11

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

I spent time on this thinking there was a missed UpdateDimensions call but I found that this is actually a Gaia 1.3 regression (bug 949044)

Comment 12

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

https://hg.mozilla.org/integration/b2g-inbound/rev/bb7e20d28936

Comment 13

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/bb7e20d28936

Comment 14

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/f9f15a1dceae